Soccer predictions: actual data vs predicted data

Slate 20261031 · 12 match(es)

How to read this report. Each match shows the model's headline prediction, the stage-by-stage lambda (goal-rate) build, both teams' standing/form inputs, and every starter's projected line next to their real season baseline. A moderate / large deviation badge flags a projection that strays far from the player's season form — a quick way to spot when the inputs look right but a projection does not. When the real result is known an actual-vs-predicted banner appears.

All matches overview

Match	Score	Winner	H / D / A	O2.5	BTTS	Conf	λ H/A	Actual	Result
Barcelona vs Alavés LaLiga	3-1	Barcelona	82.3% / 9.9% / 7.8%	88.7%	70.9%	0.83	3.86 / 1.29	-	pending
Levante vs Atlético Madrid LaLiga	1-2	Atlético Madrid	18.9% / 19.5% / 61.5%	67.1%	62.4%	0.77	1.19 / 2.27	-	pending
Man City vs Brighton Premier League	2-1	Man City	68.6% / 16.7% / 14.6%	72.6%	63.5%	0.79	2.62 / 1.15	-	pending
Rayo Vallecano vs Celta LaLiga	2-1	Rayo Vallecano	47.3% / 23.9% / 28.7%	57.3%	59.1%	0.73	1.70 / 1.28	-	pending
Tottenham vs Crystal Palace Premier League	1-0	Tottenham	43.6% / 25.3% / 31.1%	52.6%	56.0%	0.72	1.53 / 1.25	-	pending
Aston Villa vs Fulham Premier League	1-0	Aston Villa	55.4% / 23.4% / 21.2%	52.6%	52.7%	0.75	1.78 / 1.00	-	pending
Hull vs Ipswich Premier League	1-1	Draw	38.0% / 25.6% / 36.4%	52.6%	56.5%	0.7	1.41 / 1.37	-	pending
Bournemouth vs Leeds Premier League	1-0	Bournemouth	45.4% / 25.1% / 29.5%	52.6%	55.7%	0.73	1.57 / 1.22	-	pending
Real Betis vs Málaga LaLiga	2-0	Real Betis	67.7% / 19.1% / 13.2%	57.3%	49.9%	0.79	2.15 / 0.83	-	pending
Chelsea vs Man Utd Premier League	2-1	Chelsea	43.2% / 23.9% / 32.9%	59.5%	61.5%	0.71	1.66 / 1.42	-	pending
Brentford vs Forest Premier League	1-0	Brentford	51.4% / 24.3% / 24.3%	52.6%	54.2%	0.75	1.69 / 1.09	-	pending
Coventry vs Sunderland Premier League	1-1	Draw	37.1% / 25.6% / 37.3%	52.6%	56.5%	0.69	1.39 / 1.39	-	pending

Match detail

Barcelona vs Alavés

LaLiga · 20261031 · 4-3-3 vs 4-4-2

Predicted 3-1Winner: BarcelonaWinning H 3-1 (7.2%)Winning A 1-2 (1.9%)Confidence 0.83λ home/away 3.86 / 1.29Most-likely P 7.2%

Actual vs predicted: result not available yet (upcoming fixture).

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Barcelona vs Alavés. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 4-3-3 vs 4-4-2 (duel lambdas: Home 4.442, Away 1.429). Final calibrated Poisson lambdas: Home 3.863, Away 1.288 (most likely score 3-1).

Home Starters 10/11 XGBoost (9 FBref) Away Starters 11/11 XGBoost (8 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 4.44 / 1.43 Final λ (H/A) 3.86 / 1.29

FBref Team Priors: FC Barcelona (Gls/90: 3.6666666666666665, GA/90: 4.0 → E[λ]: 3.5) vs Deportivo Alavés (Gls/90: 1.6666666666666667, GA/90: 1.67 → E[λ]: 3.5) | Blend Weight: 0.228

Modeling Pipeline Stages (5 layers)

1. XGBoost player regressions for volume & event rates with market-value quality scaling
2. Multi-source prior blending (FBref 2026-2027 tables + Sofascore season history)
3. Antagonist formation duel matrix resolving 1v1 and zonal matchups
4. Team-level prior calibration (FBref possession and attack/defense efficiencies)
5. Bivariate Poisson matrix with Dixon-Coles low-score correction

Standing & form context

Metric	Barcelona	Alavés
League position	1.00	7.00
Points per game	2.20	1.26
Attack index	1.76	0.97
Defense (concede) index	0.88	0.88
Form multiplier	1.04	0.96
Standing λ multiplier	1.20	0.93

Lambda decomposition (how the goal rates were built)

Stage	Barcelona λ	Alavés λ
FBref Squad Prior E[λ] (weight 0.228)	3.500	3.500
Pre-mismatch	5.096	1.327
Pre-calibration (Micro-duels)	4.442	1.429
Final λ	3.863	1.288

competition calibration scale 0.85competition observed total goals 2.984quality gap applied Truemismatch score pressure applied Truecompetitive underdog floor applied Falsestage discount 1.0

Scoreline distribution & markets

Top 5 scorelines	P
3-1	7.2%
4-1	7.0%
2-1	5.6%
3-0	5.6%
4-0	5.4%

Home win	82.3%
Draw	9.9%
Away win	7.8%
Over 2.5	88.7%
BTTS yes	70.9%

Pregame fair line — favorite Barcelona (82.3%), fair total 5.15, fair spread 2.58

Projected players vs season baseline

Proj = projected for this match · Season = this season's per-game average from match data · deviation badge compares projected xG to the player's season xG/game.

Barcelona — projected line vs season baseline (4-3-3)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Joan García G 86 0 0.00 - n/a 0.0 6.9 60.9 32 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0104, quality 0.809)	`market_value_eur=46000000` `quality_factor=0.8089` `boost_factor=1.0104` `reference_quality=0.505` `history_matches=39` `history_shrink=0.1136`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0148)	`blend_weight=0.0148` `history_matches=39` `season_overlap_discount=1.0` `per90_to_match_factor=0.9561` `prior_reliability=0.2` `competition=La Liga`	saves: 2.9953 -> 2.9746
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Toni Martínez, Lucas Boyé', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.4285}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 3 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=3`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 3.86 (squad xG sum 2.6208, xG scale 1.4740, xA scale 1.6297, team xA 2.51 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=3.863` `squad_xg_sum=2.6208` `squad_xa_sum=1.5408` `team_expected_assists=2.511` `team_xa_to_xg_ratio=0.65`	shots_faced: 13.6 -> 12.2 expected_saves: 3.33 -> 2.98 high_pressure_saves: 1.14 -> 1.03
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 13 by projected rating, fifa_overall 86.0)	`match_squad_rank=13` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 3 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=3` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=4.4245`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=2.7263`	-

Brian Fariñas M 90 0 0.13 - n/a 2 0.0 6.9 - 9 Prior M_service

Step-by-step (8 steps)

Step	Description	Key params	Changes
`market_value_prior`	Market-value baseline for position M (quality 0.47)	`market_value_eur=5500000` `quality_factor=0.4683`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Pablo Ibáñez', 'matchup_zone': 'Central Mid vs Central Mid — Box-to-Box Midfield Contest', 'duel_win_rate': 0.433, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.0185, 'skill_edge': 0.0086, 'height_edge': 0.004, 'attacker_skill': 0.4758, 'defender_skill': 0.4567, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Pablo Ibáñez', 'market_value': 4099999, 'role': 'CM', 'price_index': 0.9, 'adjusted_quality': 0.4567}], 'xg_override': 0.0876}	aerialWon: 1.2 -> 1.1 totalShots: 1.3 -> 1.324 onTargetScoringAttempt: 0.6 -> 0.6111 shotOffTarget: 0.5 -> 0.5092
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 3 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=3`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 3.86 (squad xG sum 2.6208, xG scale 1.4740, xA scale 1.6297, team xA 2.51 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=3.863` `squad_xg_sum=2.6208` `squad_xa_sum=1.5408` `team_expected_assists=2.511` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0876 -> 0.1291 expected_assists: 0.133 -> 0.2167
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 14 by projected rating, fifa_overall 83.0)	`match_squad_rank=14` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 3 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=3` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1211` `goal_rate=0.0713` `scorer_weight=0.1138`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=0` `assist_probability=0.1948` `assist_rate=0.11` `provider_weight=0.1895`	-

Pau Cubarsí D 87 0 0.05 0.02 113% 0 0.2 7.0 53.4 33 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0135, quality 0.891)	`market_value_eur=94000000` `quality_factor=0.891` `boost_factor=1.0135` `reference_quality=0.505` `history_matches=38` `history_shrink=0.1163`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0101)	`blend_weight=0.0101` `history_matches=38` `season_overlap_discount=1.0` `per90_to_match_factor=0.9617` `prior_reliability=0.1333` `competition=La Liga`	wonTackle: 0.7281 -> 0.7353 interceptionWon: 1.1709 -> 1.1639 totalShots: 0.424 -> 0.4294
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Lucas Boyé', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 3 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=3`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 3.86 (squad xG sum 2.6208, xG scale 1.4740, xA scale 1.6297, team xA 2.51 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=3.863` `squad_xg_sum=2.6208` `squad_xa_sum=1.5408` `team_expected_assists=2.511` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0308 -> 0.0454 expected_assists: 0.0963 -> 0.1569
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 10 by projected rating, fifa_overall 86.0)	`match_squad_rank=10` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 3 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=3` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0444` `goal_rate=0.0081` `scorer_weight=0.0363`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=0` `assist_probability=0.1452` `assist_rate=0.2563` `provider_weight=0.1961`	-

Frenkie de Jong M 89 0 0.11 0.04 153% 1 0.2 7.0 51.8 18 XGBoost M_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0075, quality 0.660)	`market_value_eur=23000000` `quality_factor=0.6595` `boost_factor=1.0075` `reference_quality=0.505` `history_matches=26` `history_shrink=0.1613`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Pablo Ibáñez', 'matchup_zone': 'Central Mid vs Central Mid — Box-to-Box Midfield Contest', 'duel_win_rate': 0.525, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.155, 'skill_edge': 0.1014, 'height_edge': 0.004, 'attacker_skill': 0.6821, 'defender_skill': 0.4567, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Pablo Ibáñez', 'market_value': 4099999, 'role': 'CM', 'price_index': 0.9, 'adjusted_quality': 0.4567}], 'xg_override': 0.0767}	aerialWon: 0.5341 -> 1.1 totalShots: 0.6719 -> 0.776 shotOffTarget: 0.4821 -> 0.5568 onTargetScoringAttempt: 0.1898 -> 0.2192 expectedGoalsOnTarget: 0.1048 -> 0.121 goals: 0.0747 -> 0.0863 expectedGoals: 0.0664 -> 0.0767
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 3 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=3`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 3.86 (squad xG sum 2.6208, xG scale 1.4740, xA scale 1.6297, team xA 2.51 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=3.863` `squad_xg_sum=2.6208` `squad_xa_sum=1.5408` `team_expected_assists=2.511` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0767 -> 0.1131 expected_assists: 0.102 -> 0.1662
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 86.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 3 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=3` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1069` `goal_rate=0.0863` `scorer_weight=0.1217`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=0` `assist_probability=0.1531` `assist_rate=0.0851` `provider_weight=0.1477`	-

João Cancelo D 84 0 0.06 0.02 188% 1 0.6 6.8 50.4 32 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9993, quality 0.493)	`market_value_eur=7500000` `quality_factor=0.4927` `boost_factor=0.9993` `reference_quality=0.505` `history_matches=23` `history_shrink=0.1786`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0023)	`blend_weight=0.0023` `history_matches=23` `season_overlap_discount=1.0` `per90_to_match_factor=0.937` `prior_reliability=0.02` `competition=La Liga`	totalCross: 2.686 -> 2.6798
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Toni Martínez', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 3 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=3`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 3.86 (squad xG sum 2.6208, xG scale 1.4740, xA scale 1.6297, team xA 2.51 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=3.863` `squad_xg_sum=2.6208` `squad_xa_sum=1.5408` `team_expected_assists=2.511` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0392 -> 0.0578 expected_assists: 0.1491 -> 0.243
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 15 by projected rating, fifa_overall 83.0)	`match_squad_rank=15` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 3 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=3` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0562` `goal_rate=0.2556` `scorer_weight=0.0722`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=0` `assist_probability=0.2157` `assist_rate=0.0624` `provider_weight=0.1944`	-

Pedri M 83 0 0.13 0.04 223% 1 0.5 7.2 63.1 29 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0135, quality 0.900)	`market_value_eur=164000000` `quality_factor=0.8997` `boost_factor=1.0135` `reference_quality=0.505` `history_matches=39` `history_shrink=0.1136`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0113)	`blend_weight=0.0113` `history_matches=39` `season_overlap_discount=1.0` `per90_to_match_factor=0.9171` `prior_reliability=0.1533` `competition=La Liga`	wonTackle: 0.9198 -> 0.9274
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Pablo Ibáñez', 'matchup_zone': 'Central Mid vs Central Mid — Box-to-Box Midfield Contest', 'duel_win_rate': 0.599, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.2638, 'skill_edge': 0.1994, 'height_edge': -0.02, 'attacker_skill': 0.8999, 'defender_skill': 0.4567, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Pablo Ibáñez', 'market_value': 4099999, 'role': 'CM', 'price_index': 0.9, 'adjusted_quality': 0.4567}], 'xg_override': 0.091}	aerialWon: 0.2217 -> 0.9 totalShots: 0.7873 -> 0.995 shotOffTarget: 0.59 -> 0.7456 onTargetScoringAttempt: 0.195 -> 0.2464 goals: 0.0944 -> 0.1193 expectedGoals: 0.072 -> 0.091 expectedGoalsOnTarget: 0.0482 -> 0.0609 bigChanceMissed: 0.0224 -> 0.0283
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 3 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=3`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 3.86 (squad xG sum 2.6208, xG scale 1.4740, xA scale 1.6297, team xA 2.51 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=3.863` `squad_xg_sum=2.6208` `squad_xa_sum=1.5408` `team_expected_assists=2.511` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.091 -> 0.1341 expected_assists: 0.2018 -> 0.3289
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 90.0)	`match_squad_rank=4` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 3 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=3` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1255` `goal_rate=0.1193` `scorer_weight=0.161`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=1` `assist_probability=0.2803` `assist_rate=0.2228` `provider_weight=0.3563`	-

Fermín López M 77 0 0.34 0.25 36% 3 1.8 7.0 62.1 31 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0133, quality 0.896)	`market_value_eur=113000000` `quality_factor=0.8965` `boost_factor=1.0133` `reference_quality=0.505` `history_matches=39` `history_shrink=0.1136`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0113)	`blend_weight=0.0113` `history_matches=39` `season_overlap_discount=1.0` `per90_to_match_factor=0.8559` `prior_reliability=0.1533` `competition=La Liga`	onTargetScoringAttempt: 0.7314 -> 0.7441 goals: 0.1863 -> 0.1968 wasFouled: 0.8553 -> 0.8625 wonTackle: 0.4933 -> 0.5004 totalCross: 1.6194 -> 1.6137
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Pablo Ibáñez', 'matchup_zone': 'Central Mid vs Central Mid — Box-to-Box Midfield Contest', 'duel_win_rate': 0.607, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.2744, 'skill_edge': 0.1986, 'height_edge': -0.012, 'attacker_skill': 0.8981, 'defender_skill': 0.4567, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Pablo Ibáñez', 'market_value': 4099999, 'role': 'CM', 'price_index': 0.9, 'adjusted_quality': 0.4567}], 'xg_override': 0.2295}	totalShots: 2.0116 -> 2.5636 aerialWon: 0.5412 -> 0.9 shotOffTarget: 1.2675 -> 1.6153 onTargetScoringAttempt: 0.7441 -> 0.9483 expectedGoalsOnTarget: 0.4775 -> 0.6085 goals: 0.1968 -> 0.2508 expectedGoals: 0.1801 -> 0.2295 bigChanceMissed: 0.1588 -> 0.2024
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 3 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=3`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 3.86 (squad xG sum 2.6208, xG scale 1.4740, xA scale 1.6297, team xA 2.51 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=3.863` `squad_xg_sum=2.6208` `squad_xa_sum=1.5408` `team_expected_assists=2.511` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.2295 -> 0.3383 expected_assists: 0.1662 -> 0.2708
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 90.0)	`match_squad_rank=8` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 3 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=3` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.287` `goal_rate=0.2508` `scorer_weight=0.3641`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=0` `assist_probability=0.2372` `assist_rate=0.1934` `provider_weight=0.3049`	-

Rodri M 83 0 0.18 0.07 151% 1 0.7 7.3 67.7 27 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0142, quality 0.837)	`market_value_eur=54000000` `quality_factor=0.837` `boost_factor=1.0142` `reference_quality=0.505` `history_matches=30` `history_shrink=0.1429`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0037)	`blend_weight=0.0037` `history_matches=30` `season_overlap_discount=1.0` `per90_to_match_factor=0.9259` `prior_reliability=0.04` `competition=La Liga`	wonTackle: 1.3958 -> 1.3906
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Pablo Ibáñez', 'matchup_zone': 'Holding Mid vs Central Mid — Midfield Anchor Duel', 'duel_win_rate': 0.629, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.3373, 'skill_edge': 0.1815, 'height_edge': 0.048, 'attacker_skill': 0.8599, 'defender_skill': 0.4567, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Pablo Ibáñez', 'market_value': 4099999, 'role': 'CM', 'price_index': 0.9, 'adjusted_quality': 0.4567}], 'xg_override': 0.1252}	totalShots: 0.8991 -> 1.2024 aerialWon: 1.5695 -> 1.3 shotOffTarget: 0.5837 -> 0.7806 onTargetScoringAttempt: 0.3154 -> 0.4218 expectedGoalsOnTarget: 0.1386 -> 0.1853 goals: 0.1042 -> 0.1393 bigChanceMissed: 0.0953 -> 0.1274 expectedGoals: 0.0936 -> 0.1252
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 3 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=3`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 3.86 (squad xG sum 2.6208, xG scale 1.4740, xA scale 1.6297, team xA 2.51 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=3.863` `squad_xg_sum=2.6208` `squad_xa_sum=1.5408` `team_expected_assists=2.511` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1252 -> 0.1845 expected_assists: 0.1147 -> 0.1869
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 90.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 3 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=3` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1685` `goal_rate=0.1393` `scorer_weight=0.1994`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=0` `assist_probability=0.1705` `assist_rate=0.107` `provider_weight=0.1782`	-

Lamine Yamal F 83 1 0.95 0.48 96% 6 3.7 8.2 66.2 32 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0138, quality 0.900)	`market_value_eur=237000000` `quality_factor=0.9` `boost_factor=1.0138` `reference_quality=0.505` `history_matches=38` `history_shrink=0.1163`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0136)	`blend_weight=0.0136` `history_matches=38` `season_overlap_discount=1.0` `per90_to_match_factor=0.9208` `prior_reliability=0.18` `competition=La Liga`	totalOffside: 0.7127 -> 0.7308 totalCross: 2.7226 -> 2.7366 goalAssist: 0.6608 -> 0.6518 fouls: 1.3046 -> 1.2961 interceptionWon: 0.2848 -> 0.2902
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Nahuel Tenaglia, Ville Koski', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.395, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.2867, 'skill_edge': 0.197, 'height_edge': -0.002, 'attacker_skill': 0.9, 'defender_skill': 0.4622, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Nahuel Tenaglia', 'market_value': 3200000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4531}, {'player': 'Ville Koski', 'market_value': 4300000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4712}], 'xg_override': 0.6414}	totalShots: 4.8881 -> 6.2896 shotOffTarget: 3.4296 -> 4.4129 onTargetScoringAttempt: 1.4585 -> 1.8767 goals: 0.6802 -> 0.8752 expectedGoalsOnTarget: 0.5994 -> 0.7712 expectedGoals: 0.4985 -> 0.6414 bigChanceMissed: 0.4274 -> 0.5499
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 3 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=3`	goals: 1.0 -> 0 assists: 1.0 -> 0
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 3.86 (squad xG sum 2.6208, xG scale 1.4740, xA scale 1.6297, team xA 2.51 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=3.863` `squad_xg_sum=2.6208` `squad_xa_sum=1.5408` `team_expected_assists=2.511` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.6414 -> 0.9454 expected_assists: 0.2496 -> 0.4069
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 90.0)	`match_squad_rank=1` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 3 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=3` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.6115` `goal_rate=0.8752` `scorer_weight=1.1818`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=1` `assist_probability=0.3343` `assist_rate=0.6518` `provider_weight=0.5086`	-

Raphinha F 79 2 1.21 0.52 132% 5 2.3 7.9 58.9 27 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0147, quality 0.859)	`market_value_eur=63000000` `quality_factor=0.8589` `boost_factor=1.0147` `reference_quality=0.505` `history_matches=31` `history_shrink=0.1389`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0150)	`blend_weight=0.015` `history_matches=31` `season_overlap_discount=1.0` `per90_to_match_factor=0.8749` `prior_reliability=0.1667` `competition=La Liga`	totalCross: 2.6259 -> 2.639 goals: 0.9037 -> 0.9164 onTargetScoringAttempt: 1.6214 -> 1.6337 fouls: 0.2857 -> 0.2972 totalOffside: 0.5786 -> 0.5857 wonTackle: 0.361 -> 0.3556
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Jonny Otto', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle', 'duel_win_rate': 0.674, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.2847, 'skill_edge': 0.1897, 'height_edge': 0.004, 'attacker_skill': 0.8589, 'defender_skill': 0.4374, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Jonny Otto', 'market_value': 2100000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.4374}], 'xg_override': 0.8222}	totalShots: 3.7706 -> 4.8441 shotOffTarget: 2.1369 -> 2.7453 onTargetScoringAttempt: 1.6337 -> 2.0988 expectedGoalsOnTarget: 0.9631 -> 1.2373 goals: 0.9164 -> 1.1773 expectedGoals: 0.64 -> 0.8222 bigChanceMissed: 0.6265 -> 0.8049
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 3 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=3`	goals: 1.0 -> 0
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 3.86 (squad xG sum 2.6208, xG scale 1.4740, xA scale 1.6297, team xA 2.51 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=3.863` `squad_xg_sum=2.6208` `squad_xa_sum=1.5408` `team_expected_assists=2.511` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.8222 -> 1.2119 expected_assists: 0.1493 -> 0.2433
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 88.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 3 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=3` `allocation_applied=true` `goals_allocated=2` `goal_probability=0.7024` `goal_rate=1.1773` `scorer_weight=1.5149`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=0` `assist_probability=0.216` `assist_rate=0.2542` `provider_weight=0.3041`	-

Karim Adeyemi F 75 0 0.45 0.34 33% 3 1.2 7.0 55.3 14 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0281, quality 0.786)	`market_value_eur=41000000` `quality_factor=0.7859` `boost_factor=1.0281` `reference_quality=0.505` `history_matches=10` `history_shrink=0.3333`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0145)	`blend_weight=0.0145` `history_matches=10` `season_overlap_discount=1.0` `per90_to_match_factor=0.8362` `prior_reliability=0.0667` `competition=La Liga`	onTargetScoringAttempt: 0.7899 -> 0.8151 wasFouled: 2.7356 -> 2.7202 goals: 0.1975 -> 0.2067 goalAssist: 0.3397 -> 0.3469 totalShots: 2.0611 -> 2.0679 totalOffside: 0.3996 -> 0.3938
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Hugo Novoa', 'matchup_zone': 'Right Wing vs Left Back — Flank Battle', 'duel_win_rate': 0.629, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.2188, 'skill_edge': 0.1568, 'height_edge': -0.008, 'attacker_skill': 0.7859, 'defender_skill': 0.4374, 'attacker_role': 'RW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Hugo Novoa', 'market_value': 725000, 'role': 'LB', 'price_index': 0.7, 'adjusted_quality': 0.4374}], 'xg_override': 0.306}	totalShots: 2.0862 -> 2.5427 bigChanceMissed: 1.4188 -> 1.7292 shotOffTarget: 1.2711 -> 1.5492 onTargetScoringAttempt: 0.8151 -> 0.9934 expectedGoals: 0.2511 -> 0.306 goals: 0.2067 -> 0.2519 expectedGoalsOnTarget: 0.1359 -> 0.1656
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 3 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=3`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 3.86 (squad xG sum 2.6208, xG scale 1.4740, xA scale 1.6297, team xA 2.51 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=3.863` `squad_xg_sum=2.6208` `squad_xa_sum=1.5408` `team_expected_assists=2.511` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.306 -> 0.451 expected_assists: 0.1008 -> 0.1643
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 11 by projected rating, fifa_overall 82.0)	`match_squad_rank=11` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 3 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=3` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.363` `goal_rate=0.2519` `scorer_weight=0.3824`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=0` `assist_probability=0.1515` `assist_rate=0.3469` `provider_weight=0.2054`	-

Team totals 3 3.86 26 7.2

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Dominik Livaković	G	4.5m	11.1%	Market-value substitute prior
Jules Koundé	D	12.9m	50.8%	Market-value substitute prior
Alejandro Balde	D	15.2m	37.1%	Market-value substitute prior
Ferran Torres	M	23.6m	57.3%	Market-value substitute prior
Dani Olmo	M	25.5m	59.7%	Market-value substitute prior
Anthony Gordon	F	26.5m	49.8%	Market-value substitute prior

Alavés — projected line vs season baseline (4-4-2)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Antonio Sivera G 90 0 0.00 - n/a 0.0 7.0 63.3 34 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9990, quality 0.470)	`market_value_eur=5600000` `quality_factor=0.4695` `boost_factor=0.999` `reference_quality=0.505` `history_matches=48` `history_shrink=0.0943`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0123)	`blend_weight=0.0123` `history_matches=48` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.2` `competition=La Liga`	saves: 2.8757 -> 2.8813
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Raphinha, Lamine Yamal', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 4.4418}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.29 (squad xG sum 1.0060, xG scale 1.2803, xA scale 1.2803, team xA 0.96 = 0.74 x lambda)	`scoreline_reset=true` `team_lambda=1.288` `squad_xg_sum=1.006` `squad_xa_sum=0.7484` `team_expected_assists=0.9582` `team_xa_to_xg_ratio=0.7439`	shots_faced: 42.2 -> 36.7 expected_saves: 10.33 -> 8.98 high_pressure_saves: 3.55 -> 3.09
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 85.0)	`match_squad_rank=1` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=1.3792`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=0.8425`	-

Hugo Novoa D 90 0 0.05 0.00 883% 1 0.1 6.8 10.9 21 FBref Prior D_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9855, quality 0.437)	`market_value_eur=725000` `quality_factor=0.4374` `boost_factor=0.9855` `reference_quality=0.505` `history_matches=2` `history_shrink=0.7143`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Karim Adeyemi', 'matchup_zone': 'Right Wing vs Left Back — Flank Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.29 (squad xG sum 1.0060, xG scale 1.2803, xA scale 1.2803, team xA 0.96 = 0.74 x lambda)	`scoreline_reset=true` `team_lambda=1.288` `squad_xg_sum=1.006` `squad_xa_sum=0.7484` `team_expected_assists=0.9582` `team_xa_to_xg_ratio=0.7439`	expected_goals: 0.0373 -> 0.0478 expected_assists: 0.1109 -> 0.142
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 10 by projected rating, fifa_overall 64.0)	`match_squad_rank=10` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0467` `goal_rate=0.4528` `scorer_weight=0.0598`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1324` `assist_rate=0.0` `provider_weight=0.1136`	-

Nahuel Tenaglia D 87 0 0.07 0.04 70% 1 0.7 6.8 60.1 34 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9981, quality 0.440)	`market_value_eur=3200000` `quality_factor=0.4399` `boost_factor=0.9981` `reference_quality=0.505` `history_matches=46` `history_shrink=0.098`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0127)	`blend_weight=0.0127` `history_matches=46` `season_overlap_discount=1.0` `per90_to_match_factor=0.968` `prior_reliability=0.2` `competition=La Liga`	totalCross: 1.5006 -> 1.4897 wasFouled: 1.7079 -> 1.6985 interceptionWon: 1.0765 -> 1.0833 totalShots: 0.8338 -> 0.8396
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Lamine Yamal', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.29 (squad xG sum 1.0060, xG scale 1.2803, xA scale 1.2803, team xA 0.96 = 0.74 x lambda)	`scoreline_reset=true` `team_lambda=1.288` `squad_xg_sum=1.006` `squad_xa_sum=0.7484` `team_expected_assists=0.9582` `team_xa_to_xg_ratio=0.7439`	expected_goals: 0.0578 -> 0.074 expected_assists: 0.035 -> 0.0448
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 77.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0713` `goal_rate=0.0587` `scorer_weight=0.0746`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0438` `assist_rate=0.0` `provider_weight=0.0358`	-

Ville Koski D 90 0 0.12 0.04 208% 1 0.2 6.8 38.2 25 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9961, quality 0.454)	`market_value_eur=4300000` `quality_factor=0.4535` `boost_factor=0.9961` `reference_quality=0.505` `history_matches=15` `history_shrink=0.25`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0314)	`blend_weight=0.0314` `history_matches=15` `season_overlap_discount=1.0` `per90_to_match_factor=0.9947` `prior_reliability=0.1933` `competition=La Liga`	fouls: 1.0536 -> 1.0205 totalShots: 0.5405 -> 0.5235 wonTackle: 0.6826 -> 0.6719 wasFouled: 0.255 -> 0.247 onTargetScoringAttempt: 0.2383 -> 0.2308 interceptionWon: 1.2055 -> 1.2 totalCross: 0.1742 -> 0.1687
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Lamine Yamal', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.29 (squad xG sum 1.0060, xG scale 1.2803, xA scale 1.2803, team xA 0.96 = 0.74 x lambda)	`scoreline_reset=true` `team_lambda=1.288` `squad_xg_sum=1.006` `squad_xa_sum=0.7484` `team_expected_assists=0.9582` `team_xa_to_xg_ratio=0.7439`	expected_goals: 0.0943 -> 0.1207
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 11 by projected rating, fifa_overall 68.0)	`match_squad_rank=11` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1137` `goal_rate=0.1008` `scorer_weight=0.1261`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0163` `assist_rate=0.0` `provider_weight=0.0131`	-

Jonny Otto D 84 0 0.02 0.01 84% 0 0.2 6.7 50.6 33 XGBoost D_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9978, quality 0.437)	`market_value_eur=2100000` `quality_factor=0.4374` `boost_factor=0.9978` `reference_quality=0.505` `history_matches=42` `history_shrink=0.1064`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Raphinha', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.29 (squad xG sum 1.0060, xG scale 1.2803, xA scale 1.2803, team xA 0.96 = 0.74 x lambda)	`scoreline_reset=true` `team_lambda=1.288` `squad_xg_sum=1.006` `squad_xa_sum=0.7484` `team_expected_assists=0.9582` `team_xa_to_xg_ratio=0.7439`	expected_assists: 0.0265 -> 0.0339
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 13 by projected rating, fifa_overall 84.0)	`match_squad_rank=13` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0182` `goal_rate=0.0` `scorer_weight=0.0147`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0333` `assist_rate=0.0` `provider_weight=0.0271`	-

Ángel Pérez M 82 0 0.12 0.07 69% 1 1.0 6.9 60.5 33 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9976, quality 0.457)	`market_value_eur=4600000` `quality_factor=0.4572` `boost_factor=0.9976` `reference_quality=0.505` `history_matches=25` `history_shrink=0.1667`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0173)	`blend_weight=0.0173` `history_matches=25` `season_overlap_discount=1.0` `per90_to_match_factor=0.9132` `prior_reliability=0.16` `competition=La Liga`	totalCross: 4.3879 -> 4.3646 wasFouled: 0.9149 -> 0.9056 wonTackle: 0.8314 -> 0.8236 interceptionWon: 0.7113 -> 0.7188 fouls: 1.1019 -> 1.096
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Joan García', 'matchup_zone': 'Left Mid vs Right Back — Wide Channel', 'duel_win_rate': 0.19, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.6624, 'skill_edge': -0.1896, 'height_edge': -0.04, 'attacker_skill': 0.4635, 'defender_skill': 0.885, 'attacker_role': 'LM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Joan García', 'market_value': 46000000, 'role': 'GK', 'price_index': 0.55, 'adjusted_quality': 0.885}], 'xg_override': 0.0969}	aerialWon: 0.2219 -> 1.1 totalShots: 1.614 -> 1.0691 shotOffTarget: 1.0626 -> 0.7039 onTargetScoringAttempt: 0.5497 -> 0.3641 goals: 0.178 -> 0.1179 expectedGoals: 0.1463 -> 0.0969 bigChanceMissed: 0.1201 -> 0.0796 expectedGoalsOnTarget: 0.1191 -> 0.0789
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.29 (squad xG sum 1.0060, xG scale 1.2803, xA scale 1.2803, team xA 0.96 = 0.74 x lambda)	`scoreline_reset=true` `team_lambda=1.288` `squad_xg_sum=1.006` `squad_xa_sum=0.7484` `team_expected_assists=0.9582` `team_xa_to_xg_ratio=0.7439`	expected_goals: 0.0969 -> 0.1241 expected_assists: 0.1642 -> 0.2102
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 72.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1167` `goal_rate=0.1179` `scorer_weight=0.1418`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=1` `assist_probability=0.1896` `assist_rate=0.1388` `provider_weight=0.1853`	-

Pablo Ibáñez M 83 0 0.07 0.08 ok 1 1.2 6.8 61.7 33 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9984, quality 0.451)	`market_value_eur=4099999` `quality_factor=0.4511` `boost_factor=0.9984` `reference_quality=0.505` `history_matches=47` `history_shrink=0.0962`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0125)	`blend_weight=0.0125` `history_matches=47` `season_overlap_discount=1.0` `per90_to_match_factor=0.9192` `prior_reliability=0.2` `competition=La Liga`	totalCross: 1.8606 -> 1.845 wasFouled: 2.2059 -> 2.1975 wonTackle: 1.2608 -> 1.2527 totalShots: 1.0776 -> 1.0833
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Rodri, Pedri', 'matchup_zone': 'CM vs CDM + CM', 'duel_win_rate': 0.18, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.6995, 'skill_edge': -0.1904, 'height_edge': -0.014, 'attacker_skill': 0.4567, 'defender_skill': 0.8799, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Rodri', 'market_value': 54000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8599}, {'player': 'Pedri', 'market_value': 164000000, 'role': 'CM', 'price_index': 0.9, 'adjusted_quality': 0.8999}], 'xg_override': 0.0511}	totalShots: 1.0833 -> 0.7578 shotOffTarget: 0.7916 -> 0.5537 onTargetScoringAttempt: 0.2861 -> 0.2001 aerialWon: 0.9385 -> 1.0 bigChanceMissed: 0.0966 -> 0.0676 goals: 0.0901 -> 0.063 expectedGoalsOnTarget: 0.082 -> 0.0574 expectedGoals: 0.073 -> 0.0511
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.29 (squad xG sum 1.0060, xG scale 1.2803, xA scale 1.2803, team xA 0.96 = 0.74 x lambda)	`scoreline_reset=true` `team_lambda=1.288` `squad_xg_sum=1.006` `squad_xa_sum=0.7484` `team_expected_assists=0.9582` `team_xa_to_xg_ratio=0.7439`	expected_goals: 0.0511 -> 0.0654 expected_assists: 0.0678 -> 0.0868
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 73.0)	`match_squad_rank=4` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0633` `goal_rate=0.063` `scorer_weight=0.0731`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0831` `assist_rate=0.0734` `provider_weight=0.0909`	-

Antonio Blanco M 82 0 0.04 0.03 ok 1 1.0 6.9 62.6 34 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0004, quality 0.519)	`market_value_eur=9700000` `quality_factor=0.5189` `boost_factor=1.0004` `reference_quality=0.505` `history_matches=46` `history_shrink=0.098`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0127)	`blend_weight=0.0127` `history_matches=46` `season_overlap_discount=1.0` `per90_to_match_factor=0.9126` `prior_reliability=0.2` `competition=La Liga`	interceptionWon: 1.1609 -> 1.1771 fouls: 1.49 -> 1.5059 totalCross: 0.9675 -> 0.9746 totalShots: 0.6987 -> 0.7054
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Rodri, Pedri', 'matchup_zone': 'CM vs CDM + CM', 'duel_win_rate': 0.18, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.7313, 'skill_edge': -0.1568, 'height_edge': -0.026, 'attacker_skill': 0.5316, 'defender_skill': 0.8799, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Rodri', 'market_value': 54000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8599}, {'player': 'Pedri', 'market_value': 164000000, 'role': 'CM', 'price_index': 0.9, 'adjusted_quality': 0.8999}], 'xg_override': 0.031}	aerialWon: 0.4808 -> 0.9 totalShots: 0.7054 -> 0.5159 shotOffTarget: 0.4941 -> 0.3613 onTargetScoringAttempt: 0.202 -> 0.1477 goals: 0.05 -> 0.0366 expectedGoals: 0.0424 -> 0.031
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.29 (squad xG sum 1.0060, xG scale 1.2803, xA scale 1.2803, team xA 0.96 = 0.74 x lambda)	`scoreline_reset=true` `team_lambda=1.288` `squad_xg_sum=1.006` `squad_xa_sum=0.7484` `team_expected_assists=0.9582` `team_xa_to_xg_ratio=0.7439`	expected_goals: 0.031 -> 0.0397 expected_assists: 0.0651 -> 0.0834
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 78.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0389` `goal_rate=0.0366` `scorer_weight=0.0424`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.08` `assist_rate=0.0693` `provider_weight=0.0864`	-

Abderrahman Rebbach F 77 0 0.06 0.04 54% 1 0.6 6.8 50.3 44 XGBoost F_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9977, quality 0.437)	`market_value_eur=2400000` `quality_factor=0.4374` `boost_factor=0.9977` `reference_quality=0.505` `history_matches=40` `history_shrink=0.1111`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Pau Cubarsí, Joan García', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.6261, 'skill_edge': -0.2043, 'height_edge': -0.05, 'attacker_skill': 0.4374, 'defender_skill': 0.8915, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Pau Cubarsí', 'market_value': 94000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8981}, {'player': 'Joan García', 'market_value': 46000000, 'role': 'GK', 'price_index': 0.55, 'adjusted_quality': 0.885}], 'xg_override': 0.0453}	totalShots: 0.9675 -> 0.6058 shotOffTarget: 0.6409 -> 0.4013 onTargetScoringAttempt: 0.3265 -> 0.2044 expectedGoals: 0.0723 -> 0.0453 goals: 0.0709 -> 0.0444 expectedGoalsOnTarget: 0.0352 -> 0.022
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.29 (squad xG sum 1.0060, xG scale 1.2803, xA scale 1.2803, team xA 0.96 = 0.74 x lambda)	`scoreline_reset=true` `team_lambda=1.288` `squad_xg_sum=1.006` `squad_xa_sum=0.7484` `team_expected_assists=0.9582` `team_xa_to_xg_ratio=0.7439`	expected_goals: 0.0453 -> 0.058 expected_assists: 0.0782 -> 0.1001
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 12 by projected rating, fifa_overall 78.0)	`match_squad_rank=12` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0564` `goal_rate=0.0444` `scorer_weight=0.0575`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0953` `assist_rate=0.0545` `provider_weight=0.0816`	-

Lucas Boyé F 80 0 0.27 0.32 ok 2 2.0 6.8 59.6 29 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9986, quality 0.466)	`market_value_eur=5300000` `quality_factor=0.4659` `boost_factor=0.9986` `reference_quality=0.505` `history_matches=38` `history_shrink=0.1163`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0081)	`blend_weight=0.0081` `history_matches=38` `season_overlap_discount=1.0` `per90_to_match_factor=0.8842` `prior_reliability=0.1067` `competition=La Liga`	wonTackle: 0.9877 -> 1.002 totalShots: 2.1157 -> 2.1212
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Pau Cubarsí, Joan García', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.6861, 'skill_edge': -0.1916, 'height_edge': -0.022, 'attacker_skill': 0.4659, 'defender_skill': 0.8915, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Pau Cubarsí', 'market_value': 94000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8981}, {'player': 'Joan García', 'market_value': 46000000, 'role': 'GK', 'price_index': 0.55, 'adjusted_quality': 0.885}], 'xg_override': 0.2087}	totalShots: 2.1212 -> 1.4554 shotOffTarget: 1.3396 -> 0.9191 onTargetScoringAttempt: 0.7789 -> 0.5344 expectedGoalsOnTarget: 0.3887 -> 0.2667 goals: 0.3412 -> 0.2341 expectedGoals: 0.3042 -> 0.2087 bigChanceMissed: 0.1866 -> 0.128
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.29 (squad xG sum 1.0060, xG scale 1.2803, xA scale 1.2803, team xA 0.96 = 0.74 x lambda)	`scoreline_reset=true` `team_lambda=1.288` `squad_xg_sum=1.006` `squad_xa_sum=0.7484` `team_expected_assists=0.9582` `team_xa_to_xg_ratio=0.7439`	expected_goals: 0.2087 -> 0.2672 expected_assists: 0.0699 -> 0.0895
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 77.0)	`match_squad_rank=5` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2345` `goal_rate=0.2341` `scorer_weight=0.2934`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0856` `assist_rate=0.0059` `provider_weight=0.0716`	-

Toni Martínez F 75 1 0.29 0.33 ok 2 2.5 6.8 63.9 33 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9996, quality 0.490)	`market_value_eur=7300000` `quality_factor=0.4903` `boost_factor=0.9996` `reference_quality=0.505` `history_matches=46` `history_shrink=0.098`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0110)	`blend_weight=0.011` `history_matches=46` `season_overlap_discount=1.0` `per90_to_match_factor=0.8303` `prior_reliability=0.1733` `competition=La Liga`	onTargetScoringAttempt: 0.8555 -> 0.8461
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Pau Cubarsí, Joan García', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.6846, 'skill_edge': -0.1806, 'height_edge': -0.034, 'attacker_skill': 0.4903, 'defender_skill': 0.8915, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Pau Cubarsí', 'market_value': 94000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8981}, {'player': 'Joan García', 'market_value': 46000000, 'role': 'GK', 'price_index': 0.55, 'adjusted_quality': 0.885}], 'xg_override': 0.2276}	totalShots: 2.4268 -> 1.6614 shotOffTarget: 1.5734 -> 1.0771 onTargetScoringAttempt: 0.8461 -> 0.5792 goals: 0.3494 -> 0.2392 expectedGoals: 0.3325 -> 0.2276 bigChanceMissed: 0.3177 -> 0.2175 expectedGoalsOnTarget: 0.2369 -> 0.1622
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.29 (squad xG sum 1.0060, xG scale 1.2803, xA scale 1.2803, team xA 0.96 = 0.74 x lambda)	`scoreline_reset=true` `team_lambda=1.288` `squad_xg_sum=1.006` `squad_xa_sum=0.7484` `team_expected_assists=0.9582` `team_xa_to_xg_ratio=0.7439`	expected_goals: 0.2276 -> 0.2914 expected_assists: 0.0334 -> 0.0428
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 78.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.2528` `goal_rate=0.2392` `scorer_weight=0.3036`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0419` `assist_rate=0.0177` `provider_weight=0.0347`	-

Team totals 1 1.29 13 6.8

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Grégoire Swiderski	G	4.4m	10.9%	Market-value substitute prior
Youssef Enriquez	D	13.7m	30.4%	Market-value substitute prior
Victor Parada	D	13.6m	30.1%	Market-value substitute prior
Carlos Vicente	M	22.2m	50.8%	Market-value substitute prior
Carles Aleñá	M	13.2m	69.2%	Market-value substitute prior
Miguel Rodríguez	F	25.4m	56.2%	Market-value substitute prior

Antagonist / tactical summary

Home midfield dominates with ~53% possession expected (Pedri vs Antonio Blanco). Total antagonist xG — Home: 2.38, Away: 0.66. Home danger zone: Left Wing vs Right Back — Flank Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Midfield — possession 53.1% / 46.9%, dominance ratio 0.63, key duel: Pedri vs Antonio Blanco

Raphinha xG 0.82Karim Adeyemi xG 0.31Lamine Yamal xG 0.64Lucas Boyé xG 0.21Toni Martínez xG 0.23Abderrahman Rebbach xG 0.05

Summary

Antagonist matchup analysis for Barcelona vs Alavés. Most likely scoreline: 3-1. Lambdas derived from 8 home and 6 away individual player duels. Head-to-head: Barcelona 7W-2D-1L vs Alavés over 10 meetings.

Tactical parity between 4-3-3 and 4-4-2. Match outcome driven primarily by individual player quality. Home midfield dominates with ~53% possession expected (Pedri vs Antonio Blanco). Total antagonist xG — Home: 2.38, Away: 0.66. Home danger zone: Left Wing vs Right Back — Flank Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Levante vs Atlético Madrid

LaLiga · 20261031 · 4-3-3 vs 4-2-3-1

Predicted 1-2Winner: Atlético MadridWinning H 2-1 (5.1%)Winning A 1-2 (9.6%)Confidence 0.77λ home/away 1.19 / 2.27Most-likely P 9.6%

Actual vs predicted: result not available yet (upcoming fixture).

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Levante vs Atlético Madrid. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 4-3-3 vs 4-2-3-1 (duel lambdas: Home 1.396, Away 2.653). Final calibrated Poisson lambdas: Home 1.191, Away 2.266 (most likely score 1-2).

Home Starters 11/11 XGBoost (6 FBref) Away Starters 11/11 XGBoost (9 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 1.40 / 2.65 Final λ (H/A) 1.19 / 2.27

FBref Team Priors: Levante UD (Gls/90: 1.6666666666666667, GA/90: 1.67 → E[λ]: 2.877) vs Atlético Madrid (Gls/90: 2.3333333333333335, GA/90: 2.33 → E[λ]: 2.886) | Blend Weight: 0.228

Modeling Pipeline Stages (5 layers)

1. XGBoost player regressions for volume & event rates with market-value quality scaling
2. Multi-source prior blending (FBref 2026-2027 tables + Sofascore season history)
3. Antagonist formation duel matrix resolving 1v1 and zonal matchups
4. Team-level prior calibration (FBref possession and attack/defense efficiencies)
5. Bivariate Poisson matrix with Dixon-Coles low-score correction

Standing & form context

Metric	Levante	Atlético Madrid
League position	14.00	2.00
Points per game	1.14	2.07
Attack index	1.03	1.15
Defense (concede) index	0.79	1.12
Form multiplier	0.96	1.04
Standing λ multiplier	0.91	1.11

Lambda decomposition (how the goal rates were built)

Stage	Levante λ	Atlético Madrid λ
FBref Squad Prior E[λ] (weight 0.228)	2.877	2.886
Pre-mismatch	1.376	2.691
Pre-calibration (Micro-duels)	1.396	2.653
Final λ	1.191	2.266

Scoreline distribution & markets

Top 5 scorelines	P
1-2	9.6%
1-1	8.6%
0-2	8.1%
1-3	7.3%
0-1	7.1%

Home win	18.9%
Draw	19.5%
Away win	61.5%
Over 2.5	67.1%
BTTS yes	62.4%

Pregame fair line — favorite Atlético Madrid (61.5%), fair total 3.46, fair spread -1.07

Projected players vs season baseline

Proj = projected for this match · Season = this season's per-game average from match data · deviation badge compares projected xG to the player's season xG/game.

Levante — projected line vs season baseline (4-3-3)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Oriol Rey M 78 0 0.02 0.01 308% 0 0.2 6.8 35.5 32 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9970, quality 0.437)	`market_value_eur=1700000` `quality_factor=0.4374` `boost_factor=0.997` `reference_quality=0.505` `history_matches=29` `history_shrink=0.1471`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0191)	`blend_weight=0.0191` `history_matches=29` `season_overlap_discount=1.0` `per90_to_match_factor=0.8692` `prior_reliability=0.2` `competition=La Liga`	totalShots: 0.5063 -> 0.4966 totalCross: 0.4842 -> 0.4749 wasFouled: 0.6818 -> 0.6743
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Morten Hjulmand', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.227, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.6723, 'skill_edge': -0.1829, 'height_edge': -0.04, 'attacker_skill': 0.4374, 'defender_skill': 0.8439, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Morten Hjulmand', 'market_value': 48000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8439}], 'xg_override': 0.0204}	totalShots: 0.5037 -> 0.3387 shotOffTarget: 0.3702 -> 0.2489 aerialWon: 0.8459 -> 0.9 onTargetScoringAttempt: 0.1335 -> 0.0898 goals: 0.0487 -> 0.0327 expectedGoals: 0.0303 -> 0.0204
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.19 (squad xG sum 0.9866, xG scale 1.2071, xA scale 1.4295, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.191` `squad_xg_sum=0.9866` `squad_xa_sum=0.5416` `team_expected_assists=0.7742` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0334 -> 0.0477
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 73.0)	`match_squad_rank=6` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0243` `goal_rate=0.0327` `scorer_weight=0.0289`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0466` `assist_rate=0.0483` `provider_weight=0.0562`	-

Jon Ander Olasagasti M 86 0 0.07 0.08 ok 1 0.9 6.9 60.7 34 XGBoost M_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9978, quality 0.437)	`market_value_eur=2700000` `quality_factor=0.4374` `boost_factor=0.9978` `reference_quality=0.505` `history_matches=42` `history_shrink=0.1064`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Morten Hjulmand', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.231, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.6782, 'skill_edge': -0.1829, 'height_edge': -0.036, 'attacker_skill': 0.4374, 'defender_skill': 0.8439, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Morten Hjulmand', 'market_value': 48000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8439}], 'xg_override': 0.0559}	aerialWon: 0.4426 -> 0.9 totalShots: 1.1085 -> 0.7518 shotOffTarget: 0.7565 -> 0.5131 onTargetScoringAttempt: 0.352 -> 0.2387 expectedGoalsOnTarget: 0.1184 -> 0.0803 goals: 0.0954 -> 0.0647 expectedGoals: 0.0824 -> 0.0559 bigChanceMissed: 0.0565 -> 0.0383
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.19 (squad xG sum 0.9866, xG scale 1.2071, xA scale 1.4295, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.191` `squad_xg_sum=0.9866` `squad_xa_sum=0.5416` `team_expected_assists=0.7742` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0559 -> 0.0675 expected_assists: 0.1228 -> 0.1757
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 84.0)	`match_squad_rank=4` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0653` `goal_rate=0.0647` `scorer_weight=0.0731`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1611` `assist_rate=0.1028` `provider_weight=0.1554`	-

Manuel Sánchez D 84 0 0.01 0.01 ok 0 0.2 6.8 54.9 34 XGBoost D_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9992, quality 0.479)	`market_value_eur=6400000` `quality_factor=0.4793` `boost_factor=0.9992` `reference_quality=0.505` `history_matches=43` `history_shrink=0.1042`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Kang-in Lee', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.19 (squad xG sum 0.9866, xG scale 1.2071, xA scale 1.4295, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.191` `squad_xg_sum=0.9866` `squad_xa_sum=0.5416` `team_expected_assists=0.7742` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0501 -> 0.0716
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 10 by projected rating, fifa_overall 78.0)	`match_squad_rank=10` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0094` `goal_rate=0.0015` `scorer_weight=0.0084`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0691` `assist_rate=0.0154` `provider_weight=0.0573`	-

Aïssa Mandi D 87 0 0.00 0.00 ok 0 0.1 6.8 42.2 13 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9922, quality 0.437)	`market_value_eur=1100000` `quality_factor=0.4374` `boost_factor=0.9922` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0500)	`blend_weight=0.05` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=0.9642` `prior_reliability=0.2` `competition=La Liga`	wonTackle: 0.6335 -> 0.6822 fouls: 0.4473 -> 0.4249 wasFouled: 0.8556 -> 0.845
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Julián Alvarez', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.19 (squad xG sum 0.9866, xG scale 1.2071, xA scale 1.4295, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.191` `squad_xg_sum=0.9866` `squad_xa_sum=0.5416` `team_expected_assists=0.7742` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 79.0)	`match_squad_rank=5` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0037` `goal_rate=0.0` `scorer_weight=0.0035`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0091` `assist_rate=0.0` `provider_weight=0.0081`	-

Ifeanyi Ndukwe D 84 0 0.10 0.04 177% 0 0.1 6.8 33.7 12 FBref Prior D_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9911, quality 0.440)	`market_value_eur=3200000` `quality_factor=0.4399` `boost_factor=0.9911` `reference_quality=0.505` `history_matches=6` `history_shrink=0.4545`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Julián Alvarez', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.19 (squad xG sum 0.9866, xG scale 1.2071, xA scale 1.4295, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.191` `squad_xg_sum=0.9866` `squad_xa_sum=0.5416` `team_expected_assists=0.7742` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0815 -> 0.0984
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 84.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0937` `goal_rate=0.191` `scorer_weight=0.123`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0047` `assist_rate=0.0` `provider_weight=0.0044`	-

Nacho Pérez D 74 0 0.07 0.02 254% 1 0.1 6.8 21.9 34 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9940, quality 0.437)	`market_value_eur=1600000` `quality_factor=0.4374` `boost_factor=0.994` `reference_quality=0.505` `history_matches=12` `history_shrink=0.2941`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0280)	`blend_weight=0.028` `history_matches=12` `season_overlap_discount=1.0` `per90_to_match_factor=0.8272` `prior_reliability=0.1467` `competition=La Liga`	totalCross: 1.4482 -> 1.4919 wonTackle: 0.9273 -> 0.954 interceptionWon: 0.6936 -> 0.7163 totalShots: 0.5184 -> 0.5039 goals: 0.378 -> 0.3674 onTargetScoringAttempt: 0.378 -> 0.3674 fouls: 0.6664 -> 0.6583
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Alejandro Grimaldo', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.19 (squad xG sum 0.9866, xG scale 1.2071, xA scale 1.4295, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.191` `squad_xg_sum=0.9866` `squad_xa_sum=0.5416` `team_expected_assists=0.7742` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.058 -> 0.07 expected_assists: 0.0272 -> 0.0389
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 67.0)	`match_squad_rank=8` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0676` `goal_rate=0.3674` `scorer_weight=0.0875`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0382` `assist_rate=0.0` `provider_weight=0.0311`	-

Mathew Ryan G 86 0 0.00 0.00 100% 0.1 7.0 65.9 35 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9980, quality 0.437)	`market_value_eur=1700000` `quality_factor=0.4374` `boost_factor=0.998` `reference_quality=0.505` `history_matches=46` `history_shrink=0.098`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0127)	`blend_weight=0.0127` `history_matches=46` `season_overlap_discount=1.0` `per90_to_match_factor=0.9503` `prior_reliability=0.2` `competition=La Liga`	saves: 2.897 -> 2.9166
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Julián Alvarez, Alejandro Grimaldo', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 2.653}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.19 (squad xG sum 0.9866, xG scale 1.2071, xA scale 1.4295, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.191` `squad_xg_sum=0.9866` `squad_xa_sum=0.5416` `team_expected_assists=0.7742` `team_xa_to_xg_ratio=0.65`	shots_faced: 25.2 -> 21.5 expected_saves: 6.17 -> 5.26 high_pressure_saves: 2.12 -> 1.81
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 79.0)	`match_squad_rank=1` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=1.3229`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=0.7982`	-

Daniel Requena M 80 0 0.07 0.03 133% 1 0.4 6.7 40.4 13 FBref Prior M_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9916, quality 0.437)	`market_value_eur=1500000` `quality_factor=0.4374` `boost_factor=0.9916` `reference_quality=0.505` `history_matches=7` `history_shrink=0.4167`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Morten Hjulmand', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.241, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.737, 'skill_edge': -0.1829, 'height_edge': 0.004, 'attacker_skill': 0.4374, 'defender_skill': 0.8439, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Morten Hjulmand', 'market_value': 48000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8439}], 'xg_override': 0.0586}	aerialWon: 0.4894 -> 1.2 totalShots: 0.9408 -> 0.6934 shotOffTarget: 0.646 -> 0.4761 bigChanceMissed: 0.3421 -> 0.2521 onTargetScoringAttempt: 0.2947 -> 0.2172 expectedGoals: 0.0795 -> 0.0586 goals: 0.0782 -> 0.0576 expectedGoalsOnTarget: 0.0785 -> 0.0579
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.19 (squad xG sum 0.9866, xG scale 1.2071, xA scale 1.4295, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.191` `squad_xg_sum=0.9866` `squad_xa_sum=0.5416` `team_expected_assists=0.7742` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0586 -> 0.0707 expected_assists: 0.0347 -> 0.0496
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 12 by projected rating, fifa_overall 84.0)	`match_squad_rank=12` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0683` `goal_rate=0.0576` `scorer_weight=0.07`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0484` `assist_rate=0.0498` `provider_weight=0.0584`	-

Roger Brugué F 73 1 0.29 0.33 ok 2 1.8 6.8 52.3 24 XGBoost F_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9969, quality 0.437)	`market_value_eur=2800000` `quality_factor=0.4374` `boost_factor=0.9969` `reference_quality=0.505` `history_matches=28` `history_shrink=0.1515`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Jan Oblak', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle', 'duel_win_rate': 0.237, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.731, 'skill_edge': -0.123, 'height_edge': -0.06, 'attacker_skill': 0.4374, 'defender_skill': 0.7107, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Jan Oblak', 'market_value': 16000000, 'role': 'GK', 'price_index': 0.55, 'adjusted_quality': 0.7107}], 'xg_override': 0.2397}	totalShots: 2.0875 -> 1.526 shotOffTarget: 1.158 -> 0.8465 onTargetScoringAttempt: 0.9295 -> 0.6795 expectedGoalsOnTarget: 0.3961 -> 0.2895 goals: 0.3629 -> 0.2653 expectedGoals: 0.3279 -> 0.2397 bigChanceMissed: 0.2043 -> 0.1493
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.19 (squad xG sum 0.9866, xG scale 1.2071, xA scale 1.4295, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.191` `squad_xg_sum=0.9866` `squad_xa_sum=0.5416` `team_expected_assists=0.7742` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.2397 -> 0.2894 expected_assists: 0.0492 -> 0.0703
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 88.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.2513` `goal_rate=0.2653` `scorer_weight=0.315`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0679` `assist_rate=0.1267` `provider_weight=0.0879`	-

Iván Romero F 70 0 0.23 0.23 ok 2 1.6 6.8 60.7 33 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9996, quality 0.491)	`market_value_eur=7400000` `quality_factor=0.4915` `boost_factor=0.9996` `reference_quality=0.505` `history_matches=42` `history_shrink=0.1064`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0115)	`blend_weight=0.0115` `history_matches=42` `season_overlap_discount=1.0` `per90_to_match_factor=0.7786` `prior_reliability=0.1667` `competition=La Liga`	totalShots: 1.888 -> 1.8988 wasFouled: 2.035 -> 2.0259 interceptionWon: 0.4541 -> 0.4489
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Dávid Hancko, Marc Pubill', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.682, 'skill_edge': -0.1563, 'height_edge': -0.06, 'attacker_skill': 0.4915, 'defender_skill': 0.8388, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Dávid Hancko', 'market_value': 36000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8168}, {'player': 'Marc Pubill', 'market_value': 48000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8608}], 'xg_override': 0.194}	totalShots: 1.8988 -> 1.295 shotOffTarget: 1.1332 -> 0.7728 onTargetScoringAttempt: 0.7533 -> 0.5138 goals: 0.3167 -> 0.216 bigChanceMissed: 0.2904 -> 0.1981 expectedGoals: 0.2844 -> 0.194 expectedGoalsOnTarget: 0.2616 -> 0.1784
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.19 (squad xG sum 0.9866, xG scale 1.2071, xA scale 1.4295, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.191` `squad_xg_sum=0.9866` `squad_xa_sum=0.5416` `team_expected_assists=0.7742` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.194 -> 0.2342 expected_assists: 0.0553 -> 0.079
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 11 by projected rating, fifa_overall 75.0)	`match_squad_rank=11` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2088` `goal_rate=0.216` `scorer_weight=0.2553`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.076` `assist_rate=0.0034` `provider_weight=0.0632`	-

Thiago Fernández F 72 0 0.12 0.05 126% 1 0.9 6.9 54.2 35 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9976, quality 0.458)	`market_value_eur=4700000` `quality_factor=0.4585` `boost_factor=0.9976` `reference_quality=0.505` `history_matches=24` `history_shrink=0.1724`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0120)	`blend_weight=0.012` `history_matches=24` `season_overlap_discount=1.0` `per90_to_match_factor=0.7963` `prior_reliability=0.1067` `competition=La Liga`	interceptionWon: 0.303 -> 0.3113 totalShots: 1.5102 -> 1.5038 wasFouled: 0.5093 -> 0.5032 onTargetScoringAttempt: 0.4356 -> 0.4304
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Marcos Llorente', 'matchup_zone': 'Right Wing vs Left Back — Flank Battle', 'duel_win_rate': 0.327, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v1', 'xg_multiplier': 0.7757, 'skill_edge': -0.1166, 'height_edge': -0.036, 'attacker_skill': 0.4585, 'defender_skill': 0.7176, 'attacker_role': 'RW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Marcos Llorente', 'market_value': 21000000, 'role': 'LB', 'price_index': 0.7, 'adjusted_quality': 0.7176}], 'xg_override': 0.0957}	totalShots: 1.505 -> 1.1675 shotOffTarget: 1.0746 -> 0.8336 onTargetScoringAttempt: 0.4304 -> 0.3339 goals: 0.1708 -> 0.1325 expectedGoals: 0.1234 -> 0.0957 bigChanceMissed: 0.0956 -> 0.0742 expectedGoalsOnTarget: 0.0459 -> 0.0356
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.19 (squad xG sum 0.9866, xG scale 1.2071, xA scale 1.4295, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.191` `squad_xg_sum=0.9866` `squad_xa_sum=0.5416` `team_expected_assists=0.7742` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0957 -> 0.1155 expected_assists: 0.1092 -> 0.1561
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 74.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1091` `goal_rate=0.1325` `scorer_weight=0.1444`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=1` `assist_probability=0.1445` `assist_rate=0.2455` `provider_weight=0.1951`	-

Team totals 1 1.19 10 6.8

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Pablo Cunat Campos	G	4.4m	10.9%	Market-value substitute prior
Axel Tape	D	13.8m	31.1%	Market-value substitute prior
Matias Moreno	D	13.7m	30.5%	Market-value substitute prior
Carlos Álvarez	M	22.6m	31.9%	Market-value substitute prior
Enzo Bardeli	M	22.1m	50.5%	Market-value substitute prior
Etta Eyong	F	25.9m	58.5%	Market-value substitute prior

Atlético Madrid — projected line vs season baseline (4-2-3-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Julián Alvarez F 75 1 0.56 0.13 324% 3 1.5 7.2 51.0 23 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0146, quality 0.895)	`market_value_eur=107000000` `quality_factor=0.8953` `boost_factor=1.0146` `reference_quality=0.505` `history_matches=35` `history_shrink=0.125`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0016)	`blend_weight=0.0016` `history_matches=35` `season_overlap_discount=1.0` `per90_to_match_factor=0.8377` `prior_reliability=0.02` `competition=La Liga`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Aïssa Mandi, Ifeanyi Ndukwe', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.342, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.2095, 'skill_edge': 0.2025, 'height_edge': -0.06, 'attacker_skill': 0.8953, 'defender_skill': 0.4453, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Aïssa Mandi', 'market_value': 1100000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'Ifeanyi Ndukwe', 'market_value': 3200000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4531}], 'xg_override': 0.3932}	totalShots: 2.526 -> 3.0552 shotOffTarget: 1.3629 -> 1.6484 onTargetScoringAttempt: 1.1604 -> 1.4035 goals: 0.454 -> 0.5491 expectedGoalsOnTarget: 0.3427 -> 0.4145 expectedGoals: 0.3251 -> 0.3932 bigChanceMissed: 0.2983 -> 0.3608
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	goals: 1.0 -> 0
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.27 (squad xG sum 1.5979, xG scale 1.4181, xA scale 1.4181, team xA 1.65 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.266` `squad_xg_sum=1.5979` `squad_xa_sum=1.1609` `team_expected_assists=1.6463` `team_xa_to_xg_ratio=0.7265`	expected_goals: 0.3932 -> 0.5577 expected_assists: 0.145 -> 0.2056
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 86.0)	`match_squad_rank=1` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.4275` `goal_rate=0.5491` `scorer_weight=0.6971`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=1` `assist_probability=0.1858` `assist_rate=0.2168` `provider_weight=0.257`	-

Alex Baena F 75 1 0.27 0.08 219% 3 1.4 7.1 58.7 42 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0111, quality 0.817)	`market_value_eur=48000000` `quality_factor=0.8168` `boost_factor=1.0111` `reference_quality=0.505` `history_matches=37` `history_shrink=0.119`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0093)	`blend_weight=0.0093` `history_matches=37` `season_overlap_discount=1.0` `per90_to_match_factor=0.836` `prior_reliability=0.12` `competition=La Liga`	totalCross: 3.3552 -> 3.3413 goals: 0.2507 -> 0.2613 totalOffside: 0.1662 -> 0.1733 interceptionWon: 1.1417 -> 1.1354 onTargetScoringAttempt: 0.7281 -> 0.7342 wasFouled: 1.2007 -> 1.2068 wonTackle: 0.5532 -> 0.5481
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Aïssa Mandi, Ifeanyi Ndukwe', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.307, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.1576, 'skill_edge': 0.1672, 'height_edge': -0.06, 'attacker_skill': 0.8168, 'defender_skill': 0.4453, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Aïssa Mandi', 'market_value': 1100000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'Ifeanyi Ndukwe', 'market_value': 3200000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4531}], 'xg_override': 0.1888}	totalShots: 2.1297 -> 2.4653 shotOffTarget: 1.3955 -> 1.6154 onTargetScoringAttempt: 0.7342 -> 0.8499 goals: 0.2613 -> 0.3025 expectedGoalsOnTarget: 0.2148 -> 0.2487 expectedGoals: 0.1631 -> 0.1888
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.27 (squad xG sum 1.5979, xG scale 1.4181, xA scale 1.4181, team xA 1.65 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.266` `squad_xg_sum=1.5979` `squad_xa_sum=1.1609` `team_expected_assists=1.6463` `team_xa_to_xg_ratio=0.7265`	expected_goals: 0.1888 -> 0.2677 expected_assists: 0.1938 -> 0.2748
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 83.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.2349` `goal_rate=0.3025` `scorer_weight=0.3346`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.2403` `assist_rate=0.0528` `provider_weight=0.2198`	-

Kang-in Lee F 73 0 0.27 0.15 77% 3 2.1 7.2 56.3 16 XGBoost F_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0259, quality 0.764)	`market_value_eur=37000000` `quality_factor=0.7641` `boost_factor=1.0259` `reference_quality=0.505` `history_matches=10` `history_shrink=0.3333`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Aïssa Mandi, Ifeanyi Ndukwe', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.283, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.1227, 'skill_edge': 0.1435, 'height_edge': -0.06, 'attacker_skill': 0.7641, 'defender_skill': 0.4453, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Aïssa Mandi', 'market_value': 1100000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'Ifeanyi Ndukwe', 'market_value': 3200000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4531}], 'xg_override': 0.1877}	totalShots: 2.3133 -> 2.5972 shotOffTarget: 1.5842 -> 1.7786 onTargetScoringAttempt: 0.7291 -> 0.8186 expectedGoalsOnTarget: 0.2062 -> 0.2315 goals: 0.199 -> 0.2234 expectedGoals: 0.1672 -> 0.1877 bigChanceMissed: 0.143 -> 0.1605
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.27 (squad xG sum 1.5979, xG scale 1.4181, xA scale 1.4181, team xA 1.65 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.266` `squad_xg_sum=1.5979` `squad_xa_sum=1.1609` `team_expected_assists=1.6463` `team_xa_to_xg_ratio=0.7265`	expected_goals: 0.1877 -> 0.2662 expected_assists: 0.1829 -> 0.2594
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 80.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2337` `goal_rate=0.2234` `scorer_weight=0.3062`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.2285` `assist_rate=0.0798` `provider_weight=0.2075`	-

Morten Hjulmand M 83 0 0.10 0.03 199% 1 0.6 6.9 52.0 12 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0334, quality 0.817)	`market_value_eur=48000000` `quality_factor=0.8168` `boost_factor=1.0334` `reference_quality=0.505` `history_matches=9` `history_shrink=0.3571`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0325)	`blend_weight=0.0325` `history_matches=9` `season_overlap_discount=1.0` `per90_to_match_factor=0.9253` `prior_reliability=0.14` `competition=La Liga`	interceptionWon: 2.687 -> 2.6426 wonTackle: 0.7933 -> 0.8248 fouls: 0.669 -> 0.6902 totalShots: 0.8413 -> 0.8284 wasFouled: 1.4074 -> 1.4189 onTargetScoringAttempt: 0.1831 -> 0.1771
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Oriol Rey', 'matchup_zone': 'Holding Mid vs Central Mid — Midfield Anchor Duel', 'duel_win_rate': 0.623, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.3277, 'skill_edge': 0.1829, 'height_edge': 0.04, 'attacker_skill': 0.8439, 'defender_skill': 0.4374, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Oriol Rey', 'market_value': 1700000, 'role': 'CM', 'price_index': 0.9, 'adjusted_quality': 0.4374}], 'xg_override': 0.0673}	totalShots: 0.8353 -> 1.109 shotOffTarget: 0.6582 -> 0.8739 onTargetScoringAttempt: 0.1771 -> 0.2351 aerialWon: 1.1489 -> 1.2 goals: 0.0555 -> 0.0737 expectedGoals: 0.0507 -> 0.0673
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.27 (squad xG sum 1.5979, xG scale 1.4181, xA scale 1.4181, team xA 1.65 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.266` `squad_xg_sum=1.5979` `squad_xa_sum=1.1609` `team_expected_assists=1.6463` `team_xa_to_xg_ratio=0.7265`	expected_goals: 0.0673 -> 0.0954 expected_assists: 0.041 -> 0.0581
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 11 by projected rating, fifa_overall 83.0)	`match_squad_rank=11` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.091` `goal_rate=0.0737` `scorer_weight=0.1006`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0564` `assist_rate=0.0557` `provider_weight=0.0675`	-

Pablo Barrios M 70 0 0.12 0.04 238% 1 0.8 7.0 57.2 16 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0160, quality 0.879)	`market_value_eur=77000000` `quality_factor=0.8792` `boost_factor=1.016` `reference_quality=0.505` `history_matches=30` `history_shrink=0.1429`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0142)	`blend_weight=0.0142` `history_matches=30` `season_overlap_discount=1.0` `per90_to_match_factor=0.7742` `prior_reliability=0.1533` `competition=La Liga`	wasFouled: 0.6741 -> 0.6885 totalCross: 1.1708 -> 1.1589 totalShots: 1.023 -> 1.0133 interceptionWon: 1.4391 -> 1.433
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Daniel Requena', 'matchup_zone': 'Attacking Mid vs Single Holding Mid — Space Creator Duel', 'duel_win_rate': 0.64, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.2653, 'skill_edge': 0.2005, 'height_edge': -0.02, 'attacker_skill': 0.8829, 'defender_skill': 0.4374, 'attacker_role': 'CAM', 'attacker_mv_price_index': 0.95, 'defender_adjusted_parts': [{'player': 'Daniel Requena', 'market_value': 1500000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.4374}], 'xg_override': 0.0838}	aerialWon: 0.7936 -> 1.1 totalShots: 1.0236 -> 1.2952 shotOffTarget: 0.7277 -> 0.9208 onTargetScoringAttempt: 0.2959 -> 0.3744 expectedGoalsOnTarget: 0.078 -> 0.0987 expectedGoals: 0.0662 -> 0.0838 goals: 0.0657 -> 0.0831
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.27 (squad xG sum 1.5979, xG scale 1.4181, xA scale 1.4181, team xA 1.65 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.266` `squad_xg_sum=1.5979` `squad_xa_sum=1.1609` `team_expected_assists=1.6463` `team_xa_to_xg_ratio=0.7265`	expected_goals: 0.0838 -> 0.1188 expected_assists: 0.1152 -> 0.1634
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 83.0)	`match_squad_rank=8` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.112` `goal_rate=0.0831` `scorer_weight=0.1182`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1507` `assist_rate=0.1409` `provider_weight=0.1888`	-

Alejandro Grimaldo M 89 0 0.28 0.18 56% 2 1.1 7.0 53.6 20 XGBoost M_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0129, quality 0.625)	`market_value_eur=19400000` `quality_factor=0.6251` `boost_factor=1.0129` `reference_quality=0.505` `history_matches=9` `history_shrink=0.3571`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Nacho Pérez', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle', 'duel_win_rate': 0.548, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.1006, 'skill_edge': 0.0845, 'height_edge': -0.016, 'attacker_skill': 0.6251, 'defender_skill': 0.4374, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Nacho Pérez', 'market_value': 1600000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.4374}], 'xg_override': 0.1985}	aerialWon: 0.3474 -> 0.8 totalShots: 1.5977 -> 1.7584 shotOffTarget: 0.9351 -> 1.0292 onTargetScoringAttempt: 0.6625 -> 0.7291 expectedGoalsOnTarget: 0.257 -> 0.2829 expectedGoals: 0.1804 -> 0.1985 bigChanceMissed: 0.1749 -> 0.1925 goals: 0.149 -> 0.164
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.27 (squad xG sum 1.5979, xG scale 1.4181, xA scale 1.4181, team xA 1.65 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.266` `squad_xg_sum=1.5979` `squad_xa_sum=1.1609` `team_expected_assists=1.6463` `team_xa_to_xg_ratio=0.7265`	expected_goals: 0.1985 -> 0.2815 expected_assists: 0.1492 -> 0.2116
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 89.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2453` `goal_rate=0.164` `scorer_weight=0.2424`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1907` `assist_rate=0.1197` `provider_weight=0.1803`	-

Dávid Hancko D 88 0 0.09 0.05 83% 1 0.7 7.0 53.3 29 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0084, quality 0.758)	`market_value_eur=36000000` `quality_factor=0.7581` `boost_factor=1.0084` `reference_quality=0.505` `history_matches=40` `history_shrink=0.1111`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0144)	`blend_weight=0.0144` `history_matches=40` `season_overlap_discount=1.0` `per90_to_match_factor=0.9803` `prior_reliability=0.2` `competition=La Liga`	interceptionWon: 0.9553 -> 0.9698 wasFouled: 0.1822 -> 0.1937 wonTackle: 0.8219 -> 0.8289
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Iván Romero', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.27 (squad xG sum 1.5979, xG scale 1.4181, xA scale 1.4181, team xA 1.65 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.266` `squad_xg_sum=1.5979` `squad_xa_sum=1.1609` `team_expected_assists=1.6463` `team_xa_to_xg_ratio=0.7265`	expected_goals: 0.0613 -> 0.0869 expected_assists: 0.0444 -> 0.063
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 83.0)	`match_squad_rank=6` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0832` `goal_rate=0.0` `scorer_weight=0.0695`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0611` `assist_rate=0.1488` `provider_weight=0.0788`	-

Marc Pubill D 82 0 0.08 0.04 107% 1 0.4 6.8 49.3 30 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0146, quality 0.817)	`market_value_eur=48000000` `quality_factor=0.8168` `boost_factor=1.0146` `reference_quality=0.505` `history_matches=27` `history_shrink=0.1562`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0115)	`blend_weight=0.0115` `history_matches=27` `season_overlap_discount=1.0` `per90_to_match_factor=0.9074` `prior_reliability=0.1133` `competition=La Liga`	interceptionWon: 0.8389 -> 0.8292 totalShots: 0.5217 -> 0.5278 totalCross: 0.6161 -> 0.6213
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Iván Romero', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.27 (squad xG sum 1.5979, xG scale 1.4181, xA scale 1.4181, team xA 1.65 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.266` `squad_xg_sum=1.5979` `squad_xa_sum=1.1609` `team_expected_assists=1.6463` `team_xa_to_xg_ratio=0.7265`	expected_goals: 0.0538 -> 0.0763 expected_assists: 0.0201 -> 0.0285
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 13 by projected rating, fifa_overall 87.0)	`match_squad_rank=13` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0735` `goal_rate=0.1142` `scorer_weight=0.0954`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0281` `assist_rate=0.007` `provider_weight=0.0232`	-

Marcos Llorente D 84 0 0.07 0.04 83% 1 0.6 7.0 60.3 33 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0046, quality 0.641)	`market_value_eur=21000000` `quality_factor=0.6408` `boost_factor=1.0046` `reference_quality=0.505` `history_matches=39` `history_shrink=0.1136`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0074)	`blend_weight=0.0074` `history_matches=39` `season_overlap_discount=1.0` `per90_to_match_factor=0.9328` `prior_reliability=0.1` `competition=La Liga`	fouls: 0.7828 -> 0.8046 totalCross: 1.6539 -> 1.6417 interceptionWon: 1.0637 -> 1.0742 wasFouled: 0.2029 -> 0.2106
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Thiago Fernández', 'matchup_zone': 'Right Wing vs Left Back — Flank Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.27 (squad xG sum 1.5979, xG scale 1.4181, xA scale 1.4181, team xA 1.65 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.266` `squad_xg_sum=1.5979` `squad_xa_sum=1.1609` `team_expected_assists=1.6463` `team_xa_to_xg_ratio=0.7265`	expected_goals: 0.0481 -> 0.0682 expected_assists: 0.0812 -> 0.1152
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 85.0)	`match_squad_rank=5` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0659` `goal_rate=0.0` `scorer_weight=0.0546`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1088` `assist_rate=0.0923` `provider_weight=0.1249`	-

Jan Oblak G 90 0 0.00 - n/a 0.0 7.1 56.6 29 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0028, quality 0.590)	`market_value_eur=16000000` `quality_factor=0.59` `boost_factor=1.0028` `reference_quality=0.505` `history_matches=40` `history_shrink=0.1111`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0144)	`blend_weight=0.0144` `history_matches=40` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.2` `competition=La Liga`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Roger Brugué, Iván Romero', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.3957}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.27 (squad xG sum 1.5979, xG scale 1.4181, xA scale 1.4181, team xA 1.65 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.266` `squad_xg_sum=1.5979` `squad_xa_sum=1.1609` `team_expected_assists=1.6463` `team_xa_to_xg_ratio=0.7265`	shots_faced: 13.3 -> 11.3 expected_saves: 3.26 -> 2.76 high_pressure_saves: 1.12 -> 0.95
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 88.0)	`match_squad_rank=4` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=2.4769`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=1.6132`	-

Giuliano Simeone M 78 0 0.28 0.10 163% 2 0.9 7.0 58.9 32 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0096, quality 0.800)	`market_value_eur=44000000` `quality_factor=0.8002` `boost_factor=1.0096` `reference_quality=0.505` `history_matches=41` `history_shrink=0.1087`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0113)	`blend_weight=0.0113` `history_matches=41` `season_overlap_discount=1.0` `per90_to_match_factor=0.8686` `prior_reliability=0.16` `competition=La Liga`	interceptionWon: 0.5557 -> 0.5658
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Oriol Rey', 'matchup_zone': 'Holding Mid vs Central Mid — Midfield Anchor Duel', 'duel_win_rate': 0.569, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.248, 'skill_edge': 0.1767, 'height_edge': -0.008, 'attacker_skill': 0.8301, 'defender_skill': 0.4374, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Oriol Rey', 'market_value': 1700000, 'role': 'CM', 'price_index': 0.9, 'adjusted_quality': 0.4374}], 'xg_override': 0.1942}	aerialWon: 1.4898 -> 0.9 totalShots: 1.4166 -> 1.7679 shotOffTarget: 0.8679 -> 1.0831 onTargetScoringAttempt: 0.5461 -> 0.6815 goals: 0.1622 -> 0.2024 expectedGoals: 0.1556 -> 0.1942 expectedGoalsOnTarget: 0.0941 -> 0.1174 bigChanceMissed: 0.0864 -> 0.1078
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.27 (squad xG sum 1.5979, xG scale 1.4181, xA scale 1.4181, team xA 1.65 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.266` `squad_xg_sum=1.5979` `squad_xa_sum=1.1609` `team_expected_assists=1.6463` `team_xa_to_xg_ratio=0.7265`	expected_goals: 0.1942 -> 0.2754 expected_assists: 0.1235 -> 0.1751
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 81.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2407` `goal_rate=0.2024` `scorer_weight=0.2847`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1606` `assist_rate=0.1073` `provider_weight=0.1587`	-

Team totals 2 2.27 20 7.0

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Juan Musso	G	4.4m	10.9%	Market-value substitute prior
Cristian Romero	D	15.0m	36.6%	Market-value substitute prior
Robin Le Normand	D	14.3m	23.2%	Market-value substitute prior
Rodrigo Mendoza	M	13.6m	45.2%	Market-value substitute prior
Thiago Almada	M	22.8m	53.5%	Market-value substitute prior
Ademola Lookman	F	26.7m	36.3%	Market-value substitute prior

Antagonist / tactical summary

Away midfield dictates play with ~55% possession expected (Jon Ander Olasagasti vs Pablo Barrios). Total antagonist xG — Home: 0.66, Away: 1.31. Home danger zone: Left Wing vs Right Back — Flank Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Midfield — possession 45.4% / 54.6%, dominance ratio 0.35, key duel: Jon Ander Olasagasti vs Pablo Barrios

Roger Brugué xG 0.24Thiago Fernández xG 0.10Iván Romero xG 0.19Alejandro Grimaldo xG 0.20Pablo Barrios xG 0.08Julián Alvarez xG 0.39

Summary

Antagonist matchup analysis for Levante vs Atlético Madrid. Most likely scoreline: 1-2. Lambdas derived from 6 home and 7 away individual player duels. Head-to-head: Levante 2W-4D-4L vs Atlético Madrid over 10 meetings.

Tactical parity between 4-3-3 and 4-2-3-1. Match outcome driven primarily by individual player quality. Away midfield dictates play with ~55% possession expected (Jon Ander Olasagasti vs Pablo Barrios). Total antagonist xG — Home: 0.66, Away: 1.31. Home danger zone: Left Wing vs Right Back — Flank Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Man City vs Brighton

Premier League · 20261031 · 4-2-3-1 vs 4-2-3-1

Predicted 2-1Winner: Man CityWinning H 2-1 (9.1%)Winning A 1-2 (4.0%)Confidence 0.79λ home/away 2.62 / 1.15Most-likely P 9.1%

Actual vs predicted: result not available yet (upcoming fixture).

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Man City vs Brighton. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 4-2-3-1 vs 4-2-3-1 (duel lambdas: Home 2.747, Away 1.419). Final calibrated Poisson lambdas: Home 2.615, Away 1.154 (most likely score 2-1).

Home Starters 11/11 XGBoost (10 FBref) Away Starters 11/11 XGBoost (10 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 2.75 / 1.42 Final λ (H/A) 2.62 / 1.15

FBref Team Priors: Manchester City (Gls/90: 3.0, GA/90: 3.0 → E[λ]: 3.5) vs Brighton & Hove Albion (Gls/90: 2.5, GA/90: 3.5 → E[λ]: 3.5) | Blend Weight: 0.192

Modeling Pipeline Stages (5 layers)

1. XGBoost player regressions for volume & event rates with market-value quality scaling
2. Multi-source prior blending (FBref 2026-2027 tables + Sofascore season history)
3. Antagonist formation duel matrix resolving 1v1 and zonal matchups
4. Team-level prior calibration (FBref possession and attack/defense efficiencies)
5. Bivariate Poisson matrix with Dixon-Coles low-score correction

Standing & form context

Metric	Man City	Brighton
League position	1.00	3.00
Points per game	2.08	1.72
Attack index	1.32	1.43
Defense (concede) index	0.85	0.92
Form multiplier	1.02	0.98
Standing λ multiplier	1.05	1.12

Lambda decomposition (how the goal rates were built)

Stage	Man City λ	Brighton λ
FBref Squad Prior E[λ] (weight 0.192)	3.500	3.500
Pre-mismatch	3.087	1.347
Pre-calibration (Micro-duels)	2.747	1.419
Final λ	2.615	1.154

competition calibration scale 0.85competition observed total goals 2.781quality gap applied Truemismatch score pressure applied Truecompetitive underdog floor applied Falsestage discount 1.0

Scoreline distribution & markets

Top 5 scorelines	P
2-1	9.1%
3-1	7.9%
2-0	7.9%
1-1	7.0%
3-0	6.9%

Home win	68.6%
Draw	16.7%
Away win	14.6%
Over 2.5	72.6%
BTTS yes	63.5%

Pregame fair line — favorite Man City (68.6%), fair total 3.77, fair spread 1.46

Projected players vs season baseline

Proj = projected for this match · Season = this season's per-game average from match data · deviation badge compares projected xG to the player's season xG/game.

Man City — projected line vs season baseline (4-2-3-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Gianluigi Donnarumma G 90 0 0.00 - n/a 0.0 6.9 57.8 33 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0097, quality 0.809)	`market_value_eur=46000000` `quality_factor=0.8089` `boost_factor=1.0097` `reference_quality=0.505` `history_matches=42` `history_shrink=0.1064`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0092)	`blend_weight=0.0092` `history_matches=42` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	saves: 2.8368 -> 2.8291
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Charalampos Kostoulas, Diego Gómez', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.4192}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.62 (squad xG sum 1.8062, xG scale 1.4478, xA scale 1.4478, team xA 1.90 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.615` `squad_xg_sum=1.8062` `squad_xa_sum=1.3129` `team_expected_assists=1.9009` `team_xa_to_xg_ratio=0.7269`	shots_faced: 13.5 -> 11.0 expected_saves: 3.31 -> 2.7 high_pressure_saves: 1.14 -> 0.92
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 10 by projected rating, fifa_overall 89.0)	`match_squad_rank=10` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=2.9791`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=1.8886`	-

Matheus Nunes D 85 0 0.00 0.00 34% 0 0.1 7.0 50.5 33 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0099, quality 0.796)	`market_value_eur=43000000` `quality_factor=0.7957` `boost_factor=1.0099` `reference_quality=0.505` `history_matches=39` `history_shrink=0.1136`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0020)	`blend_weight=0.002` `history_matches=39` `season_overlap_discount=1.0` `per90_to_match_factor=0.9499` `prior_reliability=0.0267` `competition=Premier League`	totalCross: 1.3665 -> 1.3732
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Charalampos Kostoulas', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.62 (squad xG sum 1.8062, xG scale 1.4478, xA scale 1.4478, team xA 1.90 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.615` `squad_xg_sum=1.8062` `squad_xa_sum=1.3129` `team_expected_assists=1.9009` `team_xa_to_xg_ratio=0.7269`	expected_assists: 0.0377 -> 0.0546
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 83.0)	`match_squad_rank=8` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0036` `goal_rate=0.0006` `scorer_weight=0.0035`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0531` `assist_rate=0.0496` `provider_weight=0.062`	-

Abdukodir Khusanov D 89 0 0.04 0.02 87% 1 0.2 7.0 45.6 32 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0149, quality 0.813)	`market_value_eur=47000000` `quality_factor=0.8129` `boost_factor=1.0149` `reference_quality=0.505` `history_matches=26` `history_shrink=0.1613`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0140)	`blend_weight=0.014` `history_matches=26` `season_overlap_discount=1.0` `per90_to_match_factor=0.9854` `prior_reliability=0.1333` `competition=Premier League`	wonTackle: 1.052 -> 1.0717 fouls: 1.3378 -> 1.3191 totalShots: 0.5687 -> 0.5608 wasFouled: 0.5684 -> 0.5605
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=true`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Pascal Groß', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.62 (squad xG sum 1.8062, xG scale 1.4478, xA scale 1.4478, team xA 1.90 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.615` `squad_xg_sum=1.8062` `squad_xa_sum=1.3129` `team_expected_assists=1.9009` `team_xa_to_xg_ratio=0.7269`	expected_goals: 0.0284 -> 0.0411 expected_assists: 0.0462 -> 0.0669
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 82.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0403` `goal_rate=0.0` `scorer_weight=0.0329`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0647` `assist_rate=0.0` `provider_weight=0.0535`	-

Joško Gvardiol D 85 0 0.18 0.09 109% 1 0.6 6.8 44.6 19 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0176, quality 0.869)	`market_value_eur=69000000` `quality_factor=0.8692` `boost_factor=1.0176` `reference_quality=0.505` `history_matches=26` `history_shrink=0.1613`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0126)	`blend_weight=0.0126` `history_matches=26` `season_overlap_discount=1.0` `per90_to_match_factor=0.9423` `prior_reliability=0.12` `competition=Premier League`	wasFouled: 1.3226 -> 1.306 totalCross: 0.2822 -> 0.2918 wonTackle: 1.1964 -> 1.1879 goals: 0.115 -> 0.1201
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Charalampos Kostoulas', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.62 (squad xG sum 1.8062, xG scale 1.4478, xA scale 1.4478, team xA 1.90 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.615` `squad_xg_sum=1.8062` `squad_xa_sum=1.3129` `team_expected_assists=1.9009` `team_xa_to_xg_ratio=0.7269`	expected_goals: 0.1263 -> 0.1829 expected_assists: 0.1147 -> 0.1661
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 15 by projected rating, fifa_overall 85.0)	`match_squad_rank=15` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1671` `goal_rate=0.1201` `scorer_weight=0.1765`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.153` `assist_rate=0.2699` `provider_weight=0.2076`	-

Elliot Anderson M 87 0 0.09 0.04 134% 1 1.0 7.3 61.0 33 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0115, quality 0.896)	`market_value_eur=109000000` `quality_factor=0.8957` `boost_factor=1.0115` `reference_quality=0.505` `history_matches=46` `history_shrink=0.098`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0068)	`blend_weight=0.0068` `history_matches=46` `season_overlap_discount=1.0` `per90_to_match_factor=0.9616` `prior_reliability=0.1067` `competition=Premier League`	interceptionWon: 1.4187 -> 1.4295 wasFouled: 3.0994 -> 3.0906 wonTackle: 1.2715 -> 1.2792 fouls: 1.423 -> 1.4297
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Chema Andrés', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.468, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v1', 'xg_multiplier': 1.0701, 'skill_edge': 0.0917, 'height_edge': -0.044, 'attacker_skill': 0.8984, 'defender_skill': 0.6946, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Chema Andrés', 'market_value': 23000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.6946}], 'xg_override': 0.0641}	aerialWon: 1.2368 -> 1.0 totalShots: 0.8367 -> 0.8954 shotOffTarget: 0.6438 -> 0.6889 onTargetScoringAttempt: 0.1929 -> 0.2064
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.62 (squad xG sum 1.8062, xG scale 1.4478, xA scale 1.4478, team xA 1.90 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.615` `squad_xg_sum=1.8062` `squad_xa_sum=1.3129` `team_expected_assists=1.9009` `team_xa_to_xg_ratio=0.7269`	expected_goals: 0.0641 -> 0.0928 expected_assists: 0.1348 -> 0.1952
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 84.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0886` `goal_rate=0.0708` `scorer_weight=0.0982`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1773` `assist_rate=0.0906` `provider_weight=0.1562`	-

Enzo Fernández M 86 0 0.23 0.17 33% 2 1.4 7.1 58.1 31 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0122, quality 0.897)	`market_value_eur=116000000` `quality_factor=0.897` `boost_factor=1.0122` `reference_quality=0.505` `history_matches=43` `history_shrink=0.1042`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0014)	`blend_weight=0.0014` `history_matches=43` `season_overlap_discount=1.0` `per90_to_match_factor=0.9598` `prior_reliability=0.02` `competition=Premier League`	totalCross: 3.9545 -> 3.9491
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Chema Andrés', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.46, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v1', 'xg_multiplier': 1.0587, 'skill_edge': 0.0919, 'height_edge': -0.052, 'attacker_skill': 0.8989, 'defender_skill': 0.6946, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Chema Andrés', 'market_value': 23000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.6946}], 'xg_override': 0.1584}	aerialWon: 0.7711 -> 1.0 totalShots: 1.5695 -> 1.6616 shotOffTarget: 1.0209 -> 1.0808 onTargetScoringAttempt: 0.5486 -> 0.5808 goals: 0.1997 -> 0.2114 expectedGoals: 0.1496 -> 0.1584 expectedGoalsOnTarget: 0.1288 -> 0.1364 bigChanceMissed: 0.1196 -> 0.1266
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.62 (squad xG sum 1.8062, xG scale 1.4478, xA scale 1.4478, team xA 1.90 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.615` `squad_xg_sum=1.8062` `squad_xa_sum=1.3129` `team_expected_assists=1.9009` `team_xa_to_xg_ratio=0.7269`	expected_goals: 0.1584 -> 0.2293 expected_assists: 0.1963 -> 0.2842
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 86.0)	`match_squad_rank=5` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2049` `goal_rate=0.2114` `scorer_weight=0.2866`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.2474` `assist_rate=0.1361` `provider_weight=0.2274`	-

Iliman Ndiaye F 81 0 0.29 0.13 117% 2 1.1 7.1 49.8 40 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0118, quality 0.859)	`market_value_eur=63000000` `quality_factor=0.8589` `boost_factor=1.0118` `reference_quality=0.505` `history_matches=40` `history_shrink=0.1111`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0096)	`blend_weight=0.0096` `history_matches=40` `season_overlap_discount=1.0` `per90_to_match_factor=0.9038` `prior_reliability=0.1333` `competition=Premier League`	totalCross: 2.4705 -> 2.4554 wasFouled: 1.3133 -> 1.305 onTargetScoringAttempt: 0.7696 -> 0.7752 wonTackle: 0.985 -> 0.9799
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Luka Vušković, Lewis Dunk', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.237, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 1.0541, 'skill_edge': 0.0828, 'height_edge': -0.046, 'attacker_skill': 0.8589, 'defender_skill': 0.6749, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Luka Vušković', 'market_value': 69000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.89}, {'player': 'Lewis Dunk', 'market_value': 3600000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4597}], 'xg_override': 0.2005}	totalShots: 1.7902 -> 1.887 shotOffTarget: 1.015 -> 1.0699 onTargetScoringAttempt: 0.7752 -> 0.8171 bigChanceMissed: 0.2078 -> 0.219 goals: 0.1962 -> 0.2068 expectedGoals: 0.1902 -> 0.2005 expectedGoalsOnTarget: 0.1401 -> 0.1477
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.62 (squad xG sum 1.8062, xG scale 1.4478, xA scale 1.4478, team xA 1.90 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.615` `squad_xg_sum=1.8062` `squad_xa_sum=1.3129` `team_expected_assists=1.9009` `team_xa_to_xg_ratio=0.7269`	expected_goals: 0.2005 -> 0.2903 expected_assists: 0.1542 -> 0.2233
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 82.0)	`match_squad_rank=6` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.252` `goal_rate=0.2068` `scorer_weight=0.2976`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.2001` `assist_rate=0.3545` `provider_weight=0.2791`	-

Rayan Cherki M 73 0 0.38 0.13 185% 3 1.6 7.2 57.6 34 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0131, quality 0.898)	`market_value_eur=120000000` `quality_factor=0.8975` `boost_factor=1.0131` `reference_quality=0.505` `history_matches=40` `history_shrink=0.1111`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0058)	`blend_weight=0.0058` `history_matches=40` `season_overlap_discount=1.0` `per90_to_match_factor=0.8119` `prior_reliability=0.08` `competition=Premier League`	totalCross: 2.2824 -> 2.2927 fouls: 0.9872 -> 0.9971 goalAssist: 0.1766 -> 0.1834 onTargetScoringAttempt: 0.8688 -> 0.8754 goals: 0.2798 -> 0.286 wasFouled: 0.8828 -> 0.8777
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Maxim De Cuyper', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle', 'duel_win_rate': 0.501, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v1', 'xg_multiplier': 1.0312, 'skill_edge': 0.0492, 'height_edge': -0.028, 'attacker_skill': 0.8975, 'defender_skill': 0.7881, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Maxim De Cuyper', 'market_value': 29000000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.7881}], 'xg_override': 0.2649}	aerialWon: 0.0 -> 0.9 totalShots: 2.3878 -> 2.4623 shotOffTarget: 1.5124 -> 1.5596 onTargetScoringAttempt: 0.8754 -> 0.9027 expectedGoalsOnTarget: 0.4315 -> 0.445 goals: 0.286 -> 0.2949 expectedGoals: 0.2569 -> 0.2649 bigChanceMissed: 0.1987 -> 0.2049
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.62 (squad xG sum 1.8062, xG scale 1.4478, xA scale 1.4478, team xA 1.90 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.615` `squad_xg_sum=1.8062` `squad_xa_sum=1.3129` `team_expected_assists=1.9009` `team_xa_to_xg_ratio=0.7269`	expected_goals: 0.2649 -> 0.3835 expected_assists: 0.2075 -> 0.3004
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 86.0)	`match_squad_rank=4` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.3185` `goal_rate=0.2949` `scorer_weight=0.42`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.2595` `assist_rate=0.1834` `provider_weight=0.2723`	-

Jérémy Doku F 83 0 0.25 0.06 307% 3 1.1 7.3 46.7 32 XGBoost F_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0141, quality 0.871)	`market_value_eur=70000000` `quality_factor=0.8707` `boost_factor=1.0141` `reference_quality=0.505` `history_matches=34` `history_shrink=0.1282`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Luka Vušković, Lewis Dunk', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.228, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 1.0413, 'skill_edge': 0.0881, 'height_edge': -0.06, 'attacker_skill': 0.8707, 'defender_skill': 0.6749, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Luka Vušković', 'market_value': 69000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.89}, {'player': 'Lewis Dunk', 'market_value': 3600000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4597}], 'xg_override': 0.175}	totalShots: 2.5283 -> 2.6328 shotOffTarget: 1.8612 -> 1.9381 onTargetScoringAttempt: 0.6671 -> 0.6947 expectedGoalsOnTarget: 0.3153 -> 0.3283 goals: 0.2134 -> 0.2222 expectedGoals: 0.1681 -> 0.175
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.62 (squad xG sum 1.8062, xG scale 1.4478, xA scale 1.4478, team xA 1.90 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.615` `squad_xg_sum=1.8062` `squad_xa_sum=1.3129` `team_expected_assists=1.9009` `team_xa_to_xg_ratio=0.7269`	expected_goals: 0.175 -> 0.2534 expected_assists: 0.2184 -> 0.316
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 84.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2238` `goal_rate=0.2222` `scorer_weight=0.3066`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=1` `assist_probability=0.2709` `assist_rate=0.2541` `provider_weight=0.358`	-

Erling Haaland F 88 2 0.92 0.48 92% 4 2.8 7.4 56.4 32 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0123, quality 0.900)	`market_value_eur=240000000` `quality_factor=0.9` `boost_factor=1.0123` `reference_quality=0.505` `history_matches=43` `history_shrink=0.1042`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0090)	`blend_weight=0.009` `history_matches=43` `season_overlap_discount=1.0` `per90_to_match_factor=0.9777` `prior_reliability=0.1333` `competition=Premier League`	totalShots: 2.9982 -> 3.0153 onTargetScoringAttempt: 1.4368 -> 1.4459 fouls: 0.769 -> 0.7621
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Luka Vušković, Lewis Dunk', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.315, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 1.1695, 'skill_edge': 0.1013, 'height_edge': 0.014, 'attacker_skill': 0.9, 'defender_skill': 0.6749, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Luka Vušković', 'market_value': 69000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.89}, {'player': 'Lewis Dunk', 'market_value': 3600000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4597}], 'xg_override': 0.636}	totalShots: 3.0153 -> 3.5264 shotOffTarget: 1.5614 -> 1.8261 onTargetScoringAttempt: 1.4459 -> 1.691 expectedGoalsOnTarget: 0.9417 -> 1.1013 goals: 0.7908 -> 0.9248 bigChanceMissed: 0.651 -> 0.7613 expectedGoals: 0.5438 -> 0.636
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	goals: 1.0 -> 0
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.62 (squad xG sum 1.8062, xG scale 1.4478, xA scale 1.4478, team xA 1.90 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.615` `squad_xg_sum=1.8062` `squad_xa_sum=1.3129` `team_expected_assists=1.9009` `team_xa_to_xg_ratio=0.7269`	expected_goals: 0.636 -> 0.9208 expected_assists: 0.0608 -> 0.088
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 91.0)	`match_squad_rank=1` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=2` `goal_probability=0.6018` `goal_rate=0.9248` `scorer_weight=1.151`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0842` `assist_rate=0.0672` `provider_weight=0.0931`	-

Rúben Dias D 89 0 0.08 0.03 232% 1 0.3 7.0 43.4 26 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0129, quality 0.840)	`market_value_eur=55000000` `quality_factor=0.8399` `boost_factor=1.0129` `reference_quality=0.505` `history_matches=34` `history_shrink=0.1282`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0111)	`blend_weight=0.0111` `history_matches=34` `season_overlap_discount=1.0` `per90_to_match_factor=0.9864` `prior_reliability=0.1333` `competition=Premier League`	fouls: 1.6517 -> 1.6334
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Charalampos Kostoulas', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.62 (squad xG sum 1.8062, xG scale 1.4478, xA scale 1.4478, team xA 1.90 = 0.73 x lambda)	`scoreline_reset=true` `team_lambda=2.615` `squad_xg_sum=1.8062` `squad_xa_sum=1.3129` `team_expected_assists=1.9009` `team_xa_to_xg_ratio=0.7269`	expected_goals: 0.0579 -> 0.0838 expected_assists: 0.0669 -> 0.0969
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 87.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0804` `goal_rate=0.0134` `scorer_weight=0.067`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0924` `assist_rate=0.0` `provider_weight=0.0775`	-

Team totals 2 2.62 21 7.1

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Ederson	G	4.7m	12.2%	Market-value substitute prior
Nico O'Reilly	D	15.3m	25.3%	Market-value substitute prior
Marc Guéhi	D	15.3m	25.2%	Market-value substitute prior
Ayyoub Bouaddi	M	15.7m	72.9%	Market-value substitute prior
Phil Foden	M	23.7m	34.2%	Market-value substitute prior
Antoine Semenyo	F	27.1m	37.0%	Market-value substitute prior

Brighton — projected line vs season baseline (4-2-3-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Bart Verbruggen G 90 0 0.00 - n/a 0.0 7.0 57.7 33 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0076, quality 0.764)	`market_value_eur=37000000` `quality_factor=0.7641` `boost_factor=1.0076` `reference_quality=0.505` `history_matches=46` `history_shrink=0.098`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0085)	`blend_weight=0.0085` `history_matches=46` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	saves: 2.9074 -> 2.887
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Erling Haaland, Rayan Cherki', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 2.7469}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.15 (squad xG sum 1.2390, xG scale 0.9314, xA scale 0.9314, team xA 0.91 = 0.78 x lambda)	`scoreline_reset=true` `team_lambda=1.154` `squad_xg_sum=1.239` `squad_xa_sum=0.9722` `team_expected_assists=0.9055` `team_xa_to_xg_ratio=0.7846`	shots_faced: 26.1 -> 24.8 expected_saves: 6.39 -> 6.06 high_pressure_saves: 2.2 -> 2.09
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 81.0)	`match_squad_rank=4` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=1.2134`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=0.9064`	-

Costinha D 90 0 0.02 0.01 70% 1 0.2 6.7 34.1 14 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0009, quality 0.512)	`market_value_eur=9100000` `quality_factor=0.5118` `boost_factor=1.0009` `reference_quality=0.505` `history_matches=7` `history_shrink=0.4167`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0072)	`blend_weight=0.0072` `history_matches=7` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.0267` `competition=Premier League`	totalShots: 0.5274 -> 0.5579 onTargetScoringAttempt: 0.0834 -> 0.0999 interceptionWon: 0.7669 -> 0.7794 wonTackle: 0.9032 -> 0.9147 fouls: 1.0028 -> 1.0136 totalCross: 1.1349 -> 1.1267
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Erling Haaland', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.15 (squad xG sum 1.2390, xG scale 0.9314, xA scale 0.9314, team xA 0.91 = 0.78 x lambda)	`scoreline_reset=true` `team_lambda=1.154` `squad_xg_sum=1.239` `squad_xa_sum=0.9722` `team_expected_assists=0.9055` `team_xa_to_xg_ratio=0.7846`	expected_assists: 0.1088 -> 0.1013
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 16 by projected rating, fifa_overall 75.0)	`match_squad_rank=16` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0227` `goal_rate=0.0` `scorer_weight=0.0184`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0963` `assist_rate=0.0` `provider_weight=0.081`	-

Luka Vušković D 89 0 0.17 0.09 91% 1 1.1 7.1 42.4 14 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0420, quality 0.869)	`market_value_eur=69000000` `quality_factor=0.8692` `boost_factor=1.042` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0333)	`blend_weight=0.0333` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=0.9886` `prior_reliability=0.1333` `competition=Premier League`	interceptionWon: 1.2228 -> 1.1821 wasFouled: 1.0365 -> 1.002 totalShots: 1.3477 -> 1.3687 fouls: 1.5357 -> 1.5175 totalCross: 0.4055 -> 0.392 onTargetScoringAttempt: 0.2779 -> 0.2686 wonTackle: 0.7113 -> 0.7041
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Erling Haaland', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.15 (squad xG sum 1.2390, xG scale 0.9314, xA scale 0.9314, team xA 0.91 = 0.78 x lambda)	`scoreline_reset=true` `team_lambda=1.154` `squad_xg_sum=1.239` `squad_xa_sum=0.9722` `team_expected_assists=0.9055` `team_xa_to_xg_ratio=0.7846`	expected_goals: 0.1823 -> 0.1698
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 81.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1562` `goal_rate=0.0921` `scorer_weight=0.1358`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0246` `assist_rate=0.0` `provider_weight=0.0199`	-

Lewis Dunk D 90 0 0.03 0.04 ok 1 0.5 7.1 53.0 30 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9980, quality 0.445)	`market_value_eur=3600000` `quality_factor=0.4449` `boost_factor=0.998` `reference_quality=0.505` `history_matches=40` `history_shrink=0.1111`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0096)	`blend_weight=0.0096` `history_matches=40` `season_overlap_discount=1.0` `per90_to_match_factor=0.9987` `prior_reliability=0.1333` `competition=Premier League`	fouls: 0.466 -> 0.4759 totalCross: 0.1765 -> 0.1844
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Erling Haaland', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.15 (squad xG sum 1.2390, xG scale 0.9314, xA scale 0.9314, team xA 0.91 = 0.78 x lambda)	`scoreline_reset=true` `team_lambda=1.154` `squad_xg_sum=1.239` `squad_xa_sum=0.9722` `team_expected_assists=0.9055` `team_xa_to_xg_ratio=0.7846`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 77.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0322` `goal_rate=0.1509` `scorer_weight=0.0409`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0166` `assist_rate=0.0` `provider_weight=0.0134`	-

Olivier Boscagli D 79 0 0.01 0.01 122% 1 0.2 7.0 37.1 34 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0067, quality 0.616)	`market_value_eur=18500000` `quality_factor=0.6161` `boost_factor=1.0067` `reference_quality=0.505` `history_matches=20` `history_shrink=0.2`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0173)	`blend_weight=0.0173` `history_matches=20` `season_overlap_discount=1.0` `per90_to_match_factor=0.8746` `prior_reliability=0.1333` `competition=Premier League`	fouls: 1.0602 -> 1.0721 totalCross: 2.708 -> 2.699 onTargetScoringAttempt: 0.0608 -> 0.0673
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Iliman Ndiaye', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.15 (squad xG sum 1.2390, xG scale 0.9314, xA scale 0.9314, team xA 0.91 = 0.78 x lambda)	`scoreline_reset=true` `team_lambda=1.154` `squad_xg_sum=1.239` `squad_xa_sum=0.9722` `team_expected_assists=0.9055` `team_xa_to_xg_ratio=0.7846`	expected_assists: 0.0788 -> 0.0734
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 78.0)	`match_squad_rank=5` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0147` `goal_rate=0.0` `scorer_weight=0.0118`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0708` `assist_rate=0.1508` `provider_weight=0.0918`	-

Pascal Groß M 87 0 0.11 0.11 ok 1 0.9 7.2 58.1 33 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9968, quality 0.437)	`market_value_eur=2300000` `quality_factor=0.4374` `boost_factor=0.9968` `reference_quality=0.505` `history_matches=27` `history_shrink=0.1562`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0135)	`blend_weight=0.0135` `history_matches=27` `season_overlap_discount=1.0` `per90_to_match_factor=0.9713` `prior_reliability=0.1333` `competition=Premier League`	totalCross: 6.4988 -> 6.5292 totalShots: 1.6308 -> 1.6153 interceptionWon: 0.4566 -> 0.4504 wasFouled: 0.3917 -> 0.3864
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Abdukodir Khusanov', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle', 'duel_win_rate': 0.267, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.6869, 'skill_edge': -0.193, 'height_edge': -0.02, 'attacker_skill': 0.4374, 'defender_skill': 0.8663, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Abdukodir Khusanov', 'market_value': 47000000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.8663}], 'xg_override': 0.118}	aerialWon: 0.5298 -> 1.1 totalShots: 1.6298 -> 1.1195 shotOffTarget: 1.0727 -> 0.7368 onTargetScoringAttempt: 0.5571 -> 0.3827 goals: 0.1787 -> 0.1227 expectedGoals: 0.1718 -> 0.118 expectedGoalsOnTarget: 0.1062 -> 0.0729 bigChanceMissed: 0.0704 -> 0.0484
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.15 (squad xG sum 1.2390, xG scale 0.9314, xA scale 0.9314, team xA 0.91 = 0.78 x lambda)	`scoreline_reset=true` `team_lambda=1.154` `squad_xg_sum=1.239` `squad_xa_sum=0.9722` `team_expected_assists=0.9055` `team_xa_to_xg_ratio=0.7846`	expected_goals: 0.118 -> 0.1099 expected_assists: 0.2031 -> 0.1891
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 79.0)	`match_squad_rank=1` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1041` `goal_rate=0.1227` `scorer_weight=0.113`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1723` `assist_rate=0.1704` `provider_weight=0.1647`	-

Chema Andrés M 80 0 0.06 0.04 ok 1 0.3 6.8 29.2 12 XGBoost M_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0232, quality 0.660)	`market_value_eur=23000000` `quality_factor=0.6595` `boost_factor=1.0232` `reference_quality=0.505` `history_matches=5` `history_shrink=0.5`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Elliot Anderson', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.372, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.9299, 'skill_edge': -0.0917, 'height_edge': 0.044, 'attacker_skill': 0.6946, 'defender_skill': 0.8984, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Elliot Anderson', 'market_value': 109000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8984}], 'xg_override': 0.0601}	totalShots: 0.7737 -> 0.7195 aerialWon: 1.2588 -> 1.3 shotOffTarget: 0.5131 -> 0.4771 onTargetScoringAttempt: 0.2605 -> 0.2422 expectedGoalsOnTarget: 0.1833 -> 0.1705
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.15 (squad xG sum 1.2390, xG scale 0.9314, xA scale 0.9314, team xA 0.91 = 0.78 x lambda)	`scoreline_reset=true` `team_lambda=1.154` `squad_xg_sum=1.239` `squad_xa_sum=0.9722` `team_expected_assists=0.9055` `team_xa_to_xg_ratio=0.7846`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 86.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0545` `goal_rate=0.0609` `scorer_weight=0.0564`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0365` `assist_rate=0.0526` `provider_weight=0.0425`	-

Diego Gómez M 84 0 0.16 0.13 ok 2 1.4 7.0 54.1 29 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0076, quality 0.732)	`market_value_eur=32000000` `quality_factor=0.732` `boost_factor=1.0076` `reference_quality=0.505` `history_matches=40` `history_shrink=0.1111`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0092)	`blend_weight=0.0092` `history_matches=40` `season_overlap_discount=1.0` `per90_to_match_factor=0.9307` `prior_reliability=0.1267` `competition=Premier League`	totalShots: 2.1177 -> 2.1333 wonTackle: 1.1542 -> 1.1616 onTargetScoringAttempt: 0.6991 -> 0.7058
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Elliot Anderson', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.385, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.949, 'skill_edge': -0.0587, 'height_edge': 0.024, 'attacker_skill': 0.7679, 'defender_skill': 0.8984, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Elliot Anderson', 'market_value': 109000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8984}], 'xg_override': 0.1734}	aerialWon: 2.9534 -> 1.2 totalShots: 2.1333 -> 2.0245 shotOffTarget: 1.4186 -> 1.3463 onTargetScoringAttempt: 0.7058 -> 0.6698 bigChanceMissed: 0.2704 -> 0.2566 goals: 0.1947 -> 0.1848 expectedGoals: 0.1827 -> 0.1734 expectedGoalsOnTarget: 0.1091 -> 0.1035
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.15 (squad xG sum 1.2390, xG scale 0.9314, xA scale 0.9314, team xA 0.91 = 0.78 x lambda)	`scoreline_reset=true` `team_lambda=1.154` `squad_xg_sum=1.239` `squad_xa_sum=0.9722` `team_expected_assists=0.9055` `team_xa_to_xg_ratio=0.7846`	expected_goals: 0.1734 -> 0.1615 expected_assists: 0.1042 -> 0.097
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 78.0)	`match_squad_rank=6` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1491` `goal_rate=0.1848` `scorer_weight=0.1697`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0924` `assist_rate=0.1222` `provider_weight=0.1083`	-

Yasin Ayari M 84 0 0.13 0.07 90% 2 0.9 6.9 48.5 27 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0103, quality 0.786)	`market_value_eur=41000000` `quality_factor=0.7859` `boost_factor=1.0103` `reference_quality=0.505` `history_matches=36` `history_shrink=0.122`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0037)	`blend_weight=0.0037` `history_matches=36` `season_overlap_discount=1.0` `per90_to_match_factor=0.9353` `prior_reliability=0.0467` `competition=Premier League`	fouls: 0.5457 -> 0.5536
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Elliot Anderson, Enzo Fernández', 'matchup_zone': 'Attacking Mid vs Double Pivot — Tight Space Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.8971, 'skill_edge': -0.046, 'height_edge': -0.024, 'attacker_skill': 0.7964, 'defender_skill': 0.8986, 'attacker_role': 'CAM', 'attacker_mv_price_index': 0.95, 'defender_adjusted_parts': [{'player': 'Elliot Anderson', 'market_value': 109000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8984}, {'player': 'Enzo Fernández', 'market_value': 116000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8989}], 'xg_override': 0.1409}	aerialWon: 0.302 -> 0.8 totalShots: 1.8621 -> 1.6705 shotOffTarget: 1.3961 -> 1.2524 onTargetScoringAttempt: 0.4613 -> 0.4138 expectedGoalsOnTarget: 0.2052 -> 0.1841 goals: 0.1597 -> 0.1433 expectedGoals: 0.1571 -> 0.1409 bigChanceMissed: 0.154 -> 0.1382
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.15 (squad xG sum 1.2390, xG scale 0.9314, xA scale 0.9314, team xA 0.91 = 0.78 x lambda)	`scoreline_reset=true` `team_lambda=1.154` `squad_xg_sum=1.239` `squad_xa_sum=0.9722` `team_expected_assists=0.9055` `team_xa_to_xg_ratio=0.7846`	expected_goals: 0.1409 -> 0.1312
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 79.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.123` `goal_rate=0.1433` `scorer_weight=0.1329`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0583` `assist_rate=0.0686` `provider_weight=0.0623`	-

Maxim De Cuyper D 82 0 0.12 0.10 ok 1 0.6 6.8 49.6 38 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0072, quality 0.710)	`market_value_eur=29000000` `quality_factor=0.71` `boost_factor=1.0072` `reference_quality=0.505` `history_matches=38` `history_shrink=0.1163`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0096)	`blend_weight=0.0096` `history_matches=38` `season_overlap_discount=1.0` `per90_to_match_factor=0.9079` `prior_reliability=0.1267` `competition=Premier League`	totalCross: 4.2843 -> 4.3028 totalShots: 0.8661 -> 0.8812 wasFouled: 1.0788 -> 1.0685 fouls: 1.3271 -> 1.3327
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Rayan Cherki', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.15 (squad xG sum 1.2390, xG scale 0.9314, xA scale 0.9314, team xA 0.91 = 0.78 x lambda)	`scoreline_reset=true` `team_lambda=1.154` `squad_xg_sum=1.239` `squad_xa_sum=0.9722` `team_expected_assists=0.9055` `team_xa_to_xg_ratio=0.7846`	expected_goals: 0.1293 -> 0.1204 expected_assists: 0.1854 -> 0.1727
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 78.0)	`match_squad_rank=8` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1134` `goal_rate=0.1723` `scorer_weight=0.1493`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=1` `assist_probability=0.1586` `assist_rate=0.207` `provider_weight=0.1886`	-

Charalampos Kostoulas F 78 1 0.26 0.13 104% 2 1.1 6.8 43.2 30 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0128, quality 0.796)	`market_value_eur=43000000` `quality_factor=0.7957` `boost_factor=1.0128` `reference_quality=0.505` `history_matches=29` `history_shrink=0.1471`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0076)	`blend_weight=0.0076` `history_matches=29` `season_overlap_discount=1.0` `per90_to_match_factor=0.8714` `prior_reliability=0.08` `competition=Premier League`	wasFouled: 1.4458 -> 1.4347 wonTackle: 0.6546 -> 0.6496
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Joško Gvardiol, Rúben Dias', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.9308, 'skill_edge': -0.0391, 'height_edge': -0.008, 'attacker_skill': 0.7957, 'defender_skill': 0.8826, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Joško Gvardiol', 'market_value': 69000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.89}, {'player': 'Rúben Dias', 'market_value': 55000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8751}], 'xg_override': 0.2778}	totalShots: 2.2516 -> 2.0958 shotOffTarget: 1.464 -> 1.3627 onTargetScoringAttempt: 0.7873 -> 0.7328 goals: 0.3737 -> 0.3478 expectedGoalsOnTarget: 0.331 -> 0.3081 bigChanceMissed: 0.3193 -> 0.2972 expectedGoals: 0.2985 -> 0.2778
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.15 (squad xG sum 1.2390, xG scale 0.9314, xA scale 0.9314, team xA 0.91 = 0.78 x lambda)	`scoreline_reset=true` `team_lambda=1.154` `squad_xg_sum=1.239` `squad_xa_sum=0.9722` `team_expected_assists=0.9055` `team_xa_to_xg_ratio=0.7846`	expected_goals: 0.2778 -> 0.2587
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 11 by projected rating, fifa_overall 73.0)	`match_squad_rank=11` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.2279` `goal_rate=0.3478` `scorer_weight=0.3139`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0592` `assist_rate=0.3407` `provider_weight=0.0762`	-

Team totals 1 1.15 15 6.9

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Tom McGill	G	4.4m	10.9%	Market-value substitute prior
Ferdi Kadıoğlu	D	14.9m	36.1%	Market-value substitute prior
Mats Wieffer	D	14.5m	23.6%	Market-value substitute prior
Jack Hinshelwood	M	23.2m	33.2%	Market-value substitute prior
Georginio Rutter	M	23.1m	33.0%	Market-value substitute prior
Yankuba Minteh	F	26.8m	62.9%	Market-value substitute prior

Antagonist / tactical summary

Home midfield dominates with ~52% possession expected (Rayan Cherki vs Yasin Ayari). Total antagonist xG — Home: 1.50, Away: 0.77. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Midfield — possession 51.6% / 48.4%, dominance ratio 0.57, key duel: Rayan Cherki vs Yasin Ayari

Rayan Cherki xG 0.26Erling Haaland xG 0.64Iliman Ndiaye xG 0.20Pascal Groß xG 0.12Yasin Ayari xG 0.14Charalampos Kostoulas xG 0.28

Summary

Antagonist matchup analysis for Man City vs Brighton. Most likely scoreline: 2-1. Lambdas derived from 6 home and 5 away individual player duels. Head-to-head: Man City 5W-3D-2L vs Brighton over 10 meetings.

Tactical parity between 4-2-3-1 and 4-2-3-1. Match outcome driven primarily by individual player quality. Home midfield dominates with ~52% possession expected (Rayan Cherki vs Yasin Ayari). Total antagonist xG — Home: 1.50, Away: 0.77. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Rayo Vallecano vs Celta

LaLiga · 20261031 · 4-2-3-1 vs 3-4-3

Predicted 2-1Winner: Rayo VallecanoWinning H 2-1 (9.4%)Winning A 1-2 (7.1%)Confidence 0.73λ home/away 1.70 / 1.28Most-likely P 9.4%

Actual vs predicted: result not available yet (upcoming fixture).

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Rayo Vallecano vs Celta. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 4-2-3-1 vs 3-4-3 (duel lambdas: Home 1.027, Away 0.982). Final calibrated Poisson lambdas: Home 1.704, Away 1.280 (most likely score 2-1).

Home Starters 11/11 XGBoost (7 FBref) Away Starters 11/11 XGBoost (10 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 1.03 / 0.98 Final λ (H/A) 1.70 / 1.28

FBref Team Priors: Rayo Vallecano (Gls/90: 1.3333333333333333, GA/90: 1.33 → E[λ]: 0.247) vs Celta Vigo (Gls/90: 0.25, GA/90: 0.25 → E[λ]: 0.246) | Blend Weight: 0.228

Modeling Pipeline Stages (5 layers)

1. XGBoost player regressions for volume & event rates with market-value quality scaling
2. Multi-source prior blending (FBref 2026-2027 tables + Sofascore season history)
3. Antagonist formation duel matrix resolving 1v1 and zonal matchups
4. Team-level prior calibration (FBref possession and attack/defense efficiencies)
5. Bivariate Poisson matrix with Dixon-Coles low-score correction

Standing & form context

Metric	Rayo Vallecano	Celta
League position	17.00	9.00
Points per game	1.15	1.26
Attack index	1.00	0.85
Defense (concede) index	0.76	1.27
Form multiplier	0.99	1.01
Standing λ multiplier	0.90	1.03

Lambda decomposition (how the goal rates were built)

Stage	Rayo Vallecano λ	Celta λ
FBref Squad Prior E[λ] (weight 0.228)	0.247	0.246
Pre-competitive-floor	1.326	1.038
Pre-calibration (Micro-duels)	1.027	0.982
Final λ	1.704	1.280

competition calibration scale 1.262competition observed total goals 2.984quality gap applied Truemismatch score pressure applied Falsecompetitive underdog floor applied Falsestage discount 1.0

Scoreline distribution & markets

Top 5 scorelines	P
1-1	11.1%
2-1	9.4%
1-0	8.5%
2-0	7.3%
1-2	7.1%

Home win	47.3%
Draw	23.9%
Away win	28.7%
Over 2.5	57.3%
BTTS yes	59.1%

Pregame fair line — favorite Rayo Vallecano (47.3%), fair total 2.98, fair spread 0.42

Projected players vs season baseline

Proj = projected for this match · Season = this season's per-game average from match data · deviation badge compares projected xG to the player's season xG/game.

Rayo Vallecano — projected line vs season baseline (4-2-3-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Emil Audero G 90 0 0.00 - n/a 0.0 7.0 34.4 10 XGBoost G_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9899, quality 0.437)	`market_value_eur=2700000` `quality_factor=0.4374` `boost_factor=0.9899` `reference_quality=0.505` `history_matches=5` `history_shrink=0.5`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Ferrán Jutglà, Pablo Durán', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 0.982}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.70 (squad xG sum 1.1313, xG scale 1.5063, xA scale 1.5063, team xA 1.21 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.704` `squad_xg_sum=1.1313` `squad_xa_sum=0.8` `team_expected_assists=1.2051` `team_xa_to_xg_ratio=0.7072`	shots_faced: 9.3 -> 12.2 expected_saves: 2.27 -> 2.99 high_pressure_saves: 0.79 -> 1.02
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 79.0)	`match_squad_rank=4` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=1.8323`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=1.1517`	-

Iván Balliu D 79 0 0.03 0.00 3462% 0 0.0 6.7 28.8 30 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9962, quality 0.437)	`market_value_eur=1100000` `quality_factor=0.4374` `boost_factor=0.9962` `reference_quality=0.505` `history_matches=22` `history_shrink=0.1852`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0160)	`blend_weight=0.016` `history_matches=22` `season_overlap_discount=1.0` `per90_to_match_factor=0.8808` `prior_reliability=0.1333` `competition=La Liga`	totalCross: 0.8873 -> 0.8801 totalShots: 0.3318 -> 0.3265
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Ferrán Jutglà', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.70 (squad xG sum 1.1313, xG scale 1.5063, xA scale 1.5063, team xA 1.21 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.704` `squad_xg_sum=1.1313` `squad_xa_sum=0.8` `team_expected_assists=1.2051` `team_xa_to_xg_ratio=0.7072`	expected_goals: 0.0189 -> 0.0285 expected_assists: 0.0311 -> 0.0468
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 15 by projected rating, fifa_overall 74.0)	`match_squad_rank=15` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0281` `goal_rate=0.0` `scorer_weight=0.0228`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0457` `assist_rate=0.0` `provider_weight=0.0374`	-

Florian Lejeune D 90 0 0.11 0.06 89% 1 1.4 7.0 59.5 33 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9980, quality 0.437)	`market_value_eur=1700000` `quality_factor=0.4374` `boost_factor=0.998` `reference_quality=0.505` `history_matches=45` `history_shrink=0.1`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0130)	`blend_weight=0.013` `history_matches=45` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.2` `competition=La Liga`	interceptionWon: 1.1913 -> 1.2105 wonTackle: 0.9052 -> 0.9194 fouls: 1.4816 -> 1.4753
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Ferrán Jutglà', 'matchup_zone': 'Shadow Striker vs Centre Back — Half-Space Threat'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.70 (squad xG sum 1.1313, xG scale 1.5063, xA scale 1.5063, team xA 1.21 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.704` `squad_xg_sum=1.1313` `squad_xa_sum=0.8` `team_expected_assists=1.2051` `team_xa_to_xg_ratio=0.7072`	expected_goals: 0.0741 -> 0.1116 expected_assists: 0.0169 -> 0.0255
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 79.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1056` `goal_rate=0.0471` `scorer_weight=0.0893`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0252` `assist_rate=0.0094` `provider_weight=0.0226`	-

Mujaid Sadick D 86 0 0.05 0.01 370% 0 0.1 6.8 33.1 10 XGBoost D_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9930, quality 0.458)	`market_value_eur=4700000` `quality_factor=0.4585` `boost_factor=0.993` `reference_quality=0.505` `history_matches=5` `history_shrink=0.5`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Pablo Durán', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.70 (squad xG sum 1.1313, xG scale 1.5063, xA scale 1.5063, team xA 1.21 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.704` `squad_xg_sum=1.1313` `squad_xa_sum=0.8` `team_expected_assists=1.2051` `team_xa_to_xg_ratio=0.7072`	expected_goals: 0.0338 -> 0.0509
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 84.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0496` `goal_rate=0.0` `scorer_weight=0.0407`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0087` `assist_rate=0.0` `provider_weight=0.0078`	-

Adrià Pedrosa D 82 0 0.05 0.01 258% 1 0.4 6.7 44.7 32 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9974, quality 0.437)	`market_value_eur=2700000` `quality_factor=0.4374` `boost_factor=0.9974` `reference_quality=0.505` `history_matches=34` `history_shrink=0.1282`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0039)	`blend_weight=0.0039` `history_matches=34` `season_overlap_discount=1.0` `per90_to_match_factor=0.9127` `prior_reliability=0.0467` `competition=La Liga`	fouls: 1.3042 -> 1.3093
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Aleix Febas', 'matchup_zone': 'Left Wingback vs Right Back — Overlapping Run'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.70 (squad xG sum 1.1313, xG scale 1.5063, xA scale 1.5063, team xA 1.21 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.704` `squad_xg_sum=1.1313` `squad_xa_sum=0.8` `team_expected_assists=1.2051` `team_xa_to_xg_ratio=0.7072`	expected_goals: 0.0354 -> 0.0533 expected_assists: 0.1358 -> 0.2046
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 13 by projected rating, fifa_overall 74.0)	`match_squad_rank=13` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0519` `goal_rate=0.102` `scorer_weight=0.0666`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=1` `assist_probability=0.185` `assist_rate=0.12` `provider_weight=0.1872`	-

Óscar Valentín M 72 0 0.05 0.02 109% 0 0.2 6.9 53.6 32 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9978, quality 0.437)	`market_value_eur=1900000` `quality_factor=0.4374` `boost_factor=0.9978` `reference_quality=0.505` `history_matches=41` `history_shrink=0.1087`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0108)	`blend_weight=0.0108` `history_matches=41` `season_overlap_discount=1.0` `per90_to_match_factor=0.7946` `prior_reliability=0.1533` `competition=La Liga`	totalCross: 0.1696 -> 0.1752 fouls: 1.2291 -> 1.2345 totalShots: 0.484 -> 0.4788 interceptionWon: 0.9159 -> 0.9209
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Ilaix Moriba, Javier Rueda', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.70 (squad xG sum 1.1313, xG scale 1.5063, xA scale 1.5063, team xA 1.21 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.704` `squad_xg_sum=1.1313` `squad_xa_sum=0.8` `team_expected_assists=1.2051` `team_xa_to_xg_ratio=0.7072`	expected_goals: 0.0345 -> 0.052 expected_assists: 0.044 -> 0.0663
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 77.0)	`match_squad_rank=5` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0507` `goal_rate=0.0518` `scorer_weight=0.0626`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0641` `assist_rate=0.052` `provider_weight=0.0719`	-

Unai López M 81 0 0.09 0.03 257% 1 0.6 7.1 50.7 30 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9976, quality 0.437)	`market_value_eur=2600000` `quality_factor=0.4374` `boost_factor=0.9976` `reference_quality=0.505` `history_matches=37` `history_shrink=0.119`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0139)	`blend_weight=0.0139` `history_matches=37` `season_overlap_discount=1.0` `per90_to_match_factor=0.9047` `prior_reliability=0.18` `competition=La Liga`	totalShots: 1.1063 -> 1.0956 fouls: 0.9474 -> 0.9575 interceptionWon: 1.1698 -> 1.1628
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Javier Rueda', 'matchup_zone': 'Midfield lane'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.70 (squad xG sum 1.1313, xG scale 1.5063, xA scale 1.5063, team xA 1.21 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.704` `squad_xg_sum=1.1313` `squad_xa_sum=0.8` `team_expected_assists=1.2051` `team_xa_to_xg_ratio=0.7072`	expected_goals: 0.0614 -> 0.0925 expected_assists: 0.1446 -> 0.2178
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 76.0)	`match_squad_rank=1` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0884` `goal_rate=0.0752` `scorer_weight=0.104`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1957` `assist_rate=0.1061` `provider_weight=0.1747`	-

Jorge de Frutos F 69 0 0.20 0.19 ok 1 1.1 6.7 60.1 36 XGBoost F_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0009, quality 0.533)	`market_value_eur=10900000` `quality_factor=0.533` `boost_factor=1.0009` `reference_quality=0.505` `history_matches=43` `history_shrink=0.1042`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=true`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Javi Rodríguez, Marcos Alonso', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 0.9123, 'skill_edge': -0.0077, 'height_edge': -0.052, 'attacker_skill': 0.533, 'defender_skill': 0.55, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Javi Rodríguez', 'market_value': 17500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.6625}, {'player': 'Marcos Alonso', 'market_value': 1300000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.1334}	totalShots: 1.2907 -> 1.1775 shotOffTarget: 0.7939 -> 0.7243 onTargetScoringAttempt: 0.4968 -> 0.4532 expectedGoalsOnTarget: 0.1755 -> 0.1601 goals: 0.154 -> 0.1405 expectedGoals: 0.1462 -> 0.1334 bigChanceMissed: 0.1059 -> 0.0966
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.70 (squad xG sum 1.1313, xG scale 1.5063, xA scale 1.5063, team xA 1.21 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.704` `squad_xg_sum=1.1313` `squad_xa_sum=0.8` `team_expected_assists=1.2051` `team_xa_to_xg_ratio=0.7072`	expected_goals: 0.1334 -> 0.2009 expected_assists: 0.0633 -> 0.0953
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 12 by projected rating, fifa_overall 81.0)	`match_squad_rank=12` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.182` `goal_rate=0.1405` `scorer_weight=0.2087`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0909` `assist_rate=0.195` `provider_weight=0.1191`	-

Isi Palazón M 67 0 0.13 0.10 35% 1 1.1 6.8 53.8 24 XGBoost M_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9975, quality 0.437)	`market_value_eur=3000000` `quality_factor=0.4374` `boost_factor=0.9975` `reference_quality=0.505` `history_matches=35` `history_shrink=0.125`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Javi Rodríguez', 'matchup_zone': 'LW vs CB', 'duel_win_rate': 0.259, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.7629, 'skill_edge': -0.1013, 'height_edge': -0.06, 'attacker_skill': 0.4374, 'defender_skill': 0.6625, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Javi Rodríguez', 'market_value': 17500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.6625}], 'xg_override': 0.0869}	totalShots: 1.5131 -> 1.1543 shotOffTarget: 1.0627 -> 0.8107 onTargetScoringAttempt: 0.4503 -> 0.3435 aerialWon: 0.8712 -> 0.8 bigChanceMissed: 0.1481 -> 0.113 goals: 0.1245 -> 0.095 expectedGoals: 0.1139 -> 0.0869 expectedGoalsOnTarget: 0.0729 -> 0.0556
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.70 (squad xG sum 1.1313, xG scale 1.5063, xA scale 1.5063, team xA 1.21 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.704` `squad_xg_sum=1.1313` `squad_xa_sum=0.8` `team_expected_assists=1.2051` `team_xa_to_xg_ratio=0.7072`	expected_goals: 0.0869 -> 0.1309 expected_assists: 0.1174 -> 0.1768
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 79.0)	`match_squad_rank=8` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1227` `goal_rate=0.095` `scorer_weight=0.1386`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1621` `assist_rate=0.094` `provider_weight=0.1521`	-

Álvaro García F 82 1 0.35 0.16 114% 2 1.0 6.9 53.4 38 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9978, quality 0.437)	`market_value_eur=1500000` `quality_factor=0.4374` `boost_factor=0.9978` `reference_quality=0.505` `history_matches=41` `history_shrink=0.1087`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0104)	`blend_weight=0.0104` `history_matches=41` `season_overlap_discount=1.0` `per90_to_match_factor=0.9089` `prior_reliability=0.1467` `competition=La Liga`	totalCross: 4.0359 -> 4.0198 totalShots: 2.0446 -> 2.0321 fouls: 0.5481 -> 0.5553
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Javi Rodríguez, Marcos Alonso', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 0.8373, 'skill_edge': -0.0507, 'height_edge': -0.06, 'attacker_skill': 0.4374, 'defender_skill': 0.55, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Javi Rodríguez', 'market_value': 17500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.6625}, {'player': 'Marcos Alonso', 'market_value': 1300000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.2308}	totalShots: 2.0415 -> 1.7094 shotOffTarget: 1.3279 -> 1.1119 onTargetScoringAttempt: 0.7136 -> 0.5975 goals: 0.3139 -> 0.2628 expectedGoalsOnTarget: 0.2847 -> 0.2384 expectedGoals: 0.2756 -> 0.2308 bigChanceMissed: 0.2525 -> 0.2114
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.70 (squad xG sum 1.1313, xG scale 1.5063, xA scale 1.5063, team xA 1.21 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.704` `squad_xg_sum=1.1313` `squad_xa_sum=0.8` `team_expected_assists=1.2051` `team_xa_to_xg_ratio=0.7072`	expected_goals: 0.2308 -> 0.3476 expected_assists: 0.085 -> 0.128
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 81.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.2936` `goal_rate=0.2628` `scorer_weight=0.3872`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1201` `assist_rate=0.2127` `provider_weight=0.16`	-

Sergio Camello F 71 1 0.44 0.17 154% 2 1.4 7.0 46.9 31 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0012, quality 0.533)	`market_value_eur=10900000` `quality_factor=0.533` `boost_factor=1.0012` `reference_quality=0.505` `history_matches=31` `history_shrink=0.1389`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0126)	`blend_weight=0.0126` `history_matches=31` `season_overlap_discount=1.0` `per90_to_match_factor=0.7933` `prior_reliability=0.14` `competition=La Liga`	goals: 0.3602 -> 0.37 wasFouled: 1.5246 -> 1.534 onTargetScoringAttempt: 0.8541 -> 0.8622 totalCross: 0.1612 -> 0.1687
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Javi Rodríguez, Marcos Alonso', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 0.9358, 'skill_edge': -0.0077, 'height_edge': -0.036, 'attacker_skill': 0.533, 'defender_skill': 0.55, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Javi Rodríguez', 'market_value': 17500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.6625}, {'player': 'Marcos Alonso', 'market_value': 1300000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.2928}	totalShots: 2.2529 -> 2.1083 shotOffTarget: 1.3907 -> 1.3014 onTargetScoringAttempt: 0.8622 -> 0.8068 bigChanceMissed: 0.4454 -> 0.4168 expectedGoalsOnTarget: 0.4036 -> 0.3777 goals: 0.37 -> 0.3462 expectedGoals: 0.3129 -> 0.2928
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.70 (squad xG sum 1.1313, xG scale 1.5063, xA scale 1.5063, team xA 1.21 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.704` `squad_xg_sum=1.1313` `squad_xa_sum=0.8` `team_expected_assists=1.2051` `team_xa_to_xg_ratio=0.7072`	expected_goals: 0.2928 -> 0.4411 expected_assists: 0.092 -> 0.1386
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 74.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.3567` `goal_rate=0.3462` `scorer_weight=0.5098`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1294` `assist_rate=0.0711` `provider_weight=0.1182`	-

Team totals 2 1.70 12 6.9

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Augusto Batalla	G	4.5m	13.4%	Market-value substitute prior
Andrei Rațiu	D	14.4m	23.4%	Market-value substitute prior
Abdul Mumin	D	13.6m	29.9%	Market-value substitute prior
Pedro Díaz	M	23.1m	69.2%	Market-value substitute prior
Pathé Ismaël Ciss	M	22.0m	49.9%	Market-value substitute prior
Alemão	F	25.4m	56.2%	Market-value substitute prior

Celta — projected line vs season baseline (3-4-3)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Ionuț Radu G 90 0 0.00 - n/a 0.0 7.1 63.9 34 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0004, quality 0.519)	`market_value_eur=9700000` `quality_factor=0.5189` `boost_factor=1.0004` `reference_quality=0.505` `history_matches=48` `history_shrink=0.0943`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0164)	`blend_weight=0.0164` `history_matches=48` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.2667` `competition=La Liga`	saves: 2.8675 -> 2.8778
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Sergio Camello, Álvaro García', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.0273}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.28 (squad xG sum 0.9213, xG scale 1.3894, xA scale 1.3894, team xA 0.91 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.28` `squad_xg_sum=0.9213` `squad_xa_sum=0.6541` `team_expected_assists=0.9087` `team_xa_to_xg_ratio=0.7099`	shots_faced: 9.8 -> 16.2 expected_saves: 2.4 -> 3.97 high_pressure_saves: 0.82 -> 1.36
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 78.0)	`match_squad_rank=1` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=1.4379`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=0.9297`	-

Javi Rodríguez D 84 0 0.03 0.02 64% 1 0.3 6.9 52.0 33 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0034, quality 0.606)	`market_value_eur=17500000` `quality_factor=0.6058` `boost_factor=1.0034` `reference_quality=0.505` `history_matches=39` `history_shrink=0.1136`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0128)	`blend_weight=0.0128` `history_matches=39` `season_overlap_discount=1.0` `per90_to_match_factor=0.936` `prior_reliability=0.1733` `competition=La Liga`	wasFouled: 1.1099 -> 1.1003
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Isi Palazón', 'matchup_zone': 'LW vs CB'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.28 (squad xG sum 0.9213, xG scale 1.3894, xA scale 1.3894, team xA 0.91 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.28` `squad_xg_sum=0.9213` `squad_xa_sum=0.6541` `team_expected_assists=0.9087` `team_xa_to_xg_ratio=0.7099`	expected_goals: 0.0198 -> 0.0275 expected_assists: 0.0623 -> 0.0866
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 76.0)	`match_squad_rank=5` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0271` `goal_rate=0.1388` `scorer_weight=0.0344`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.083` `assist_rate=0.1449` `provider_weight=0.1082`	-

Yoel Lago D 89 0 0.02 0.01 322% 0 0.1 6.9 42.7 29 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9990, quality 0.487)	`market_value_eur=7000000` `quality_factor=0.4866` `boost_factor=0.999` `reference_quality=0.505` `history_matches=22` `history_shrink=0.1852`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0281)	`blend_weight=0.0281` `history_matches=22` `season_overlap_discount=1.0` `per90_to_match_factor=0.9937` `prior_reliability=0.2333` `competition=La Liga`	totalShots: 0.3882 -> 0.3773 totalCross: 0.0 -> 0.008 wonTackle: 0.8261 -> 0.8188 interceptionWon: 1.177 -> 1.1838 wasFouled: 1.3346 -> 1.329
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Álvaro García', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.28 (squad xG sum 0.9213, xG scale 1.3894, xA scale 1.3894, team xA 0.91 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.28` `squad_xg_sum=0.9213` `squad_xa_sum=0.6541` `team_expected_assists=0.9087` `team_xa_to_xg_ratio=0.7099`	expected_goals: 0.0178 -> 0.0247
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 71.0)	`match_squad_rank=6` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0244` `goal_rate=0.0` `scorer_weight=0.0198`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0149` `assist_rate=0.0` `provider_weight=0.0123`	-

Marcos Alonso D 87 0 0.07 0.07 ok 1 0.7 7.0 60.7 30 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9977, quality 0.437)	`market_value_eur=1300000` `quality_factor=0.4374` `boost_factor=0.9977` `reference_quality=0.505` `history_matches=40` `history_shrink=0.1111`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0125)	`blend_weight=0.0125` `history_matches=40` `season_overlap_discount=1.0` `per90_to_match_factor=0.9684` `prior_reliability=0.1733` `competition=La Liga`	interceptionWon: 1.0576 -> 1.0817 wonTackle: 0.8087 -> 0.8172
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Sergio Camello', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.28 (squad xG sum 0.9213, xG scale 1.3894, xA scale 1.3894, team xA 0.91 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.28` `squad_xg_sum=0.9213` `squad_xa_sum=0.6541` `team_expected_assists=0.9087` `team_xa_to_xg_ratio=0.7099`	expected_goals: 0.0514 -> 0.0714 expected_assists: 0.0377 -> 0.0524
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 79.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0689` `goal_rate=0.0048` `scorer_weight=0.0571`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0511` `assist_rate=0.0` `provider_weight=0.0419`	-

Javier Rueda M 73 0 0.07 0.02 209% 1 0.3 6.8 50.1 36 XGBoost M_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9990, quality 0.477)	`market_value_eur=6200000` `quality_factor=0.4769` `boost_factor=0.999` `reference_quality=0.505` `history_matches=36` `history_shrink=0.122`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Óscar Valentín', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.468, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.0258, 'skill_edge': 0.0215, 'height_edge': -0.004, 'attacker_skill': 0.4853, 'defender_skill': 0.4374, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Óscar Valentín', 'market_value': 1900000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.4374}], 'xg_override': 0.0526}	aerialWon: 0.5866 -> 0.9 totalShots: 0.817 -> 0.8381 shotOffTarget: 0.5797 -> 0.5947 onTargetScoringAttempt: 0.2372 -> 0.2433
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.28 (squad xG sum 0.9213, xG scale 1.3894, xA scale 1.3894, team xA 0.91 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.28` `squad_xg_sum=0.9213` `squad_xa_sum=0.6541` `team_expected_assists=0.9087` `team_xa_to_xg_ratio=0.7099`	expected_goals: 0.0526 -> 0.0731 expected_assists: 0.0914 -> 0.127
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 12 by projected rating, fifa_overall 81.0)	`match_squad_rank=12` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0705` `goal_rate=0.0555` `scorer_weight=0.0752`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1193` `assist_rate=0.093` `provider_weight=0.1284`	-

Aleix Febas M 84 0 0.06 0.03 82% 1 0.4 6.9 58.3 29 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9987, quality 0.465)	`market_value_eur=5200000` `quality_factor=0.4646` `boost_factor=0.9987` `reference_quality=0.505` `history_matches=41` `history_shrink=0.1087`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0085)	`blend_weight=0.0085` `history_matches=41` `season_overlap_discount=1.0` `per90_to_match_factor=0.935` `prior_reliability=0.12` `competition=La Liga`	fouls: 1.6338 -> 1.642
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Adrià Pedrosa', 'matchup_zone': 'Left Wingback vs Right Back — Overlapping Run', 'duel_win_rate': 0.388, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9819, 'skill_edge': 0.0197, 'height_edge': -0.032, 'attacker_skill': 0.4918, 'defender_skill': 0.4481, 'attacker_role': 'LWB', 'attacker_mv_price_index': 0.7, 'defender_adjusted_parts': [{'player': 'Adrià Pedrosa', 'market_value': 2700000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.4481}], 'xg_override': 0.0425}	aerialWon: 0.0927 -> 0.8 totalShots: 0.6092 -> 0.5982 shotOffTarget: 0.4007 -> 0.3934
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.28 (squad xG sum 0.9213, xG scale 1.3894, xA scale 1.3894, team xA 0.91 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.28` `squad_xg_sum=0.9213` `squad_xa_sum=0.6541` `team_expected_assists=0.9087` `team_xa_to_xg_ratio=0.7099`	expected_goals: 0.0425 -> 0.059 expected_assists: 0.0529 -> 0.0735
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 79.0)	`match_squad_rank=8` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0573` `goal_rate=0.0505` `scorer_weight=0.0641`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0709` `assist_rate=0.0629` `provider_weight=0.0806`	-

Ilaix Moriba M 81 0 0.07 0.03 113% 1 0.4 6.9 62.1 32 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0018, quality 0.565)	`market_value_eur=13700000` `quality_factor=0.5649` `boost_factor=1.0018` `reference_quality=0.505` `history_matches=46` `history_shrink=0.098`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0115)	`blend_weight=0.0115` `history_matches=46` `season_overlap_discount=1.0` `per90_to_match_factor=0.8981` `prior_reliability=0.18` `competition=La Liga`	fouls: 1.2362 -> 1.222 interceptionWon: 1.1237 -> 1.1184
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Óscar Valentín', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.539, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.1306, 'skill_edge': 0.0649, 'height_edge': 0.024, 'attacker_skill': 0.5816, 'defender_skill': 0.4374, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Óscar Valentín', 'market_value': 1900000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.4374}], 'xg_override': 0.0477}	aerialWon: 0.1933 -> 1.1 totalShots: 0.598 -> 0.6761 shotOffTarget: 0.457 -> 0.5167 onTargetScoringAttempt: 0.1384 -> 0.1565 expectedGoalsOnTarget: 0.0859 -> 0.0971 goals: 0.0458 -> 0.0518 expectedGoals: 0.0422 -> 0.0477
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.28 (squad xG sum 0.9213, xG scale 1.3894, xA scale 1.3894, team xA 0.91 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.28` `squad_xg_sum=0.9213` `squad_xa_sum=0.6541` `team_expected_assists=0.9087` `team_xa_to_xg_ratio=0.7099`	expected_goals: 0.0477 -> 0.0663 expected_assists: 0.043 -> 0.0597
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 78.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0641` `goal_rate=0.0518` `scorer_weight=0.0691`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.058` `assist_rate=0.048` `provider_weight=0.0629`	-

Javi Galán D 79 0 0.01 0.01 ok 0 0.2 6.8 57.4 37 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9973, quality 0.437)	`market_value_eur=1700000` `quality_factor=0.4374` `boost_factor=0.9973` `reference_quality=0.505` `history_matches=33` `history_shrink=0.1316`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0086)	`blend_weight=0.0086` `history_matches=33` `season_overlap_discount=1.0` `per90_to_match_factor=0.8782` `prior_reliability=0.1` `competition=La Liga`	totalCross: 4.1626 -> 4.157
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Jorge de Frutos', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.28 (squad xG sum 0.9213, xG scale 1.3894, xA scale 1.3894, team xA 0.91 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.28` `squad_xg_sum=0.9213` `squad_xa_sum=0.6541` `team_expected_assists=0.9087` `team_xa_to_xg_ratio=0.7099`	expected_assists: 0.1094 -> 0.152
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 11 by projected rating, fifa_overall 77.0)	`match_squad_rank=11` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0058` `goal_rate=0.0` `scorer_weight=0.0054`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.141` `assist_rate=0.005` `provider_weight=0.1216`	-

Pablo Durán F 67 0 0.32 0.11 193% 2 0.9 6.7 50.6 31 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9981, quality 0.451)	`market_value_eur=4099999` `quality_factor=0.4511` `boost_factor=0.9981` `reference_quality=0.505` `history_matches=37` `history_shrink=0.119`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0077)	`blend_weight=0.0077` `history_matches=37` `season_overlap_discount=1.0` `per90_to_match_factor=0.7407` `prior_reliability=0.1` `competition=La Liga`	totalShots: 1.5928 -> 1.5844
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=true`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Florian Lejeune, Mujaid Sadick', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 0.9281, 'skill_edge': -0.0029, 'height_edge': -0.046, 'attacker_skill': 0.4511, 'defender_skill': 0.4576, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Florian Lejeune', 'market_value': 1700000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'Mujaid Sadick', 'market_value': 4700000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4777}], 'xg_override': 0.2278}	totalShots: 1.5912 -> 1.4768 shotOffTarget: 0.8778 -> 0.8147 onTargetScoringAttempt: 0.7134 -> 0.6621 goals: 0.3452 -> 0.3204 bigChanceMissed: 0.3281 -> 0.3045 expectedGoals: 0.2454 -> 0.2278 expectedGoalsOnTarget: 0.132 -> 0.1225
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.28 (squad xG sum 0.9213, xG scale 1.3894, xA scale 1.3894, team xA 0.91 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.28` `squad_xg_sum=0.9213` `squad_xa_sum=0.6541` `team_expected_assists=0.9087` `team_xa_to_xg_ratio=0.7099`	expected_goals: 0.2278 -> 0.3165 expected_assists: 0.049 -> 0.0681
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 14 by projected rating, fifa_overall 74.0)	`match_squad_rank=14` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2713` `goal_rate=0.3204` `scorer_weight=0.3956`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0658` `assist_rate=0.0` `provider_weight=0.0545`	-

Ferrán Jutglà F 75 1 0.33 0.18 84% 2 1.1 6.7 58.0 31 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0009, quality 0.531)	`market_value_eur=10700000` `quality_factor=0.5306` `boost_factor=1.0009` `reference_quality=0.505` `history_matches=40` `history_shrink=0.1111`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0125)	`blend_weight=0.0125` `history_matches=40` `season_overlap_discount=1.0` `per90_to_match_factor=0.8369` `prior_reliability=0.1733` `competition=La Liga`	totalShots: 1.5481 -> 1.5328 onTargetScoringAttempt: 0.6527 -> 0.6445 wasFouled: 1.6186 -> 1.6265 totalOffside: 0.4596 -> 0.4538 fouls: 1.3732 -> 1.3681
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Florian Lejeune', 'matchup_zone': 'Shadow Striker vs Centre Back — Half-Space Threat', 'duel_win_rate': 0.432, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9735, 'skill_edge': 0.0419, 'height_edge': -0.06, 'attacker_skill': 0.5306, 'defender_skill': 0.4374, 'attacker_role': 'SS', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Florian Lejeune', 'market_value': 1700000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.2344}	totalShots: 1.5398 -> 1.499 shotOffTarget: 0.8953 -> 0.8716 onTargetScoringAttempt: 0.6445 -> 0.6274 expectedGoalsOnTarget: 0.3861 -> 0.3759 goals: 0.3079 -> 0.2997 expectedGoals: 0.2408 -> 0.2344
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.28 (squad xG sum 0.9213, xG scale 1.3894, xA scale 1.3894, team xA 0.91 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.28` `squad_xg_sum=0.9213` `squad_xa_sum=0.6541` `team_expected_assists=0.9087` `team_xa_to_xg_ratio=0.7099`	expected_goals: 0.2344 -> 0.3258 expected_assists: 0.0502 -> 0.0697
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 16 by projected rating, fifa_overall 75.0)	`match_squad_rank=16` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.2781` `goal_rate=0.2997` `scorer_weight=0.4006`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0673` `assist_rate=0.0619` `provider_weight=0.0778`	-

Hugo Álvarez F 73 0 0.19 0.07 151% 1 0.4 6.8 51.7 42 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9997, quality 0.496)	`market_value_eur=7800000` `quality_factor=0.4963` `boost_factor=0.9997` `reference_quality=0.505` `history_matches=36` `history_shrink=0.122`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0069)	`blend_weight=0.0069` `history_matches=36` `season_overlap_discount=1.0` `per90_to_match_factor=0.8066` `prior_reliability=0.0867` `competition=La Liga`	totalCross: 1.2633 -> 1.2802 totalShots: 1.1549 -> 1.147
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=true`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Florian Lejeune', 'matchup_zone': 'Shadow Striker vs Centre Back — Half-Space Threat', 'duel_win_rate': 0.42, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9566, 'skill_edge': 0.0265, 'height_edge': -0.056, 'attacker_skill': 0.4963, 'defender_skill': 0.4374, 'attacker_role': 'SS', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Florian Lejeune', 'market_value': 1700000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.1349}	totalShots: 1.152 -> 1.102 shotOffTarget: 0.7327 -> 0.7009 onTargetScoringAttempt: 0.4193 -> 0.4011 goals: 0.1496 -> 0.1431 expectedGoals: 0.141 -> 0.1349
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.28 (squad xG sum 0.9213, xG scale 1.3894, xA scale 1.3894, team xA 0.91 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.28` `squad_xg_sum=0.9213` `squad_xa_sum=0.6541` `team_expected_assists=0.9087` `team_xa_to_xg_ratio=0.7099`	expected_goals: 0.1349 -> 0.1874 expected_assists: 0.0895 -> 0.1243
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 75.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1709` `goal_rate=0.1431` `scorer_weight=0.1957`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=1` `assist_probability=0.1169` `assist_rate=0.2569` `provider_weight=0.1554`	-

Team totals 1 1.28 12 6.9

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Altay Bayındır	G	4.5m	11.3%	Market-value substitute prior
Abdoulaye Faye	D	13.6m	21.8%	Market-value substitute prior
Carl Starfelt	D	27.6m	41.1%	Market-value substitute prior
Fer López	M	22.6m	52.9%	Market-value substitute prior
Miguel Román	M	34.9m	41.4%	Market-value substitute prior
Williot Swedberg	F	30.4m	44.5%	Market-value substitute prior

Antagonist / tactical summary

Midfield is evenly contested — Isi Palazón vs Ilaix Moriba will be the pivotal duel. Total antagonist xG — Home: 0.74, Away: 0.74. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Shadow Striker vs Centre Back — Half-Space Threat.

Midfield — possession 49.0% / 51.0%, dominance ratio 0.47, key duel: Isi Palazón vs Ilaix Moriba

Isi Palazón xG 0.09Sergio Camello xG 0.29Jorge de Frutos xG 0.13Aleix Febas xG 0.04Ilaix Moriba xG 0.05Javier Rueda xG 0.05

Summary

Antagonist matchup analysis for Rayo Vallecano vs Celta. Most likely scoreline: 2-1. Lambdas derived from 4 home and 6 away individual player duels. Head-to-head: Rayo Vallecano 2W-5D-3L vs Celta over 10 meetings.

Tactical parity between 4-2-3-1 and 3-4-3. Match outcome driven primarily by individual player quality. Midfield is evenly contested — Isi Palazón vs Ilaix Moriba will be the pivotal duel. Total antagonist xG — Home: 0.74, Away: 0.74. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Shadow Striker vs Centre Back — Half-Space Threat.

Tottenham vs Crystal Palace

Premier League · 20261031 · 4-2-3-1 vs 3-4-2-1

Predicted 1-0Winner: TottenhamWinning H 1-0 (9.4%)Winning A 0-1 (7.6%)Confidence 0.72λ home/away 1.53 / 1.25Most-likely P 9.4%

Actual vs predicted: result not available yet (upcoming fixture).

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Tottenham vs Crystal Palace. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 4-2-3-1 vs 3-4-2-1 (duel lambdas: Home 1.495, Away 1.079). Final calibrated Poisson lambdas: Home 1.529, Away 1.252 (most likely score 1-0).

Home Starters 11/11 XGBoost (10 FBref) Away Starters 11/11 XGBoost (11 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 1.50 / 1.08 Final λ (H/A) 1.53 / 1.25

Modeling Pipeline Stages (5 layers)

1. XGBoost player regressions for volume & event rates with market-value quality scaling
2. Multi-source prior blending (FBref 2026-2027 tables + Sofascore season history)
3. Antagonist formation duel matrix resolving 1v1 and zonal matchups
4. Team-level prior calibration (FBref possession and attack/defense efficiencies)
5. Bivariate Poisson matrix with Dixon-Coles low-score correction

Standing & form context

Metric	Tottenham	Crystal Palace
League position	14.00	17.00
Points per game	1.22	1.08
Attack index	0.89	0.92
Defense (concede) index	1.28	0.96
Form multiplier	1.01	0.99
Standing λ multiplier	1.06	0.95

Lambda decomposition (how the goal rates were built)

Stage	Tottenham λ	Crystal Palace λ
Pre-competitive-floor	1.471	1.170
Pre-calibration (Micro-duels)	1.495	1.079
Final λ	1.529	1.252

competition calibration scale 1.053competition observed total goals 2.781quality gap applied Truemismatch score pressure applied Falsecompetitive underdog floor applied Falsestage discount 1.0

Scoreline distribution & markets

Top 5 scorelines	P
1-1	12.0%
1-0	9.4%
2-1	9.1%
0-1	7.6%
1-2	7.4%

Home win	43.6%
Draw	25.3%
Away win	31.1%
Over 2.5	52.6%
BTTS yes	56.0%

Pregame fair line — favorite Tottenham (43.6%), fair total 2.78, fair spread 0.28

Projected players vs season baseline

Proj = projected for this match · Season = this season's per-game average from match data · deviation badge compares projected xG to the player's season xG/game.

Tottenham — projected line vs season baseline (4-2-3-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Antonín Kinský G 90 0 0.00 - n/a 0.0 7.0 30.6 33 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0086, quality 0.620)	`market_value_eur=18900000` `quality_factor=0.6201` `boost_factor=1.0086` `reference_quality=0.505` `history_matches=15` `history_shrink=0.25`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0217)	`blend_weight=0.0217` `history_matches=15` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	saves: 2.8874 -> 2.879 wasFouled: 0.2345 -> 0.2294
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Jean-Philippe Mateta, Yéremy Pino', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.079}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.53 (squad xG sum 1.0797, xG scale 1.4161, xA scale 1.2320, team xA 1.22 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.529` `squad_xg_sum=1.0797` `squad_xa_sum=0.9929` `team_expected_assists=1.2232` `team_xa_to_xg_ratio=0.8`	shots_faced: 10.3 -> 11.9 expected_saves: 2.53 -> 2.91 high_pressure_saves: 0.86 -> 1.0
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 77.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=1.6563`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=1.1846`	-

Pedro Porro D 79 0 0.10 0.06 78% 1 1.3 6.8 53.7 27 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0100, quality 0.791)	`market_value_eur=42000000` `quality_factor=0.7909` `boost_factor=1.01` `reference_quality=0.505` `history_matches=38` `history_shrink=0.1163`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0050)	`blend_weight=0.005` `history_matches=38` `season_overlap_discount=1.0` `per90_to_match_factor=0.8756` `prior_reliability=0.0667` `competition=Premier League`	totalShots: 1.1245 -> 1.1365 totalCross: 5.8317 -> 5.82 wonTackle: 1.3378 -> 1.3311 interceptionWon: 1.003 -> 0.9979
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Tyrick Mitchell', 'matchup_zone': 'Right Wingback vs Left Back — Overlapping Run'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.53 (squad xG sum 1.0797, xG scale 1.4161, xA scale 1.2320, team xA 1.22 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.529` `squad_xg_sum=1.0797` `squad_xa_sum=0.9929` `team_expected_assists=1.2232` `team_xa_to_xg_ratio=0.8`	expected_goals: 0.0714 -> 0.1011 expected_assists: 0.1348 -> 0.1661
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 12 by projected rating, fifa_overall 83.0)	`match_squad_rank=12` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0962` `goal_rate=0.0478` `scorer_weight=0.0814`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.153` `assist_rate=0.0151` `provider_weight=0.1329`	-

Jan Paul van Hecke D 90 0 0.14 0.07 100% 1 0.5 7.3 58.9 32 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0109, quality 0.863)	`market_value_eur=65000000` `quality_factor=0.8627` `boost_factor=1.0109` `reference_quality=0.505` `history_matches=44` `history_shrink=0.102`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0088)	`blend_weight=0.0088` `history_matches=44` `season_overlap_discount=1.0` `per90_to_match_factor=0.9982` `prior_reliability=0.1333` `competition=Premier League`	interceptionWon: 1.196 -> 1.1898
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Jean-Philippe Mateta', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.53 (squad xG sum 1.0797, xG scale 1.4161, xA scale 1.2320, team xA 1.22 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.529` `squad_xg_sum=1.0797` `squad_xa_sum=0.9929` `team_expected_assists=1.2232` `team_xa_to_xg_ratio=0.8`	expected_goals: 0.0976 -> 0.1382 expected_assists: 0.0636 -> 0.0784
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 81.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1291` `goal_rate=0.112` `scorer_weight=0.1517`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0754` `assist_rate=0.0232` `provider_weight=0.0627`	-

Micky van de Ven D 88 0 0.09 0.05 94% 1 0.3 6.9 54.5 31 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0101, quality 0.821)	`market_value_eur=49000000` `quality_factor=0.8206` `boost_factor=1.0101` `reference_quality=0.505` `history_matches=42` `history_shrink=0.1064`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0046)	`blend_weight=0.0046` `history_matches=42` `season_overlap_discount=1.0` `per90_to_match_factor=0.9828` `prior_reliability=0.0667` `competition=Premier League`	fouls: 0.8891 -> 0.8941
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Jean-Philippe Mateta', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.53 (squad xG sum 1.0797, xG scale 1.4161, xA scale 1.2320, team xA 1.22 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.529` `squad_xg_sum=1.0797` `squad_xa_sum=0.9929` `team_expected_assists=1.2232` `team_xa_to_xg_ratio=0.8`	expected_goals: 0.0663 -> 0.0939
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 81.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0896` `goal_rate=0.0108` `scorer_weight=0.0751`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0114` `assist_rate=0.0042` `provider_weight=0.0105`	-

Andy Robertson D 79 0 0.07 0.03 130% 1 0.3 6.8 40.8 33 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9995, quality 0.493)	`market_value_eur=7500000` `quality_factor=0.4927` `boost_factor=0.9995` `reference_quality=0.505` `history_matches=31` `history_shrink=0.1389`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0102)	`blend_weight=0.0102` `history_matches=31` `season_overlap_discount=1.0` `per90_to_match_factor=0.8807` `prior_reliability=0.1133` `competition=Premier League`	wasFouled: 0.1791 -> 0.1879
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Daichi Kamada', 'matchup_zone': 'Left Wingback vs Right Back — Overlapping Run'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.53 (squad xG sum 1.0797, xG scale 1.4161, xA scale 1.2320, team xA 1.22 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.529` `squad_xg_sum=1.0797` `squad_xa_sum=0.9929` `team_expected_assists=1.2232` `team_xa_to_xg_ratio=0.8`	expected_goals: 0.0528 -> 0.0748 expected_assists: 0.0904 -> 0.1114
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 14 by projected rating, fifa_overall 85.0)	`match_squad_rank=14` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0721` `goal_rate=0.0489` `scorer_weight=0.072`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1054` `assist_rate=0.2491` `provider_weight=0.1392`	-

Mateus Fernandes M 81 0 0.10 0.04 155% 1 0.8 7.0 58.0 33 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0111, quality 0.868)	`market_value_eur=68000000` `quality_factor=0.8677` `boost_factor=1.0111` `reference_quality=0.505` `history_matches=44` `history_shrink=0.102`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0066)	`blend_weight=0.0066` `history_matches=44` `season_overlap_discount=1.0` `per90_to_match_factor=0.9026` `prior_reliability=0.1` `competition=Premier League`	totalCross: 2.2436 -> 2.2567 wasFouled: 0.9461 -> 0.9398 interceptionWon: 0.8793 -> 0.8735
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Dwight McNeil, Yéremy Pino, Adam Wharton', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.53 (squad xG sum 1.0797, xG scale 1.4161, xA scale 1.2320, team xA 1.22 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.529` `squad_xg_sum=1.0797` `squad_xa_sum=0.9929` `team_expected_assists=1.2232` `team_xa_to_xg_ratio=0.8`	expected_goals: 0.0725 -> 0.1027 expected_assists: 0.1026 -> 0.1264
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 80.0)	`match_squad_rank=5` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0976` `goal_rate=0.065` `scorer_weight=0.0964`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1187` `assist_rate=0.089` `provider_weight=0.1151`	-

Sandro Tonali M 86 0 0.09 0.05 67% 1 1.2 7.3 56.1 30 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0119, quality 0.884)	`market_value_eur=83000000` `quality_factor=0.8845` `boost_factor=1.0119` `reference_quality=0.505` `history_matches=43` `history_shrink=0.1042`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0081)	`blend_weight=0.0081` `history_matches=43` `season_overlap_discount=1.0` `per90_to_match_factor=0.9576` `prior_reliability=0.12` `competition=Premier League`	wonTackle: 1.1574 -> 1.1523
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Anan Khalaili', 'matchup_zone': 'Midfield lane'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.53 (squad xG sum 1.0797, xG scale 1.4161, xA scale 1.2320, team xA 1.22 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.529` `squad_xg_sum=1.0797` `squad_xa_sum=0.9929` `team_expected_assists=1.2232` `team_xa_to_xg_ratio=0.8`	expected_goals: 0.0639 -> 0.0905 expected_assists: 0.1364 -> 0.1679
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 85.0)	`match_squad_rank=1` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0865` `goal_rate=0.095` `scorer_weight=0.1131`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=1` `assist_probability=0.1546` `assist_rate=0.1124` `provider_weight=0.1463`	-

Mohammed Kudus F 77 0 0.17 0.03 393% 1 0.7 6.9 38.5 25 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0153, quality 0.821)	`market_value_eur=49000000` `quality_factor=0.8206` `boost_factor=1.0153` `reference_quality=0.505` `history_matches=26` `history_shrink=0.1613`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0021)	`blend_weight=0.0021` `history_matches=26` `season_overlap_discount=1.0` `per90_to_match_factor=0.8504` `prior_reliability=0.02` `competition=Premier League`	totalCross: 4.4763 -> 4.4966
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Jaydee Canvot, Chris Richards', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 0.9645, 'skill_edge': 0.0238, 'height_edge': -0.048, 'attacker_skill': 0.8206, 'defender_skill': 0.7676, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Jaydee Canvot', 'market_value': 32000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7941}, {'player': 'Chris Richards', 'market_value': 25000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7411}], 'xg_override': 0.1174}	totalShots: 1.5292 -> 1.4749 shotOffTarget: 1.0823 -> 1.0439 onTargetScoringAttempt: 0.4422 -> 0.4265 goals: 0.1498 -> 0.1445
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.53 (squad xG sum 1.0797, xG scale 1.4161, xA scale 1.2320, team xA 1.22 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.529` `squad_xg_sum=1.0797` `squad_xa_sum=0.9929` `team_expected_assists=1.2232` `team_xa_to_xg_ratio=0.8`	expected_goals: 0.1174 -> 0.1663 expected_assists: 0.0948 -> 0.1168
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 11 by projected rating, fifa_overall 81.0)	`match_squad_rank=11` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1532` `goal_rate=0.1445` `scorer_weight=0.1932`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1102` `assist_rate=0.203` `provider_weight=0.146`	-

James Maddison M 73 0 0.08 0.02 295% 1 0.3 6.9 34.4 18 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0096, quality 0.594)	`market_value_eur=16399999` `quality_factor=0.5942` `boost_factor=1.0096` `reference_quality=0.505` `history_matches=9` `history_shrink=0.3571`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0046)	`blend_weight=0.0046` `history_matches=9` `season_overlap_discount=1.0` `per90_to_match_factor=0.8071` `prior_reliability=0.02` `competition=Premier League`	wasFouled: 1.5563 -> 1.5741 totalCross: 1.0992 -> 1.1066
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Jaydee Canvot', 'matchup_zone': 'LW vs CB', 'duel_win_rate': 0.278, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.7913, 'skill_edge': -0.0899, 'height_edge': -0.052, 'attacker_skill': 0.5942, 'defender_skill': 0.7941, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Jaydee Canvot', 'market_value': 32000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7941}], 'xg_override': 0.056}	aerialWon: 0.1211 -> 0.9 totalShots: 0.8769 -> 0.6939 shotOffTarget: 0.6039 -> 0.4779 onTargetScoringAttempt: 0.273 -> 0.216 expectedGoalsOnTarget: 0.1083 -> 0.0857 goals: 0.0928 -> 0.0734 bigChanceMissed: 0.0757 -> 0.0599 expectedGoals: 0.0708 -> 0.056
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.53 (squad xG sum 1.0797, xG scale 1.4161, xA scale 1.2320, team xA 1.22 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.529` `squad_xg_sum=1.0797` `squad_xa_sum=0.9929` `team_expected_assists=1.2232` `team_xa_to_xg_ratio=0.8`	expected_goals: 0.056 -> 0.0793 expected_assists: 0.0705 -> 0.0869
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 82.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0762` `goal_rate=0.0734` `scorer_weight=0.0923`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0832` `assist_rate=0.0898` `provider_weight=0.1008`	-

Mathys Tel F 80 0 0.20 0.07 174% 1 1.2 6.9 54.4 37 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0055, quality 0.660)	`market_value_eur=23000000` `quality_factor=0.6595` `boost_factor=1.0055` `reference_quality=0.505` `history_matches=37` `history_shrink=0.119`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0088)	`blend_weight=0.0088` `history_matches=37` `season_overlap_discount=1.0` `per90_to_match_factor=0.8902` `prior_reliability=0.1133` `competition=Premier League`	totalCross: 4.6563 -> 4.6338 totalShots: 1.6441 -> 1.6521 wasFouled: 0.9125 -> 0.9045
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Jaydee Canvot, Chris Richards', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 0.8991, 'skill_edge': -0.0486, 'height_edge': -0.02, 'attacker_skill': 0.6595, 'defender_skill': 0.7676, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Jaydee Canvot', 'market_value': 32000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7941}, {'player': 'Chris Richards', 'market_value': 25000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7411}], 'xg_override': 0.141}	totalShots: 1.6521 -> 1.4854 shotOffTarget: 1.1089 -> 0.997 onTargetScoringAttempt: 0.535 -> 0.481 goals: 0.2411 -> 0.2168 expectedGoals: 0.1568 -> 0.141 expectedGoalsOnTarget: 0.1241 -> 0.1116 bigChanceMissed: 0.0908 -> 0.0816
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.53 (squad xG sum 1.0797, xG scale 1.4161, xA scale 1.2320, team xA 1.22 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.529` `squad_xg_sum=1.0797` `squad_xa_sum=0.9929` `team_expected_assists=1.2232` `team_xa_to_xg_ratio=0.8`	expected_goals: 0.141 -> 0.1997 expected_assists: 0.1028 -> 0.1266
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 10 by projected rating, fifa_overall 78.0)	`match_squad_rank=10` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.181` `goal_rate=0.2168` `scorer_weight=0.2496`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1189` `assist_rate=0.0511` `provider_weight=0.1013`	-

Omar Marmoush F 77 1 0.26 0.07 266% 2 1.1 6.8 40.1 32 XGBoost F_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0152, quality 0.840)	`market_value_eur=55000000` `quality_factor=0.8399` `boost_factor=1.0152` `reference_quality=0.505` `history_matches=28` `history_shrink=0.1515`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Jaydee Canvot, Chris Richards', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.213, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 1.0184, 'skill_edge': 0.0325, 'height_edge': -0.02, 'attacker_skill': 0.8399, 'defender_skill': 0.7676, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Jaydee Canvot', 'market_value': 32000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7941}, {'player': 'Chris Richards', 'market_value': 25000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7411}], 'xg_override': 0.1845}	totalShots: 1.9409 -> 1.9766 shotOffTarget: 1.2793 -> 1.3028 onTargetScoringAttempt: 0.6616 -> 0.6738 bigChanceMissed: 0.3936 -> 0.4008
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.53 (squad xG sum 1.0797, xG scale 1.4161, xA scale 1.2320, team xA 1.22 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.529` `squad_xg_sum=1.0797` `squad_xa_sum=0.9929` `team_expected_assists=1.2232` `team_xa_to_xg_ratio=0.8`	expected_goals: 0.1845 -> 0.2612 expected_assists: 0.0558 -> 0.0687
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 13 by projected rating, fifa_overall 82.0)	`match_squad_rank=13` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.2299` `goal_rate=0.2499` `scorer_weight=0.3265`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0664` `assist_rate=0.1179` `provider_weight=0.0859`	-

Team totals 1 1.53 13 7.0

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Guglielmo Vicario	G	5.0m	13.9%	Market-value substitute prior
Destiny Udogie	D	14.3m	62.3%	Market-value substitute prior
Marcos Senesi	D	14.4m	23.5%	Market-value substitute prior
Xavi Simons	M	23.4m	56.3%	Market-value substitute prior
Lucas Bergvall	M	23.4m	33.5%	Market-value substitute prior
Sávio	F	26.9m	63.0%	Market-value substitute prior

Crystal Palace — projected line vs season baseline (3-4-2-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Daichi Kamada M 84 0 0.10 0.07 49% 1 0.5 6.8 55.0 31 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0008, quality 0.526)	`market_value_eur=10300000` `quality_factor=0.526` `boost_factor=1.0008` `reference_quality=0.505` `history_matches=36` `history_shrink=0.122`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0106)	`blend_weight=0.0106` `history_matches=36` `season_overlap_discount=1.0` `per90_to_match_factor=0.9384` `prior_reliability=0.1333` `competition=Premier League`	wonTackle: 1.3478 -> 1.3683 totalCross: 0.8234 -> 0.8345 interceptionWon: 0.3894 -> 0.3952 totalShots: 0.9515 -> 0.9464
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Andy Robertson', 'matchup_zone': 'Left Wingback vs Right Back — Overlapping Run', 'duel_win_rate': 0.444, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.0652, 'skill_edge': 0.0204, 'height_edge': 0.024, 'attacker_skill': 0.5761, 'defender_skill': 0.5308, 'attacker_role': 'LWB', 'attacker_mv_price_index': 0.7, 'defender_adjusted_parts': [{'player': 'Andy Robertson', 'market_value': 7500000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.5308}], 'xg_override': 0.0962}	aerialWon: 0.3443 -> 1.1 totalShots: 0.9477 -> 1.0095 shotOffTarget: 0.5915 -> 0.6301 onTargetScoringAttempt: 0.3562 -> 0.3794 bigChanceMissed: 0.1081 -> 0.1151 expectedGoalsOnTarget: 0.0907 -> 0.0966 expectedGoals: 0.0903 -> 0.0962 goals: 0.0857 -> 0.0913
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.25 (squad xG sum 1.1947, xG scale 1.0480, xA scale 1.0543, team xA 0.81 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.252` `squad_xg_sum=1.1947` `squad_xa_sum=0.7719` `team_expected_assists=0.8138` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 79.0)	`match_squad_rank=6` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0959` `goal_rate=0.0913` `scorer_weight=0.0974`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.078` `assist_rate=0.0848` `provider_weight=0.0862`	-

Tyrick Mitchell M 87 0 0.06 0.06 ok 1 0.6 6.9 56.7 34 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0045, quality 0.660)	`market_value_eur=23000000` `quality_factor=0.6595` `boost_factor=1.0045` `reference_quality=0.505` `history_matches=46` `history_shrink=0.098`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0085)	`blend_weight=0.0085` `history_matches=46` `season_overlap_discount=1.0` `per90_to_match_factor=0.9623` `prior_reliability=0.1333` `competition=Premier League`	wonTackle: 1.743 -> 1.7364 fouls: 0.3165 -> 0.322
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Pedro Porro', 'matchup_zone': 'Right Wingback vs Left Back — Overlapping Run', 'duel_win_rate': 0.352, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.93, 'skill_edge': -0.0516, 'height_edge': 0.004, 'attacker_skill': 0.7379, 'defender_skill': 0.8526, 'attacker_role': 'RWB', 'attacker_mv_price_index': 0.7, 'defender_adjusted_parts': [{'player': 'Pedro Porro', 'market_value': 42000000, 'role': 'LB', 'price_index': 0.7, 'adjusted_quality': 0.8526}], 'xg_override': 0.0612}	aerialWon: 0.7394 -> 0.9 totalShots: 0.9613 -> 0.894 shotOffTarget: 0.7089 -> 0.6593 onTargetScoringAttempt: 0.2503 -> 0.2328 bigChanceMissed: 0.0991 -> 0.0922 expectedGoalsOnTarget: 0.0915 -> 0.0851
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.25 (squad xG sum 1.1947, xG scale 1.0480, xA scale 1.0543, team xA 0.81 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.252` `squad_xg_sum=1.1947` `squad_xa_sum=0.7719` `team_expected_assists=0.8138` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 79.0)	`match_squad_rank=5` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0621` `goal_rate=0.0647` `scorer_weight=0.0661`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0667` `assist_rate=0.0715` `provider_weight=0.0727`	-

Dwight McNeil F 73 0 0.10 0.03 212% 1 0.6 6.7 30.6 35 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0072, quality 0.660)	`market_value_eur=23000000` `quality_factor=0.6595` `boost_factor=1.0072` `reference_quality=0.505` `history_matches=27` `history_shrink=0.1562`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0061)	`blend_weight=0.0061` `history_matches=27` `season_overlap_discount=1.0` `per90_to_match_factor=0.8099` `prior_reliability=0.06` `competition=Premier League`	totalCross: 1.8259 -> 1.8148 wonTackle: 1.0036 -> 0.9975 wasFouled: 0.8574 -> 0.8522
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=true`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Mateus Fernandes, Jan Paul van Hecke', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.8482, 'skill_edge': -0.1013, 'height_edge': -0.002, 'attacker_skill': 0.6595, 'defender_skill': 0.8845, 'attacker_role': 'SS', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Mateus Fernandes', 'market_value': 68000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.882}, {'player': 'Jan Paul van Hecke', 'market_value': 65000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.887}], 'xg_override': 0.0932}	totalShots: 1.1661 -> 0.9891 shotOffTarget: 0.8515 -> 0.7222 onTargetScoringAttempt: 0.3146 -> 0.2668 goals: 0.1169 -> 0.0992 expectedGoals: 0.1099 -> 0.0932 bigChanceMissed: 0.0636 -> 0.0539 expectedGoalsOnTarget: 0.0441 -> 0.0374
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.25 (squad xG sum 1.1947, xG scale 1.0480, xA scale 1.0543, team xA 0.81 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.252` `squad_xg_sum=1.1947` `squad_xa_sum=0.7719` `team_expected_assists=0.8138` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 13 by projected rating, fifa_overall 76.0)	`match_squad_rank=13` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0931` `goal_rate=0.0992` `scorer_weight=0.1018`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0889` `assist_rate=0.0196` `provider_weight=0.0745`	-

Yéremy Pino F 72 0 0.12 0.08 60% 1 0.9 6.9 52.5 35 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0058, quality 0.694)	`market_value_eur=27000000` `quality_factor=0.6941` `boost_factor=1.0058` `reference_quality=0.505` `history_matches=44` `history_shrink=0.102`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0049)	`blend_weight=0.0049` `history_matches=44` `season_overlap_discount=1.0` `per90_to_match_factor=0.8006` `prior_reliability=0.0733` `competition=Premier League`	totalCross: 3.1498 -> 3.1663 fouls: 0.634 -> 0.6415 wasFouled: 1.0142 -> 1.0199 totalOffside: 0.2951 -> 0.3007
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Mateus Fernandes, Jan Paul van Hecke', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.8064, 'skill_edge': -0.0857, 'height_edge': -0.046, 'attacker_skill': 0.6941, 'defender_skill': 0.8845, 'attacker_role': 'SS', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Mateus Fernandes', 'market_value': 68000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.882}, {'player': 'Jan Paul van Hecke', 'market_value': 65000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.887}], 'xg_override': 0.1164}	totalShots: 1.3663 -> 1.1018 shotOffTarget: 0.8188 -> 0.6603 onTargetScoringAttempt: 0.5413 -> 0.4365 expectedGoalsOnTarget: 0.1915 -> 0.1544 goals: 0.19 -> 0.1532 expectedGoals: 0.1443 -> 0.1164 bigChanceMissed: 0.0814 -> 0.0656
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.25 (squad xG sum 1.1947, xG scale 1.0480, xA scale 1.0543, team xA 0.81 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.252` `squad_xg_sum=1.1947` `squad_xa_sum=0.7719` `team_expected_assists=0.8138` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1164 -> 0.122 expected_assists: 0.1304 -> 0.1375
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 79.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.1149` `goal_rate=0.1532` `scorer_weight=0.149`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.1285` `assist_rate=0.0049` `provider_weight=0.11`	-

Jean-Philippe Mateta F 74 0 0.33 0.27 ok 2 0.9 6.7 49.0 23 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0077, quality 0.710)	`market_value_eur=29000000` `quality_factor=0.71` `boost_factor=1.0077` `reference_quality=0.505` `history_matches=35` `history_shrink=0.125`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0038)	`blend_weight=0.0038` `history_matches=35` `season_overlap_discount=1.0` `per90_to_match_factor=0.8262` `prior_reliability=0.0467` `competition=Premier League`	fouls: 0.8695 -> 0.8752
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Jan Paul van Hecke, Micky van de Ven', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.8966, 'skill_edge': -0.0743, 'height_edge': 0.004, 'attacker_skill': 0.71, 'defender_skill': 0.8752, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Jan Paul van Hecke', 'market_value': 65000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.887}, {'player': 'Micky van de Ven', 'market_value': 49000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8633}], 'xg_override': 0.3109}	totalShots: 1.887 -> 1.6919 bigChanceMissed: 1.1915 -> 1.0683 shotOffTarget: 1.1052 -> 0.9909 onTargetScoringAttempt: 0.7769 -> 0.6966 goals: 0.3615 -> 0.3241 expectedGoals: 0.3468 -> 0.3109 expectedGoalsOnTarget: 0.2749 -> 0.2465
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.25 (squad xG sum 1.1947, xG scale 1.0480, xA scale 1.0543, team xA 0.81 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.252` `squad_xg_sum=1.1947` `squad_xa_sum=0.7719` `team_expected_assists=0.8138` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.3109 -> 0.3258
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 16 by projected rating, fifa_overall 81.0)	`match_squad_rank=16` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.2781` `goal_rate=0.3241` `scorer_weight=0.3377`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.031` `assist_rate=0.0` `provider_weight=0.0252`	-

Adam Wharton M 88 0 0.06 0.03 123% 1 0.7 7.0 56.0 32 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0120, quality 0.881)	`market_value_eur=79000000` `quality_factor=0.8811` `boost_factor=1.012` `reference_quality=0.505` `history_matches=42` `history_shrink=0.1064`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0092)	`blend_weight=0.0092` `history_matches=42` `season_overlap_discount=1.0` `per90_to_match_factor=0.9744` `prior_reliability=0.1333` `competition=Premier League`	totalCross: 3.1919 -> 3.167 interceptionWon: 0.7815 -> 0.7923 totalShots: 0.961 -> 0.9521 fouls: 0.706 -> 0.6995 wasFouled: 1.2921 -> 1.2982 wonTackle: 0.8606 -> 0.8661
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Mateus Fernandes', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.469, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 1.0273, 'skill_edge': 0.0026, 'height_edge': 0.016, 'attacker_skill': 0.8877, 'defender_skill': 0.882, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Mateus Fernandes', 'market_value': 68000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.882}], 'xg_override': 0.0556}	totalShots: 0.9582 -> 0.9844 shotOffTarget: 0.6586 -> 0.6766 onTargetScoringAttempt: 0.2996 -> 0.3078
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.25 (squad xG sum 1.1947, xG scale 1.0480, xA scale 1.0543, team xA 0.81 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.252` `squad_xg_sum=1.1947` `squad_xa_sum=0.7719` `team_expected_assists=0.8138` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.1493 -> 0.1574
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 82.0)	`match_squad_rank=1` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0566` `goal_rate=0.0494` `scorer_weight=0.0549`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.1456` `assist_rate=0.0935` `provider_weight=0.1259`	-

Anan Khalaili M 88 0 0.10 0.08 ok 1 0.7 6.9 42.4 15 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0189, quality 0.669)	`market_value_eur=24000000` `quality_factor=0.6685` `boost_factor=1.0189` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0200)	`blend_weight=0.02` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=0.9793` `prior_reliability=0.08` `competition=Premier League`	wonTackle: 1.5728 -> 1.6882 totalCross: 3.2568 -> 3.3059 interceptionWon: 1.536 -> 1.5542 totalShots: 1.1885 -> 1.1808 onTargetScoringAttempt: 0.3507 -> 0.3437 wasFouled: 0.501 -> 0.5073 fouls: 1.3353 -> 1.3412
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Mateus Fernandes', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.384, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.9033, 'skill_edge': -0.0858, 'height_edge': 0.02, 'attacker_skill': 0.6914, 'defender_skill': 0.882, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Mateus Fernandes', 'market_value': 68000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.882}], 'xg_override': 0.0925}	aerialWon: 1.2429 -> 1.1 totalShots: 1.1814 -> 1.0672 shotOffTarget: 0.8377 -> 0.7567 onTargetScoringAttempt: 0.3437 -> 0.3105 bigChanceMissed: 0.2172 -> 0.1962 goals: 0.1177 -> 0.1063 expectedGoals: 0.1024 -> 0.0925
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.25 (squad xG sum 1.1947, xG scale 1.0480, xA scale 1.0543, team xA 0.81 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.252` `squad_xg_sum=1.1947` `squad_xa_sum=0.7719` `team_expected_assists=0.8138` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0966 -> 0.1018
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 78.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0924` `goal_rate=0.1063` `scorer_weight=0.1063`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0968` `assist_rate=0.0862` `provider_weight=0.0946`	-

Jaydee Canvot D 90 0 0.03 0.05 33% 0 0.2 6.7 45.9 32 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0106, quality 0.732)	`market_value_eur=32000000` `quality_factor=0.732` `boost_factor=1.0106` `reference_quality=0.505` `history_matches=27` `history_shrink=0.1562`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0135)	`blend_weight=0.0135` `history_matches=27` `season_overlap_discount=1.0` `per90_to_match_factor=0.9962` `prior_reliability=0.1333` `competition=Premier League`	wonTackle: 1.103 -> 1.115 fouls: 0.7699 -> 0.7595 interceptionWon: 1.1879 -> 1.1786
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'James Maddison', 'matchup_zone': 'LW vs CB'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.25 (squad xG sum 1.1947, xG scale 1.0480, xA scale 1.0543, team xA 0.81 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.252` `squad_xg_sum=1.1947` `squad_xa_sum=0.7719` `team_expected_assists=0.8138` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 12 by projected rating, fifa_overall 76.0)	`match_squad_rank=12` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0307` `goal_rate=0.0` `scorer_weight=0.025`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0223` `assist_rate=0.0` `provider_weight=0.0181`	-

Chris Richards D 90 0 0.06 0.07 ok 1 0.5 6.8 52.5 31 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0056, quality 0.677)	`market_value_eur=25000000` `quality_factor=0.6773` `boost_factor=1.0056` `reference_quality=0.505` `history_matches=41` `history_shrink=0.1087`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0094)	`blend_weight=0.0094` `history_matches=41` `season_overlap_discount=1.0` `per90_to_match_factor=0.9999` `prior_reliability=0.1333` `competition=Premier League`	interceptionWon: 1.1829 -> 1.1906 fouls: 0.6883 -> 0.6818
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Omar Marmoush', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.25 (squad xG sum 1.1947, xG scale 1.0480, xA scale 1.0543, team xA 0.81 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.252` `squad_xg_sum=1.1947` `squad_xa_sum=0.7719` `team_expected_assists=0.8138` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 10 by projected rating, fifa_overall 80.0)	`match_squad_rank=10` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0628` `goal_rate=0.0067` `scorer_weight=0.0519`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0197` `assist_rate=0.0` `provider_weight=0.0159`	-

Dean Henderson G 90 0 0.00 - n/a 0.0 6.9 53.8 28 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0068, quality 0.710)	`market_value_eur=29000000` `quality_factor=0.71` `boost_factor=1.0068` `reference_quality=0.505` `history_matches=40` `history_shrink=0.1111`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0096)	`blend_weight=0.0096` `history_matches=40` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Omar Marmoush, Mathys Tel', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.495}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.25 (squad xG sum 1.1947, xG scale 1.0480, xA scale 1.0543, team xA 0.81 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.252` `squad_xg_sum=1.1947` `squad_xa_sum=0.7719` `team_expected_assists=0.8138` `team_xa_to_xg_ratio=0.65`	shots_faced: 14.2 -> 14.5 expected_saves: 3.47 -> 3.55 high_pressure_saves: 1.2 -> 1.22
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 82.0)	`match_squad_rank=4` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=1.3073`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=0.7286`	-

Axel Disasi D 83 0 0.08 0.06 37% 1 0.4 6.8 50.2 24 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0052, quality 0.585)	`market_value_eur=15500000` `quality_factor=0.5846` `boost_factor=1.0052` `reference_quality=0.505` `history_matches=18` `history_shrink=0.2174`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0019)	`blend_weight=0.0019` `history_matches=18` `season_overlap_discount=1.0` `per90_to_match_factor=0.9207` `prior_reliability=0.0133` `competition=Premier League`	fouls: 0.8837 -> 0.8907
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Mathys Tel', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.25 (squad xG sum 1.1947, xG scale 1.0480, xA scale 1.0543, team xA 0.81 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.252` `squad_xg_sum=1.1947` `squad_xa_sum=0.7719` `team_expected_assists=0.8138` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 77.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0734` `goal_rate=0.0` `scorer_weight=0.061`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.009` `assist_rate=0.0` `provider_weight=0.0077`	-

Team totals 0 1.25 13 6.8

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Walter Benítez	G	4.4m	11.0%	Market-value substitute prior
Honest Ahanor	D	14.9m	36.1%	Market-value substitute prior
Chadi Riad	D	14.3m	23.2%	Market-value substitute prior
Quinten Timber	M	22.9m	54.0%	Market-value substitute prior
Óscar Mingueza	M	19.0m	70.7%	Market-value substitute prior
Ismaïla Sarr	F	26.7m	62.3%	Market-value substitute prior

Antagonist / tactical summary

Midfield is evenly contested — Sandro Tonali vs Adam Wharton will be the pivotal duel. Total antagonist xG — Home: 0.50, Away: 0.83. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Midfield — possession 50.8% / 49.2%, dominance ratio 0.53, key duel: Sandro Tonali vs Adam Wharton

James Maddison xG 0.06Omar Marmoush xG 0.18Mohammed Kudus xG 0.12Jean-Philippe Mateta xG 0.31Dwight McNeil xG 0.09Yéremy Pino xG 0.12

Summary

Antagonist matchup analysis for Tottenham vs Crystal Palace. Most likely scoreline: 1-0. Lambdas derived from 4 home and 7 away individual player duels. Head-to-head: Tottenham 6W-0D-4L vs Crystal Palace over 10 meetings.

Tactical parity between 4-2-3-1 and 3-4-2-1. Match outcome driven primarily by individual player quality. Midfield is evenly contested — Sandro Tonali vs Adam Wharton will be the pivotal duel. Total antagonist xG — Home: 0.50, Away: 0.83. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Aston Villa vs Fulham

Premier League · 20261031 · 4-2-3-1 vs 4-2-3-1

Predicted 1-0Winner: Aston VillaWinning H 1-0 (10.9%)Winning A 0-1 (6.2%)Confidence 0.75λ home/away 1.78 / 1.00Most-likely P 10.9%

Actual vs predicted: result not available yet (upcoming fixture).

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Aston Villa vs Fulham. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 4-2-3-1 vs 4-2-3-1 (duel lambdas: Home 1.702, Away 1.048). Final calibrated Poisson lambdas: Home 1.776, Away 1.005 (most likely score 1-0).

Home Starters 11/11 XGBoost (10 FBref) Away Starters 11/11 XGBoost (10 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 1.70 / 1.05 Final λ (H/A) 1.78 / 1.00

Modeling Pipeline Stages (5 layers)

1. XGBoost player regressions for volume & event rates with market-value quality scaling
2. Multi-source prior blending (FBref 2026-2027 tables + Sofascore season history)
3. Antagonist formation duel matrix resolving 1v1 and zonal matchups
4. Team-level prior calibration (FBref possession and attack/defense efficiencies)
5. Bivariate Poisson matrix with Dixon-Coles low-score correction

Standing & form context

Metric	Aston Villa	Fulham
League position	9.00	15.00
Points per game	1.37	1.15
Attack index	0.92	0.89
Defense (concede) index	0.96	1.03
Form multiplier	1.01	0.99
Standing λ multiplier	0.96	0.96

Lambda decomposition (how the goal rates were built)

Stage	Aston Villa λ	Fulham λ
Pre-competitive-floor	1.952	0.973
Pre-calibration (Micro-duels)	1.702	1.048
Final λ	1.776	1.005

competition calibration scale 0.951competition observed total goals 2.781quality gap applied Falsemismatch score pressure applied Falsecompetitive underdog floor applied Falsestage discount 1.0

Scoreline distribution & markets

Top 5 scorelines	P
1-1	11.1%
1-0	10.9%
2-1	9.8%
2-0	9.8%
0-0	6.3%

Home win	55.4%
Draw	23.4%
Away win	21.2%
Over 2.5	52.6%
BTTS yes	52.7%

Pregame fair line — favorite Aston Villa (55.4%), fair total 2.78, fair spread 0.77

Projected players vs season baseline

Proj = projected for this match · Season = this season's per-game average from match data · deviation badge compares projected xG to the player's season xG/game.

Aston Villa — projected line vs season baseline (4-2-3-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Zion Suzuki G 90 0 0.00 - n/a 0.0 7.1 36.1 14 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0275, quality 0.725)	`market_value_eur=31000000` `quality_factor=0.7249` `boost_factor=1.0275` `reference_quality=0.505` `history_matches=7` `history_shrink=0.4167`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0181)	`blend_weight=0.0181` `history_matches=7` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.0667` `competition=Premier League`	saves: 3.298 -> 3.2565 wasFouled: 0.5758 -> 0.5654
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Gonzalo García, Josh King', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.0478}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.78 (squad xG sum 1.3619, xG scale 1.3041, xA scale 1.7623, team xA 1.15 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.776` `squad_xg_sum=1.3619` `squad_xa_sum=0.6551` `team_expected_assists=1.1544` `team_xa_to_xg_ratio=0.65`	shots_faced: 10.0 -> 9.5 expected_saves: 2.45 -> 2.32 high_pressure_saves: 0.84 -> 0.8
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 76.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=1.8892`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=1.1787`	-

Matty Cash D 87 0 0.04 0.02 48% 1 0.6 6.7 53.8 30 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0040, quality 0.629)	`market_value_eur=19800000` `quality_factor=0.6291` `boost_factor=1.004` `reference_quality=0.505` `history_matches=42` `history_shrink=0.1064`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0088)	`blend_weight=0.0088` `history_matches=42` `season_overlap_discount=1.0` `per90_to_match_factor=0.9704` `prior_reliability=0.1267` `competition=Premier League`	totalCross: 2.2687 -> 2.2488 wasFouled: 1.0367 -> 1.0455 wonTackle: 1.5722 -> 1.5808
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Gonzalo García', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.78 (squad xG sum 1.3619, xG scale 1.3041, xA scale 1.7623, team xA 1.15 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.776` `squad_xg_sum=1.3619` `squad_xa_sum=0.6551` `team_expected_assists=1.1544` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0279 -> 0.0364 expected_assists: 0.0328 -> 0.0578
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 13 by projected rating, fifa_overall 81.0)	`match_squad_rank=13` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0357` `goal_rate=0.0045` `scorer_weight=0.0291`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0562` `assist_rate=0.0479` `provider_weight=0.0683`	-

Victor Lindelöf D 87 0 0.05 0.02 144% 0 0.1 6.8 35.7 32 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9977, quality 0.460)	`market_value_eur=4800000` `quality_factor=0.4597` `boost_factor=0.9977` `reference_quality=0.505` `history_matches=25` `history_shrink=0.1667`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0144)	`blend_weight=0.0144` `history_matches=25` `season_overlap_discount=1.0` `per90_to_match_factor=0.9712` `prior_reliability=0.1333` `competition=Premier League`	interceptionWon: 0.7725 -> 0.7613 wonTackle: 0.5702 -> 0.562 totalShots: 0.3805 -> 0.375
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Gonzalo García', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.78 (squad xG sum 1.3619, xG scale 1.3041, xA scale 1.7623, team xA 1.15 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.776` `squad_xg_sum=1.3619` `squad_xa_sum=0.6551` `team_expected_assists=1.1544` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0419 -> 0.0546 expected_assists: 0.0142 -> 0.025
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 77.0)	`match_squad_rank=8` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0531` `goal_rate=0.0` `scorer_weight=0.0437`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0247` `assist_rate=0.0083` `provider_weight=0.0227`	-

Pau Torres D 80 0 0.08 0.04 110% 0 0.2 6.7 31.3 30 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0066, quality 0.641)	`market_value_eur=21000000` `quality_factor=0.6408` `boost_factor=1.0066` `reference_quality=0.505` `history_matches=26` `history_shrink=0.1613`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0126)	`blend_weight=0.0126` `history_matches=26` `season_overlap_discount=1.0` `per90_to_match_factor=0.8906` `prior_reliability=0.12` `competition=Premier League`	wonTackle: 0.5809 -> 0.5736 fouls: 0.4777 -> 0.4841
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Gonzalo García', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.78 (squad xG sum 1.3619, xG scale 1.3041, xA scale 1.7623, team xA 1.15 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.776` `squad_xg_sum=1.3619` `squad_xa_sum=0.6551` `team_expected_assists=1.1544` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0597 -> 0.0779 expected_assists: 0.0097 -> 0.0171
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 10 by projected rating, fifa_overall 81.0)	`match_squad_rank=10` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0749` `goal_rate=0.0` `scorer_weight=0.0623`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.017` `assist_rate=0.0` `provider_weight=0.0143`	-

Ian Maatsen D 84 0 0.07 0.04 100% 1 0.6 6.7 47.9 30 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0088, quality 0.739)	`market_value_eur=33000000` `quality_factor=0.7389` `boost_factor=1.0088` `reference_quality=0.505` `history_matches=35` `history_shrink=0.125`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0103)	`blend_weight=0.0103` `history_matches=35` `season_overlap_discount=1.0` `per90_to_match_factor=0.9382` `prior_reliability=0.1267` `competition=Premier League`	fouls: 1.153 -> 1.1665 totalCross: 1.7395 -> 1.7267 wonTackle: 1.4938 -> 1.4835 interceptionWon: 1.1595 -> 1.1679 totalShots: 0.7954 -> 0.8025 wasFouled: 0.4998 -> 0.4947
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Alex Iwobi', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.78 (squad xG sum 1.3619, xG scale 1.3041, xA scale 1.7623, team xA 1.15 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.776` `squad_xg_sum=1.3619` `squad_xa_sum=0.6551` `team_expected_assists=1.1544` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0567 -> 0.0739 expected_assists: 0.1063 -> 0.1873
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 17 by projected rating, fifa_overall 79.0)	`match_squad_rank=17` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0712` `goal_rate=0.0` `scorer_weight=0.0591`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1708` `assist_rate=0.0421` `provider_weight=0.1498`	-

João Gomes M 82 0 0.05 0.02 104% 1 0.4 6.8 51.7 29 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0084, quality 0.758)	`market_value_eur=36000000` `quality_factor=0.7581` `boost_factor=1.0084` `reference_quality=0.505` `history_matches=40` `history_shrink=0.1111`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0019)	`blend_weight=0.0019` `history_matches=40` `season_overlap_discount=1.0` `per90_to_match_factor=0.9076` `prior_reliability=0.0267` `competition=Premier League`	wonTackle: 1.7998 -> 1.8051
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Shea Charles', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.408, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v1', 'xg_multiplier': 0.9821, 'skill_edge': 0.0399, 'height_edge': -0.052, 'attacker_skill': 0.7926, 'defender_skill': 0.704, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Shea Charles', 'market_value': 24000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.704}], 'xg_override': 0.0353}	aerialWon: 0.5536 -> 0.9 totalShots: 0.5126 -> 0.5035 shotOffTarget: 0.3593 -> 0.3529
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.78 (squad xG sum 1.3619, xG scale 1.3041, xA scale 1.7623, team xA 1.15 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.776` `squad_xg_sum=1.3619` `squad_xa_sum=0.6551` `team_expected_assists=1.1544` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0353 -> 0.046 expected_assists: 0.0533 -> 0.0939
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 78.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.045` `goal_rate=0.0419` `scorer_weight=0.0496`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0896` `assist_rate=0.074` `provider_weight=0.1127`	-

Alysson F 66 0 0.21 0.01 1363% 1 0.3 6.7 33.2 22 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0087, quality 0.592)	`market_value_eur=16200000` `quality_factor=0.5921` `boost_factor=1.0087` `reference_quality=0.505` `history_matches=10` `history_shrink=0.3333`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0043)	`blend_weight=0.0043` `history_matches=10` `season_overlap_discount=1.0` `per90_to_match_factor=0.7364` `prior_reliability=0.02` `competition=Premier League`	wonTackle: 0.6633 -> 0.6817 totalShots: 1.3766 -> 1.3706 fouls: 1.1896 -> 1.1844
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Timothy Castagne, Calvin Bassey', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 0.9566, 'skill_edge': -0.0255, 'height_edge': -0.004, 'attacker_skill': 0.5921, 'defender_skill': 0.6488, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Timothy Castagne', 'market_value': 7600000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.524}, {'player': 'Calvin Bassey', 'market_value': 29000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7736}], 'xg_override': 0.1606}	totalShots: 1.3747 -> 1.315 shotOffTarget: 0.9344 -> 0.8938 onTargetScoringAttempt: 0.4403 -> 0.4212 expectedGoals: 0.1679 -> 0.1606 goals: 0.1184 -> 0.1133
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.78 (squad xG sum 1.3619, xG scale 1.3041, xA scale 1.7623, team xA 1.15 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.776` `squad_xg_sum=1.3619` `squad_xa_sum=0.6551` `team_expected_assists=1.1544` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1606 -> 0.2094 expected_assists: 0.0437 -> 0.077
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 16 by projected rating, fifa_overall 70.0)	`match_squad_rank=16` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1889` `goal_rate=0.1133` `scorer_weight=0.1675`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0741` `assist_rate=0.0` `provider_weight=0.0616`	-

Johan Manzambi M 78 0 0.35 0.14 155% 3 1.4 6.9 33.0 11 XGBoost M_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0551, quality 0.872)	`market_value_eur=71000000` `quality_factor=0.8721` `boost_factor=1.0551` `reference_quality=0.505` `history_matches=5` `history_shrink=0.5`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'David Affengruber', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle', 'duel_win_rate': 0.552, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.1064, 'skill_edge': 0.0844, 'height_edge': -0.012, 'attacker_skill': 0.8721, 'defender_skill': 0.6846, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'David Affengruber', 'market_value': 18100000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.6846}], 'xg_override': 0.2669}	aerialWon: 0.0 -> 1.1 totalShots: 2.487 -> 2.7517 shotOffTarget: 1.6594 -> 1.836 onTargetScoringAttempt: 0.8276 -> 0.9157 expectedGoalsOnTarget: 0.3855 -> 0.4265 expectedGoals: 0.2412 -> 0.2669 goals: 0.2362 -> 0.2613
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.78 (squad xG sum 1.3619, xG scale 1.3041, xA scale 1.7623, team xA 1.15 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.776` `squad_xg_sum=1.3619` `squad_xa_sum=0.6551` `team_expected_assists=1.1544` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.2669 -> 0.3481 expected_assists: 0.1312 -> 0.2314
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 80.0)	`match_squad_rank=6` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.294` `goal_rate=0.2613` `scorer_weight=0.3419`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=1` `assist_probability=0.2066` `assist_rate=0.1369` `provider_weight=0.2387`	-

John McGinn F 81 0 0.18 0.09 91% 1 0.6 6.8 47.9 36 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0017, quality 0.552)	`market_value_eur=12600000` `quality_factor=0.5525` `boost_factor=1.0017` `reference_quality=0.505` `history_matches=38` `history_shrink=0.1163`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0096)	`blend_weight=0.0096` `history_matches=38` `season_overlap_discount=1.0` `per90_to_match_factor=0.8983` `prior_reliability=0.1267` `competition=Premier League`	fouls: 1.024 -> 1.0323
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Timothy Castagne, Calvin Bassey', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 0.8951, 'skill_edge': -0.0433, 'height_edge': -0.028, 'attacker_skill': 0.5525, 'defender_skill': 0.6488, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Timothy Castagne', 'market_value': 7600000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.524}, {'player': 'Calvin Bassey', 'market_value': 29000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7736}], 'xg_override': 0.1355}	totalShots: 1.2212 -> 1.0931 shotOffTarget: 0.7302 -> 0.6536 onTargetScoringAttempt: 0.4843 -> 0.4335 bigChanceMissed: 0.2572 -> 0.2302 expectedGoalsOnTarget: 0.2099 -> 0.1879 goals: 0.1972 -> 0.1765 expectedGoals: 0.1514 -> 0.1355
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.78 (squad xG sum 1.3619, xG scale 1.3041, xA scale 1.7623, team xA 1.15 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.776` `squad_xg_sum=1.3619` `squad_xa_sum=0.6551` `team_expected_assists=1.1544` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1355 -> 0.1767 expected_assists: 0.0949 -> 0.1672
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 83.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.162` `goal_rate=0.1765` `scorer_weight=0.2158`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.154` `assist_rate=0.4311` `provider_weight=0.209`	-

Nicolas Jackson F 81 1 0.48 0.14 244% 3 1.5 7.2 40.1 13 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0369, quality 0.800)	`market_value_eur=44000000` `quality_factor=0.8002` `boost_factor=1.0369` `reference_quality=0.505` `history_matches=7` `history_shrink=0.4167`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0144)	`blend_weight=0.0144` `history_matches=7` `season_overlap_discount=1.0` `per90_to_match_factor=0.895` `prior_reliability=0.0533` `competition=Premier League`	totalShots: 2.7932 -> 2.7529 wasFouled: 1.0943 -> 1.0785 onTargetScoringAttempt: 1.0519 -> 1.0367 totalCross: 0.3285 -> 0.3399 totalOffside: 0.6117 -> 0.6029 goals: 0.556 -> 0.548 fouls: 0.489 -> 0.4819 interceptionWon: 0.4235 -> 0.4174
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Timothy Castagne, Calvin Bassey', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.272, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.106, 'skill_edge': 0.0681, 'height_edge': 0.004, 'attacker_skill': 0.8002, 'defender_skill': 0.6488, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Timothy Castagne', 'market_value': 7600000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.524}, {'player': 'Calvin Bassey', 'market_value': 29000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7736}], 'xg_override': 0.3708}	totalShots: 2.778 -> 3.0725 shotOffTarget: 1.7413 -> 1.9259 onTargetScoringAttempt: 1.0367 -> 1.1466 goals: 0.548 -> 0.6061 expectedGoals: 0.3353 -> 0.3708 bigChanceMissed: 0.3351 -> 0.3706 expectedGoalsOnTarget: 0.3083 -> 0.341
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	goals: 1.0 -> 0
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.78 (squad xG sum 1.3619, xG scale 1.3041, xA scale 1.7623, team xA 1.15 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.776` `squad_xg_sum=1.3619` `squad_xa_sum=0.6551` `team_expected_assists=1.1544` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.3708 -> 0.4836 expected_assists: 0.0789 -> 0.139
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 79.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.3834` `goal_rate=0.6061` `scorer_weight=0.6045`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1298` `assist_rate=0.0` `provider_weight=0.1112`	-

Boubacar Kamara M 88 0 0.06 0.01 307% 1 0.2 6.9 44.0 16 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0153, quality 0.821)	`market_value_eur=49000000` `quality_factor=0.8206` `boost_factor=1.0153` `reference_quality=0.505` `history_matches=26` `history_shrink=0.1613`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0140)	`blend_weight=0.014` `history_matches=26` `season_overlap_discount=1.0` `per90_to_match_factor=0.9773` `prior_reliability=0.1333` `competition=Premier League`	wasFouled: 2.331 -> 2.3053 wonTackle: 1.7348 -> 1.7242 totalShots: 0.5827 -> 0.5746 interceptionWon: 1.0269 -> 1.0194 totalCross: 0.1166 -> 0.1218
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Shea Charles', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.464, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v1', 'xg_multiplier': 1.0651, 'skill_edge': 0.0643, 'height_edge': -0.02, 'attacker_skill': 0.847, 'defender_skill': 0.704, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Shea Charles', 'market_value': 24000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.704}], 'xg_override': 0.0451}	aerialWon: 1.4338 -> 1.1 totalShots: 0.5805 -> 0.6183 shotOffTarget: 0.4286 -> 0.4565 onTargetScoringAttempt: 0.1519 -> 0.1618
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.78 (squad xG sum 1.3619, xG scale 1.3041, xA scale 1.7623, team xA 1.15 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.776` `squad_xg_sum=1.3619` `squad_xa_sum=0.6551` `team_expected_assists=1.1544` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0451 -> 0.0588 expected_assists: 0.0332 -> 0.0585
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 84.0)	`match_squad_rank=5` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0571` `goal_rate=0.0493` `scorer_weight=0.0614`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0568` `assist_rate=0.0614` `provider_weight=0.0731`	-

Team totals 1 1.78 14 6.8

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Marco Bizot	G	4.4m	13.1%	Market-value substitute prior
Matteo Ruggeri	D	11.5m	62.3%	Market-value substitute prior
Taylor Harwood-Bellis	D	14.6m	34.5%	Market-value substitute prior
Amadou Onana	M	23.4m	56.1%	Market-value substitute prior
Douglas Luiz	M	22.7m	53.2%	Market-value substitute prior
Donyell Malen	F	26.8m	62.7%	Market-value substitute prior

Fulham — projected line vs season baseline (4-2-3-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Josh King M 78 0 0.14 0.10 38% 2 1.0 7.0 48.8 31 XGBoost M_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0094, quality 0.786)	`market_value_eur=41000000` `quality_factor=0.7859` `boost_factor=1.0094` `reference_quality=0.505` `history_matches=40` `history_shrink=0.1111`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'João Gomes, Boubacar Kamara', 'matchup_zone': 'Attacking Mid vs Double Pivot — Tight Space Battle', 'duel_win_rate': 0.197, 'pressing_faced': 'HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.9375, 'skill_edge': -0.0105, 'height_edge': -0.032, 'attacker_skill': 0.7964, 'defender_skill': 0.8198, 'attacker_role': 'CAM', 'attacker_mv_price_index': 0.95, 'defender_adjusted_parts': [{'player': 'João Gomes', 'market_value': 36000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.7926}, {'player': 'Boubacar Kamara', 'market_value': 49000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.847}], 'xg_override': 0.1769}	aerialWon: 0.1121 -> 0.8 totalShots: 2.251 -> 2.1103 shotOffTarget: 1.4382 -> 1.3483 onTargetScoringAttempt: 0.8128 -> 0.762 bigChanceMissed: 0.3249 -> 0.3046 expectedGoalsOnTarget: 0.2136 -> 0.2003 goals: 0.2102 -> 0.1971 expectedGoals: 0.1887 -> 0.1769
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.00 (squad xG sum 1.2631, xG scale 0.7957, xA scale 0.7957, team xA 0.76 = 0.76 x lambda)	`scoreline_reset=true` `team_lambda=1.005` `squad_xg_sum=1.2631` `squad_xa_sum=0.9586` `team_expected_assists=0.7627` `team_xa_to_xg_ratio=0.7589`	expected_goals: 0.1769 -> 0.1408 expected_assists: 0.1378 -> 0.1096
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 78.0)	`match_squad_rank=8` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.1313` `goal_rate=0.1971` `scorer_weight=0.1533`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.1038` `assist_rate=0.1361` `provider_weight=0.1086`	-

Alex Iwobi M 82 0 0.11 0.06 75% 1 1.1 7.3 50.9 37 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0048, quality 0.641)	`market_value_eur=21000000` `quality_factor=0.6408` `boost_factor=1.0048` `reference_quality=0.505` `history_matches=37` `history_shrink=0.119`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0103)	`blend_weight=0.0103` `history_matches=37` `season_overlap_discount=1.0` `per90_to_match_factor=0.9097` `prior_reliability=0.1333` `competition=Premier League`	totalCross: 4.1844 -> 4.1647 totalShots: 1.6273 -> 1.6152 interceptionWon: 0.3004 -> 0.3114
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Ian Maatsen', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle', 'duel_win_rate': 0.422, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.9151, 'skill_edge': -0.0777, 'height_edge': 0.02, 'attacker_skill': 0.6408, 'defender_skill': 0.8135, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Ian Maatsen', 'market_value': 33000000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.8135}], 'xg_override': 0.134}	aerialWon: 0.0708 -> 1.1 totalShots: 1.6223 -> 1.4846 shotOffTarget: 1.1539 -> 1.0559 onTargetScoringAttempt: 0.4684 -> 0.4286 bigChanceMissed: 0.2309 -> 0.2113 goals: 0.1653 -> 0.1513 expectedGoals: 0.1464 -> 0.134 expectedGoalsOnTarget: 0.0782 -> 0.0716
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.00 (squad xG sum 1.2631, xG scale 0.7957, xA scale 0.7957, team xA 0.76 = 0.76 x lambda)	`scoreline_reset=true` `team_lambda=1.005` `squad_xg_sum=1.2631` `squad_xa_sum=0.9586` `team_expected_assists=0.7627` `team_xa_to_xg_ratio=0.7589`	expected_goals: 0.134 -> 0.1066 expected_assists: 0.2124 -> 0.1692
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 80.0)	`match_squad_rank=1` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.1011` `goal_rate=0.1513` `scorer_weight=0.1166`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.1557` `assist_rate=0.1217` `provider_weight=0.1354`	-

Gonzalo García F 81 0 0.25 0.23 ok 2 1.1 6.7 48.8 34 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0088, quality 0.758)	`market_value_eur=36000000` `quality_factor=0.7581` `boost_factor=1.0088` `reference_quality=0.505` `history_matches=38` `history_shrink=0.1163`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0101)	`blend_weight=0.0101` `history_matches=38` `season_overlap_discount=1.0` `per90_to_match_factor=0.9001` `prior_reliability=0.1333` `competition=Premier League`	fouls: 0.8512 -> 0.8608 wasFouled: 0.3114 -> 0.3173 totalShots: 2.1583 -> 2.1638
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Matty Cash, Pau Torres', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.202, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 1.0031, 'skill_edge': 0.0281, 'height_edge': -0.026, 'attacker_skill': 0.7581, 'defender_skill': 0.6957, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Matty Cash', 'market_value': 19800000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.6892}, {'player': 'Pau Torres', 'market_value': 21000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7022}], 'xg_override': 0.3197}	totalShots: 2.1638 -> 2.1705
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.00 (squad xG sum 1.2631, xG scale 0.7957, xA scale 0.7957, team xA 0.76 = 0.76 x lambda)	`scoreline_reset=true` `team_lambda=1.005` `squad_xg_sum=1.2631` `squad_xa_sum=0.9586` `team_expected_assists=0.7627` `team_xa_to_xg_ratio=0.7589`	expected_goals: 0.3197 -> 0.2543 expected_assists: 0.0644 -> 0.0512
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 13 by projected rating, fifa_overall 78.0)	`match_squad_rank=13` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.2245` `goal_rate=0.2526` `scorer_weight=0.2081`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0499` `assist_rate=0.0` `provider_weight=0.041`	-

Oscar Bobb F 74 0 0.12 0.08 52% 2 0.6 7.0 49.4 39 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0079, quality 0.694)	`market_value_eur=27000000` `quality_factor=0.6941` `boost_factor=1.0079` `reference_quality=0.505` `history_matches=31` `history_shrink=0.1389`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0090)	`blend_weight=0.009` `history_matches=31` `season_overlap_discount=1.0` `per90_to_match_factor=0.819` `prior_reliability=0.1` `competition=Premier League`	totalShots: 1.6665 -> 1.6515 fouls: 0.1828 -> 0.191 totalCross: 2.7634 -> 2.7582 interceptionWon: 0.5357 -> 0.5407
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=true`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Matty Cash, Pau Torres', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 0.9137, 'skill_edge': -0.0007, 'height_edge': -0.058, 'attacker_skill': 0.6941, 'defender_skill': 0.6957, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Matty Cash', 'market_value': 19800000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.6892}, {'player': 'Pau Torres', 'market_value': 21000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7022}], 'xg_override': 0.1479}	totalShots: 1.6615 -> 1.5181 shotOffTarget: 1.1284 -> 1.031 onTargetScoringAttempt: 0.5331 -> 0.4871 goals: 0.1824 -> 0.1667 expectedGoalsOnTarget: 0.1679 -> 0.1534 expectedGoals: 0.1619 -> 0.1479 bigChanceMissed: 0.1318 -> 0.1204
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.00 (squad xG sum 1.2631, xG scale 0.7957, xA scale 0.7957, team xA 0.76 = 0.76 x lambda)	`scoreline_reset=true` `team_lambda=1.005` `squad_xg_sum=1.2631` `squad_xa_sum=0.9586` `team_expected_assists=0.7627` `team_xa_to_xg_ratio=0.7589`	expected_goals: 0.1479 -> 0.1177 expected_assists: 0.1192 -> 0.0948
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 76.0)	`match_squad_rank=5` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.111` `goal_rate=0.1667` `scorer_weight=0.1289`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0904` `assist_rate=0.0084` `provider_weight=0.0758`	-

Shea Charles M 86 0 0.04 0.01 205% 1 0.4 7.1 40.6 14 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0189, quality 0.669)	`market_value_eur=24000000` `quality_factor=0.6685` `boost_factor=1.0189` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0083)	`blend_weight=0.0083` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=0.9516` `prior_reliability=0.0333` `competition=Premier League`	totalShots: 0.7461 -> 0.7702 totalCross: 0.1507 -> 0.1653 wasFouled: 1.329 -> 1.3179 wonTackle: 1.1554 -> 1.1616
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'João Gomes', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.432, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 1.0179, 'skill_edge': -0.0399, 'height_edge': 0.052, 'attacker_skill': 0.704, 'defender_skill': 0.7926, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'João Gomes', 'market_value': 36000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.7926}], 'xg_override': 0.0488}	aerialWon: 1.364 -> 1.3 totalShots: 0.7702 -> 0.784 shotOffTarget: 0.5327 -> 0.5422
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.00 (squad xG sum 1.2631, xG scale 0.7957, xA scale 0.7957, team xA 0.76 = 0.76 x lambda)	`scoreline_reset=true` `team_lambda=1.005` `squad_xg_sum=1.2631` `squad_xa_sum=0.9586` `team_expected_assists=0.7627` `team_xa_to_xg_ratio=0.7589`	expected_goals: 0.0488 -> 0.0388 expected_assists: 0.0816 -> 0.0649
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 73.0)	`match_squad_rank=4` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0381` `goal_rate=0.0459` `scorer_weight=0.0377`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0628` `assist_rate=0.0517` `provider_weight=0.0519`	-

Sander Berge M 84 0 0.05 0.04 35% 1 0.5 7.0 52.5 33 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0057, quality 0.686)	`market_value_eur=26000000` `quality_factor=0.6858` `boost_factor=1.0057` `reference_quality=0.505` `history_matches=43` `history_shrink=0.1042`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0077)	`blend_weight=0.0077` `history_matches=43` `season_overlap_discount=1.0` `per90_to_match_factor=0.9333` `prior_reliability=0.1133` `competition=Premier League`	wonTackle: 0.8443 -> 0.8378
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'João Gomes', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.448, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 1.0414, 'skill_edge': -0.0318, 'height_edge': 0.06, 'attacker_skill': 0.7219, 'defender_skill': 0.7926, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'João Gomes', 'market_value': 36000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.7926}], 'xg_override': 0.0614}	aerialWon: 0.9696 -> 1.4 totalShots: 0.7554 -> 0.7867 shotOffTarget: 0.5262 -> 0.548 onTargetScoringAttempt: 0.2291 -> 0.2386
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.00 (squad xG sum 1.2631, xG scale 0.7957, xA scale 0.7957, team xA 0.76 = 0.76 x lambda)	`scoreline_reset=true` `team_lambda=1.005` `squad_xg_sum=1.2631` `squad_xa_sum=0.9586` `team_expected_assists=0.7627` `team_xa_to_xg_ratio=0.7589`	expected_goals: 0.0614 -> 0.0489 expected_assists: 0.06 -> 0.0477
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 79.0)	`match_squad_rank=6` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0477` `goal_rate=0.0681` `scorer_weight=0.0518`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0466` `assist_rate=0.0663` `provider_weight=0.0504`	-

Antonee Robinson D 87 0 0.04 0.03 ok 1 0.2 6.8 48.3 32 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0072, quality 0.669)	`market_value_eur=24000000` `quality_factor=0.6685` `boost_factor=1.0072` `reference_quality=0.505` `history_matches=29` `history_shrink=0.1471`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0121)	`blend_weight=0.0121` `history_matches=29` `season_overlap_discount=1.0` `per90_to_match_factor=0.9666` `prior_reliability=0.1267` `competition=Premier League`	fouls: 1.4783 -> 1.4604 wonTackle: 1.7037 -> 1.7139 interceptionWon: 1.1728 -> 1.1648
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'John McGinn', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.00 (squad xG sum 1.2631, xG scale 0.7957, xA scale 0.7957, team xA 0.76 = 0.76 x lambda)	`scoreline_reset=true` `team_lambda=1.005` `squad_xg_sum=1.2631` `squad_xa_sum=0.9586` `team_expected_assists=0.7627` `team_xa_to_xg_ratio=0.7589`	expected_goals: 0.0556 -> 0.0442 expected_assists: 0.0536 -> 0.0426
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 12 by projected rating, fifa_overall 80.0)	`match_squad_rank=12` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0432` `goal_rate=0.0415` `scorer_weight=0.0383`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0417` `assist_rate=0.0` `provider_weight=0.0341`	-

Bernd Leno G 90 0 0.00 - n/a 0.0 7.2 56.8 34 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9989, quality 0.468)	`market_value_eur=5500000` `quality_factor=0.4683` `boost_factor=0.9989` `reference_quality=0.505` `history_matches=46` `history_shrink=0.098`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0085)	`blend_weight=0.0085` `history_matches=46` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Nicolas Jackson, Johan Manzambi', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.702}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.00 (squad xG sum 1.2631, xG scale 0.7957, xA scale 0.7957, team xA 0.76 = 0.76 x lambda)	`scoreline_reset=true` `team_lambda=1.005` `squad_xg_sum=1.2631` `squad_xa_sum=0.9586` `team_expected_assists=0.7627` `team_xa_to_xg_ratio=0.7589`	shots_faced: 16.2 -> 16.9 expected_saves: 3.97 -> 4.14 high_pressure_saves: 1.36 -> 1.42
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 79.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=0.9696`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=0.655`	-

Timothy Castagne D 87 0 0.06 0.05 ok 1 0.4 6.8 48.2 34 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9996, quality 0.494)	`market_value_eur=7600000` `quality_factor=0.4939` `boost_factor=0.9996` `reference_quality=0.505` `history_matches=37` `history_shrink=0.119`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0088)	`blend_weight=0.0088` `history_matches=37` `season_overlap_discount=1.0` `per90_to_match_factor=0.9623` `prior_reliability=0.1133` `competition=Premier League`	fouls: 1.1789 -> 1.1934 wasFouled: 0.6533 -> 0.6674 totalCross: 2.5089 -> 2.5018
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Nicolas Jackson', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.00 (squad xG sum 1.2631, xG scale 0.7957, xA scale 0.7957, team xA 0.76 = 0.76 x lambda)	`scoreline_reset=true` `team_lambda=1.005` `squad_xg_sum=1.2631` `squad_xa_sum=0.9586` `team_expected_assists=0.7627` `team_xa_to_xg_ratio=0.7589`	expected_goals: 0.0723 -> 0.0575 expected_assists: 0.0764 -> 0.0608
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 11 by projected rating, fifa_overall 75.0)	`match_squad_rank=11` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0559` `goal_rate=0.0` `scorer_weight=0.046`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.059` `assist_rate=0.0618` `provider_weight=0.0538`	-

David Affengruber D 89 0 0.03 0.03 ok 1 0.5 7.0 54.8 35 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0033, quality 0.612)	`market_value_eur=18100000` `quality_factor=0.612` `boost_factor=1.0033` `reference_quality=0.505` `history_matches=44` `history_shrink=0.102`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0133)	`blend_weight=0.0133` `history_matches=44` `season_overlap_discount=1.0` `per90_to_match_factor=0.9898` `prior_reliability=0.2` `competition=La Liga`	totalShots: 0.521 -> 0.536 fouls: 1.127 -> 1.1164 wasFouled: 0.5804 -> 0.5902 interceptionWon: 1.1216 -> 1.1286
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Johan Manzambi', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.00 (squad xG sum 1.2631, xG scale 0.7957, xA scale 0.7957, team xA 0.76 = 0.76 x lambda)	`scoreline_reset=true` `team_lambda=1.005` `squad_xg_sum=1.2631` `squad_xa_sum=0.9586` `team_expected_assists=0.7627` `team_xa_to_xg_ratio=0.7589`	expected_goals: 0.0357 -> 0.0284 expected_assists: 0.031 -> 0.0247
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 76.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.028` `goal_rate=0.0166` `scorer_weight=0.0227`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0244` `assist_rate=0.0153` `provider_weight=0.0199`	-

Calvin Bassey D 90 0 0.04 0.04 ok 1 0.4 7.1 51.4 30 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0072, quality 0.710)	`market_value_eur=29000000` `quality_factor=0.71` `boost_factor=1.0072` `reference_quality=0.505` `history_matches=38` `history_shrink=0.1163`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0101)	`blend_weight=0.0101` `history_matches=38` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	fouls: 1.3228 -> 1.3145
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Nicolas Jackson', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.00 (squad xG sum 1.2631, xG scale 0.7957, xA scale 0.7957, team xA 0.76 = 0.76 x lambda)	`scoreline_reset=true` `team_lambda=1.005` `squad_xg_sum=1.2631` `squad_xa_sum=0.9586` `team_expected_assists=0.7627` `team_xa_to_xg_ratio=0.7589`	expected_goals: 0.0466 -> 0.0371
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 78.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0364` `goal_rate=0.0055` `scorer_weight=0.0297`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0127` `assist_rate=0.0` `provider_weight=0.0102`	-

Team totals 0 1.00 17 7.0

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Benjamin Lecomte	G	4.4m	10.9%	Market-value substitute prior
Ryan Sessegnon	D	18.0m	61.8%	Market-value substitute prior
Joachim Andersen	D	14.4m	23.3%	Market-value substitute prior
Hugo Larsson	M	23.1m	54.9%	Market-value substitute prior
Emile Smith Rowe	M	18.4m	70.6%	Market-value substitute prior
Kevin	F	19.0m	73.9%	Market-value substitute prior

Antagonist / tactical summary

Midfield is evenly contested — Johan Manzambi vs Josh King will be the pivotal duel. Total antagonist xG — Home: 1.01, Away: 0.89. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Midfield — possession 51.9% / 48.1%, dominance ratio 0.53, key duel: Johan Manzambi vs Josh King

Johan Manzambi xG 0.27Nicolas Jackson xG 0.37Alysson xG 0.16Alex Iwobi xG 0.13Josh King xG 0.18Gonzalo García xG 0.32

Summary

Antagonist matchup analysis for Aston Villa vs Fulham. Most likely scoreline: 1-0. Lambdas derived from 6 home and 6 away individual player duels. Head-to-head: Aston Villa 8W-0D-2L vs Fulham over 10 meetings.

Tactical parity between 4-2-3-1 and 4-2-3-1. Match outcome driven primarily by individual player quality. Midfield is evenly contested — Johan Manzambi vs Josh King will be the pivotal duel. Total antagonist xG — Home: 1.01, Away: 0.89. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Hull vs Ipswich

Premier League · 20261031 · 5-4-1 vs 4-2-3-1

Predicted 1-1Winner: DrawWinning H 1-0 (8.6%)Winning A 0-1 (8.4%)Confidence 0.7λ home/away 1.41 / 1.37Most-likely P 12.1%

Actual vs predicted: result not available yet (upcoming fixture).

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Hull vs Ipswich. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 5-4-1 vs 4-2-3-1 (duel lambdas: Home 1.398, Away 1.442). Final calibrated Poisson lambdas: Home 1.409, Away 1.372 (most likely score 1-1).

Home Starters 11/11 XGBoost (9 FBref) Away Starters 11/11 XGBoost (8 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 1.40 / 1.44 Final λ (H/A) 1.41 / 1.37

FBref Team Priors: Hull City (Gls/90: 1.5, GA/90: 1.5 → E[λ]: 2.069) vs Ipswich Town (Gls/90: 2.0, GA/90: 2.0 → E[λ]: 2.069) | Blend Weight: 0.192

Modeling Pipeline Stages (5 layers)

1. XGBoost player regressions for volume & event rates with market-value quality scaling
2. Multi-source prior blending (FBref 2026-2027 tables + Sofascore season history)
3. Antagonist formation duel matrix resolving 1v1 and zonal matchups
4. Team-level prior calibration (FBref possession and attack/defense efficiencies)
5. Bivariate Poisson matrix with Dixon-Coles low-score correction

Standing & form context

Metric	Hull	Ipswich
League position	10.00	12.00
Points per game	1.30	1.30
Attack index	0.92	1.03
Defense (concede) index	1.18	0.89
Form multiplier	1.00	1.00
Standing λ multiplier	1.03	0.97

Lambda decomposition (how the goal rates were built)

Stage	Hull λ	Ipswich λ
FBref Squad Prior E[λ] (weight 0.192)	2.069	2.069
Pre-competitive-floor	1.398	1.442
Pre-calibration (Micro-duels)	1.398	1.442
Final λ	1.409	1.372

competition calibration scale 0.979competition observed total goals 2.781quality gap applied Truemismatch score pressure applied Falsecompetitive underdog floor applied Falsestage discount 1.0

Scoreline distribution & markets

Top 5 scorelines	P
1-1	12.1%
1-0	8.6%
2-1	8.4%
0-1	8.4%
1-2	8.2%

Home win	38.0%
Draw	25.6%
Away win	36.4%
Over 2.5	52.6%
BTTS yes	56.5%

Pregame fair line — favorite Hull (38.0%), fair total 2.78, fair spread 0.04

Projected players vs season baseline

Proj = projected for this match · Season = this season's per-game average from match data · deviation badge compares projected xG to the player's season xG/game.

Hull — projected line vs season baseline (5-4-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Lewie Coyle D 79 0 0.04 0.01 183% 0 0.1 6.6 37.1 14 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9922, quality 0.437)	`market_value_eur=2700000` `quality_factor=0.4374` `boost_factor=0.9922` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0217)	`blend_weight=0.0217` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=0.8826` `prior_reliability=0.0867` `competition=Premier League`	wasFouled: 0.4333 -> 0.4828 totalCross: 2.7116 -> 2.6823 fouls: 0.2905 -> 0.3136 totalShots: 0.3598 -> 0.352 wonTackle: 0.9627 -> 0.9565
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': "Julio Enciso, Dara O'Shea", 'matchup_zone': 'RWB vs LW + LB'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.41 (squad xG sum 0.8743, xG scale 1.6116, xA scale 2.0289, team xA 0.92 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.409` `squad_xg_sum=0.8743` `squad_xa_sum=0.4514` `team_expected_assists=0.9159` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0236 -> 0.038 expected_assists: 0.0107 -> 0.0217
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 16 by projected rating, fifa_overall 70.0)	`match_squad_rank=16` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0373` `goal_rate=0.0` `scorer_weight=0.0304`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0215` `assist_rate=0.0` `provider_weight=0.0179`	-

Semi Ajayi D 89 0 0.08 0.05 67% 0 0.2 6.7 38.8 14 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9922, quality 0.437)	`market_value_eur=930000` `quality_factor=0.4374` `boost_factor=0.9922` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0283)	`blend_weight=0.0283` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=0.9848` `prior_reliability=0.1133` `competition=Premier League`	wonTackle: 1.3912 -> 1.3682 goals: 0.0656 -> 0.0802 onTargetScoringAttempt: 0.0656 -> 0.0802 wasFouled: 0.0772 -> 0.0914 fouls: 0.2954 -> 0.287 totalCross: 0.842 -> 0.8346 totalShots: 0.3495 -> 0.3561
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Julio Enciso', 'matchup_zone': 'LW vs CB'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.41 (squad xG sum 0.8743, xG scale 1.6116, xA scale 2.0289, team xA 0.92 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.409` `squad_xg_sum=0.8743` `squad_xa_sum=0.4514` `team_expected_assists=0.9159` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0489 -> 0.0788 expected_assists: 0.0059 -> 0.012
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 12 by projected rating, fifa_overall 70.0)	`match_squad_rank=12` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0758` `goal_rate=0.0802` `scorer_weight=0.0985`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0119` `assist_rate=0.0` `provider_weight=0.0107`	-

John Egan D 89 0 0.03 0.01 247% 1 0.4 6.7 39.4 14 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9922, quality 0.437)	`market_value_eur=1500000` `quality_factor=0.4374` `boost_factor=0.9922` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0333)	`blend_weight=0.0333` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=0.9864` `prior_reliability=0.1333` `competition=Premier League`	totalShots: 0.5231 -> 0.5057 wonTackle: 0.8262 -> 0.8151 fouls: 0.3043 -> 0.2942
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Emersonn', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.41 (squad xG sum 0.8743, xG scale 1.6116, xA scale 2.0289, team xA 0.92 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.409` `squad_xg_sum=0.8743` `squad_xa_sum=0.4514` `team_expected_assists=0.9159` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0213 -> 0.0343 expected_assists: 0.0253 -> 0.0513
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 13 by projected rating, fifa_overall 72.0)	`match_squad_rank=13` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0337` `goal_rate=0.0` `scorer_weight=0.0274`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.05` `assist_rate=0.0` `provider_weight=0.041`	-

Nobel Mendy D 75 0 0.13 0.04 246% 1 0.3 6.8 46.7 29 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0060, quality 0.641)	`market_value_eur=21000000` `quality_factor=0.6408` `boost_factor=1.006` `reference_quality=0.505` `history_matches=29` `history_shrink=0.1471`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0083)	`blend_weight=0.0083` `history_matches=29` `season_overlap_discount=1.0` `per90_to_match_factor=0.834` `prior_reliability=0.0867` `competition=Premier League`	interceptionWon: 0.9842 -> 0.976 onTargetScoringAttempt: 0.3047 -> 0.3125 totalShots: 0.582 -> 0.5875 fouls: 1.2394 -> 1.2344
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Jaden Philogene-Bidace', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.41 (squad xG sum 0.8743, xG scale 1.6116, xA scale 2.0289, team xA 0.92 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.409` `squad_xg_sum=0.8743` `squad_xa_sum=0.4514` `team_expected_assists=0.9159` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0837 -> 0.1349 expected_assists: 0.0082 -> 0.0166
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 75.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1262` `goal_rate=0.1192` `scorer_weight=0.1686`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0165` `assist_rate=0.0` `provider_weight=0.0143`	-

Ryan Giles D 82 0 0.00 0.00 80% 0 0.1 6.6 37.9 14 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0079, quality 0.574)	`market_value_eur=14500000` `quality_factor=0.5737` `boost_factor=1.0079` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0317)	`blend_weight=0.0317` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=0.9066` `prior_reliability=0.1267` `competition=Premier League`	totalCross: 7.7099 -> 7.6622 interceptionWon: 0.678 -> 0.6716 fouls: 0.1608 -> 0.1557
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': "Christian Walton, Dara O'Shea", 'matchup_zone': 'LWB vs RW + RB'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.41 (squad xG sum 0.8743, xG scale 1.6116, xA scale 2.0289, team xA 0.92 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.409` `squad_xg_sum=0.8743` `squad_xa_sum=0.4514` `team_expected_assists=0.9159` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0599 -> 0.1215
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 18 by projected rating, fifa_overall 72.0)	`match_squad_rank=18` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0027` `goal_rate=0.0` `scorer_weight=0.0026`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1144` `assist_rate=0.1324` `provider_weight=0.1519`	-

Mohamed Belloumi F 82 0 0.27 0.09 207% 2 1.6 7.2 41.6 25 XGBoost F_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0111, quality 0.602)	`market_value_eur=17100000` `quality_factor=0.6016` `boost_factor=1.0111` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Issa Diop, Leif Davis', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 0.8983, 'skill_edge': -0.0252, 'height_edge': -0.044, 'attacker_skill': 0.6016, 'defender_skill': 0.6575, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Issa Diop', 'market_value': 10300000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.5652}, {'player': 'Leif Davis', 'market_value': 26000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7498}], 'xg_override': 0.166}	totalShots: 2.2932 -> 2.06 shotOffTarget: 1.5434 -> 1.3864 onTargetScoringAttempt: 0.7497 -> 0.6735 goals: 0.2605 -> 0.234 expectedGoals: 0.1848 -> 0.166 expectedGoalsOnTarget: 0.1846 -> 0.1658 bigChanceMissed: 0.1613 -> 0.1449
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.41 (squad xG sum 0.8743, xG scale 1.6116, xA scale 2.0289, team xA 0.92 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.409` `squad_xg_sum=0.8743` `squad_xa_sum=0.4514` `team_expected_assists=0.9159` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.166 -> 0.2675 expected_assists: 0.1093 -> 0.2218
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 73.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2347` `goal_rate=0.234` `scorer_weight=0.3344`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=1` `assist_probability=0.1989` `assist_rate=0.2604` `provider_weight=0.2772`	-

Tim Iroegbunam M 81 0 0.08 0.02 324% 1 0.4 6.8 41.7 29 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0030, quality 0.585)	`market_value_eur=15500000` `quality_factor=0.5846` `boost_factor=1.003` `reference_quality=0.505` `history_matches=35` `history_shrink=0.125`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0005)	`blend_weight=0.0005` `history_matches=35` `season_overlap_discount=1.0` `per90_to_match_factor=0.9038` `prior_reliability=0.0067` `competition=Premier League`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=true`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Florentino Luís', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.437, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9813, 'skill_edge': -0.0047, 'height_edge': -0.008, 'attacker_skill': 0.6029, 'defender_skill': 0.6134, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Florentino Luís', 'market_value': 15500000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.6134}], 'xg_override': 0.051}	aerialWon: 1.3302 -> 1.1 totalShots: 0.8058 -> 0.7908 shotOffTarget: 0.6278 -> 0.6161
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.41 (squad xG sum 0.8743, xG scale 1.6116, xA scale 2.0289, team xA 0.92 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.409` `squad_xg_sum=0.8743` `squad_xa_sum=0.4514` `team_expected_assists=0.9159` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.051 -> 0.0822 expected_assists: 0.0357 -> 0.0724
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 76.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0789` `goal_rate=0.0462` `scorer_weight=0.0783`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0698` `assist_rate=0.049` `provider_weight=0.0836`	-

Regan Slater M 84 0 0.11 0.02 392% 1 0.5 6.8 38.9 14 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0009, quality 0.513)	`market_value_eur=9200000` `quality_factor=0.513` `boost_factor=1.0009` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0333)	`blend_weight=0.0333` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=0.9343` `prior_reliability=0.1333` `competition=Premier League`	totalShots: 1.0307 -> 0.9964 fouls: 0.5999 -> 0.6266 interceptionWon: 0.5667 -> 0.5478 goalAssist: 0.0824 -> 0.0952 onTargetScoringAttempt: 0.3359 -> 0.3247 wonTackle: 1.1471 -> 1.14 wasFouled: 0.647 -> 0.641
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Florentino Luís', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.366, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.8769, 'skill_edge': -0.0397, 'height_edge': -0.044, 'attacker_skill': 0.5251, 'defender_skill': 0.6134, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Florentino Luís', 'market_value': 15500000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.6134}], 'xg_override': 0.0669}	aerialWon: 1.1442 -> 0.9 totalShots: 1.0195 -> 0.894 shotOffTarget: 0.6948 -> 0.6093 onTargetScoringAttempt: 0.3247 -> 0.2847 expectedGoals: 0.0763 -> 0.0669 goals: 0.0585 -> 0.0513
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.41 (squad xG sum 0.8743, xG scale 1.6116, xA scale 2.0289, team xA 0.92 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.409` `squad_xg_sum=0.8743` `squad_xa_sum=0.4514` `team_expected_assists=0.9159` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0669 -> 0.1078 expected_assists: 0.072 -> 0.1461
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 72.0)	`match_squad_rank=6` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1022` `goal_rate=0.0513` `scorer_weight=0.0934`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1359` `assist_rate=0.0952` `provider_weight=0.1739`	-

Elliot Stroud M 75 0 0.07 0.00 1437% 1 0.2 6.7 28.4 14 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9998, quality 0.503)	`market_value_eur=8400000` `quality_factor=0.5035` `boost_factor=0.9998` `reference_quality=0.505` `history_matches=6` `history_shrink=0.4545`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0276)	`blend_weight=0.0276` `history_matches=6` `season_overlap_discount=1.0` `per90_to_match_factor=0.829` `prior_reliability=0.0933` `competition=Premier League`	totalCross: 0.6418 -> 0.6731 interceptionWon: 0.3062 -> 0.3304 wonTackle: 0.7801 -> 0.7912 wasFouled: 0.2695 -> 0.2621 onTargetScoringAttempt: 0.1943 -> 0.1889
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Christian Walton', 'matchup_zone': 'Left Mid vs Right Back — Wide Channel', 'duel_win_rate': 0.401, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9722, 'skill_edge': 0.0251, 'height_edge': -0.044, 'attacker_skill': 0.5146, 'defender_skill': 0.4588, 'attacker_role': 'LM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Christian Walton', 'market_value': 2600000, 'role': 'GK', 'price_index': 0.55, 'adjusted_quality': 0.4588}], 'xg_override': 0.0447}	aerialWon: 0.3698 -> 1.2 totalShots: 0.6577 -> 0.6394 shotOffTarget: 0.4656 -> 0.4527 onTargetScoringAttempt: 0.1889 -> 0.1836
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.41 (squad xG sum 0.8743, xG scale 1.6116, xA scale 2.0289, team xA 0.92 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.409` `squad_xg_sum=0.8743` `squad_xa_sum=0.4514` `team_expected_assists=0.9159` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0447 -> 0.072 expected_assists: 0.0502 -> 0.1018
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 11 by projected rating, fifa_overall 74.0)	`match_squad_rank=11` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0695` `goal_rate=0.0506` `scorer_weight=0.0765`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0968` `assist_rate=0.0704` `provider_weight=0.1223`	-

Konstantinos Tzolakis G 90 0 0.00 - n/a 0.0 7.0 40.5 14 XGBoost G_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0157, quality 0.641)	`market_value_eur=21000000` `quality_factor=0.6408` `boost_factor=1.0157` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Emersonn, Julio Enciso', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.4418}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.41 (squad xG sum 0.8743, xG scale 1.6116, xA scale 2.0289, team xA 0.92 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.409` `squad_xg_sum=0.8743` `squad_xa_sum=0.4514` `team_expected_assists=0.9159` `team_xa_to_xg_ratio=0.65`	shots_faced: 13.7 -> 13.0 expected_saves: 3.35 -> 3.18 high_pressure_saves: 1.15 -> 1.1
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 79.0)	`match_squad_rank=4` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=1.4546`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=1.054`	-

Oliver McBurnie F 87 1 0.42 0.14 192% 2 0.8 6.7 38.3 14 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9987, quality 0.494)	`market_value_eur=7600000` `quality_factor=0.4939` `boost_factor=0.9987` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0333)	`blend_weight=0.0333` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=0.9628` `prior_reliability=0.1333` `competition=Premier League`	totalShots: 1.8358 -> 1.8067 wasFouled: 1.7674 -> 1.7887 interceptionWon: 0.3324 -> 0.3213 goals: 0.2865 -> 0.277 onTargetScoringAttempt: 0.7104 -> 0.7188 totalCross: 0.2059 -> 0.199 fouls: 1.2944 -> 1.2994
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Issa Diop, Leif Davis', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 0.8917, 'skill_edge': -0.0736, 'height_edge': 0.0, 'attacker_skill': 0.4939, 'defender_skill': 0.6575, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Issa Diop', 'market_value': 10300000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.5652}, {'player': 'Leif Davis', 'market_value': 26000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7498}], 'xg_override': 0.2612}	totalShots: 1.8441 -> 1.6444 shotOffTarget: 1.1253 -> 1.0034 onTargetScoringAttempt: 0.7188 -> 0.641 expectedGoals: 0.2929 -> 0.2612 goals: 0.277 -> 0.247 bigChanceMissed: 0.2649 -> 0.2362 expectedGoalsOnTarget: 0.2228 -> 0.1987
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.41 (squad xG sum 0.8743, xG scale 1.6116, xA scale 2.0289, team xA 0.92 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.409` `squad_xg_sum=0.8743` `squad_xa_sum=0.4514` `team_expected_assists=0.9159` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.2612 -> 0.4212 expected_assists: 0.0379 -> 0.0769
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 10 by projected rating, fifa_overall 75.0)	`match_squad_rank=10` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.3437` `goal_rate=0.247` `scorer_weight=0.402`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.074` `assist_rate=0.1039` `provider_weight=0.0961`	-

Team totals 1 1.41 11 6.8

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Jack Butland	G	4.4m	10.9%	Market-value substitute prior
Lucas Herrington	D	22.9m	61.5%	Market-value substitute prior
Charlie Hughes	D	13.9m	31.2%	Market-value substitute prior
Christos Mouzakitis	M	23.0m	54.3%	Market-value substitute prior
Brooke Norton-Cuffy	M	22.6m	52.5%	Market-value substitute prior
Robinio Vaz	F	26.0m	59.1%	Market-value substitute prior

Ipswich — projected line vs season baseline (4-2-3-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Christian Walton G 90 0 0.00 - n/a 0.0 6.9 18.8 14 FBref Prior G_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9887, quality 0.437)	`market_value_eur=2600000` `quality_factor=0.4374` `boost_factor=0.9887` `reference_quality=0.505` `history_matches=4` `history_shrink=0.5556`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Oliver McBurnie, Mohamed Belloumi', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.398}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.37 (squad xG sum 1.2999, xG scale 1.0555, xA scale 1.3513, team xA 0.89 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.372` `squad_xg_sum=1.2999` `squad_xa_sum=0.66` `team_expected_assists=0.8918` `team_xa_to_xg_ratio=0.65`	shots_faced: 13.3 -> 13.4 expected_saves: 3.26 -> 3.28 high_pressure_saves: 1.12 -> 1.13
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 73.0)	`match_squad_rank=4` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=1.5041`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=1.011`	-

Dara O'Shea D 90 0 0.04 0.02 98% 0 0.1 6.8 38.3 14 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0157, quality 0.641)	`market_value_eur=21000000` `quality_factor=0.6408` `boost_factor=1.0157` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0333)	`blend_weight=0.0333` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	fouls: 0.7176 -> 0.7437 interceptionWon: 1.2246 -> 1.2005 wonTackle: 1.109 -> 1.0887
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Ryan Giles', 'matchup_zone': 'LWB vs RW + RB'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.37 (squad xG sum 1.2999, xG scale 1.0555, xA scale 1.3513, team xA 0.89 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.372` `squad_xg_sum=1.2999` `squad_xa_sum=0.66` `team_expected_assists=0.8918` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0305 -> 0.0412
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 75.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0424` `goal_rate=0.1463` `scorer_weight=0.0541`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0404` `assist_rate=0.0` `provider_weight=0.033`	-

Issa Diop D 90 0 0.05 0.03 66% 0 0.1 6.8 34.7 31 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0012, quality 0.526)	`market_value_eur=10300000` `quality_factor=0.526` `boost_factor=1.0012` `reference_quality=0.505` `history_matches=21` `history_shrink=0.1923`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0167)	`blend_weight=0.0167` `history_matches=21` `season_overlap_discount=1.0` `per90_to_match_factor=0.9977` `prior_reliability=0.1333` `competition=Premier League`	fouls: 0.8352 -> 0.8462 totalShots: 0.4379 -> 0.4306 interceptionWon: 1.0828 -> 1.0897
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=true`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Mohamed Belloumi', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.37 (squad xG sum 1.2999, xG scale 1.0555, xA scale 1.3513, team xA 0.89 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.372` `squad_xg_sum=1.2999` `squad_xa_sum=0.66` `team_expected_assists=0.8918` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0166 -> 0.0224
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 75.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0471` `goal_rate=0.0436` `scorer_weight=0.0471`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0222` `assist_rate=0.0` `provider_weight=0.0179`	-

Leif Davis D 90 0 0.18 0.12 55% 1 0.5 6.8 39.0 14 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0209, quality 0.686)	`market_value_eur=26000000` `quality_factor=0.6858` `boost_factor=1.0209` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0333)	`blend_weight=0.0333` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	totalCross: 2.8755 -> 2.8297 wonTackle: 0.796 -> 0.7695 onTargetScoringAttempt: 0.2687 -> 0.2931 fouls: 0.6392 -> 0.6179 totalShots: 0.5978 -> 0.6112 goals: 0.1893 -> 0.1997
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=true`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Mohamed Belloumi', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.37 (squad xG sum 1.2999, xG scale 1.0555, xA scale 1.3513, team xA 0.89 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.372` `squad_xg_sum=1.2999` `squad_xa_sum=0.66` `team_expected_assists=0.8918` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1724 -> 0.182 expected_assists: 0.0784 -> 0.1059
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 10 by projected rating, fifa_overall 75.0)	`match_squad_rank=10` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1664` `goal_rate=0.1997` `scorer_weight=0.2043`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1005` `assist_rate=0.1061` `provider_weight=0.1287`	-

Florentino Luís M 77 0 0.04 0.00 760% 1 0.2 6.8 44.9 31 XGBoost M_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0029, quality 0.585)	`market_value_eur=15500000` `quality_factor=0.5846` `boost_factor=1.0029` `reference_quality=0.505` `history_matches=36` `history_shrink=0.122`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Tim Iroegbunam, Regan Slater', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.37 (squad xG sum 1.2999, xG scale 1.0555, xA scale 1.3513, team xA 0.89 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.372` `squad_xg_sum=1.2999` `squad_xa_sum=0.66` `team_expected_assists=0.8918` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0367 -> 0.0496
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 80.0)	`match_squad_rank=6` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0362` `goal_rate=0.0432` `scorer_weight=0.0405`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0484` `assist_rate=0.0556` `provider_weight=0.0604`	-

Exequiel Palacios M 86 0 0.04 - n/a 1 0.0 7.0 37.5 13 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0226, quality 0.686)	`market_value_eur=26000000` `quality_factor=0.6858` `boost_factor=1.0226` `reference_quality=0.505` `history_matches=7` `history_shrink=0.4167`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0054)	`blend_weight=0.0054` `history_matches=7` `season_overlap_discount=1.0` `per90_to_match_factor=0.96` `prior_reliability=0.02` `competition=Premier League`	interceptionWon: 1.1252 -> 1.1538 fouls: 0.9162 -> 0.9286 wasFouled: 1.0507 -> 1.0623 wonTackle: 1.6353 -> 1.6438
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Tim Iroegbunam', 'matchup_zone': 'CAM vs CM', 'duel_win_rate': 0.442, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.0329, 'skill_edge': 0.0424, 'height_edge': -0.02, 'attacker_skill': 0.6971, 'defender_skill': 0.6029, 'attacker_role': 'CAM', 'attacker_mv_price_index': 0.95, 'defender_adjusted_parts': [{'player': 'Tim Iroegbunam', 'market_value': 15500000, 'role': 'CM', 'price_index': 0.9, 'adjusted_quality': 0.6029}], 'xg_override': 0.0371}	aerialWon: 0.0 -> 1.0 totalShots: 0.4923 -> 0.5085 shotOffTarget: 0.3505 -> 0.362
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.37 (squad xG sum 1.2999, xG scale 1.0555, xA scale 1.3513, team xA 0.89 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.372` `squad_xg_sum=1.2999` `squad_xa_sum=0.66` `team_expected_assists=0.8918` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0597 -> 0.0807
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 83.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0384` `goal_rate=0.0468` `scorer_weight=0.0436`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0775` `assist_rate=0.0606` `provider_weight=0.0814`	-

Abdul Fatawu Issahaku F 73 0 0.11 0.02 539% 1 0.4 6.9 36.1 24 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0204, quality 0.669)	`market_value_eur=24000000` `quality_factor=0.6685` `boost_factor=1.0204` `reference_quality=0.505` `history_matches=7` `history_shrink=0.4167`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0289)	`blend_weight=0.0289` `history_matches=7` `season_overlap_discount=1.0` `per90_to_match_factor=0.816` `prior_reliability=0.1067` `competition=Premier League`	totalShots: 1.3562 -> 1.317 wonTackle: 0.8257 -> 0.8461 totalCross: 4.3969 -> 4.4172 fouls: 1.6938 -> 1.6743 wasFouled: 1.0154 -> 1.0303 onTargetScoringAttempt: 0.3899 -> 0.3786 goalAssist: 0.1283 -> 0.1393 interceptionWon: 0.8553 -> 0.8453
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Semi Ajayi, John Egan', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.268, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.0999, 'skill_edge': 0.104, 'height_edge': -0.036, 'attacker_skill': 0.6685, 'defender_skill': 0.4374, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Semi Ajayi', 'market_value': 930000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'John Egan', 'market_value': 1500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.1014}	totalShots: 1.3449 -> 1.4793 shotOffTarget: 0.9663 -> 1.0628 onTargetScoringAttempt: 0.3786 -> 0.4164 goals: 0.1151 -> 0.1266 expectedGoals: 0.0922 -> 0.1014
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.37 (squad xG sum 1.2999, xG scale 1.0555, xA scale 1.3513, team xA 0.89 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.372` `squad_xg_sum=1.2999` `squad_xa_sum=0.66` `team_expected_assists=0.8918` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1014 -> 0.107 expected_assists: 0.1027 -> 0.1388
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 82.0)	`match_squad_rank=5` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1015` `goal_rate=0.1266` `scorer_weight=0.1248`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1296` `assist_rate=0.1393` `provider_weight=0.1721`	-

Julio Enciso M 82 0 0.21 0.12 81% 3 1.6 6.9 40.1 14 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0228, quality 0.702)	`market_value_eur=28000000` `quality_factor=0.7022` `boost_factor=1.0228` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0300)	`blend_weight=0.03` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=0.9108` `prior_reliability=0.12` `competition=Premier League`	totalCross: 1.0427 -> 1.0114 goalAssist: 0.1359 -> 0.1622 totalShots: 2.4031 -> 2.4234 wonTackle: 0.5373 -> 0.5515 wasFouled: 1.5845 -> 1.5977 interceptionWon: 0.434 -> 0.421 onTargetScoringAttempt: 0.8168 -> 0.8076 goals: 0.2261 -> 0.2193
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Semi Ajayi', 'matchup_zone': 'LW vs CB', 'duel_win_rate': 0.479, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.0869, 'skill_edge': 0.1191, 'height_edge': -0.06, 'attacker_skill': 0.7022, 'defender_skill': 0.4374, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Semi Ajayi', 'market_value': 930000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.2008}`	aerialWon: 0.1101 -> 0.9 totalShots: 2.4234 -> 2.634 shotOffTarget: 1.5863 -> 1.7241 onTargetScoringAttempt: 0.8076 -> 0.8778 expectedGoalsOnTarget: 0.3137 -> 0.341 goals: 0.2193 -> 0.2384 expectedGoals: 0.1847 -> 0.2008 bigChanceMissed: 0.0876 -> 0.0952
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.37 (squad xG sum 1.2999, xG scale 1.0555, xA scale 1.3513, team xA 0.89 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.372` `squad_xg_sum=1.2999` `squad_xa_sum=0.66` `team_expected_assists=0.8918` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.2008 -> 0.2119 expected_assists: 0.1269 -> 0.1714
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 78.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.191` `goal_rate=0.2384` `scorer_weight=0.2437`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=1` `assist_probability=0.1575` `assist_rate=0.1622` `provider_weight=0.2056`	-

Jaden Philogene-Bidace F 71 0 0.14 0.01 923% 2 0.3 6.8 18.9 17 XGBoost F_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0190, quality 0.619)	`market_value_eur=18800000` `quality_factor=0.6191` `boost_factor=1.019` `reference_quality=0.505` `history_matches=4` `history_shrink=0.5556`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Semi Ajayi, John Egan', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.25, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.0731, 'skill_edge': 0.0818, 'height_edge': -0.032, 'attacker_skill': 0.6191, 'defender_skill': 0.4374, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Semi Ajayi', 'market_value': 930000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'John Egan', 'market_value': 1500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.1365}	totalShots: 1.5343 -> 1.6465 shotOffTarget: 0.9826 -> 1.0544 onTargetScoringAttempt: 0.5517 -> 0.592 goals: 0.1469 -> 0.1576 expectedGoals: 0.1272 -> 0.1365 expectedGoalsOnTarget: 0.0915 -> 0.0982
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.37 (squad xG sum 1.2999, xG scale 1.0555, xA scale 1.3513, team xA 0.89 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.372` `squad_xg_sum=1.2999` `squad_xa_sum=0.66` `team_expected_assists=0.8918` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1365 -> 0.1441 expected_assists: 0.0555 -> 0.075
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 13 by projected rating, fifa_overall 85.0)	`match_squad_rank=13` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1342` `goal_rate=0.1576` `scorer_weight=0.1604`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0723` `assist_rate=0.0` `provider_weight=0.06`	-

Emersonn F 70 1 0.36 0.18 103% 2 1.0 6.7 38.6 14 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0189, quality 0.669)	`market_value_eur=24000000` `quality_factor=0.6685` `boost_factor=1.0189` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0200)	`blend_weight=0.02` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=0.7782` `prior_reliability=0.08` `competition=Premier League`	totalCross: 0.0673 -> 0.0789 wasFouled: 1.8271 -> 1.8165 onTargetScoringAttempt: 0.7708 -> 0.7809 totalOffside: 0.365 -> 0.3577 wonTackle: 0.2896 -> 0.2968 goals: 0.2978 -> 0.3048
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Semi Ajayi, John Egan', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.292, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.1352, 'skill_edge': 0.104, 'height_edge': -0.012, 'attacker_skill': 0.6685, 'defender_skill': 0.4374, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Semi Ajayi', 'market_value': 930000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'John Egan', 'market_value': 1500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.3409}	totalShots: 2.0983 -> 2.382 shotOffTarget: 1.3174 -> 1.4955 onTargetScoringAttempt: 0.7809 -> 0.8865 bigChanceMissed: 0.3636 -> 0.4128 goals: 0.3048 -> 0.346 expectedGoals: 0.3003 -> 0.3409 expectedGoalsOnTarget: 0.1665 -> 0.189
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.37 (squad xG sum 1.2999, xG scale 1.0555, xA scale 1.3513, team xA 0.89 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.372` `squad_xg_sum=1.2999` `squad_xa_sum=0.66` `team_expected_assists=0.8918` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.3409 -> 0.3598 expected_assists: 0.0482 -> 0.0651
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 14 by projected rating, fifa_overall 74.0)	`match_squad_rank=14` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.3022` `goal_rate=0.346` `scorer_weight=0.3645`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.063` `assist_rate=0.1074` `provider_weight=0.0814`	-

Abdoul Ouattara M 89 0 0.04 0.01 431% 1 0.2 6.8 47.2 20 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0104, quality 0.595)	`market_value_eur=16500000` `quality_factor=0.5953` `boost_factor=1.0104` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0050)	`blend_weight=0.005` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=0.9852` `prior_reliability=0.02` `competition=Premier League`	totalCross: 1.4958 -> 1.4883 interceptionWon: 1.4403 -> 1.4331 wonTackle: 1.0236 -> 1.0185
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Elliot Stroud', 'matchup_zone': 'Midfield lane'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.37 (squad xG sum 1.2999, xG scale 1.0555, xA scale 1.3513, team xA 0.89 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.372` `squad_xg_sum=1.2999` `squad_xa_sum=0.66` `team_expected_assists=0.8918` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0568 -> 0.0768
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 73.0)	`match_squad_rank=8` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0436` `goal_rate=0.0574` `scorer_weight=0.0519`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0739` `assist_rate=0.0951` `provider_weight=0.096`	-

Team totals 1 1.37 14 6.8

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Kjell Scherpen	G	4.8m	14.0%	Market-value substitute prior
Jacob Greaves	D	14.2m	23.0%	Market-value substitute prior
Cédric Kipré	D	13.6m	29.9%	Market-value substitute prior
Marcelino Núñez	M	22.5m	31.8%	Market-value substitute prior
Azor Matusiwa	M	22.4m	52.0%	Market-value substitute prior
Jack Clarke	F	22.7m	73.5%	Market-value substitute prior

Antagonist / tactical summary

Midfield is evenly contested — Tim Iroegbunam vs Julio Enciso will be the pivotal duel. Total antagonist xG — Home: 0.61, Away: 0.82. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Midfield — possession 48.3% / 51.7%, dominance ratio 0.44, key duel: Tim Iroegbunam vs Julio Enciso

Mohamed Belloumi xG 0.17Oliver McBurnie xG 0.26Elliot Stroud xG 0.04Julio Enciso xG 0.20Exequiel Palacios xG 0.04Emersonn xG 0.34

Summary

Antagonist matchup analysis for Hull vs Ipswich. Most likely scoreline: 1-1. Lambdas derived from 7 home and 5 away individual player duels. Head-to-head: Hull 4W-2D-4L vs Ipswich over 10 meetings.

Tactical parity between 5-4-1 and 4-2-3-1. Match outcome driven primarily by individual player quality. Midfield is evenly contested — Tim Iroegbunam vs Julio Enciso will be the pivotal duel. Total antagonist xG — Home: 0.61, Away: 0.82. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Bournemouth vs Leeds

Premier League · 20261031 · 4-2-3-1 vs 3-4-2-1

Predicted 1-0Winner: BournemouthWinning H 1-0 (9.6%)Winning A 0-1 (7.4%)Confidence 0.73λ home/away 1.57 / 1.22Most-likely P 9.6%

Actual vs predicted: result not available yet (upcoming fixture).

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Bournemouth vs Leeds. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 4-2-3-1 vs 3-4-2-1 (duel lambdas: Home 1.482, Away 1.059). Final calibrated Poisson lambdas: Home 1.566, Away 1.215 (most likely score 1-0).

Home Starters 11/11 XGBoost (10 FBref) Away Starters 11/11 XGBoost (10 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 1.48 / 1.06 Final λ (H/A) 1.57 / 1.22

FBref Team Priors: Bournemouth (Gls/90: 1.0, GA/90: 1.0 → E[λ]: 0.69) vs Leeds United (Gls/90: 1.0, GA/90: 1.0 → E[λ]: 0.69) | Blend Weight: 0.192

Modeling Pipeline Stages (5 layers)

1. XGBoost player regressions for volume & event rates with market-value quality scaling
2. Multi-source prior blending (FBref 2026-2027 tables + Sofascore season history)
3. Antagonist formation duel matrix resolving 1v1 and zonal matchups
4. Team-level prior calibration (FBref possession and attack/defense efficiencies)
5. Bivariate Poisson matrix with Dixon-Coles low-score correction

Standing & form context

Metric	Bournemouth	Leeds
League position	13.00	7.00
Points per game	1.22	1.37
Attack index	1.03	0.92
Defense (concede) index	0.82	1.07
Form multiplier	0.99	1.01
Standing λ multiplier	0.93	1.00

Lambda decomposition (how the goal rates were built)

Stage	Bournemouth λ	Leeds λ
FBref Squad Prior E[λ] (weight 0.192)	0.690	0.690
Pre-competitive-floor	1.457	1.077
Pre-calibration (Micro-duels)	1.482	1.059
Final λ	1.566	1.215

competition calibration scale 1.097competition observed total goals 2.781quality gap applied Falsemismatch score pressure applied Falsecompetitive underdog floor applied Falsestage discount 1.0

Scoreline distribution & markets

Top 5 scorelines	P
1-1	11.9%
1-0	9.6%
2-1	9.2%
2-0	7.6%
0-1	7.4%

Home win	45.4%
Draw	25.1%
Away win	29.5%
Over 2.5	52.6%
BTTS yes	55.7%

Pregame fair line — favorite Bournemouth (45.4%), fair total 2.78, fair spread 0.35

Projected players vs season baseline

Proj = projected for this match · Season = this season's per-game average from match data · deviation badge compares projected xG to the player's season xG/game.

Bournemouth — projected line vs season baseline (4-2-3-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Marcus Tavernier F 83 0 0.30 0.28 ok 2 1.6 7.0 51.8 40 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0060, quality 0.694)	`market_value_eur=27000000` `quality_factor=0.6941` `boost_factor=1.006` `reference_quality=0.505` `history_matches=42` `history_shrink=0.1064`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0092)	`blend_weight=0.0092` `history_matches=42` `season_overlap_discount=1.0` `per90_to_match_factor=0.9251` `prior_reliability=0.1333` `competition=Premier League`	wasFouled: 1.0301 -> 1.0377 onTargetScoringAttempt: 1.051 -> 1.0456
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'James Justin, Nico Elvedi', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.207, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.0096, 'skill_edge': 0.0366, 'height_edge': -0.03, 'attacker_skill': 0.6941, 'defender_skill': 0.6129, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'James Justin', 'market_value': 16500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.6503}, {'player': 'Nico Elvedi', 'market_value': 11000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.5755}], 'xg_override': 0.3592}	totalShots: 2.0349 -> 2.0544 onTargetScoringAttempt: 1.0456 -> 1.0556 shotOffTarget: 0.9856 -> 0.9951 expectedGoalsOnTarget: 0.5595 -> 0.5649
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.57 (squad xG sum 1.8872, xG scale 0.8298, xA scale 1.0336, team xA 1.02 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.566` `squad_xg_sum=1.8872` `squad_xa_sum=0.9848` `team_expected_assists=1.0179` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.3592 -> 0.2981 expected_assists: 0.1692 -> 0.1748
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 79.0)	`match_squad_rank=4` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2578` `goal_rate=0.3502` `scorer_weight=0.2915`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1604` `assist_rate=0.0172` `provider_weight=0.1398`	-

Evanilson F 84 0 0.30 0.28 ok 3 1.8 7.0 55.4 33 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0072, quality 0.739)	`market_value_eur=33000000` `quality_factor=0.7389` `boost_factor=1.0072` `reference_quality=0.505` `history_matches=44` `history_shrink=0.102`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0075)	`blend_weight=0.0075` `history_matches=44` `season_overlap_discount=1.0` `per90_to_match_factor=0.9287` `prior_reliability=0.1133` `competition=Premier League`	fouls: 1.3847 -> 1.3948 wasFouled: 1.6989 -> 1.7067 interceptionWon: 0.7007 -> 0.7077 onTargetScoringAttempt: 0.7937 -> 0.7877 totalShots: 2.4108 -> 2.4049 goalAssist: 0.3288 -> 0.3345 wonTackle: 0.7552 -> 0.7495 totalOffside: 0.7507 -> 0.7451
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'James Justin, Nico Elvedi', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.239, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.0569, 'skill_edge': 0.0567, 'height_edge': -0.018, 'attacker_skill': 0.7389, 'defender_skill': 0.6129, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'James Justin', 'market_value': 16500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.6503}, {'player': 'Nico Elvedi', 'market_value': 11000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.5755}], 'xg_override': 0.3622}	totalShots: 2.4049 -> 2.5417 shotOffTarget: 1.6171 -> 1.7091 onTargetScoringAttempt: 0.7877 -> 0.8325 expectedGoalsOnTarget: 0.4304 -> 0.4549 goals: 0.3713 -> 0.3924 expectedGoals: 0.3427 -> 0.3622 bigChanceMissed: 0.3238 -> 0.3422
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.57 (squad xG sum 1.8872, xG scale 0.8298, xA scale 1.0336, team xA 1.02 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.566` `squad_xg_sum=1.8872` `squad_xa_sum=0.9848` `team_expected_assists=1.0179` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.3622 -> 0.3006
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 79.0)	`match_squad_rank=5` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2596` `goal_rate=0.3924` `scorer_weight=0.3226`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0941` `assist_rate=0.3345` `provider_weight=0.1235`	-

Justin Kluivert M 74 0 0.15 0.09 68% 2 1.0 6.8 37.2 16 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0079, quality 0.669)	`market_value_eur=24000000` `quality_factor=0.6685` `boost_factor=1.0079` `reference_quality=0.505` `history_matches=26` `history_shrink=0.1613`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0119)	`blend_weight=0.0119` `history_matches=26` `season_overlap_discount=1.0` `per90_to_match_factor=0.8226` `prior_reliability=0.1133` `competition=Premier League`	totalCross: 2.2905 -> 2.2805 onTargetScoringAttempt: 0.6017 -> 0.6117 totalShots: 1.7204 -> 1.7285 fouls: 1.6498 -> 1.6417 wonTackle: 0.4886 -> 0.4943 interceptionWon: 0.9165 -> 0.9114
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'James Justin', 'matchup_zone': 'LW vs CB', 'duel_win_rate': 0.38, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9415, 'skill_edge': 0.0082, 'height_edge': -0.048, 'attacker_skill': 0.6685, 'defender_skill': 0.6503, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'James Justin', 'market_value': 16500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.6503}], 'xg_override': 0.177}`	aerialWon: 0.3635 -> 0.8 totalShots: 1.7304 -> 1.6292 shotOffTarget: 1.1187 -> 1.0533 onTargetScoringAttempt: 0.6117 -> 0.5759 expectedGoalsOnTarget: 0.2968 -> 0.2794 expectedGoals: 0.188 -> 0.177 goals: 0.1776 -> 0.1672 bigChanceMissed: 0.1126 -> 0.106
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.57 (squad xG sum 1.8872, xG scale 0.8298, xA scale 1.0336, team xA 1.02 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.566` `squad_xg_sum=1.8872` `squad_xa_sum=0.9848` `team_expected_assists=1.0179` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.177 -> 0.1469
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 13 by projected rating, fifa_overall 79.0)	`match_squad_rank=13` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1366` `goal_rate=0.1672` `scorer_weight=0.1406`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1188` `assist_rate=0.1338` `provider_weight=0.1349`	-

Rayan F 86 1 0.31 0.18 75% 3 1.8 7.2 50.3 40 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0184, quality 0.848)	`market_value_eur=58000000` `quality_factor=0.8478` `boost_factor=1.0184` `reference_quality=0.505` `history_matches=23` `history_shrink=0.1786`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0155)	`blend_weight=0.0155` `history_matches=23` `season_overlap_discount=1.0` `per90_to_match_factor=0.9559` `prior_reliability=0.1333` `competition=Premier League`	totalCross: 1.835 -> 1.8066 wasFouled: 1.3869 -> 1.3654 fouls: 0.9579 -> 0.9431 wonTackle: 0.658 -> 0.6478 onTargetScoringAttempt: 0.8848 -> 0.8936 interceptionWon: 0.9909 -> 0.983 goals: 0.367 -> 0.3613
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'James Justin, Nico Elvedi', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.304, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.1525, 'skill_edge': 0.1057, 'height_edge': -0.002, 'attacker_skill': 0.8478, 'defender_skill': 0.6129, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'James Justin', 'market_value': 16500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.6503}, {'player': 'Nico Elvedi', 'market_value': 11000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.5755}], 'xg_override': 0.375}	totalShots: 2.8038 -> 3.2314 shotOffTarget: 1.9102 -> 2.2015 bigChanceMissed: 0.9239 -> 1.0648 onTargetScoringAttempt: 0.8936 -> 1.0299 goals: 0.3613 -> 0.4164 expectedGoalsOnTarget: 0.3315 -> 0.3821 expectedGoals: 0.3254 -> 0.375
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.57 (squad xG sum 1.8872, xG scale 0.8298, xA scale 1.0336, team xA 1.02 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.566` `squad_xg_sum=1.8872` `squad_xa_sum=0.9848` `team_expected_assists=1.0179` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.375 -> 0.3111
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 79.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.2674` `goal_rate=0.4164` `scorer_weight=0.3414`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0832` `assist_rate=0.0222` `provider_weight=0.0695`	-

Tyler Adams M 83 0 0.03 0.01 260% 1 0.2 6.9 48.7 25 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0087, quality 0.725)	`market_value_eur=31000000` `quality_factor=0.7249` `boost_factor=1.0087` `reference_quality=0.505` `history_matches=33` `history_shrink=0.1316`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0023)	`blend_weight=0.0023` `history_matches=33` `season_overlap_discount=1.0` `per90_to_match_factor=0.9277` `prior_reliability=0.0267` `competition=Premier League`	interceptionWon: 1.4265 -> 1.4338
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Harry Wilson, Noah Okafor, Ethan Ampadu', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.57 (squad xG sum 1.8872, xG scale 0.8298, xA scale 1.0336, team xA 1.02 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.566` `squad_xg_sum=1.8872` `squad_xa_sum=0.9848` `team_expected_assists=1.0179` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0378 -> 0.0314
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 79.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0309` `goal_rate=0.0485` `scorer_weight=0.0352`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0351` `assist_rate=0.0473` `provider_weight=0.0411`	-

Đorđe Petrović G 90 0 0.00 - n/a 0.0 6.8 56.1 33 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0060, quality 0.710)	`market_value_eur=29000000` `quality_factor=0.71` `boost_factor=1.006` `reference_quality=0.505` `history_matches=46` `history_shrink=0.098`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0085)	`blend_weight=0.0085` `history_matches=46` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	saves: 2.7746 -> 2.7893
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Dominic Calvert-Lewin, Noah Okafor', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.0586}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.57 (squad xG sum 1.8872, xG scale 0.8298, xA scale 1.0336, team xA 1.02 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.566` `squad_xg_sum=1.8872` `squad_xa_sum=0.9848` `team_expected_assists=1.0179` `team_xa_to_xg_ratio=0.65`	shots_faced: 10.1 -> 11.5 expected_saves: 2.48 -> 2.81 high_pressure_saves: 0.85 -> 0.97
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 10 by projected rating, fifa_overall 79.0)	`match_squad_rank=10` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=1.5747`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=0.9927`	-

Adam Smith D 84 0 0.06 0.02 254% 0 0.1 6.7 35.8 33 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9971, quality 0.437)	`market_value_eur=310000` `quality_factor=0.4374` `boost_factor=0.9971` `reference_quality=0.505` `history_matches=30` `history_shrink=0.1429`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0124)	`blend_weight=0.0124` `history_matches=30` `season_overlap_discount=1.0` `per90_to_match_factor=0.9321` `prior_reliability=0.1333` `competition=Premier League`	interceptionWon: 0.784 -> 0.7916
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Ao Tanaka', 'matchup_zone': 'Right Wingback vs Left Back — Overlapping Run'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.57 (squad xG sum 1.8872, xG scale 0.8298, xA scale 1.0336, team xA 1.02 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.566` `squad_xg_sum=1.8872` `squad_xa_sum=0.9848` `team_expected_assists=1.0179` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0714 -> 0.0592
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 15 by projected rating, fifa_overall 73.0)	`match_squad_rank=15` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0575` `goal_rate=0.0` `scorer_weight=0.0474`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0567` `assist_rate=0.0433` `provider_weight=0.0512`	-

James Hill D 90 0 0.01 0.02 42% 1 0.3 7.5 58.8 33 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0081, quality 0.732)	`market_value_eur=32000000` `quality_factor=0.732` `boost_factor=1.0081` `reference_quality=0.505` `history_matches=37` `history_shrink=0.119`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0103)	`blend_weight=0.0103` `history_matches=37` `season_overlap_discount=1.0` `per90_to_match_factor=0.9994` `prior_reliability=0.1333` `competition=Premier League`	fouls: 0.8168 -> 0.8238 totalShots: 0.5793 -> 0.5733
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Dominic Calvert-Lewin', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.57 (squad xG sum 1.8872, xG scale 0.8298, xA scale 1.0336, team xA 1.02 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.566` `squad_xg_sum=1.8872` `squad_xa_sum=0.9848` `team_expected_assists=1.0179` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 78.0)	`match_squad_rank=1` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0136` `goal_rate=0.0` `scorer_weight=0.011`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0703` `assist_rate=0.0144` `provider_weight=0.0583`	-

António Silva D 90 0 0.13 0.09 38% 1 0.4 7.1 39.8 14 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0262, quality 0.732)	`market_value_eur=32000000` `quality_factor=0.732` `boost_factor=1.0262` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0333)	`blend_weight=0.0333` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	fouls: 1.0596 -> 1.041 onTargetScoringAttempt: 0.2468 -> 0.2386 totalShots: 0.7302 -> 0.7225 interceptionWon: 1.2084 -> 1.2015 wonTackle: 0.7968 -> 0.8036
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Dominic Calvert-Lewin', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.57 (squad xG sum 1.8872, xG scale 0.8298, xA scale 1.0336, team xA 1.02 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.566` `squad_xg_sum=1.8872` `squad_xa_sum=0.9848` `team_expected_assists=1.0179` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1546 -> 0.1283
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 77.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1204` `goal_rate=0.0` `scorer_weight=0.1026`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0415` `assist_rate=0.0` `provider_weight=0.0339`	-

Adrien Truffert D 89 0 0.03 0.03 ok 1 0.4 6.8 56.5 33 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0069, quality 0.739)	`market_value_eur=33000000` `quality_factor=0.7389` `boost_factor=1.0069` `reference_quality=0.505` `history_matches=46` `history_shrink=0.098`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0085)	`blend_weight=0.0085` `history_matches=46` `season_overlap_discount=1.0` `per90_to_match_factor=0.9927` `prior_reliability=0.1333` `competition=Premier League`	totalCross: 3.938 -> 3.9256 wonTackle: 1.4115 -> 1.4037 totalShots: 0.6653 -> 0.6596
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Jayden Bogle', 'matchup_zone': 'Left Wingback vs Right Back — Overlapping Run'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.57 (squad xG sum 1.8872, xG scale 0.8298, xA scale 1.0336, team xA 1.02 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.566` `squad_xg_sum=1.8872` `squad_xa_sum=0.9848` `team_expected_assists=1.0179` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0336 -> 0.0279
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 11 by projected rating, fifa_overall 80.0)	`match_squad_rank=11` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0275` `goal_rate=0.0186` `scorer_weight=0.0229`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=1` `assist_probability=0.1402` `assist_rate=0.2445` `provider_weight=0.1889`	-

Ryan Christie M 73 0 0.09 0.06 46% 1 0.9 6.9 50.7 28 FBref Prior M_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0000, quality 0.506)	`market_value_eur=8600000` `quality_factor=0.5059` `boost_factor=1.0` `reference_quality=0.505` `history_matches=33` `history_shrink=0.1316`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Gabriel Gudmundsson', 'matchup_zone': 'Midfield lane'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.57 (squad xG sum 1.8872, xG scale 0.8298, xA scale 1.0336, team xA 1.02 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.566` `squad_xg_sum=1.8872` `squad_xa_sum=0.9848` `team_expected_assists=1.0179` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1044 -> 0.0866
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 78.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.083` `goal_rate=0.1018` `scorer_weight=0.0851`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0885` `assist_rate=0.0761` `provider_weight=0.0837`	-

Team totals 1 1.57 18 7.0

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Michele Di Gregorio	G	4.7m	12.3%	Market-value substitute prior
Bafodé Diakité	D	14.4m	33.9%	Market-value substitute prior
Álex Jiménez	D	14.3m	33.3%	Market-value substitute prior
Alex Scott	M	23.7m	34.3%	Market-value substitute prior
Hamed Junior Traorè	M	22.5m	52.2%	Market-value substitute prior
Eli Junior Kroupi	F	27.0m	63.7%	Market-value substitute prior

Leeds — projected line vs season baseline (3-4-2-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

James Trafford G 90 0 0.00 - n/a 0.0 7.2 19.1 34 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0234, quality 0.770)	`market_value_eur=38000000` `quality_factor=0.7699` `boost_factor=1.0234` `reference_quality=0.505` `history_matches=12` `history_shrink=0.2941`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0255)	`blend_weight=0.0255` `history_matches=12` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Rayan, Evanilson', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.4818}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.22 (squad xG sum 1.2773, xG scale 0.9512, xA scale 1.0781, team xA 0.79 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.215` `squad_xg_sum=1.2773` `squad_xa_sum=0.7326` `team_expected_assists=0.7898` `team_xa_to_xg_ratio=0.65`	shots_faced: 14.1 -> 14.9 expected_saves: 3.45 -> 3.65 high_pressure_saves: 1.19 -> 1.25
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 79.0)	`match_squad_rank=1` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=1.2696`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=0.772`	-

James Justin D 88 0 0.06 0.06 ok 1 0.9 6.8 59.0 36 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0032, quality 0.595)	`market_value_eur=16500000` `quality_factor=0.5953` `boost_factor=1.0032` `reference_quality=0.505` `history_matches=37` `history_shrink=0.119`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0103)	`blend_weight=0.0103` `history_matches=37` `season_overlap_discount=1.0` `per90_to_match_factor=0.9833` `prior_reliability=0.1333` `competition=Premier League`	totalCross: 2.8865 -> 2.8719 wonTackle: 1.3046 -> 1.3165
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Justin Kluivert', 'matchup_zone': 'LW vs CB'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.22 (squad xG sum 1.2773, xG scale 0.9512, xA scale 1.0781, team xA 0.79 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.215` `squad_xg_sum=1.2773` `squad_xa_sum=0.7326` `team_expected_assists=0.7898` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 10 by projected rating, fifa_overall 76.0)	`match_squad_rank=10` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.057` `goal_rate=0.148` `scorer_weight=0.0734`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0655` `assist_rate=0.0834` `provider_weight=0.0801`	-

Nico Elvedi D 89 0 0.05 0.02 180% 1 0.2 6.9 31.8 13 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0040, quality 0.534)	`market_value_eur=11000000` `quality_factor=0.5341` `boost_factor=1.004` `reference_quality=0.505` `history_matches=6` `history_shrink=0.4545`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0118)	`blend_weight=0.0118` `history_matches=6` `season_overlap_discount=1.0` `per90_to_match_factor=0.9928` `prior_reliability=0.04` `competition=Premier League`	wasFouled: 0.0 -> 0.0391 interceptionWon: 0.9417 -> 0.9501 wonTackle: 0.6195 -> 0.6122 totalShots: 0.5275 -> 0.5213 onTargetScoringAttempt: 0.4664 -> 0.4609
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Rayan', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.22 (squad xG sum 1.2773, xG scale 0.9512, xA scale 1.0781, team xA 0.79 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.215` `squad_xg_sum=1.2773` `squad_xa_sum=0.7326` `team_expected_assists=0.7898` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0893 -> 0.0963
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 78.0)	`match_squad_rank=6` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0495` `goal_rate=0.0` `scorer_weight=0.0406`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0918` `assist_rate=0.0` `provider_weight=0.077`	-

Tarik Muharemović D 90 0 0.08 0.04 82% 1 0.5 7.1 41.0 14 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0335, quality 0.796)	`market_value_eur=43000000` `quality_factor=0.7957` `boost_factor=1.0335` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0333)	`blend_weight=0.0333` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	interceptionWon: 1.1552 -> 1.1834 totalCross: 0.0804 -> 0.0944 wasFouled: 0.9182 -> 0.9043 totalShots: 0.8668 -> 0.8546 goalAssist: 0.2924 -> 0.2827 fouls: 0.7794 -> 0.7701 onTargetScoringAttempt: 0.2814 -> 0.2887
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=true`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Marcus Tavernier', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.22 (squad xG sum 1.2773, xG scale 0.9512, xA scale 1.0781, team xA 0.79 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.215` `squad_xg_sum=1.2773` `squad_xa_sum=0.7326` `team_expected_assists=0.7898` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 75.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0783` `goal_rate=0.0` `scorer_weight=0.0652`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0379` `assist_rate=0.2827` `provider_weight=0.0483`	-

Jayden Bogle M 82 0 0.12 0.08 53% 1 0.6 6.9 54.8 30 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0038, quality 0.623)	`market_value_eur=19200000` `quality_factor=0.6231` `boost_factor=1.0038` `reference_quality=0.505` `history_matches=42` `history_shrink=0.1064`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0078)	`blend_weight=0.0078` `history_matches=42` `season_overlap_discount=1.0` `per90_to_match_factor=0.9076` `prior_reliability=0.1133` `competition=Premier League`	wasFouled: 1.2588 -> 1.2699 totalShots: 1.2749 -> 1.2649 totalCross: 2.7667 -> 2.7576
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Adrien Truffert', 'matchup_zone': 'Left Wingback vs Right Back — Overlapping Run', 'duel_win_rate': 0.356, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.9351, 'skill_edge': -0.0522, 'height_edge': 0.008, 'attacker_skill': 0.6976, 'defender_skill': 0.8135, 'attacker_role': 'LWB', 'attacker_mv_price_index': 0.7, 'defender_adjusted_parts': [{'player': 'Adrien Truffert', 'market_value': 33000000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.8135}], 'xg_override': 0.1269}	totalShots: 1.2715 -> 1.189 shotOffTarget: 0.8385 -> 0.7841 onTargetScoringAttempt: 0.433 -> 0.4049 goals: 0.1915 -> 0.1791 expectedGoalsOnTarget: 0.1689 -> 0.1579 aerialWon: 1.0108 -> 1.0 expectedGoals: 0.1357 -> 0.1269 bigChanceMissed: 0.1299 -> 0.1215
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.22 (squad xG sum 1.2773, xG scale 0.9512, xA scale 1.0781, team xA 0.79 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.215` `squad_xg_sum=1.2773` `squad_xa_sum=0.7326` `team_expected_assists=0.7898` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1269 -> 0.1207 expected_assists: 0.1143 -> 0.123
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 76.0)	`match_squad_rank=5` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.1137` `goal_rate=0.1791` `scorer_weight=0.1509`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.1157` `assist_rate=0.0949` `provider_weight=0.1085`	-

Ethan Ampadu M 90 0 0.04 0.05 ok 1 0.5 7.1 58.9 33 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0073, quality 0.739)	`market_value_eur=33000000` `quality_factor=0.7389` `boost_factor=1.0073` `reference_quality=0.505` `history_matches=43` `history_shrink=0.1042`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0090)	`blend_weight=0.009` `history_matches=43` `season_overlap_discount=1.0` `per90_to_match_factor=0.9979` `prior_reliability=0.1333` `competition=Premier League`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Tyler Adams', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.479, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 1.0419, 'skill_edge': 0.0005, 'height_edge': 0.028, 'attacker_skill': 0.7622, 'defender_skill': 0.761, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Tyler Adams', 'market_value': 31000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.761}], 'xg_override': 0.0399}	aerialWon: 1.1434 -> 1.1 totalShots: 0.6128 -> 0.6385 shotOffTarget: 0.4629 -> 0.4823 onTargetScoringAttempt: 0.1495 -> 0.1558
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.22 (squad xG sum 1.2773, xG scale 0.9512, xA scale 1.0781, team xA 0.79 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.215` `squad_xg_sum=1.2773` `squad_xa_sum=0.7326` `team_expected_assists=0.7898` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 80.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0373` `goal_rate=0.0481` `scorer_weight=0.0415`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0618` `assist_rate=0.0685` `provider_weight=0.0692`	-

Ao Tanaka M 78 0 0.10 0.05 105% 1 0.4 6.8 36.2 34 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0033, quality 0.592)	`market_value_eur=16200000` `quality_factor=0.5921` `boost_factor=1.0033` `reference_quality=0.505` `history_matches=35` `history_shrink=0.125`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0027)	`blend_weight=0.0027` `history_matches=35` `season_overlap_discount=1.0` `per90_to_match_factor=0.8642` `prior_reliability=0.0333` `competition=Premier League`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Adam Smith', 'matchup_zone': 'Right Wingback vs Left Back — Overlapping Run', 'duel_win_rate': 0.501, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.1478, 'skill_edge': 0.1005, 'height_edge': 0.0, 'attacker_skill': 0.6608, 'defender_skill': 0.4374, 'attacker_role': 'RWB', 'attacker_mv_price_index': 0.7, 'defender_adjusted_parts': [{'player': 'Adam Smith', 'market_value': 310000, 'role': 'LB', 'price_index': 0.7, 'adjusted_quality': 0.4374}], 'xg_override': 0.1072}	aerialWon: 0.2305 -> 1.1 totalShots: 1.0613 -> 1.2182 shotOffTarget: 0.7653 -> 0.8784 onTargetScoringAttempt: 0.296 -> 0.3397 bigChanceMissed: 0.0952 -> 0.1093 expectedGoals: 0.0934 -> 0.1072 goals: 0.0896 -> 0.1028 expectedGoalsOnTarget: 0.0401 -> 0.046
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.22 (squad xG sum 1.2773, xG scale 0.9512, xA scale 1.0781, team xA 0.79 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.215` `squad_xg_sum=1.2773` `squad_xa_sum=0.7326` `team_expected_assists=0.7898` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1072 -> 0.102
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 76.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.097` `goal_rate=0.1028` `scorer_weight=0.0991`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0586` `assist_rate=0.0537` `provider_weight=0.0591`	-

Gabriel Gudmundsson M 86 0 0.04 0.01 309% 1 0.3 6.8 40.0 34 FBref Prior M_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0063, quality 0.686)	`market_value_eur=26000000` `quality_factor=0.6858` `boost_factor=1.0063` `reference_quality=0.505` `history_matches=38` `history_shrink=0.1163`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Tyler Adams', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.451, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 1.001, 'skill_edge': -0.0233, 'height_edge': 0.024, 'attacker_skill': 0.7091, 'defender_skill': 0.761, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Tyler Adams', 'market_value': 31000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.761}], 'xg_override': 0.0409}	aerialWon: 0.5098 -> 1.1
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.22 (squad xG sum 1.2773, xG scale 0.9512, xA scale 1.0781, team xA 0.79 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.215` `squad_xg_sum=1.2773` `squad_xa_sum=0.7326` `team_expected_assists=0.7898` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0737 -> 0.0795
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 11 by projected rating, fifa_overall 78.0)	`match_squad_rank=11` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0382` `goal_rate=0.0484` `scorer_weight=0.0421`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0764` `assist_rate=0.0768` `provider_weight=0.0815`	-

Harry Wilson F 76 0 0.11 0.09 ok 1 1.2 6.7 45.8 38 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0052, quality 0.669)	`market_value_eur=24000000` `quality_factor=0.6685` `boost_factor=1.0052` `reference_quality=0.505` `history_matches=42` `history_shrink=0.1064`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0055)	`blend_weight=0.0055` `history_matches=42` `season_overlap_discount=1.0` `per90_to_match_factor=0.8424` `prior_reliability=0.08` `competition=Premier League`	interceptionWon: 0.3185 -> 0.3245
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Tyler Adams, António Silva', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle', 'duel_win_rate': 0.189, 'pressing_faced': 'HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.8808, 'skill_edge': -0.0491, 'height_edge': -0.032, 'attacker_skill': 0.6685, 'defender_skill': 0.7775, 'attacker_role': 'SS', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Tyler Adams', 'market_value': 31000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.761}, {'player': 'António Silva', 'market_value': 32000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7941}], 'xg_override': 0.1108}	totalShots: 1.317 -> 1.1601 shotOffTarget: 0.9299 -> 0.8191 onTargetScoringAttempt: 0.3871 -> 0.341 expectedGoals: 0.1258 -> 0.1108 goals: 0.1254 -> 0.1105 bigChanceMissed: 0.0772 -> 0.068 expectedGoalsOnTarget: 0.0758 -> 0.0668
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.22 (squad xG sum 1.2773, xG scale 0.9512, xA scale 1.0781, team xA 0.79 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.215` `squad_xg_sum=1.2773` `squad_xa_sum=0.7326` `team_expected_assists=0.7898` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1108 -> 0.1054 expected_assists: 0.0725 -> 0.0782
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 15 by projected rating, fifa_overall 81.0)	`match_squad_rank=15` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.1` `goal_rate=0.1105` `scorer_weight=0.1052`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0752` `assist_rate=0.0848` `provider_weight=0.0861`	-

Noah Okafor F 72 0 0.16 0.11 48% 2 0.9 6.8 54.2 27 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0090, quality 0.752)	`market_value_eur=35000000` `quality_factor=0.752` `boost_factor=1.009` `reference_quality=0.505` `history_matches=36` `history_shrink=0.122`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0053)	`blend_weight=0.0053` `history_matches=36` `season_overlap_discount=1.0` `per90_to_match_factor=0.7998` `prior_reliability=0.0667` `competition=Premier League`	totalCross: 1.9211 -> 1.9321 wonTackle: 0.697 -> 0.706
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Tyler Adams, António Silva', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle', 'duel_win_rate': 0.274, 'pressing_faced': 'HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 1.0066, 'skill_edge': -0.0115, 'height_edge': 0.016, 'attacker_skill': 0.752, 'defender_skill': 0.7775, 'attacker_role': 'SS', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Tyler Adams', 'market_value': 31000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.761}, {'player': 'António Silva', 'market_value': 32000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7941}], 'xg_override': 0.1709}	totalShots: 1.7167 -> 1.728 shotOffTarget: 1.09 -> 1.0972
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.22 (squad xG sum 1.2773, xG scale 0.9512, xA scale 1.0781, team xA 0.79 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.215` `squad_xg_sum=1.2773` `squad_xa_sum=0.7326` `team_expected_assists=0.7898` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1709 -> 0.1626 expected_assists: 0.0706 -> 0.0761
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 77.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.1501` `goal_rate=0.2311` `scorer_weight=0.2032`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0733` `assist_rate=0.0418` `provider_weight=0.0609`	-

Dominic Calvert-Lewin F 86 0 0.33 0.39 ok 2 1.6 6.8 54.7 33 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0059, quality 0.694)	`market_value_eur=27000000` `quality_factor=0.6941` `boost_factor=1.0059` `reference_quality=0.505` `history_matches=43` `history_shrink=0.1042`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0090)	`blend_weight=0.009` `history_matches=43` `season_overlap_discount=1.0` `per90_to_match_factor=0.9541` `prior_reliability=0.1333` `competition=Premier League`	totalOffside: 0.7592 -> 0.7653
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'António Silva, James Hill', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.9427, 'skill_edge': -0.045, 'height_edge': 0.006, 'attacker_skill': 0.6941, 'defender_skill': 0.7941, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'António Silva', 'market_value': 32000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7941}, {'player': 'James Hill', 'market_value': 32000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7941}], 'xg_override': 0.3436}	totalShots: 2.1225 -> 2.0009 shotOffTarget: 1.2461 -> 1.1747 onTargetScoringAttempt: 0.8748 -> 0.8247 bigChanceMissed: 0.53 -> 0.4996 expectedGoalsOnTarget: 0.4716 -> 0.4446 expectedGoals: 0.3645 -> 0.3436 goals: 0.3324 -> 0.3134
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.22 (squad xG sum 1.2773, xG scale 0.9512, xA scale 1.0781, team xA 0.79 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.215` `squad_xg_sum=1.2773` `squad_xa_sum=0.7326` `team_expected_assists=0.7898` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.3436 -> 0.3268
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 13 by projected rating, fifa_overall 79.0)	`match_squad_rank=13` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.2788` `goal_rate=0.3134` `scorer_weight=0.3017`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0387` `assist_rate=0.003` `provider_weight=0.0316`	-

Team totals 0 1.22 15 6.9

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Lucas Perri	G	4.7m	12.2%	Market-value substitute prior
Joe Rodon	D	14.3m	23.2%	Market-value substitute prior
Jaka Bijol	D	14.3m	23.1%	Market-value substitute prior
Anton Stach	M	23.1m	33.0%	Market-value substitute prior
Sean Longstaff	M	15.1m	70.3%	Market-value substitute prior
Jean-Mattéo Bahoya	F	26.2m	60.0%	Market-value substitute prior

Antagonist / tactical summary

Midfield is evenly contested — Tyler Adams vs Ethan Ampadu will be the pivotal duel. Total antagonist xG — Home: 1.27, Away: 0.94. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Midfield — possession 49.4% / 50.6%, dominance ratio 0.48, key duel: Tyler Adams vs Ethan Ampadu

Justin Kluivert xG 0.18Rayan xG 0.38Marcus Tavernier xG 0.36Dominic Calvert-Lewin xG 0.34Harry Wilson xG 0.11Noah Okafor xG 0.17

Summary

Antagonist matchup analysis for Bournemouth vs Leeds. Most likely scoreline: 1-0. Lambdas derived from 4 home and 7 away individual player duels. Head-to-head: Bournemouth 2W-2D-4L vs Leeds over 8 meetings.

Tactical parity between 4-2-3-1 and 3-4-2-1. Match outcome driven primarily by individual player quality. Midfield is evenly contested — Tyler Adams vs Ethan Ampadu will be the pivotal duel. Total antagonist xG — Home: 1.27, Away: 0.94. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Real Betis vs Málaga

LaLiga · 20261031 · 4-2-3-1 vs 3-4-2-1

Predicted 2-0Winner: Real BetisWinning H 2-0 (11.7%)Winning A 0-1 (4.2%)Confidence 0.79λ home/away 2.15 / 0.83Most-likely P 11.7%

Actual vs predicted: result not available yet (upcoming fixture).

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Real Betis vs Málaga. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 4-2-3-1 vs 3-4-2-1 (duel lambdas: Home 1.815, Away 0.718). Final calibrated Poisson lambdas: Home 2.153, Away 0.831 (most likely score 2-0).

Home Starters 11/11 XGBoost (10 FBref) Away Starters 11/11 XGBoost (5 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 1.81 / 0.72 Final λ (H/A) 2.15 / 0.83

FBref Team Priors: Real Betis (Gls/90: 1.3333333333333333, GA/90: 1.33 → E[λ]: 0.326) vs Málaga CF (Gls/90: 0.3333333333333333, GA/90: 0.33 → E[λ]: 0.328) | Blend Weight: 0.228

Modeling Pipeline Stages (5 layers)

1. XGBoost player regressions for volume & event rates with market-value quality scaling
2. Multi-source prior blending (FBref 2026-2027 tables + Sofascore season history)
3. Antagonist formation duel matrix resolving 1v1 and zonal matchups
4. Team-level prior calibration (FBref possession and attack/defense efficiencies)
5. Bivariate Poisson matrix with Dixon-Coles low-score correction

Standing & form context

Metric	Real Betis	Málaga
League position	4.00	20.00
Points per game	1.76	0.82
Attack index	1.03	0.76
Defense (concede) index	1.12	0.97
Form multiplier	1.04	0.96
Standing λ multiplier	1.07	0.88

Lambda decomposition (how the goal rates were built)

Stage	Real Betis λ	Málaga λ
FBref Squad Prior E[λ] (weight 0.228)	0.326	0.328
Pre-mismatch	1.841	0.708
Pre-calibration (Micro-duels)	1.815	0.718
Final λ	2.153	0.831

competition calibration scale 1.171competition observed total goals 2.984quality gap applied Truemismatch score pressure applied Truecompetitive underdog floor applied Falsestage discount 1.0

Scoreline distribution & markets

Top 5 scorelines	P
2-0	11.7%
1-0	10.8%
2-1	9.7%
1-1	9.1%
3-0	8.4%

Home win	67.7%
Draw	19.1%
Away win	13.2%
Over 2.5	57.3%
BTTS yes	49.9%

Pregame fair line — favorite Real Betis (67.7%), fair total 2.98, fair spread 1.32

Projected players vs season baseline

Proj = projected for this match · Season = this season's per-game average from match data · deviation badge compares projected xG to the player's season xG/game.

Real Betis — projected line vs season baseline (4-2-3-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Fran García D 89 0 0.02 0.02 ok 0 0.2 6.7 30.8 33 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9989, quality 0.485)	`market_value_eur=6900000` `quality_factor=0.4854` `boost_factor=0.9989` `reference_quality=0.505` `history_matches=21` `history_shrink=0.1923`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0200)	`blend_weight=0.02` `history_matches=21` `season_overlap_discount=1.0` `per90_to_match_factor=0.9862` `prior_reliability=0.16` `competition=La Liga`	totalCross: 2.6533 -> 2.6906 totalShots: 0.4127 -> 0.4044 wonTackle: 1.6366 -> 1.6285 goalAssist: 0.0268 -> 0.0345
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'David Larrubia', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.15 (squad xG sum 1.3311, xG scale 1.6175, xA scale 1.6175, team xA 1.53 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=2.153` `squad_xg_sum=1.3311` `squad_xa_sum=0.9447` `team_expected_assists=1.528` `team_xa_to_xg_ratio=0.7097`	expected_goals: 0.0117 -> 0.0189 expected_assists: 0.0498 -> 0.0806
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 16 by projected rating, fifa_overall 77.0)	`match_squad_rank=16` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0187` `goal_rate=0.0` `scorer_weight=0.0153`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0774` `assist_rate=0.0345` `provider_weight=0.0682`	-

Marc Roca M 86 0 0.07 0.02 185% 1 0.5 7.0 48.9 34 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9981, quality 0.452)	`market_value_eur=4200000` `quality_factor=0.4523` `boost_factor=0.9981` `reference_quality=0.505` `history_matches=36` `history_shrink=0.122`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0090)	`blend_weight=0.009` `history_matches=36` `season_overlap_discount=1.0` `per90_to_match_factor=0.9509` `prior_reliability=0.1133` `competition=La Liga`	totalShots: 0.5697 -> 0.5794 wonTackle: 0.8805 -> 0.8877 fouls: 1.8208 -> 1.8145
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'David Larrubia, Adrián Niño, Dani Lorenzo', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.15 (squad xG sum 1.3311, xG scale 1.6175, xA scale 1.6175, team xA 1.53 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=2.153` `squad_xg_sum=1.3311` `squad_xa_sum=0.9447` `team_expected_assists=1.528` `team_xa_to_xg_ratio=0.7097`	expected_goals: 0.0406 -> 0.0657 expected_assists: 0.071 -> 0.1148
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 78.0)	`match_squad_rank=6` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0636` `goal_rate=0.0439` `scorer_weight=0.0681`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1085` `assist_rate=0.0636` `provider_weight=0.1078`	-

Pablo Fornals M 80 0 0.12 0.06 101% 1 0.9 7.1 59.9 34 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0000, quality 0.503)	`market_value_eur=8400000` `quality_factor=0.5035` `boost_factor=1.0` `reference_quality=0.505` `history_matches=46` `history_shrink=0.098`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0093)	`blend_weight=0.0093` `history_matches=46` `season_overlap_discount=1.0` `per90_to_match_factor=0.884` `prior_reliability=0.1467` `competition=La Liga`	totalCross: 2.7633 -> 2.775 totalShots: 1.1304 -> 1.1419 interceptionWon: 0.6804 -> 0.6891 wasFouled: 0.8037 -> 0.7962
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Carlos Dotor', 'matchup_zone': 'Midfield lane'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.15 (squad xG sum 1.3311, xG scale 1.6175, xA scale 1.6175, team xA 1.53 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=2.153` `squad_xg_sum=1.3311` `squad_xa_sum=0.9447` `team_expected_assists=1.528` `team_xa_to_xg_ratio=0.7097`	expected_goals: 0.0744 -> 0.1203 expected_assists: 0.1803 -> 0.2916
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 85.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1133` `goal_rate=0.0943` `scorer_weight=0.1395`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.2529` `assist_rate=0.1056` `provider_weight=0.2333`	-

Antony F 77 0 0.30 0.20 48% 3 2.7 7.1 58.7 41 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0084, quality 0.764)	`market_value_eur=37000000` `quality_factor=0.7641` `boost_factor=1.0084` `reference_quality=0.505` `history_matches=41` `history_shrink=0.1087`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0141)	`blend_weight=0.0141` `history_matches=41` `season_overlap_discount=1.0` `per90_to_match_factor=0.8556` `prior_reliability=0.2` `competition=La Liga`	totalShots: 2.4151 -> 2.4414 totalCross: 1.5948 -> 1.6085 onTargetScoringAttempt: 0.9028 -> 0.9102 wasFouled: 1.2882 -> 1.2942
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Einar Galilea, Fernando Calero', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.305, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.1544, 'skill_edge': 0.147, 'height_edge': -0.042, 'attacker_skill': 0.7641, 'defender_skill': 0.4374, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Einar Galilea', 'market_value': 365000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'Fernando Calero', 'market_value': 2100000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.1875}	totalShots: 2.4414 -> 2.8184 shotOffTarget: 1.5123 -> 1.7458 onTargetScoringAttempt: 0.9102 -> 1.0507 bigChanceMissed: 0.3415 -> 0.3942 expectedGoalsOnTarget: 0.2475 -> 0.2857 goals: 0.2053 -> 0.237 expectedGoals: 0.1624 -> 0.1875
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.15 (squad xG sum 1.3311, xG scale 1.6175, xA scale 1.6175, team xA 1.53 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=2.153` `squad_xg_sum=1.3311` `squad_xa_sum=0.9447` `team_expected_assists=1.528` `team_xa_to_xg_ratio=0.7097`	expected_goals: 0.1875 -> 0.3033 expected_assists: 0.1197 -> 0.1936
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 81.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2616` `goal_rate=0.237` `scorer_weight=0.3665`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.176` `assist_rate=0.0446` `provider_weight=0.1549`	-

Isco M 68 0 0.26 0.18 42% 2 1.1 7.1 55.1 20 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9959, quality 0.445)	`market_value_eur=3600000` `quality_factor=0.4449` `boost_factor=0.9959` `reference_quality=0.505` `history_matches=17` `history_shrink=0.2273`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0128)	`blend_weight=0.0128` `history_matches=17` `season_overlap_discount=1.0` `per90_to_match_factor=0.7599` `prior_reliability=0.0867` `competition=La Liga`	fouls: 0.5226 -> 0.5459 totalCross: 2.8878 -> 2.9032 onTargetScoringAttempt: 0.5218 -> 0.5368
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Einar Galilea', 'matchup_zone': 'LW vs CB', 'duel_win_rate': 0.399, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9696, 'skill_edge': 0.0034, 'height_edge': -0.024, 'attacker_skill': 0.4449, 'defender_skill': 0.4374, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Einar Galilea', 'market_value': 365000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.1619}`	aerialWon: 0.0 -> 0.9 totalShots: 1.6959 -> 1.6444 shotOffTarget: 1.1591 -> 1.1239 onTargetScoringAttempt: 0.5368 -> 0.5205 expectedGoalsOnTarget: 0.4394 -> 0.426 bigChanceMissed: 0.1956 -> 0.1897 expectedGoals: 0.167 -> 0.1619
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.15 (squad xG sum 1.3311, xG scale 1.6175, xA scale 1.6175, team xA 1.53 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=2.153` `squad_xg_sum=1.3311` `squad_xa_sum=0.9447` `team_expected_assists=1.528` `team_xa_to_xg_ratio=0.7097`	expected_goals: 0.1619 -> 0.2619 expected_assists: 0.1602 -> 0.2591
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 83.0)	`match_squad_rank=5` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2304` `goal_rate=0.1473` `scorer_weight=0.2439`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.2283` `assist_rate=0.1198` `provider_weight=0.2093`	-

Abdessamad Ezzalzouli F 84 1 0.61 0.24 154% 3 1.9 7.2 61.7 32 XGBoost F_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0106, quality 0.796)	`market_value_eur=43000000` `quality_factor=0.7957` `boost_factor=1.0106` `reference_quality=0.505` `history_matches=36` `history_shrink=0.122`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Einar Galilea, Fernando Calero', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.339, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.2046, 'skill_edge': 0.1612, 'height_edge': -0.022, 'attacker_skill': 0.7957, 'defender_skill': 0.4374, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Einar Galilea', 'market_value': 365000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'Fernando Calero', 'market_value': 2100000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.3778}	totalShots: 2.2172 -> 2.6708 onTargetScoringAttempt: 1.1331 -> 1.3649 shotOffTarget: 1.0841 -> 1.3059 expectedGoalsOnTarget: 0.4353 -> 0.5244 goals: 0.3357 -> 0.4044 expectedGoals: 0.3136 -> 0.3778 bigChanceMissed: 0.1607 -> 0.1936
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.15 (squad xG sum 1.3311, xG scale 1.6175, xA scale 1.6175, team xA 1.53 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=2.153` `squad_xg_sum=1.3311` `squad_xa_sum=0.9447` `team_expected_assists=1.528` `team_xa_to_xg_ratio=0.7097`	expected_goals: 0.3778 -> 0.6112 expected_assists: 0.1931 -> 0.3125
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 86.0)	`match_squad_rank=1` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.4573` `goal_rate=0.4044` `scorer_weight=0.6492`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=1` `assist_probability=0.2684` `assist_rate=0.0952` `provider_weight=0.25`	-

Troy Parrott F 73 1 0.44 0.14 214% 2 0.7 6.7 38.2 12 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0138, quality 0.615)	`market_value_eur=18400000` `quality_factor=0.615` `boost_factor=1.0138` `reference_quality=0.505` `history_matches=7` `history_shrink=0.4167`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0144)	`blend_weight=0.0144` `history_matches=7` `season_overlap_discount=1.0` `per90_to_match_factor=0.8061` `prior_reliability=0.0533` `competition=La Liga`	wonTackle: 0.2831 -> 0.3227 totalShots: 1.5424 -> 1.5614 fouls: 0.0994 -> 0.1125 totalOffside: 0.0994 -> 0.1125 onTargetScoringAttempt: 0.6424 -> 0.6331
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Einar Galilea, Fernando Calero', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.29, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.1322, 'skill_edge': 0.0799, 'height_edge': 0.01, 'attacker_skill': 0.615, 'defender_skill': 0.4374, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Einar Galilea', 'market_value': 365000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'Fernando Calero', 'market_value': 2100000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.2727}	totalShots: 1.5614 -> 1.7678 shotOffTarget: 0.9 -> 1.019 onTargetScoringAttempt: 0.6331 -> 0.7168 bigChanceMissed: 0.4849 -> 0.549 goals: 0.2582 -> 0.2923 expectedGoals: 0.2409 -> 0.2727 expectedGoalsOnTarget: 0.0973 -> 0.1102
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.15 (squad xG sum 1.3311, xG scale 1.6175, xA scale 1.6175, team xA 1.53 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=2.153` `squad_xg_sum=1.3311` `squad_xa_sum=0.9447` `team_expected_assists=1.528` `team_xa_to_xg_ratio=0.7097`	expected_goals: 0.2727 -> 0.4411 expected_assists: 0.0588 -> 0.0951
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 17 by projected rating, fifa_overall 78.0)	`match_squad_rank=17` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.3567` `goal_rate=0.2923` `scorer_weight=0.4679`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0907` `assist_rate=0.1355` `provider_weight=0.1189`	-

Natan D 89 0 0.12 0.02 404% 1 0.2 6.9 49.5 33 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0074, quality 0.732)	`market_value_eur=32000000` `quality_factor=0.732` `boost_factor=1.0074` `reference_quality=0.505` `history_matches=41` `history_shrink=0.1087`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0094)	`blend_weight=0.0094` `history_matches=41` `season_overlap_discount=1.0` `per90_to_match_factor=0.9917` `prior_reliability=0.1333` `competition=La Liga`	wonTackle: 0.709 -> 0.721 interceptionWon: 0.8995 -> 0.891 wasFouled: 0.6377 -> 0.6317
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'David Larrubia', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.15 (squad xG sum 1.3311, xG scale 1.6175, xA scale 1.6175, team xA 1.53 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=2.153` `squad_xg_sum=1.3311` `squad_xa_sum=0.9447` `team_expected_assists=1.528` `team_xa_to_xg_ratio=0.7097`	expected_goals: 0.0729 -> 0.1179 expected_assists: 0.0115 -> 0.0186
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 79.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1112` `goal_rate=0.0786` `scorer_weight=0.1234`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0184` `assist_rate=0.0` `provider_weight=0.0151`	-

Diego Llorente D 81 0 0.02 0.01 107% 0 0.1 6.8 49.8 28 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9964, quality 0.437)	`market_value_eur=2600000` `quality_factor=0.4374` `boost_factor=0.9964` `reference_quality=0.505` `history_matches=23` `history_shrink=0.1786`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0077)	`blend_weight=0.0077` `history_matches=23` `season_overlap_discount=1.0` `per90_to_match_factor=0.8962` `prior_reliability=0.0667` `competition=La Liga`	interceptionWon: 1.1667 -> 1.1924 wonTackle: 0.7537 -> 0.7479
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=true`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Adrián Niño', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.15 (squad xG sum 1.3311, xG scale 1.6175, xA scale 1.6175, team xA 1.53 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=2.153` `squad_xg_sum=1.3311` `squad_xa_sum=0.9447` `team_expected_assists=1.528` `team_xa_to_xg_ratio=0.7097`	expected_goals: 0.0135 -> 0.0218
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 79.0)	`match_squad_rank=8` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0216` `goal_rate=0.0` `scorer_weight=0.0174`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0055` `assist_rate=0.0` `provider_weight=0.0051`	-

Angel Ortiz D 73 0 0.02 0.00 419% 0 0.1 6.8 29.8 25 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9984, quality 0.482)	`market_value_eur=6600000` `quality_factor=0.4818` `boost_factor=0.9984` `reference_quality=0.505` `history_matches=17` `history_shrink=0.2273`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0079)	`blend_weight=0.0079` `history_matches=17` `season_overlap_discount=1.0` `per90_to_match_factor=0.8158` `prior_reliability=0.0533` `competition=La Liga`	totalCross: 1.9026 -> 1.8876 interceptionWon: 0.8756 -> 0.8687 wasFouled: 0.6816 -> 0.6762
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Ramón Enríquez', 'matchup_zone': 'Left Wingback vs Right Back — Overlapping Run'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.15 (squad xG sum 1.3311, xG scale 1.6175, xA scale 1.6175, team xA 1.53 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=2.153` `squad_xg_sum=1.3311` `squad_xa_sum=0.9447` `team_expected_assists=1.528` `team_xa_to_xg_ratio=0.7097`	expected_goals: 0.0099 -> 0.016 expected_assists: 0.0553 -> 0.0894
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 13 by projected rating, fifa_overall 70.0)	`match_squad_rank=13` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0159` `goal_rate=0.0` `scorer_weight=0.0134`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0855` `assist_rate=0.0` `provider_weight=0.0715`	-

Álvaro Valles G 90 0 0.00 - n/a 0.0 7.1 61.1 34 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0003, quality 0.514)	`market_value_eur=9300000` `quality_factor=0.5142` `boost_factor=1.0003` `reference_quality=0.505` `history_matches=38` `history_shrink=0.1163`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0151)	`blend_weight=0.0151` `history_matches=38` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.2` `competition=La Liga`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Adrián Niño, David Larrubia', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 0.7182}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.15 (squad xG sum 1.3311, xG scale 1.6175, xA scale 1.6175, team xA 1.53 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=2.153` `squad_xg_sum=1.3311` `squad_xa_sum=0.9447` `team_expected_assists=1.528` `team_xa_to_xg_ratio=0.7097`	shots_faced: 6.8 -> 7.9 expected_saves: 1.66 -> 1.93 high_pressure_saves: 0.57 -> 0.66
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 77.0)	`match_squad_rank=4` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=2.2809`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=1.2992`	-

Team totals 2 2.15 15 7.0

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Diego Conde	G	4.4m	10.9%	Market-value substitute prior
Valentín Gómez	D	14.0m	22.5%	Market-value substitute prior
Aitor Ruibal	D	13.6m	30.1%	Market-value substitute prior
Sergi Altimira	M	22.7m	53.2%	Market-value substitute prior
Nelson Deossa	M	19.9m	57.0%	Market-value substitute prior
Cucho Hernández	F	25.9m	58.7%	Market-value substitute prior

Málaga — projected line vs season baseline (3-4-2-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Alfonso Herrero G 90 0 0.00 - n/a 0.0 7.0 54.4 13 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9932, quality 0.437)	`market_value_eur=420000` `quality_factor=0.4374` `boost_factor=0.9932` `reference_quality=0.505` `history_matches=10` `history_shrink=0.3333`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0433)	`blend_weight=0.0433` `history_matches=10` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.2` `competition=La Liga`	saves: 2.8871 -> 2.9353 wasFouled: 0.1963 -> 0.1878
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Abdessamad Ezzalzouli, Troy Parrott', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.8149}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 0.83 (squad xG sum 0.7402, xG scale 1.1227, xA scale 1.1227, team xA 0.66 = 0.79 x lambda)	`scoreline_reset=true` `team_lambda=0.831` `squad_xg_sum=0.7402` `squad_xa_sum=0.5844` `team_expected_assists=0.6561` `team_xa_to_xg_ratio=0.7895`	shots_faced: 17.2 -> 20.5 expected_saves: 4.21 -> 5.02 high_pressure_saves: 1.45 -> 1.72
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 75.0)	`match_squad_rank=1` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=0.8662`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=0.5932`	-

Einar Galilea D 82 0 0.00 - n/a 0 0.0 6.9 36.5 13 XGBoost D_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9916, quality 0.437)	`market_value_eur=365000` `quality_factor=0.4374` `boost_factor=0.9916` `reference_quality=0.505` `history_matches=7` `history_shrink=0.4167`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Isco', 'matchup_zone': 'LW vs CB'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 0.83 (squad xG sum 0.7402, xG scale 1.1227, xA scale 1.1227, team xA 0.66 = 0.79 x lambda)	`scoreline_reset=true` `team_lambda=0.831` `squad_xg_sum=0.7402` `squad_xa_sum=0.5844` `team_expected_assists=0.6561` `team_xa_to_xg_ratio=0.7895`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 69.0)	`match_squad_rank=4` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0` `goal_rate=0.0` `scorer_weight=0.001`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0063` `assist_rate=0.0` `provider_weight=0.0057`	-

Fernando Calero D 85 0 0.04 0.02 67% 0 0.3 6.8 49.8 23 XGBoost D_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9971, quality 0.437)	`market_value_eur=2100000` `quality_factor=0.4374` `boost_factor=0.9971` `reference_quality=0.505` `history_matches=30` `history_shrink=0.1429`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Abdessamad Ezzalzouli', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 0.83 (squad xG sum 0.7402, xG scale 1.1227, xA scale 1.1227, team xA 0.66 = 0.79 x lambda)	`scoreline_reset=true` `team_lambda=0.831` `squad_xg_sum=0.7402` `squad_xa_sum=0.5844` `team_expected_assists=0.6561` `team_xa_to_xg_ratio=0.7895`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 11 by projected rating, fifa_overall 81.0)	`match_squad_rank=11` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0385` `goal_rate=0.0` `scorer_weight=0.0314`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0131` `assist_rate=0.0` `provider_weight=0.0107`	-

Carlos Puga D 77 0 0.00 0.00 76% 0 0.1 6.8 48.0 13 XGBoost D_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9928, quality 0.437)	`market_value_eur=1000000` `quality_factor=0.4374` `boost_factor=0.9928` `reference_quality=0.505` `history_matches=9` `history_shrink=0.3571`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Abdessamad Ezzalzouli', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 0.83 (squad xG sum 0.7402, xG scale 1.1227, xA scale 1.1227, team xA 0.66 = 0.79 x lambda)	`scoreline_reset=true` `team_lambda=0.831` `squad_xg_sum=0.7402` `squad_xa_sum=0.5844` `team_expected_assists=0.6561` `team_xa_to_xg_ratio=0.7895`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 78.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0021` `goal_rate=0.0` `scorer_weight=0.002`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0309` `assist_rate=0.0` `provider_weight=0.0251`	-

Carlos Dotor M 80 0 0.07 0.04 58% 1 0.5 6.9 52.8 13 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9928, quality 0.437)	`market_value_eur=1200000` `quality_factor=0.4374` `boost_factor=0.9928` `reference_quality=0.505` `history_matches=9` `history_shrink=0.3571`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0387)	`blend_weight=0.0387` `history_matches=9` `season_overlap_discount=1.0` `per90_to_match_factor=0.8833` `prior_reliability=0.1667` `competition=La Liga`	fouls: 1.3689 -> 1.3296 wasFouled: 0.8002 -> 0.7692 wonTackle: 1.3961 -> 1.4241 totalShots: 0.8266 -> 0.8079 totalCross: 0.4866 -> 0.4951
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Marc Roca', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.423, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9606, 'skill_edge': -0.0108, 'height_edge': -0.016, 'attacker_skill': 0.4374, 'defender_skill': 0.4615, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Marc Roca', 'market_value': 4200000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.4615}], 'xg_override': 0.0618}	aerialWon: 0.2137 -> 1.1 totalShots: 0.8298 -> 0.7971 shotOffTarget: 0.5637 -> 0.5415 onTargetScoringAttempt: 0.2661 -> 0.2556 expectedGoalsOnTarget: 0.1542 -> 0.1481
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 0.83 (squad xG sum 0.7402, xG scale 1.1227, xA scale 1.1227, team xA 0.66 = 0.79 x lambda)	`scoreline_reset=true` `team_lambda=0.831` `squad_xg_sum=0.7402` `squad_xa_sum=0.5844` `team_expected_assists=0.6561` `team_xa_to_xg_ratio=0.7895`	expected_goals: 0.0618 -> 0.0694 expected_assists: 0.0623 -> 0.0699
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 85.0)	`match_squad_rank=6` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.067` `goal_rate=0.0702` `scorer_weight=0.0746`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0675` `assist_rate=0.0594` `provider_weight=0.0681`	-

Ramón Enríquez M 80 0 0.04 - n/a 0 0.0 6.9 41.9 13 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9928, quality 0.437)	`market_value_eur=380000` `quality_factor=0.4374` `boost_factor=0.9928` `reference_quality=0.505` `history_matches=9` `history_shrink=0.3571`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0031)	`blend_weight=0.0031` `history_matches=9` `season_overlap_discount=1.0` `per90_to_match_factor=0.8889` `prior_reliability=0.0133` `competition=La Liga`	fouls: 0.1719 -> 0.1851 wasFouled: 1.2587 -> 1.2685
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Angel Ortiz', 'matchup_zone': 'Left Wingback vs Right Back — Overlapping Run', 'duel_win_rate': 0.389, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9835, 'skill_edge': -0.0352, 'height_edge': 0.024, 'attacker_skill': 0.4374, 'defender_skill': 0.5157, 'attacker_role': 'LWB', 'attacker_mv_price_index': 0.7, 'defender_adjusted_parts': [{'player': 'Angel Ortiz', 'market_value': 6600000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.5157}], 'xg_override': 0.0349}	aerialWon: 0.1938 -> 0.9 totalShots: 0.4079 -> 0.4012
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 0.83 (squad xG sum 0.7402, xG scale 1.1227, xA scale 1.1227, team xA 0.66 = 0.79 x lambda)	`scoreline_reset=true` `team_lambda=0.831` `squad_xg_sum=0.7402` `squad_xa_sum=0.5844` `team_expected_assists=0.6561` `team_xa_to_xg_ratio=0.7895`	expected_assists: 0.0431 -> 0.0484
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 65.0)	`match_squad_rank=5` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0384` `goal_rate=0.0595` `scorer_weight=0.049`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0472` `assist_rate=0.0679` `provider_weight=0.0605`	-

Adam Aznou D 83 0 0.07 0.01 407% 1 0.1 6.7 18.0 14 XGBoost D_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9947, quality 0.477)	`market_value_eur=6200000` `quality_factor=0.4769` `boost_factor=0.9947` `reference_quality=0.505` `history_matches=3` `history_shrink=0.625`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Troy Parrott', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 0.83 (squad xG sum 0.7402, xG scale 1.1227, xA scale 1.1227, team xA 0.66 = 0.79 x lambda)	`scoreline_reset=true` `team_lambda=0.831` `squad_xg_sum=0.7402` `squad_xa_sum=0.5844` `team_expected_assists=0.6561` `team_xa_to_xg_ratio=0.7895`	expected_goals: 0.0636 -> 0.0714 expected_assists: 0.0622 -> 0.0698
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 15 by projected rating, fifa_overall 67.0)	`match_squad_rank=15` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0689` `goal_rate=0.0` `scorer_weight=0.0571`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0674` `assist_rate=0.0` `provider_weight=0.0558`	-

David Larrubia F 85 0 0.08 0.05 70% 1 1.1 7.0 51.0 21 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9973, quality 0.478)	`market_value_eur=6300000` `quality_factor=0.4781` `boost_factor=0.9973` `reference_quality=0.505` `history_matches=10` `history_shrink=0.3333`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0347)	`blend_weight=0.0347` `history_matches=10` `season_overlap_discount=1.0` `per90_to_match_factor=0.9467` `prior_reliability=0.16` `competition=La Liga`	totalCross: 1.5693 -> 1.5149 totalShots: 1.2176 -> 1.2715 wonTackle: 1.8489 -> 1.8805 onTargetScoringAttempt: 0.4049 -> 0.4184 totalOffside: 0.0624 -> 0.0739 fouls: 0.7026 -> 0.6919 interceptionWon: 0.5347 -> 0.5435 wasFouled: 1.039 -> 1.0303
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Marc Roca, Natan', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 0.8187, 'skill_edge': -0.0674, 'height_edge': -0.056, 'attacker_skill': 0.4781, 'defender_skill': 0.6278, 'attacker_role': 'SS', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Marc Roca', 'market_value': 4200000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.4615}, {'player': 'Natan', 'market_value': 32000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7941}], 'xg_override': 0.0726}	totalShots: 1.2715 -> 1.041 shotOffTarget: 0.8127 -> 0.6654 onTargetScoringAttempt: 0.4184 -> 0.3425 goals: 0.1101 -> 0.0901 expectedGoals: 0.0887 -> 0.0726 expectedGoalsOnTarget: 0.0425 -> 0.0348
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 0.83 (squad xG sum 0.7402, xG scale 1.1227, xA scale 1.1227, team xA 0.66 = 0.79 x lambda)	`scoreline_reset=true` `team_lambda=0.831` `squad_xg_sum=0.7402` `squad_xa_sum=0.5844` `team_expected_assists=0.6561` `team_xa_to_xg_ratio=0.7895`	expected_goals: 0.0726 -> 0.0815 expected_assists: 0.1934 -> 0.2174
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 81.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0783` `goal_rate=0.0901` `scorer_weight=0.0931`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.1954` `assist_rate=0.0` `provider_weight=0.1739`	-

Dani Lorenzo M 81 0 0.05 0.02 119% 1 0.2 6.9 51.6 12 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9928, quality 0.437)	`market_value_eur=2300000` `quality_factor=0.4374` `boost_factor=0.9928` `reference_quality=0.505` `history_matches=9` `history_shrink=0.3571`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0279)	`blend_weight=0.0279` `history_matches=9` `season_overlap_discount=1.0` `per90_to_match_factor=0.8971` `prior_reliability=0.12` `competition=La Liga`	fouls: 1.3615 -> 1.3236 wonTackle: 1.73 -> 1.6957 totalShots: 0.6176 -> 0.6004 interceptionWon: 0.4863 -> 0.5005 totalCross: 0.3776 -> 0.3671 onTargetScoringAttempt: 0.21 -> 0.2041
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Marc Roca', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.407, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9371, 'skill_edge': -0.0108, 'height_edge': -0.032, 'attacker_skill': 0.4374, 'defender_skill': 0.4615, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Marc Roca', 'market_value': 4200000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.4615}], 'xg_override': 0.043}	aerialWon: 0.1863 -> 0.9 totalShots: 0.6117 -> 0.5733 shotOffTarget: 0.4076 -> 0.382 onTargetScoringAttempt: 0.2041 -> 0.1913 bigChanceMissed: 0.1361 -> 0.1275
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 0.83 (squad xG sum 0.7402, xG scale 1.1227, xA scale 1.1227, team xA 0.66 = 0.79 x lambda)	`scoreline_reset=true` `team_lambda=0.831` `squad_xg_sum=0.7402` `squad_xa_sum=0.5844` `team_expected_assists=0.6561` `team_xa_to_xg_ratio=0.7895`	expected_goals: 0.043 -> 0.0483 expected_assists: 0.0582 -> 0.0653
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 72.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0472` `goal_rate=0.0516` `scorer_weight=0.0528`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0632` `assist_rate=0.0647` `provider_weight=0.0692`	-

Adrián Niño F 80 0 0.30 0.11 182% 2 1.2 6.8 48.3 9 XGBoost F_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9908, quality 0.437)	`market_value_eur=1900000` `quality_factor=0.4374` `boost_factor=0.9908` `reference_quality=0.505` `history_matches=6` `history_shrink=0.4545`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Marc Roca, Natan', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle', 'duel_win_rate': 0.184, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 0.8741, 'skill_edge': -0.0857, 'height_edge': 0.0, 'attacker_skill': 0.4374, 'defender_skill': 0.6278, 'attacker_role': 'SS', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Marc Roca', 'market_value': 4200000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.4615}, {'player': 'Natan', 'market_value': 32000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7941}], 'xg_override': 0.2684}	totalShots: 2.3916 -> 2.0905 shotOffTarget: 1.4548 -> 1.2716 onTargetScoringAttempt: 0.9368 -> 0.8189 expectedGoalsOnTarget: 0.4226 -> 0.3694 goals: 0.3289 -> 0.2875 expectedGoals: 0.3071 -> 0.2684 bigChanceMissed: 0.238 -> 0.208
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 0.83 (squad xG sum 0.7402, xG scale 1.1227, xA scale 1.1227, team xA 0.66 = 0.79 x lambda)	`scoreline_reset=true` `team_lambda=0.831` `squad_xg_sum=0.7402` `squad_xa_sum=0.5844` `team_expected_assists=0.6561` `team_xa_to_xg_ratio=0.7895`	expected_goals: 0.2684 -> 0.3014
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 10 by projected rating, fifa_overall 72.0)	`match_squad_rank=10` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.2602` `goal_rate=0.2875` `scorer_weight=0.3195`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0424` `assist_rate=0.0` `provider_weight=0.0346`	-

Rafa Garrido D 88 0 0.01 0.00 78% 0 0.2 6.8 51.2 13 XGBoost D_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9932, quality 0.437)	`market_value_eur=1400000` `quality_factor=0.4374` `boost_factor=0.9932` `reference_quality=0.505` `history_matches=10` `history_shrink=0.3333`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Antony', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 0.83 (squad xG sum 0.7402, xG scale 1.1227, xA scale 1.1227, team xA 0.66 = 0.79 x lambda)	`scoreline_reset=true` `team_lambda=0.831` `squad_xg_sum=0.7402` `squad_xa_sum=0.5844` `team_expected_assists=0.6561` `team_xa_to_xg_ratio=0.7895`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 12 by projected rating, fifa_overall 79.0)	`match_squad_rank=12` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0083` `goal_rate=0.0` `scorer_weight=0.0072`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0267` `assist_rate=0.14` `provider_weight=0.0339`	-

Team totals 0 0.83 9 6.9

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Carlos Lopez	G	4.4m	10.9%	Market-value substitute prior
Jose Salinas	D	13.6m	29.9%	Market-value substitute prior
Diego Murillo	D	13.6m	29.9%	Market-value substitute prior
Jens-Lys Cajuste	M	22.1m	50.7%	Market-value substitute prior
Izan Merino	M	22.1m	50.3%	Market-value substitute prior
Chupete	F	25.7m	57.6%	Market-value substitute prior

Antagonist / tactical summary

Midfield is evenly contested — Pablo Fornals vs Dani Lorenzo will be the pivotal duel. Total antagonist xG — Home: 1.00, Away: 0.48. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Shadow Striker vs CDM + CB — Zone 14 Battle.

Midfield — possession 50.4% / 49.6%, dominance ratio 0.52, key duel: Pablo Fornals vs Dani Lorenzo

Isco xG 0.16Abdessamad Ezzalzouli xG 0.38Antony xG 0.19David Larrubia xG 0.07Adrián Niño xG 0.27Ramón Enríquez xG 0.03

Summary

Antagonist matchup analysis for Real Betis vs Málaga. Most likely scoreline: 2-0. Lambdas derived from 4 home and 5 away individual player duels. Head-to-head: Real Betis 6W-0D-4L vs Málaga over 10 meetings.

Tactical parity between 4-2-3-1 and 3-4-2-1. Match outcome driven primarily by individual player quality. Midfield is evenly contested — Pablo Fornals vs Dani Lorenzo will be the pivotal duel. Total antagonist xG — Home: 1.00, Away: 0.48. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Shadow Striker vs CDM + CB — Zone 14 Battle.

Chelsea vs Man Utd

Premier League · 20261031 · 4-2-3-1 vs 4-2-3-1

Predicted 2-1Winner: ChelseaWinning H 2-1 (9.0%)Winning A 1-2 (7.7%)Confidence 0.71λ home/away 1.66 / 1.42Most-likely P 9.0%

Actual vs predicted: result not available yet (upcoming fixture).

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Chelsea vs Man Utd. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 4-2-3-1 vs 4-2-3-1 (duel lambdas: Home 1.810, Away 1.692). Final calibrated Poisson lambdas: Home 1.661, Away 1.422 (most likely score 2-1).

Home Starters 11/11 XGBoost (7 FBref) Away Starters 11/11 XGBoost (8 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 1.81 / 1.69 Final λ (H/A) 1.66 / 1.42

FBref Team Priors: Chelsea (Gls/90: 3.5, GA/90: 3.5 → E[λ]: 3.5) vs Manchester United (Gls/90: 2.0, GA/90: 2.5 → E[λ]: 3.5) | Blend Weight: 0.192

Modeling Pipeline Stages (5 layers)

1. XGBoost player regressions for volume & event rates with market-value quality scaling
2. Multi-source prior blending (FBref 2026-2027 tables + Sofascore season history)
3. Antagonist formation duel matrix resolving 1v1 and zonal matchups
4. Team-level prior calibration (FBref possession and attack/defense efficiencies)
5. Bivariate Poisson matrix with Dixon-Coles low-score correction

Standing & form context

Metric	Chelsea	Man Utd
League position	5.00	16.00
Points per game	1.51	1.08
Attack index	1.21	0.96
Defense (concede) index	1.07	1.14
Form multiplier	1.02	0.98
Standing λ multiplier	1.12	1.03

Lambda decomposition (how the goal rates were built)

Stage	Chelsea λ	Man Utd λ
FBref Squad Prior E[λ] (weight 0.192)	3.500	3.500
Pre-competitive-floor	1.963	1.664
Pre-calibration (Micro-duels)	1.810	1.692
Final λ	1.661	1.422

competition calibration scale 0.85competition observed total goals 2.781quality gap applied Falsemismatch score pressure applied Falsecompetitive underdog floor applied Falsestage discount 1.0

Scoreline distribution & markets

Top 5 scorelines	P
1-1	10.9%
2-1	9.0%
1-2	7.7%
1-0	7.5%
0-1	6.4%

Home win	43.2%
Draw	23.9%
Away win	32.9%
Over 2.5	59.5%
BTTS yes	61.5%

Pregame fair line — favorite Chelsea (43.2%), fair total 3.08, fair spread 0.24

Projected players vs season baseline

Proj = projected for this match · Season = this season's per-game average from match data · deviation badge compares projected xG to the player's season xG/game.

Chelsea — projected line vs season baseline (4-2-3-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Mike Penders G 90 0 0.00 - n/a 0.0 6.8 9.1 14 FBref Prior G_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0350, quality 0.669)	`market_value_eur=24000000` `quality_factor=0.6685` `boost_factor=1.035` `reference_quality=0.505` `history_matches=2` `history_shrink=0.7143`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Bryan Mbeumo, Bruno Fernandes', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.6917}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.66 (squad xG sum 0.8758, xG scale 1.8965, xA scale 1.8965, team xA 1.28 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.661` `squad_xg_sum=0.8758` `squad_xa_sum=0.6759` `team_expected_assists=1.2818` `team_xa_to_xg_ratio=0.7717`	shots_faced: 16.1 -> 13.5 expected_saves: 3.94 -> 3.3 high_pressure_saves: 1.35 -> 1.14
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 78.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=1.766`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=1.3609`	-

Josh Acheampong D 84 0 0.04 0.00 5166% 0 0.0 6.6 35.5 31 XGBoost D_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0106, quality 0.710)	`market_value_eur=29000000` `quality_factor=0.71` `boost_factor=1.0106` `reference_quality=0.505` `history_matches=24` `history_shrink=0.1724`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Bryan Mbeumo', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.66 (squad xG sum 0.8758, xG scale 1.8965, xA scale 1.8965, team xA 1.28 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.661` `squad_xg_sum=0.8758` `squad_xa_sum=0.6759` `team_expected_assists=1.2818` `team_xa_to_xg_ratio=0.7717`	expected_goals: 0.0196 -> 0.0372 expected_assists: 0.0089 -> 0.0169
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 17 by projected rating, fifa_overall 75.0)	`match_squad_rank=17` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0365` `goal_rate=0.0348` `scorer_weight=0.0453`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=0` `assist_probability=0.0168` `assist_rate=0.0` `provider_weight=0.0143`	-

Danny Welbeck F 75 0 0.28 0.16 72% 1 0.9 6.6 51.7 32 FBref Prior F_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9979, quality 0.439)	`market_value_eur=3100000` `quality_factor=0.4387` `boost_factor=0.9979` `reference_quality=0.505` `history_matches=43` `history_shrink=0.1042`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Ayden Heaven, Leny Yoro', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.7224, 'skill_edge': -0.1709, 'height_edge': -0.018, 'attacker_skill': 0.4387, 'defender_skill': 0.8184, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Ayden Heaven', 'market_value': 32000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7941}, {'player': 'Leny Yoro', 'market_value': 42000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8427}], 'xg_override': 0.1476}	totalShots: 1.2233 -> 0.8837 shotOffTarget: 0.7025 -> 0.5075 onTargetScoringAttempt: 0.5208 -> 0.3762 expectedGoalsOnTarget: 0.3095 -> 0.2236 bigChanceMissed: 0.3011 -> 0.2175 goals: 0.2111 -> 0.1525 expectedGoals: 0.2043 -> 0.1476
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.66 (squad xG sum 0.8758, xG scale 1.8965, xA scale 1.8965, team xA 1.28 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.661` `squad_xg_sum=0.8758` `squad_xa_sum=0.6759` `team_expected_assists=1.2818` `team_xa_to_xg_ratio=0.7717`	expected_goals: 0.1476 -> 0.2799 expected_assists: 0.0387 -> 0.0734
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 15 by projected rating, fifa_overall 80.0)	`match_squad_rank=15` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2441` `goal_rate=0.1525` `scorer_weight=0.2868`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=0` `assist_probability=0.0708` `assist_rate=0.0791` `provider_weight=0.0918`	-

Cole Palmer M 87 1 0.40 0.31 30% 3 2.4 7.1 59.7 35 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0149, quality 0.894)	`market_value_eur=101000000` `quality_factor=0.8936` `boost_factor=1.0149` `reference_quality=0.505` `history_matches=34` `history_shrink=0.1282`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0106)	`blend_weight=0.0106` `history_matches=34` `season_overlap_discount=1.0` `per90_to_match_factor=0.9677` `prior_reliability=0.1267` `competition=Premier League`	totalShots: 2.3789 -> 2.4019 goals: 0.2276 -> 0.236 onTargetScoringAttempt: 0.746 -> 0.7542 fouls: 1.1524 -> 1.1456
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Diogo Dalot', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle', 'duel_win_rate': 0.535, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v1', 'xg_multiplier': 1.0805, 'skill_edge': 0.0508, 'height_edge': 0.004, 'attacker_skill': 0.8936, 'defender_skill': 0.7808, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Diogo Dalot', 'market_value': 28000000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.7808}], 'xg_override': 0.2104}	aerialWon: 0.2108 -> 1.2 totalShots: 2.4019 -> 2.5953 shotOffTarget: 1.6329 -> 1.7643 onTargetScoringAttempt: 0.7542 -> 0.8149 goals: 0.236 -> 0.255 expectedGoals: 0.1947 -> 0.2104 expectedGoalsOnTarget: 0.1608 -> 0.1737 bigChanceMissed: 0.1485 -> 0.1605
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.66 (squad xG sum 0.8758, xG scale 1.8965, xA scale 1.8965, team xA 1.28 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.661` `squad_xg_sum=0.8758` `squad_xa_sum=0.6759` `team_expected_assists=1.2818` `team_xa_to_xg_ratio=0.7717`	expected_goals: 0.2104 -> 0.399 expected_assists: 0.1542 -> 0.2925
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 85.0)	`match_squad_rank=1` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.329` `goal_rate=0.255` `scorer_weight=0.4673`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=1` `assist_probability=0.2536` `assist_rate=0.1865` `provider_weight=0.3388`	-

Geovany Quenda F 77 0 0.29 0.00 11448% 1 0.1 6.7 16.7 23 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0499, quality 0.805)	`market_value_eur=45000000` `quality_factor=0.8047` `boost_factor=1.0499` `reference_quality=0.505` `history_matches=4` `history_shrink=0.5556`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0024)	`blend_weight=0.0024` `history_matches=4` `season_overlap_discount=1.0` `per90_to_match_factor=0.8592` `prior_reliability=0.0067` `competition=Premier League`	totalOffside: 0.2023 -> 0.2226
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Ayden Heaven, Leny Yoro', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.9027, 'skill_edge': -0.0062, 'height_edge': -0.06, 'attacker_skill': 0.8047, 'defender_skill': 0.8184, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Ayden Heaven', 'market_value': 32000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7941}, {'player': 'Leny Yoro', 'market_value': 42000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8427}], 'xg_override': 0.1525}	totalShots: 1.3809 -> 1.2466 shotOffTarget: 0.8893 -> 0.8028 onTargetScoringAttempt: 0.4916 -> 0.4438 expectedGoals: 0.1689 -> 0.1525 goals: 0.1062 -> 0.0959
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.66 (squad xG sum 0.8758, xG scale 1.8965, xA scale 1.8965, team xA 1.28 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.661` `squad_xg_sum=0.8758` `squad_xa_sum=0.6759` `team_expected_assists=1.2818` `team_xa_to_xg_ratio=0.7717`	expected_goals: 0.1525 -> 0.2892 expected_assists: 0.0388 -> 0.0736
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 14 by projected rating, fifa_overall 76.0)	`match_squad_rank=14` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2511` `goal_rate=0.0959` `scorer_weight=0.2314`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=0` `assist_probability=0.071` `assist_rate=0.0` `provider_weight=0.0589`	-

Moisés Caicedo M 86 0 0.07 0.03 128% 1 0.6 7.0 55.3 28 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0139, quality 0.895)	`market_value_eur=104000000` `quality_factor=0.8945` `boost_factor=1.0139` `reference_quality=0.505` `history_matches=37` `history_shrink=0.119`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0010)	`blend_weight=0.001` `history_matches=37` `season_overlap_discount=1.0` `per90_to_match_factor=0.9533` `prior_reliability=0.0133` `competition=Premier League`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Carlos Baleba', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.426, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 1.0085, 'skill_edge': 0.0098, 'height_edge': -0.004, 'attacker_skill': 0.8978, 'defender_skill': 0.876, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Carlos Baleba', 'market_value': 63000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.876}], 'xg_override': 0.0357}	aerialWon: 0.5487 -> 1.0
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.66 (squad xG sum 0.8758, xG scale 1.8965, xA scale 1.8965, team xA 1.28 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.661` `squad_xg_sum=0.8758` `squad_xa_sum=0.6759` `team_expected_assists=1.2818` `team_xa_to_xg_ratio=0.7717`	expected_goals: 0.0357 -> 0.0677 expected_assists: 0.0401 -> 0.076
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 86.0)	`match_squad_rank=6` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0655` `goal_rate=0.0418` `scorer_weight=0.0725`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=0` `assist_probability=0.0732` `assist_rate=0.0549` `provider_weight=0.0885`	-

Jordan Henderson M 83 0 0.11 0.02 459% 1 0.2 7.0 52.2 25 XGBoost M_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9975, quality 0.437)	`market_value_eur=1300000` `quality_factor=0.4374` `boost_factor=0.9975` `reference_quality=0.505` `history_matches=36` `history_shrink=0.122`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Carlos Baleba', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.239, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.7334, 'skill_edge': -0.1974, 'height_edge': 0.016, 'attacker_skill': 0.4374, 'defender_skill': 0.876, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Carlos Baleba', 'market_value': 63000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.876}], 'xg_override': 0.0571}	aerialWon: 0.6958 -> 1.1 totalShots: 0.9681 -> 0.71 shotOffTarget: 0.6236 -> 0.4573 onTargetScoringAttempt: 0.3444 -> 0.2526 goals: 0.0994 -> 0.0729 expectedGoals: 0.0779 -> 0.0571 expectedGoalsOnTarget: 0.0772 -> 0.0566
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.66 (squad xG sum 0.8758, xG scale 1.8965, xA scale 1.8965, team xA 1.28 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.661` `squad_xg_sum=0.8758` `squad_xa_sum=0.6759` `team_expected_assists=1.2818` `team_xa_to_xg_ratio=0.7717`	expected_goals: 0.0571 -> 0.1083 expected_assists: 0.126 -> 0.239
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 78.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1026` `goal_rate=0.0729` `scorer_weight=0.1243`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=0` `assist_probability=0.2126` `assist_rate=0.0997` `provider_weight=0.2033`	-

Jorrel Hato D 74 0 0.00 0.01 64% 0 0.2 6.7 51.1 35 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0140, quality 0.813)	`market_value_eur=47000000` `quality_factor=0.8129` `boost_factor=1.014` `reference_quality=0.505` `history_matches=28` `history_shrink=0.1515`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0112)	`blend_weight=0.0112` `history_matches=28` `season_overlap_discount=1.0` `per90_to_match_factor=0.8236` `prior_reliability=0.1133` `competition=Premier League`	wonTackle: 1.2781 -> 1.2692 wasFouled: 0.1883 -> 0.197 fouls: 0.6436 -> 0.6364 interceptionWon: 0.9822 -> 0.9766
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Bryan Mbeumo', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.66 (squad xG sum 0.8758, xG scale 1.8965, xA scale 1.8965, team xA 1.28 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.661` `squad_xg_sum=0.8758` `squad_xa_sum=0.6759` `team_expected_assists=1.2818` `team_xa_to_xg_ratio=0.7717`	expected_assists: 0.0336 -> 0.0637
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 13 by projected rating, fifa_overall 78.0)	`match_squad_rank=13` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0034` `goal_rate=0.0` `scorer_weight=0.0033`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=0` `assist_probability=0.0617` `assist_rate=0.186` `provider_weight=0.0796`	-

Levi Colwill D 89 0 0.00 0.00 147% 0 0.1 6.8 39.2 17 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0335, quality 0.840)	`market_value_eur=55000000` `quality_factor=0.8399` `boost_factor=1.0335` `reference_quality=0.505` `history_matches=10` `history_shrink=0.3333`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0260)	`blend_weight=0.026` `history_matches=10` `season_overlap_discount=1.0` `per90_to_match_factor=0.9858` `prior_reliability=0.12` `competition=Premier League`	wonTackle: 0.7396 -> 0.7631 fouls: 0.9971 -> 0.9854
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Bryan Mbeumo', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.66 (squad xG sum 0.8758, xG scale 1.8965, xA scale 1.8965, team xA 1.28 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.661` `squad_xg_sum=0.8758` `squad_xa_sum=0.6759` `team_expected_assists=1.2818` `team_xa_to_xg_ratio=0.7717`	expected_assists: 0.0479 -> 0.0908
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 10 by projected rating, fifa_overall 80.0)	`match_squad_rank=10` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0019` `goal_rate=0.0` `scorer_weight=0.0019`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=0` `assist_probability=0.0868` `assist_rate=0.0` `provider_weight=0.0726`	-

Wesley Fofana D 79 0 0.03 0.02 82% 0 0.2 6.8 42.8 30 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0103, quality 0.739)	`market_value_eur=33000000` `quality_factor=0.7389` `boost_factor=1.0103` `reference_quality=0.505` `history_matches=29` `history_shrink=0.1471`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0045)	`blend_weight=0.0045` `history_matches=29` `season_overlap_discount=1.0` `per90_to_match_factor=0.877` `prior_reliability=0.0467` `competition=Premier League`	fouls: 1.2436 -> 1.238 wasFouled: 1.2027 -> 1.1973
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Bruno Fernandes', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.66 (squad xG sum 0.8758, xG scale 1.8965, xA scale 1.8965, team xA 1.28 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.661` `squad_xg_sum=0.8758` `squad_xa_sum=0.6759` `team_expected_assists=1.2818` `team_xa_to_xg_ratio=0.7717`	expected_goals: 0.0152 -> 0.0288 expected_assists: 0.0126 -> 0.0239
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 11 by projected rating, fifa_overall 79.0)	`match_squad_rank=11` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0284` `goal_rate=0.0` `scorer_weight=0.023`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=0` `assist_probability=0.0236` `assist_rate=0.0` `provider_weight=0.0191`	-

Pedro Neto F 84 1 0.31 0.10 228% 2 1.0 7.1 49.1 39 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0109, quality 0.848)	`market_value_eur=58000000` `quality_factor=0.8478` `boost_factor=1.0109` `reference_quality=0.505` `history_matches=42` `history_shrink=0.1064`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0041)	`blend_weight=0.0041` `history_matches=42` `season_overlap_discount=1.0` `per90_to_match_factor=0.9307` `prior_reliability=0.06` `competition=Premier League`	onTargetScoringAttempt: 0.7649 -> 0.7702
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Ayden Heaven, Leny Yoro', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.9313, 'skill_edge': 0.0133, 'height_edge': -0.06, 'attacker_skill': 0.8478, 'defender_skill': 0.8184, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Ayden Heaven', 'market_value': 32000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7941}, {'player': 'Leny Yoro', 'market_value': 42000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8427}], 'xg_override': 0.1659}	totalShots: 1.9323 -> 1.7996 shotOffTarget: 1.1621 -> 1.0823 onTargetScoringAttempt: 0.7702 -> 0.7173 expectedGoalsOnTarget: 0.3172 -> 0.2954 goals: 0.2154 -> 0.2006 bigChanceMissed: 0.195 -> 0.1816 expectedGoals: 0.1781 -> 0.1659
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 2 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=2`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.66 (squad xG sum 0.8758, xG scale 1.8965, xA scale 1.8965, team xA 1.28 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.661` `squad_xg_sum=0.8758` `squad_xa_sum=0.6759` `team_expected_assists=1.2818` `team_xa_to_xg_ratio=0.7717`	expected_goals: 0.1659 -> 0.3146 expected_assists: 0.1417 -> 0.2687
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 81.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.2699` `goal_rate=0.2006` `scorer_weight=0.3652`	-
`integer_assist_projection`	Team scoreline allocation: 2 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=2` `assists_allocated=1` `assist_probability=0.2356` `assist_rate=0.2219` `provider_weight=0.3359`	-

Team totals 2 1.66 10 6.8

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Robert Sánchez	G	5.2m	14.7%	Market-value substitute prior
Maxence Lacroix	D	15.2m	25.1%	Market-value substitute prior
Reece James	D	19.5m	44.4%	Market-value substitute prior
Morgan Rogers	M	23.8m	34.4%	Market-value substitute prior
Marco Palestra	M	23.5m	56.6%	Market-value substitute prior
João Pedro	F	27.1m	37.1%	Market-value substitute prior

Man Utd — projected line vs season baseline (4-2-3-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Bryan Mbeumo F 87 1 0.38 0.39 ok 2 2.1 6.9 56.3 36 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0121, quality 0.875)	`market_value_eur=73000000` `quality_factor=0.8747` `boost_factor=1.0121` `reference_quality=0.505` `history_matches=41` `history_shrink=0.1087`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0094)	`blend_weight=0.0094` `history_matches=41` `season_overlap_discount=1.0` `per90_to_match_factor=0.9621` `prior_reliability=0.1333` `competition=Premier League`	totalCross: 2.4798 -> 2.5244 totalShots: 2.3954 -> 2.4136 wonTackle: 0.3951 -> 0.4004
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Josh Acheampong, Jorrel Hato', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.9624, 'skill_edge': 0.0264, 'height_edge': -0.052, 'attacker_skill': 0.8747, 'defender_skill': 0.8159, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Josh Acheampong', 'market_value': 29000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7736}, {'player': 'Jorrel Hato', 'market_value': 47000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8582}], 'xg_override': 0.3415}	totalShots: 2.4136 -> 2.3228 shotOffTarget: 1.4701 -> 1.4148 onTargetScoringAttempt: 0.9302 -> 0.8952 bigChanceMissed: 0.6451 -> 0.6208 expectedGoalsOnTarget: 0.4838 -> 0.4656 goals: 0.3744 -> 0.3603 expectedGoals: 0.3548 -> 0.3415
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.42 (squad xG sum 1.2754, xG scale 1.1149, xA scale 1.1149, team xA 1.01 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.422` `squad_xg_sum=1.2754` `squad_xa_sum=0.9076` `team_expected_assists=1.0119` `team_xa_to_xg_ratio=0.7116`	expected_goals: 0.3415 -> 0.3808 expected_assists: 0.113 -> 0.126
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 84.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.3167` `goal_rate=0.3603` `scorer_weight=0.3991`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1184` `assist_rate=0.0184` `provider_weight=0.1008`	-

Mason Mount M 77 0 0.10 0.01 795% 1 0.3 6.9 26.7 23 XGBoost M_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0075, quality 0.660)	`market_value_eur=23000000` `quality_factor=0.6595` `boost_factor=1.0075` `reference_quality=0.505` `history_matches=26` `history_shrink=0.1613`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Moisés Caicedo, Jordan Henderson', 'matchup_zone': 'Attacking Mid vs Double Pivot — Tight Space Battle', 'duel_win_rate': 0.243, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 1.0048, 'skill_edge': 0.0012, 'height_edge': 0.002, 'attacker_skill': 0.6704, 'defender_skill': 0.6676, 'attacker_role': 'CAM', 'attacker_mv_price_index': 0.95, 'defender_adjusted_parts': [{'player': 'Moisés Caicedo', 'market_value': 104000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8978}, {'player': 'Jordan Henderson', 'market_value': 1300000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.4374}], 'xg_override': 0.0877}	aerialWon: 0.2664 -> 1.1 totalShots: 1.1627 -> 1.1683
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.42 (squad xG sum 1.2754, xG scale 1.1149, xA scale 1.1149, team xA 1.01 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.422` `squad_xg_sum=1.2754` `squad_xa_sum=0.9076` `team_expected_assists=1.0119` `team_xa_to_xg_ratio=0.7116`	expected_goals: 0.0877 -> 0.0978 expected_assists: 0.1093 -> 0.1219
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 78.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0932` `goal_rate=0.0714` `scorer_weight=0.0862`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1148` `assist_rate=0.0906` `provider_weight=0.1076`	-

Bruno Fernandes M 90 0 0.23 0.20 ok 2 1.9 7.1 59.0 32 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0072, quality 0.732)	`market_value_eur=32000000` `quality_factor=0.732` `boost_factor=1.0072` `reference_quality=0.505` `history_matches=42` `history_shrink=0.1064`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0092)	`blend_weight=0.0092` `history_matches=42` `season_overlap_discount=1.0` `per90_to_match_factor=0.9954` `prior_reliability=0.1333` `competition=Premier League`	totalShots: 2.3028 -> 2.332 onTargetScoringAttempt: 0.6891 -> 0.7011 goals: 0.2898 -> 0.3009 totalCross: 6.2853 -> 6.2916 wasFouled: 0.5979 -> 0.5924
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Wesley Fofana', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle', 'duel_win_rate': 0.415, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.9049, 'skill_edge': -0.0367, 'height_edge': -0.028, 'attacker_skill': 0.732, 'defender_skill': 0.8135, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Wesley Fofana', 'market_value': 33000000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.8135}], 'xg_override': 0.21}	aerialWon: 0.2447 -> 1.0 totalShots: 2.332 -> 2.1102 shotOffTarget: 1.6136 -> 1.4601 onTargetScoringAttempt: 0.7011 -> 0.6344 goals: 0.3009 -> 0.2723 expectedGoals: 0.2321 -> 0.21 expectedGoalsOnTarget: 0.2223 -> 0.2012 bigChanceMissed: 0.1394 -> 0.1261
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.42 (squad xG sum 1.2754, xG scale 1.1149, xA scale 1.1149, team xA 1.01 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.422` `squad_xg_sum=1.2754` `squad_xa_sum=0.9076` `team_expected_assists=1.0119` `team_xa_to_xg_ratio=0.7116`	expected_goals: 0.21 -> 0.2341 expected_assists: 0.1996 -> 0.2225
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 89.0)	`match_squad_rank=1` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2087` `goal_rate=0.2723` `scorer_weight=0.2902`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=1` `assist_probability=0.1995` `assist_rate=0.2571` `provider_weight=0.2738`	-

Amad Diallo F 79 0 0.14 0.06 111% 2 1.0 6.9 44.2 32 XGBoost F_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0111, quality 0.800)	`market_value_eur=44000000` `quality_factor=0.8002` `boost_factor=1.0111` `reference_quality=0.505` `history_matches=35` `history_shrink=0.125`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=true`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Josh Acheampong, Jorrel Hato', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.9132, 'skill_edge': -0.0071, 'height_edge': -0.052, 'attacker_skill': 0.8002, 'defender_skill': 0.8159, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Josh Acheampong', 'market_value': 29000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7736}, {'player': 'Jorrel Hato', 'market_value': 47000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8582}], 'xg_override': 0.1215}	totalShots: 1.7589 -> 1.6063 shotOffTarget: 1.216 -> 1.1105 onTargetScoringAttempt: 0.5429 -> 0.4958 goals: 0.1878 -> 0.1715 bigChanceMissed: 0.1357 -> 0.1239 expectedGoals: 0.1331 -> 0.1215 expectedGoalsOnTarget: 0.1158 -> 0.1057
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.42 (squad xG sum 1.2754, xG scale 1.1149, xA scale 1.1149, team xA 1.01 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.422` `squad_xg_sum=1.2754` `squad_xa_sum=0.9076` `team_expected_assists=1.0119` `team_xa_to_xg_ratio=0.7116`	expected_goals: 0.1215 -> 0.1355 expected_assists: 0.0818 -> 0.0912
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 11 by projected rating, fifa_overall 80.0)	`match_squad_rank=11` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1267` `goal_rate=0.1715` `scorer_weight=0.1694`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0872` `assist_rate=0.102` `provider_weight=0.1052`	-

Carlos Baleba M 74 0 0.06 0.03 152% 1 0.5 6.8 44.2 26 XGBoost M_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0136, quality 0.859)	`market_value_eur=63000000` `quality_factor=0.8589` `boost_factor=1.0136` `reference_quality=0.505` `history_matches=34` `history_shrink=0.1282`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Moisés Caicedo', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.414, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.9915, 'skill_edge': -0.0098, 'height_edge': 0.004, 'attacker_skill': 0.876, 'defender_skill': 0.8978, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Moisés Caicedo', 'market_value': 104000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8978}], 'xg_override': 0.0567}	aerialWon: 1.3513 -> 1.0 totalShots: 0.9028 -> 0.8952 shotOffTarget: 0.6981 -> 0.6922
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.42 (squad xG sum 1.2754, xG scale 1.1149, xA scale 1.1149, team xA 1.01 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.422` `squad_xg_sum=1.2754` `squad_xa_sum=0.9076` `team_expected_assists=1.0119` `team_xa_to_xg_ratio=0.7116`	expected_goals: 0.0567 -> 0.0632
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 12 by projected rating, fifa_overall 80.0)	`match_squad_rank=12` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0612` `goal_rate=0.0547` `scorer_weight=0.062`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.043` `assist_rate=0.052` `provider_weight=0.0501`	-

Youri Tielemans M 76 0 0.06 0.04 45% 1 0.7 7.1 55.7 26 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0115, quality 0.796)	`market_value_eur=43000000` `quality_factor=0.7957` `boost_factor=1.0115` `reference_quality=0.505` `history_matches=33` `history_shrink=0.1316`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0097)	`blend_weight=0.0097` `history_matches=33` `season_overlap_discount=1.0` `per90_to_match_factor=0.8427` `prior_reliability=0.1133` `competition=Premier League`	totalShots: 0.8928 -> 0.903 wasFouled: 1.9305 -> 1.9406 totalCross: 1.4576 -> 1.4483 onTargetScoringAttempt: 0.277 -> 0.2837 wonTackle: 1.1708 -> 1.1643
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Moisés Caicedo', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.384, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.9467, 'skill_edge': -0.0322, 'height_edge': -0.004, 'attacker_skill': 0.8262, 'defender_skill': 0.8978, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Moisés Caicedo', 'market_value': 104000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8978}], 'xg_override': 0.0569}	totalShots: 0.903 -> 0.8549 shotOffTarget: 0.6158 -> 0.583 onTargetScoringAttempt: 0.2837 -> 0.2686 aerialWon: 1.0115 -> 1.0
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.42 (squad xG sum 1.2754, xG scale 1.1149, xA scale 1.1149, team xA 1.01 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.422` `squad_xg_sum=1.2754` `squad_xa_sum=0.9076` `team_expected_assists=1.0119` `team_xa_to_xg_ratio=0.7116`	expected_goals: 0.0569 -> 0.0634 expected_assists: 0.1138 -> 0.1269
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 85.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0614` `goal_rate=0.0569` `scorer_weight=0.0634`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1192` `assist_rate=0.0858` `provider_weight=0.1052`	-

Ayden Heaven D 82 0 0.05 0.01 258% 0 0.1 6.8 20.3 30 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0136, quality 0.732)	`market_value_eur=32000000` `quality_factor=0.732` `boost_factor=1.0136` `reference_quality=0.505` `history_matches=20` `history_shrink=0.2`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0087)	`blend_weight=0.0087` `history_matches=20` `season_overlap_discount=1.0` `per90_to_match_factor=0.9136` `prior_reliability=0.0667` `competition=Premier League`	totalShots: 0.4087 -> 0.421 interceptionWon: 0.8332 -> 0.826 wonTackle: 0.7836 -> 0.7768 wasFouled: 1.1588 -> 1.1646
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Pedro Neto', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.42 (squad xG sum 1.2754, xG scale 1.1149, xA scale 1.1149, team xA 1.01 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.422` `squad_xg_sum=1.2754` `squad_xa_sum=0.9076` `team_expected_assists=1.0119` `team_xa_to_xg_ratio=0.7116`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 13 by projected rating, fifa_overall 75.0)	`match_squad_rank=13` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0456` `goal_rate=0.0` `scorer_weight=0.0374`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0128` `assist_rate=0.0` `provider_weight=0.0105`	-

Harry Maguire D 89 0 0.04 0.04 ok 1 0.6 7.0 53.3 30 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0000, quality 0.506)	`market_value_eur=8600000` `quality_factor=0.5059` `boost_factor=1.0` `reference_quality=0.505` `history_matches=31` `history_shrink=0.1389`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0120)	`blend_weight=0.012` `history_matches=31` `season_overlap_discount=1.0` `per90_to_match_factor=0.9936` `prior_reliability=0.1333` `competition=Premier League`	totalShots: 0.6157 -> 0.6262 wonTackle: 0.7268 -> 0.7181 totalCross: 0.3638 -> 0.3714
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Danny Welbeck', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.42 (squad xG sum 1.2754, xG scale 1.1149, xA scale 1.1149, team xA 1.01 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.422` `squad_xg_sum=1.2754` `squad_xa_sum=0.9076` `team_expected_assists=1.0119` `team_xa_to_xg_ratio=0.7116`	expected_assists: 0.0448 -> 0.0499
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 82.0)	`match_squad_rank=4` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0435` `goal_rate=0.0032` `scorer_weight=0.0356`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0487` `assist_rate=0.0235` `provider_weight=0.0399`	-

Leny Yoro D 84 0 0.10 0.04 168% 1 0.2 6.7 33.6 32 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0102, quality 0.791)	`market_value_eur=42000000` `quality_factor=0.7909` `boost_factor=1.0102` `reference_quality=0.505` `history_matches=37` `history_shrink=0.119`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0010)	`blend_weight=0.001` `history_matches=37` `season_overlap_discount=1.0` `per90_to_match_factor=0.9333` `prior_reliability=0.0133` `competition=Premier League`	interceptionWon: 1.3362 -> 1.3444
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Pedro Neto', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.42 (squad xG sum 1.2754, xG scale 1.1149, xA scale 1.1149, team xA 1.01 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.422` `squad_xg_sum=1.2754` `squad_xa_sum=0.9076` `team_expected_assists=1.0119` `team_xa_to_xg_ratio=0.7116`	expected_goals: 0.0928 -> 0.1035 expected_assists: 0.0655 -> 0.073
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 17 by projected rating, fifa_overall 78.0)	`match_squad_rank=17` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0983` `goal_rate=0.0` `scorer_weight=0.0828`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0704` `assist_rate=0.0` `provider_weight=0.0584`	-

Diogo Dalot D 88 0 0.06 0.03 85% 1 0.7 6.8 53.0 32 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0063, quality 0.702)	`market_value_eur=28000000` `quality_factor=0.7022` `boost_factor=1.0063` `reference_quality=0.505` `history_matches=42` `history_shrink=0.1064`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0041)	`blend_weight=0.0041` `history_matches=42` `season_overlap_discount=1.0` `per90_to_match_factor=0.9806` `prior_reliability=0.06` `competition=Premier League`	totalShots: 0.733 -> 0.7474 totalCross: 2.6469 -> 2.6585
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Cole Palmer', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.42 (squad xG sum 1.2754, xG scale 1.1149, xA scale 1.1149, team xA 1.01 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.422` `squad_xg_sum=1.2754` `squad_xa_sum=0.9076` `team_expected_assists=1.0119` `team_xa_to_xg_ratio=0.7116`	expected_goals: 0.0552 -> 0.0615 expected_assists: 0.0546 -> 0.0609
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 15 by projected rating, fifa_overall 78.0)	`match_squad_rank=15` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0596` `goal_rate=0.0022` `scorer_weight=0.0492`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0591` `assist_rate=0.0164` `provider_weight=0.0487`	-

Senne Lammens G 90 0 0.00 - n/a 0.0 6.9 57.7 33 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0088, quality 0.770)	`market_value_eur=38000000` `quality_factor=0.7699` `boost_factor=1.0088` `reference_quality=0.505` `history_matches=40` `history_shrink=0.1111`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0096)	`blend_weight=0.0096` `history_matches=40` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Cole Palmer, Pedro Neto', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.8102}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.42 (squad xG sum 1.2754, xG scale 1.1149, xA scale 1.1149, team xA 1.01 = 0.71 x lambda)	`scoreline_reset=true` `team_lambda=1.422` `squad_xg_sum=1.2754` `squad_xa_sum=0.9076` `team_expected_assists=1.0119` `team_xa_to_xg_ratio=0.7116`	shots_faced: 17.2 -> 15.8 expected_saves: 4.21 -> 3.87 high_pressure_saves: 1.45 -> 1.33
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 10 by projected rating, fifa_overall 82.0)	`match_squad_rank=10` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=1.4656`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=0.9934`	-

Team totals 1 1.42 14 6.9

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
André Onana	G	4.7m	12.3%	Market-value substitute prior
Lisandro Martínez	D	15.1m	24.8%	Market-value substitute prior
Matthijs de Ligt	D	14.7m	35.2%	Market-value substitute prior
Kobbie Mainoo	M	23.9m	53.5%	Market-value substitute prior
Andrey Santos	M	16.6m	52.9%	Market-value substitute prior
Matheus Cunha	F	27.1m	37.0%	Market-value substitute prior

Antagonist / tactical summary

Midfield is evenly contested — Moisés Caicedo vs Carlos Baleba will be the pivotal duel. Total antagonist xG — Home: 0.77, Away: 0.87. Home danger zone: Left Wing vs Right Back — Flank Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Midfield — possession 50.2% / 49.8%, dominance ratio 0.48, key duel: Moisés Caicedo vs Carlos Baleba

Cole Palmer xG 0.21Pedro Neto xG 0.17Danny Welbeck xG 0.15Bruno Fernandes xG 0.21Mason Mount xG 0.09Bryan Mbeumo xG 0.34

Summary

Antagonist matchup analysis for Chelsea vs Man Utd. Most likely scoreline: 2-1. Lambdas derived from 6 home and 6 away individual player duels. Head-to-head: Chelsea 2W-4D-4L vs Man Utd over 10 meetings.

Tactical parity between 4-2-3-1 and 4-2-3-1. Match outcome driven primarily by individual player quality. Midfield is evenly contested — Moisés Caicedo vs Carlos Baleba will be the pivotal duel. Total antagonist xG — Home: 0.77, Away: 0.87. Home danger zone: Left Wing vs Right Back — Flank Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Brentford vs Forest

Premier League · 20261031 · 4-2-3-1 vs 3-4-2-1

Predicted 1-0Winner: BrentfordWinning H 1-0 (10.4%)Winning A 0-1 (6.7%)Confidence 0.75λ home/away 1.69 / 1.09Most-likely P 10.4%

Actual vs predicted: result not available yet (upcoming fixture).

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Brentford vs Forest. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 4-2-3-1 vs 3-4-2-1 (duel lambdas: Home 1.683, Away 1.167). Final calibrated Poisson lambdas: Home 1.692, Away 1.089 (most likely score 1-0).

Home Starters 11/11 XGBoost (10 FBref) Away Starters 11/11 XGBoost (9 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 1.68 / 1.17 Final λ (H/A) 1.69 / 1.09

FBref Team Priors: Brentford (Gls/90: 2.0, GA/90: 2.0 → E[λ]: 1.379) vs Nottingham Forest (Gls/90: 1.0, GA/90: 1.0 → E[λ]: 1.379) | Blend Weight: 0.192

Modeling Pipeline Stages (5 layers)

1. XGBoost player regressions for volume & event rates with market-value quality scaling
2. Multi-source prior blending (FBref 2026-2027 tables + Sofascore season history)
3. Antagonist formation duel matrix resolving 1v1 and zonal matchups
4. Team-level prior calibration (FBref possession and attack/defense efficiencies)
5. Bivariate Poisson matrix with Dixon-Coles low-score correction

Standing & form context

Metric	Brentford	Forest
League position	6.00	18.00
Points per game	1.37	1.01
Attack index	1.10	0.82
Defense (concede) index	0.96	0.92
Form multiplier	1.02	0.98
Standing λ multiplier	1.03	0.89

Lambda decomposition (how the goal rates were built)

Stage	Brentford λ	Forest λ
FBref Squad Prior E[λ] (weight 0.192)	1.379	1.379
Pre-competitive-floor	2.034	1.182
Pre-calibration (Micro-duels)	1.683	1.167
Final λ	1.692	1.089

competition calibration scale 0.865competition observed total goals 2.781quality gap applied Falsemismatch score pressure applied Falsecompetitive underdog floor applied Falsestage discount 1.0

Scoreline distribution & markets

Top 5 scorelines	P
1-1	11.5%
1-0	10.4%
2-1	9.7%
2-0	8.9%
0-1	6.7%

Home win	51.4%
Draw	24.3%
Away win	24.3%
Over 2.5	52.6%
BTTS yes	54.2%

Pregame fair line — favorite Brentford (51.4%), fair total 2.78, fair spread 0.60

Projected players vs season baseline

Proj = projected for this match · Season = this season's per-game average from match data · deviation badge compares projected xG to the player's season xG/game.

Brentford — projected line vs season baseline (4-2-3-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Dango Ouattara F 74 0 0.26 0.16 62% 2 1.2 6.9 49.7 39 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0084, quality 0.758)	`market_value_eur=36000000` `quality_factor=0.7581` `boost_factor=1.0084` `reference_quality=0.505` `history_matches=40` `history_shrink=0.1111`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0053)	`blend_weight=0.0053` `history_matches=40` `season_overlap_discount=1.0` `per90_to_match_factor=0.827` `prior_reliability=0.0733` `competition=Premier League`	fouls: 0.7063 -> 0.7145
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Jair, Nikola Milenković', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 0.9173, 'skill_edge': 0.0037, 'height_edge': -0.06, 'attacker_skill': 0.7581, 'defender_skill': 0.7498, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Jair', 'market_value': 26000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7498}, {'player': 'Nikola Milenković', 'market_value': 26000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7498}], 'xg_override': 0.2233}	totalShots: 1.8811 -> 1.7255 shotOffTarget: 1.2645 -> 1.1599 onTargetScoringAttempt: 0.6108 -> 0.5603 bigChanceMissed: 0.4027 -> 0.3694 expectedGoalsOnTarget: 0.3502 -> 0.3212 goals: 0.2566 -> 0.2354 expectedGoals: 0.2434 -> 0.2233
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.69 (squad xG sum 1.4531, xG scale 1.1644, xA scale 1.3015, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.692` `squad_xg_sum=1.4531` `squad_xa_sum=0.845` `team_expected_assists=1.0998` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.2233 -> 0.26 expected_assists: 0.0954 -> 0.1242
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 79.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2289` `goal_rate=0.2354` `scorer_weight=0.2715`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1168` `assist_rate=0.0648` `provider_weight=0.0994`	-

Mathias Jensen M 79 0 0.06 0.06 ok 1 0.9 7.0 56.5 30 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9999, quality 0.502)	`market_value_eur=8300000` `quality_factor=0.5023` `boost_factor=0.9999` `reference_quality=0.505` `history_matches=40` `history_shrink=0.1111`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0034)	`blend_weight=0.0034` `history_matches=40` `season_overlap_discount=1.0` `per90_to_match_factor=0.8746` `prior_reliability=0.0467` `competition=Premier League`	fouls: 0.917 -> 0.9265 totalCross: 4.651 -> 4.6564
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Jair', 'matchup_zone': 'LW vs CB', 'duel_win_rate': 0.249, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v1', 'xg_multiplier': 0.748, 'skill_edge': -0.1114, 'height_edge': -0.06, 'attacker_skill': 0.5023, 'defender_skill': 0.7498, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Jair', 'market_value': 26000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7498}], 'xg_override': 0.052}`	aerialWon: 0.1738 -> 1.1 totalShots: 1.0201 -> 0.763 shotOffTarget: 0.6702 -> 0.5013 onTargetScoringAttempt: 0.3481 -> 0.2604 bigChanceMissed: 0.1314 -> 0.0983 goals: 0.0761 -> 0.0569 expectedGoals: 0.0695 -> 0.052 expectedGoalsOnTarget: 0.0412 -> 0.0308
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.69 (squad xG sum 1.4531, xG scale 1.1644, xA scale 1.3015, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.692` `squad_xg_sum=1.4531` `squad_xa_sum=0.845` `team_expected_assists=1.0998` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.052 -> 0.0605 expected_assists: 0.1559 -> 0.2028
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 80.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0587` `goal_rate=0.0569` `scorer_weight=0.0634`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1836` `assist_rate=0.1054` `provider_weight=0.1622`	-

Kevin Schade F 78 1 0.35 0.21 66% 2 1.4 6.8 50.7 43 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0102, quality 0.831)	`market_value_eur=52000000` `quality_factor=0.8309` `boost_factor=1.0102` `reference_quality=0.505` `history_matches=43` `history_shrink=0.1042`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0090)	`blend_weight=0.009` `history_matches=43` `season_overlap_discount=1.0` `per90_to_match_factor=0.8719` `prior_reliability=0.1333` `competition=Premier League`	totalShots: 2.1824 -> 2.1947 wonTackle: 0.8686 -> 0.8804 wasFouled: 0.9059 -> 0.9135 onTargetScoringAttempt: 0.7535 -> 0.7587
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Jair, Nikola Milenković', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.182, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 0.9742, 'skill_edge': 0.0365, 'height_edge': -0.054, 'attacker_skill': 0.8309, 'defender_skill': 0.7498, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Jair', 'market_value': 26000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7498}, {'player': 'Nikola Milenković', 'market_value': 26000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7498}], 'xg_override': 0.3031}	totalShots: 2.1947 -> 2.1381 shotOffTarget: 1.4288 -> 1.3919 onTargetScoringAttempt: 0.7587 -> 0.7391 goals: 0.3757 -> 0.366 expectedGoals: 0.3111 -> 0.3031 bigChanceMissed: 0.3105 -> 0.3025 expectedGoalsOnTarget: 0.2101 -> 0.2047
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.69 (squad xG sum 1.4531, xG scale 1.1644, xA scale 1.3015, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.692` `squad_xg_sum=1.4531` `squad_xa_sum=0.845` `team_expected_assists=1.0998` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.3031 -> 0.3529 expected_assists: 0.0466 -> 0.0607
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 11 by projected rating, fifa_overall 79.0)	`match_squad_rank=11` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.2974` `goal_rate=0.366` `scorer_weight=0.4158`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0589` `assist_rate=0.0116` `provider_weight=0.0486`	-

Igor Thiago F 89 0 0.39 0.45 ok 3 1.9 6.7 55.7 34 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0107, quality 0.868)	`market_value_eur=68000000` `quality_factor=0.8677` `boost_factor=1.0107` `reference_quality=0.505` `history_matches=46` `history_shrink=0.098`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0081)	`blend_weight=0.0081` `history_matches=46` `season_overlap_discount=1.0` `per90_to_match_factor=0.9912` `prior_reliability=0.1267` `competition=Premier League`	totalShots: 2.0827 -> 2.0952 fouls: 2.0748 -> 2.0623 wonTackle: 0.6592 -> 0.6539
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Jair, Nikola Milenković', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.231, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 1.0456, 'skill_edge': 0.053, 'height_edge': -0.022, 'attacker_skill': 0.8677, 'defender_skill': 0.7498, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Jair', 'market_value': 26000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7498}, {'player': 'Nikola Milenković', 'market_value': 26000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7498}], 'xg_override': 0.333}	totalShots: 2.0952 -> 2.1907 shotOffTarget: 1.4467 -> 1.5127 onTargetScoringAttempt: 0.6393 -> 0.6685 bigChanceMissed: 0.6065 -> 0.6342 expectedGoals: 0.3185 -> 0.333 goals: 0.3046 -> 0.3185 expectedGoalsOnTarget: 0.2239 -> 0.2341
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.69 (squad xG sum 1.4531, xG scale 1.1644, xA scale 1.3015, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.692` `squad_xg_sum=1.4531` `squad_xa_sum=0.845` `team_expected_assists=1.0998` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.333 -> 0.3879 expected_assists: 0.1027 -> 0.1337
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 13 by projected rating, fifa_overall 82.0)	`match_squad_rank=13` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.3215` `goal_rate=0.3185` `scorer_weight=0.3732`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=1` `assist_probability=0.1251` `assist_rate=0.28` `provider_weight=0.1671`	-

Vitaly Janelt M 87 0 0.06 0.04 ok 1 0.4 6.9 52.3 26 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0055, quality 0.641)	`market_value_eur=21000000` `quality_factor=0.6408` `boost_factor=1.0055` `reference_quality=0.505` `history_matches=32` `history_shrink=0.1351`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0117)	`blend_weight=0.0117` `history_matches=32` `season_overlap_discount=1.0` `per90_to_match_factor=0.9662` `prior_reliability=0.1333` `competition=Premier League`	totalCross: 2.3897 -> 2.4126 interceptionWon: 2.5035 -> 2.5194 fouls: 1.3569 -> 1.3693 wonTackle: 1.2693 -> 1.2771
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Callum Hudson Odoi, Liam Delap, Morgan Gibbs-White', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.69 (squad xG sum 1.4531, xG scale 1.1644, xA scale 1.3015, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.692` `squad_xg_sum=1.4531` `squad_xa_sum=0.845` `team_expected_assists=1.0998` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0481 -> 0.056 expected_assists: 0.0817 -> 0.1063
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 78.0)	`match_squad_rank=4` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0545` `goal_rate=0.088` `scorer_weight=0.07`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1008` `assist_rate=0.0724` `provider_weight=0.0988`	-

Mamadou Sangare M 81 0 0.07 0.03 124% 1 0.4 6.9 39.7 14 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0364, quality 0.821)	`market_value_eur=49000000` `quality_factor=0.8206` `boost_factor=1.0364` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0300)	`blend_weight=0.03` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=0.9021` `prior_reliability=0.12` `competition=Premier League`	interceptionWon: 1.5675 -> 1.6107 totalShots: 0.7153 -> 0.7528 wasFouled: 0.6396 -> 0.6655 totalCross: 0.3174 -> 0.3379 wonTackle: 1.9883 -> 1.9738 goalAssist: 0.054 -> 0.0674 onTargetScoringAttempt: 0.1837 -> 0.1931
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Daniel Muñoz', 'matchup_zone': 'Midfield lane'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.69 (squad xG sum 1.4531, xG scale 1.1644, xA scale 1.3015, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.692` `squad_xg_sum=1.4531` `squad_xa_sum=0.845` `team_expected_assists=1.0998` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0567 -> 0.066 expected_assists: 0.0528 -> 0.0687
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 81.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0639` `goal_rate=0.0578` `scorer_weight=0.0667`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0664` `assist_rate=0.0674` `provider_weight=0.0785`	-

Keane Lewis-Potter D 80 0 0.15 0.10 42% 1 1.0 6.7 55.0 34 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0063, quality 0.710)	`market_value_eur=29000000` `quality_factor=0.71` `boost_factor=1.0063` `reference_quality=0.505` `history_matches=44` `history_shrink=0.102`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0088)	`blend_weight=0.0088` `history_matches=44` `season_overlap_discount=1.0` `per90_to_match_factor=0.8917` `prior_reliability=0.1333` `competition=Premier League`	wasFouled: 1.2614 -> 1.2502 totalCross: 3.2727 -> 3.2832
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Xaver Schlager', 'matchup_zone': 'Right Wingback vs Left Back — Overlapping Run'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.69 (squad xG sum 1.4531, xG scale 1.1644, xA scale 1.3015, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.692` `squad_xg_sum=1.4531` `squad_xa_sum=0.845` `team_expected_assists=1.0998` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1253 -> 0.1459 expected_assists: 0.0946 -> 0.1231
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 12 by projected rating, fifa_overall 77.0)	`match_squad_rank=12` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1358` `goal_rate=0.203` `scorer_weight=0.1824`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1158` `assist_rate=0.0972` `provider_weight=0.1253`	-

Kristoffer Ajer D 89 0 0.04 0.01 225% 1 0.2 6.8 47.1 33 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0071, quality 0.694)	`market_value_eur=27000000` `quality_factor=0.6941` `boost_factor=1.0071` `reference_quality=0.505` `history_matches=35` `history_shrink=0.125`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0108)	`blend_weight=0.0108` `history_matches=35` `season_overlap_discount=1.0` `per90_to_match_factor=0.9842` `prior_reliability=0.1333` `competition=Premier League`	wonTackle: 0.9516 -> 0.9466
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Callum Hudson Odoi', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.69 (squad xG sum 1.4531, xG scale 1.1644, xA scale 1.3015, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.692` `squad_xg_sum=1.4531` `squad_xa_sum=0.845` `team_expected_assists=1.0998` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0313 -> 0.0364 expected_assists: 0.0368 -> 0.0479
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 77.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0357` `goal_rate=0.1099` `scorer_weight=0.0455`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0468` `assist_rate=0.0` `provider_weight=0.0383`	-

Caoimhin Kelleher G 90 0 0.00 - n/a 0.0 6.9 57.3 33 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0061, quality 0.710)	`market_value_eur=29000000` `quality_factor=0.71` `boost_factor=1.0061` `reference_quality=0.505` `history_matches=45` `history_shrink=0.1`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0087)	`blend_weight=0.0087` `history_matches=45` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Liam Delap, Morgan Gibbs-White', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.1665}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.69 (squad xG sum 1.4531, xG scale 1.1644, xA scale 1.3015, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.692` `squad_xg_sum=1.4531` `squad_xa_sum=0.845` `team_expected_assists=1.0998` `team_xa_to_xg_ratio=0.65`	shots_faced: 11.1 -> 10.3 expected_saves: 2.72 -> 2.52 high_pressure_saves: 0.93 -> 0.87
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 80.0)	`match_squad_rank=6` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=1.7932`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=1.0253`	-

Michael Kayode D 84 0 0.07 0.05 38% 1 0.3 6.7 56.2 33 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0081, quality 0.775)	`market_value_eur=39000000` `quality_factor=0.7754` `boost_factor=1.0081` `reference_quality=0.505` `history_matches=45` `history_shrink=0.1`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0078)	`blend_weight=0.0078` `history_matches=45` `season_overlap_discount=1.0` `per90_to_match_factor=0.931` `prior_reliability=0.12` `competition=Premier League`	totalShots: 0.6011 -> 0.6166 wasFouled: 1.0619 -> 1.0698 wonTackle: 1.2125 -> 1.2071
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'James McAtee', 'matchup_zone': 'Left Wingback vs Right Back — Overlapping Run'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.69 (squad xG sum 1.4531, xG scale 1.1644, xA scale 1.3015, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.692` `squad_xg_sum=1.4531` `squad_xa_sum=0.845` `team_expected_assists=1.0998` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.057 -> 0.0664 expected_assists: 0.0735 -> 0.0957
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 15 by projected rating, fifa_overall 81.0)	`match_squad_rank=15` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0642` `goal_rate=0.0775` `scorer_weight=0.0791`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0913` `assist_rate=0.0182` `provider_weight=0.0766`	-

Sepp van den Berg D 88 0 0.10 0.06 77% 1 0.5 6.8 41.2 29 XGBoost D_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0090, quality 0.739)	`market_value_eur=33000000` `quality_factor=0.7389` `boost_factor=1.009` `reference_quality=0.505` `history_matches=34` `history_shrink=0.1282`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Liam Delap', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.69 (squad xG sum 1.4531, xG scale 1.1644, xA scale 1.3015, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.692` `squad_xg_sum=1.4531` `squad_xa_sum=0.845` `team_expected_assists=1.0998` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0878 -> 0.1022 expected_assists: 0.0354 -> 0.0461
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 78.0)	`match_squad_rank=8` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0972` `goal_rate=0.0` `scorer_weight=0.0818`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0451` `assist_rate=0.0217` `provider_weight=0.0387`	-

Team totals 1 1.69 16 6.8

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Hákon Rafn Valdimarsson	G	4.4m	10.9%	Market-value substitute prior
El Hadji Malick Diouf	D	14.8m	35.6%	Market-value substitute prior
Nathan Collins	D	14.7m	23.9%	Market-value substitute prior
Yehor Yarmoliuk	M	23.0m	71.5%	Market-value substitute prior
Mikkel Damsgaard	M	18.6m	52.2%	Market-value substitute prior
Jaidon Anthony	F	24.9m	54.2%	Market-value substitute prior

Forest — projected line vs season baseline (3-4-2-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Matz Sels G 90 0 0.00 - n/a 0.0 7.0 55.0 30 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9980, quality 0.446)	`market_value_eur=3700000` `quality_factor=0.4461` `boost_factor=0.998` `reference_quality=0.505` `history_matches=39` `history_shrink=0.1136`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0098)	`blend_weight=0.0098` `history_matches=39` `season_overlap_discount=1.0` `per90_to_match_factor=0.9968` `prior_reliability=0.1333` `competition=Premier League`	saves: 2.9184 -> 2.9093
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Igor Thiago, Kevin Schade', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.683}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.09 (squad xG sum 1.3706, xG scale 0.7945, xA scale 0.7945, team xA 0.76 = 0.70 x lambda)	`scoreline_reset=true` `team_lambda=1.089` `squad_xg_sum=1.3706` `squad_xa_sum=0.9608` `team_expected_assists=0.7634` `team_xa_to_xg_ratio=0.701`	shots_faced: 16.0 -> 16.1 expected_saves: 3.92 -> 3.94
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 81.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=1.0339`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=0.6878`	-

Jair D 85 0 0.03 0.02 46% 1 0.2 6.8 39.0 24 XGBoost D_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0136, quality 0.686)	`market_value_eur=26000000` `quality_factor=0.6858` `boost_factor=1.0136` `reference_quality=0.505` `history_matches=15` `history_shrink=0.25`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Mathias Jensen', 'matchup_zone': 'LW vs CB'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.09 (squad xG sum 1.3706, xG scale 0.7945, xA scale 0.7945, team xA 0.76 = 0.70 x lambda)	`scoreline_reset=true` `team_lambda=1.089` `squad_xg_sum=1.3706` `squad_xa_sum=0.9608` `team_expected_assists=0.7634` `team_xa_to_xg_ratio=0.701`	expected_goals: 0.0355 -> 0.0282
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 14 by projected rating, fifa_overall 80.0)	`match_squad_rank=14` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0278` `goal_rate=0.0` `scorer_weight=0.0226`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0058` `assist_rate=0.0` `provider_weight=0.0051`	-

Nikola Milenković D 90 0 0.02 0.01 111% 0 0.2 6.8 52.1 30 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0058, quality 0.686)	`market_value_eur=26000000` `quality_factor=0.6858` `boost_factor=1.0058` `reference_quality=0.505` `history_matches=42` `history_shrink=0.1064`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0092)	`blend_weight=0.0092` `history_matches=42` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	fouls: 1.1309 -> 1.1389
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Igor Thiago', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.09 (squad xG sum 1.3706, xG scale 0.7945, xA scale 0.7945, team xA 0.76 = 0.70 x lambda)	`scoreline_reset=true` `team_lambda=1.089` `squad_xg_sum=1.3706` `squad_xa_sum=0.9608` `team_expected_assists=0.7634` `team_xa_to_xg_ratio=0.701`	expected_goals: 0.0289 -> 0.023 expected_assists: 0.0346 -> 0.0275
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 13 by projected rating, fifa_overall 80.0)	`match_squad_rank=13` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0227` `goal_rate=0.0` `scorer_weight=0.0184`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0271` `assist_rate=0.0` `provider_weight=0.022`	-

Murillo D 88 0 0.10 0.06 75% 2 0.7 6.9 53.3 26 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0125, quality 0.821)	`market_value_eur=49000000` `quality_factor=0.8206` `boost_factor=1.0125` `reference_quality=0.505` `history_matches=33` `history_shrink=0.1316`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0114)	`blend_weight=0.0114` `history_matches=33` `season_overlap_discount=1.0` `per90_to_match_factor=0.9771` `prior_reliability=0.1333` `competition=Premier League`	totalShots: 1.3873 -> 1.3771 fouls: 1.3427 -> 1.333 interceptionWon: 1.2173 -> 1.209 wonTackle: 1.0305 -> 1.0243 onTargetScoringAttempt: 0.5214 -> 0.5155
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Dango Ouattara', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.09 (squad xG sum 1.3706, xG scale 0.7945, xA scale 0.7945, team xA 0.76 = 0.70 x lambda)	`scoreline_reset=true` `team_lambda=1.089` `squad_xg_sum=1.3706` `squad_xa_sum=0.9608` `team_expected_assists=0.7634` `team_xa_to_xg_ratio=0.701`	expected_goals: 0.1259 -> 0.1 expected_assists: 0.0533 -> 0.0423
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 82.0)	`match_squad_rank=4` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0952` `goal_rate=0.0034` `scorer_weight=0.08`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0414` `assist_rate=0.0788` `provider_weight=0.0527`	-

Daniel Muñoz M 82 0 0.10 0.06 70% 1 0.7 6.9 50.8 31 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0057, quality 0.660)	`market_value_eur=23000000` `quality_factor=0.6595` `boost_factor=1.0057` `reference_quality=0.505` `history_matches=36` `history_shrink=0.122`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0032)	`blend_weight=0.0032` `history_matches=36` `season_overlap_discount=1.0` `per90_to_match_factor=0.9069` `prior_reliability=0.04` `competition=Premier League`	totalCross: 2.1344 -> 2.1276 totalShots: 1.1667 -> 1.1726
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Vitaly Janelt', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.449, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v1', 'xg_multiplier': 0.999, 'skill_edge': 0.0033, 'height_edge': -0.004, 'attacker_skill': 0.6821, 'defender_skill': 0.6747, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Vitaly Janelt', 'market_value': 21000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.6747}], 'xg_override': 0.1212}	aerialWon: 1.9692 -> 1.1
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.09 (squad xG sum 1.3706, xG scale 0.7945, xA scale 0.7945, team xA 0.76 = 0.70 x lambda)	`scoreline_reset=true` `team_lambda=1.089` `squad_xg_sum=1.3706` `squad_xa_sum=0.9608` `team_expected_assists=0.7634` `team_xa_to_xg_ratio=0.701`	expected_goals: 0.1212 -> 0.0963 expected_assists: 0.0886 -> 0.0704
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 82.0)	`match_squad_rank=8` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0918` `goal_rate=0.119` `scorer_weight=0.0951`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.068` `assist_rate=0.0862` `provider_weight=0.0692`	-

James McAtee M 85 0 0.06 0.02 169% 1 0.5 6.9 28.9 33 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0069, quality 0.629)	`market_value_eur=19800000` `quality_factor=0.6291` `boost_factor=1.0069` `reference_quality=0.505` `history_matches=22` `history_shrink=0.1852`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0144)	`blend_weight=0.0144` `history_matches=22` `season_overlap_discount=1.0` `per90_to_match_factor=0.949` `prior_reliability=0.12` `competition=Premier League`	totalShots: 1.1809 -> 1.1638 fouls: 1.0198 -> 1.0051 interceptionWon: 0.7964 -> 0.7849 wonTackle: 1.5309 -> 1.524 totalCross: 3.3421 -> 3.3471
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Michael Kayode', 'matchup_zone': 'Left Wingback vs Right Back — Overlapping Run', 'duel_win_rate': 0.342, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.915, 'skill_edge': -0.0619, 'height_edge': 0.004, 'attacker_skill': 0.7044, 'defender_skill': 0.8419, 'attacker_role': 'LWB', 'attacker_mv_price_index': 0.7, 'defender_adjusted_parts': [{'player': 'Michael Kayode', 'market_value': 39000000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.8419}], 'xg_override': 0.0739}	aerialWon: 0.4062 -> 1.1 totalShots: 1.1761 -> 1.0762 shotOffTarget: 0.8535 -> 0.781 onTargetScoringAttempt: 0.3226 -> 0.2952 goals: 0.0877 -> 0.0802 expectedGoals: 0.0808 -> 0.0739
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.09 (squad xG sum 1.3706, xG scale 0.7945, xA scale 0.7945, team xA 0.76 = 0.70 x lambda)	`scoreline_reset=true` `team_lambda=1.089` `squad_xg_sum=1.3706` `squad_xa_sum=0.9608` `team_expected_assists=0.7634` `team_xa_to_xg_ratio=0.701`	expected_goals: 0.0739 -> 0.0587 expected_assists: 0.1292 -> 0.1027
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 73.0)	`match_squad_rank=6` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.057` `goal_rate=0.0802` `scorer_weight=0.0617`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0976` `assist_rate=0.0947` `provider_weight=0.0829`	-

Neco Williams D 89 0 0.09 0.06 50% 1 1.2 6.8 59.2 38 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0061, quality 0.710)	`market_value_eur=29000000` `quality_factor=0.71` `boost_factor=1.0061` `reference_quality=0.505` `history_matches=45` `history_shrink=0.1`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0087)	`blend_weight=0.0087` `history_matches=45` `season_overlap_discount=1.0` `per90_to_match_factor=0.9841` `prior_reliability=0.1333` `competition=Premier League`	wonTackle: 1.2354 -> 1.2545 totalCross: 5.7902 -> 5.7741 interceptionWon: 0.9884 -> 0.9798
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Igor Thiago', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.09 (squad xG sum 1.3706, xG scale 0.7945, xA scale 0.7945, team xA 0.76 = 0.70 x lambda)	`scoreline_reset=true` `team_lambda=1.089` `squad_xg_sum=1.3706` `squad_xa_sum=0.9608` `team_expected_assists=0.7634` `team_xa_to_xg_ratio=0.701`	expected_goals: 0.1178 -> 0.0936 expected_assists: 0.1415 -> 0.1124
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 10 by projected rating, fifa_overall 80.0)	`match_squad_rank=10` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0894` `goal_rate=0.1849` `scorer_weight=0.117`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.1063` `assist_rate=0.0316` `provider_weight=0.0899`	-

Morgan Gibbs-White M 89 0 0.21 0.30 ok 3 1.8 7.1 61.6 37 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0114, quality 0.878)	`market_value_eur=76000000` `quality_factor=0.8781` `boost_factor=1.0114` `reference_quality=0.505` `history_matches=44` `history_shrink=0.102`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0088)	`blend_weight=0.0088` `history_matches=44` `season_overlap_discount=1.0` `per90_to_match_factor=0.9899` `prior_reliability=0.1333` `competition=Premier League`	totalCross: 4.0121 -> 3.9941
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Vitaly Janelt', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.493, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v1', 'xg_multiplier': 1.063, 'skill_edge': 0.0949, 'height_edge': -0.052, 'attacker_skill': 0.8855, 'defender_skill': 0.6747, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Vitaly Janelt', 'market_value': 21000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.6747}], 'xg_override': 0.2622}	aerialWon: 0.47 -> 0.8 totalShots: 2.2961 -> 2.4408 shotOffTarget: 1.4796 -> 1.5728 onTargetScoringAttempt: 0.8164 -> 0.8678 bigChanceMissed: 0.4621 -> 0.4912 expectedGoals: 0.2467 -> 0.2622 goals: 0.2295 -> 0.244 expectedGoalsOnTarget: 0.216 -> 0.2296
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.09 (squad xG sum 1.3706, xG scale 0.7945, xA scale 0.7945, team xA 0.76 = 0.70 x lambda)	`scoreline_reset=true` `team_lambda=1.089` `squad_xg_sum=1.3706` `squad_xa_sum=0.9608` `team_expected_assists=0.7634` `team_xa_to_xg_ratio=0.701`	expected_goals: 0.2622 -> 0.2083 expected_assists: 0.1998 -> 0.1588
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 83.0)	`match_squad_rank=1` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.188` `goal_rate=0.244` `scorer_weight=0.1962`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.1468` `assist_rate=0.1528` `provider_weight=0.1289`	-

Callum Hudson Odoi F 73 0 0.08 0.04 132% 1 0.4 6.9 33.0 31 XGBoost F_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0070, quality 0.686)	`market_value_eur=26000000` `quality_factor=0.6858` `boost_factor=1.007` `reference_quality=0.505` `history_matches=34` `history_shrink=0.1282`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Vitaly Janelt, Kristoffer Ajer', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle', 'duel_win_rate': 0.204, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 0.9033, 'skill_edge': -0.0138, 'height_edge': -0.052, 'attacker_skill': 0.6858, 'defender_skill': 0.7164, 'attacker_role': 'SS', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Vitaly Janelt', 'market_value': 21000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.6747}, {'player': 'Kristoffer Ajer', 'market_value': 27000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7581}], 'xg_override': 0.106}	totalShots: 1.2923 -> 1.1673 shotOffTarget: 0.8943 -> 0.8078 onTargetScoringAttempt: 0.398 -> 0.3595 goals: 0.1273 -> 0.115 expectedGoals: 0.1173 -> 0.106 expectedGoalsOnTarget: 0.0947 -> 0.0855 bigChanceMissed: 0.0856 -> 0.0773
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.09 (squad xG sum 1.3706, xG scale 0.7945, xA scale 0.7945, team xA 0.76 = 0.70 x lambda)	`scoreline_reset=true` `team_lambda=1.089` `squad_xg_sum=1.3706` `squad_xa_sum=0.9608` `team_expected_assists=0.7634` `team_xa_to_xg_ratio=0.701`	expected_goals: 0.106 -> 0.0842 expected_assists: 0.0874 -> 0.0694
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 78.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0808` `goal_rate=0.115` `scorer_weight=0.0891`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.067` `assist_rate=0.305` `provider_weight=0.0868`	-

Liam Delap F 82 0 0.27 0.18 50% 3 1.3 6.8 51.1 32 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0097, quality 0.764)	`market_value_eur=37000000` `quality_factor=0.7641` `boost_factor=1.0097` `reference_quality=0.505` `history_matches=35` `history_shrink=0.125`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0016)	`blend_weight=0.0016` `history_matches=35` `season_overlap_discount=1.0` `per90_to_match_factor=0.9063` `prior_reliability=0.02` `competition=Premier League`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Vitaly Janelt, Kristoffer Ajer', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle', 'duel_win_rate': 0.275, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 1.008, 'skill_edge': 0.0215, 'height_edge': -0.016, 'attacker_skill': 0.7641, 'defender_skill': 0.7164, 'attacker_role': 'SS', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Vitaly Janelt', 'market_value': 21000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.6747}, {'player': 'Kristoffer Ajer', 'market_value': 27000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7581}], 'xg_override': 0.3392}	totalShots: 2.3631 -> 2.382 shotOffTarget: 1.545 -> 1.5574 onTargetScoringAttempt: 0.8152 -> 0.8217
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.09 (squad xG sum 1.3706, xG scale 0.7945, xA scale 0.7945, team xA 0.76 = 0.70 x lambda)	`scoreline_reset=true` `team_lambda=1.089` `squad_xg_sum=1.3706` `squad_xa_sum=0.9608` `team_expected_assists=0.7634` `team_xa_to_xg_ratio=0.701`	expected_goals: 0.3392 -> 0.2696 expected_assists: 0.062 -> 0.0493
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 77.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.2363` `goal_rate=0.2886` `scorer_weight=0.2346`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0481` `assist_rate=0.0` `provider_weight=0.0394`	-

Xaver Schlager M 86 0 0.04 0.02 73% 1 0.8 6.9 40.6 14 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0075, quality 0.570)	`market_value_eur=14200000` `quality_factor=0.5704` `boost_factor=1.0075` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0200)	`blend_weight=0.02` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=0.957` `prior_reliability=0.08` `competition=Premier League`	interceptionWon: 1.5628 -> 1.5315 fouls: 1.2287 -> 1.2041 wonTackle: 2.4026 -> 2.3864 totalShots: 1.016 -> 1.0278 totalCross: 0.4449 -> 0.452
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Keane Lewis-Potter', 'matchup_zone': 'Right Wingback vs Left Back — Overlapping Run', 'duel_win_rate': 0.347, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v1', 'xg_multiplier': 0.9215, 'skill_edge': -0.0694, 'height_edge': 0.016, 'attacker_skill': 0.6339, 'defender_skill': 0.7881, 'attacker_role': 'RWB', 'attacker_mv_price_index': 0.7, 'defender_adjusted_parts': [{'player': 'Keane Lewis-Potter', 'market_value': 29000000, 'role': 'LB', 'price_index': 0.7, 'adjusted_quality': 0.7881}], 'xg_override': 0.0536}	aerialWon: 0.6782 -> 0.9 totalShots: 1.0278 -> 0.9471 shotOffTarget: 0.798 -> 0.7354 onTargetScoringAttempt: 0.2136 -> 0.1968 bigChanceMissed: 0.1161 -> 0.107 goals: 0.0701 -> 0.0646
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 0 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=0`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.09 (squad xG sum 1.3706, xG scale 0.7945, xA scale 0.7945, team xA 0.76 = 0.70 x lambda)	`scoreline_reset=true` `team_lambda=1.089` `squad_xg_sum=1.3706` `squad_xa_sum=0.9608` `team_expected_assists=0.7634` `team_xa_to_xg_ratio=0.701`	expected_goals: 0.0536 -> 0.0426 expected_assists: 0.0789 -> 0.0627
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 78.0)	`match_squad_rank=5` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0417` `goal_rate=0.0646` `scorer_weight=0.0471`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0608` `assist_rate=0.0647` `provider_weight=0.0558`	-

Team totals 0 1.09 16 6.9

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
John Victor	G	4.5m	11.3%	Market-value substitute prior
Ousmane Diomande	D	15.0m	36.4%	Market-value substitute prior
Ola Aina	D	14.2m	23.0%	Market-value substitute prior
Omari Hutchinson	M	23.0m	54.6%	Market-value substitute prior
Ibrahim Sangaré	M	22.8m	32.3%	Market-value substitute prior
Dan Ndoye	F	26.5m	35.9%	Market-value substitute prior

Antagonist / tactical summary

Midfield is evenly contested — Mamadou Sangare vs Morgan Gibbs-White will be the pivotal duel. Total antagonist xG — Home: 0.91, Away: 0.96. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Shadow Striker vs CDM + CB — Zone 14 Battle.

Midfield — possession 50.2% / 49.8%, dominance ratio 0.48, key duel: Mamadou Sangare vs Morgan Gibbs-White

Mathias Jensen xG 0.05Igor Thiago xG 0.33Dango Ouattara xG 0.22Callum Hudson Odoi xG 0.11Liam Delap xG 0.34James McAtee xG 0.07

Summary

Antagonist matchup analysis for Brentford vs Forest. Most likely scoreline: 1-0. Lambdas derived from 4 home and 6 away individual player duels. Head-to-head: Brentford 4W-3D-3L vs Forest over 10 meetings.

Tactical parity between 4-2-3-1 and 3-4-2-1. Match outcome driven primarily by individual player quality. Midfield is evenly contested — Mamadou Sangare vs Morgan Gibbs-White will be the pivotal duel. Total antagonist xG — Home: 0.91, Away: 0.96. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Shadow Striker vs CDM + CB — Zone 14 Battle.

Coventry vs Sunderland

Premier League · 20261031 · 4-2-3-1 vs 4-2-3-1

Predicted 1-1Winner: DrawWinning H 1-0 (8.5%)Winning A 0-1 (8.5%)Confidence 0.69λ home/away 1.39 / 1.39Most-likely P 12.1%

Actual vs predicted: result not available yet (upcoming fixture).

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Coventry vs Sunderland. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 4-2-3-1 vs 4-2-3-1 (duel lambdas: Home 1.200, Away 1.347). Final calibrated Poisson lambdas: Home 1.389, Away 1.392 (most likely score 1-1).

Home Starters 11/11 XGBoost (7 FBref) Away Starters 11/11 XGBoost (11 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 1.20 / 1.35 Final λ (H/A) 1.39 / 1.39

Modeling Pipeline Stages (5 layers)

1. XGBoost player regressions for volume & event rates with market-value quality scaling
2. Multi-source prior blending (FBref 2026-2027 tables + Sofascore season history)
3. Antagonist formation duel matrix resolving 1v1 and zonal matchups
4. Team-level prior calibration (FBref possession and attack/defense efficiencies)
5. Bivariate Poisson matrix with Dixon-Coles low-score correction

Standing & form context

Metric	Coventry	Sunderland
League position	19.00	20.00
Points per game	1.01	0.94
Attack index	0.82	0.82
Defense (concede) index	1.10	1.14
Form multiplier	1.00	1.00
Standing λ multiplier	0.96	0.97

Lambda decomposition (how the goal rates were built)

Stage	Coventry λ	Sunderland λ
Pre-competitive-floor	1.226	1.318
Pre-calibration (Micro-duels)	1.200	1.347
Final λ	1.389	1.392

competition calibration scale 1.093competition observed total goals 2.781quality gap applied Truemismatch score pressure applied Falsecompetitive underdog floor applied Falsestage discount 1.0

Scoreline distribution & markets

Top 5 scorelines	P
1-1	12.1%
0-1	8.5%
1-0	8.5%
1-2	8.3%
2-1	8.3%

Home win	37.1%
Draw	25.6%
Away win	37.3%
Over 2.5	52.6%
BTTS yes	56.5%

Pregame fair line — favorite Sunderland (37.3%), fair total 2.78, fair spread -0.00

Projected players vs season baseline

Proj = projected for this match · Season = this season's per-game average from match data · deviation badge compares projected xG to the player's season xG/game.

Coventry — projected line vs season baseline (4-2-3-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Sidiki Cherif F 73 1 0.34 0.03 909% 2 0.5 6.7 24.5 11 XGBoost F_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0226, quality 0.641)	`market_value_eur=21000000` `quality_factor=0.6408` `boost_factor=1.0226` `reference_quality=0.505` `history_matches=4` `history_shrink=0.5556`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Trai Hume, Reinildo Mandava', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.26, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.0875, 'skill_edge': 0.0275, 'height_edge': 0.032, 'attacker_skill': 0.6408, 'defender_skill': 0.5797, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Trai Hume', 'market_value': 20000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.6914}, {'player': 'Reinildo Mandava', 'market_value': 4099999, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4679}], 'xg_override': 0.2302}	totalShots: 1.6308 -> 1.7735 shotOffTarget: 1.0258 -> 1.1156 onTargetScoringAttempt: 0.605 -> 0.6579 bigChanceMissed: 0.4289 -> 0.4664 goals: 0.231 -> 0.2512 expectedGoals: 0.2117 -> 0.2302 expectedGoalsOnTarget: 0.1057 -> 0.1149
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 0.9518, xG scale 1.4594, xA scale 1.5808, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.389` `squad_xg_sum=0.9518` `squad_xa_sum=0.5711` `team_expected_assists=0.9029` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.2302 -> 0.336 expected_assists: 0.0311 -> 0.0492
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 15 by projected rating, fifa_overall 72.0)	`match_squad_rank=15` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.2854` `goal_rate=0.2512` `scorer_weight=0.3612`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.048` `assist_rate=0.0` `provider_weight=0.0394`	-

Brandon Thomas-Asante F 79 0 0.25 0.05 435% 2 0.9 6.8 41.6 16 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0104, quality 0.595)	`market_value_eur=16500000` `quality_factor=0.5953` `boost_factor=1.0104` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0183)	`blend_weight=0.0183` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=0.8749` `prior_reliability=0.0733` `competition=Premier League`	totalShots: 1.7924 -> 1.7596 totalCross: 0.5064 -> 0.5263 interceptionWon: 2.2053 -> 2.2232 wasFouled: 3.091 -> 3.0781 onTargetScoringAttempt: 0.5777 -> 0.5671 fouls: 0.504 -> 0.4948 goalAssist: 0.4284 -> 0.4205
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Trai Hume, Reinildo Mandava', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.207, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.0103, 'skill_edge': 0.007, 'height_edge': 0.0, 'attacker_skill': 0.5953, 'defender_skill': 0.5797, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Trai Hume', 'market_value': 20000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.6914}, {'player': 'Reinildo Mandava', 'market_value': 4099999, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4679}], 'xg_override': 0.1695}	totalShots: 1.7818 -> 1.8002 shotOffTarget: 1.2147 -> 1.2272 onTargetScoringAttempt: 0.5671 -> 0.5729
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 0.9518, xG scale 1.4594, xA scale 1.5808, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.389` `squad_xg_sum=0.9518` `squad_xa_sum=0.5711` `team_expected_assists=0.9029` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1695 -> 0.2474 expected_assists: 0.0289 -> 0.0457
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 73.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2192` `goal_rate=0.1908` `scorer_weight=0.2714`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0447` `assist_rate=0.4205` `provider_weight=0.0571`	-

Jack Rudoni M 81 0 0.24 0.15 58% 2 0.8 6.9 45.9 17 XGBoost M_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0157, quality 0.641)	`market_value_eur=21000000` `quality_factor=0.6408` `boost_factor=1.0157` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Nordi Mukiele', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle', 'duel_win_rate': 0.466, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9791, 'skill_edge': -0.0142, 'height_edge': 0.0, 'attacker_skill': 0.6408, 'defender_skill': 0.6723, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Nordi Mukiele', 'market_value': 17100000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.6723}], 'xg_override': 0.1672}	aerialWon: 1.0917 -> 1.2 totalShots: 1.7751 -> 1.738 shotOffTarget: 1.1407 -> 1.1169 onTargetScoringAttempt: 0.6343 -> 0.621 expectedGoalsOnTarget: 0.3367 -> 0.3297 bigChanceMissed: 0.2666 -> 0.261
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 0.9518, xG scale 1.4594, xA scale 1.5808, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.389` `squad_xg_sum=0.9518` `squad_xa_sum=0.5711` `team_expected_assists=0.9029` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1672 -> 0.244 expected_assists: 0.1506 -> 0.238
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 75.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2165` `goal_rate=0.1982` `scorer_weight=0.2788`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=1` `assist_probability=0.2118` `assist_rate=0.0972` `provider_weight=0.1904`	-

Carl Rushworth G 90 0 0.00 - n/a 0.0 6.9 36.8 14 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0157, quality 0.641)	`market_value_eur=21000000` `quality_factor=0.6408` `boost_factor=1.0157` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0333)	`blend_weight=0.0333` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	saves: 2.808 -> 2.7644
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Brian Brobbey, Enzo Le Fée', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.3471}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 0.9518, xG scale 1.4594, xA scale 1.5808, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.389` `squad_xg_sum=0.9518` `squad_xa_sum=0.5711` `team_expected_assists=0.9029` `team_xa_to_xg_ratio=0.65`	shots_faced: 12.8 -> 13.2 expected_saves: 3.13 -> 3.23 high_pressure_saves: 1.08 -> 1.11
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 76.0)	`match_squad_rank=4` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=1.5573`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=0.8636`	-

Milan van Ewijk D 90 0 0.00 0.00 433% 0 0.1 6.7 37.9 16 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0193, quality 0.660)	`market_value_eur=23000000` `quality_factor=0.6595` `boost_factor=1.0193` `reference_quality=0.505` `history_matches=7` `history_shrink=0.4167`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0361)	`blend_weight=0.0361` `history_matches=7` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	wasFouled: 0.8553 -> 0.8244 wonTackle: 1.1908 -> 1.1659 interceptionWon: 1.1694 -> 1.1452 fouls: 0.5672 -> 0.5828 totalShots: 0.3901 -> 0.376
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Brian Brobbey', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 0.9518, xG scale 1.4594, xA scale 1.5808, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.389` `squad_xg_sum=0.9518` `squad_xa_sum=0.5711` `team_expected_assists=0.9029` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0583 -> 0.0922
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 11 by projected rating, fifa_overall 75.0)	`match_squad_rank=11` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0044` `goal_rate=0.0` `scorer_weight=0.0042`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0881` `assist_rate=0.0` `provider_weight=0.0738`	-

Stephen Mfuni D 90 0 0.15 0.02 603% 1 0.3 6.7 15.5 14 XGBoost D_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9873, quality 0.437)	`market_value_eur=2700000` `quality_factor=0.4374` `boost_factor=0.9873` `reference_quality=0.505` `history_matches=3` `history_shrink=0.625`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Brian Brobbey', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 0.9518, xG scale 1.4594, xA scale 1.5808, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.389` `squad_xg_sum=0.9518` `squad_xa_sum=0.5711` `team_expected_assists=0.9029` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1026 -> 0.1497
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 14 by projected rating, fifa_overall 69.0)	`match_squad_rank=14` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.139` `goal_rate=0.3262` `scorer_weight=0.1871`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0096` `assist_rate=0.0` `provider_weight=0.0086`	-

Ethan Pinnock D 88 0 0.03 0.01 441% 1 0.1 6.7 16.3 34 XGBoost D_service

Step-by-step (9 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9937, quality 0.437)	`market_value_eur=2800000` `quality_factor=0.4374` `boost_factor=0.9937` `reference_quality=0.505` `history_matches=11` `history_shrink=0.3125`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Brian Brobbey', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 0.9518, xG scale 1.4594, xA scale 1.5808, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.389` `squad_xg_sum=0.9518` `squad_xa_sum=0.5711` `team_expected_assists=0.9029` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0203 -> 0.0296 expected_assists: 0.0143 -> 0.0226
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 10 by projected rating, fifa_overall 75.0)	`match_squad_rank=10` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0292` `goal_rate=0.0` `scorer_weight=0.0237`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0223` `assist_rate=0.0` `provider_weight=0.0181`	-

Jay DaSilva D 90 0 0.08 0.05 66% 0 0.3 6.7 44.0 17 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0042, quality 0.541)	`market_value_eur=11600000` `quality_factor=0.5411` `boost_factor=1.0042` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0333)	`blend_weight=0.0333` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=0.9959` `prior_reliability=0.1333` `competition=Premier League`	interceptionWon: 0.8395 -> 0.8281 onTargetScoringAttempt: 0.2053 -> 0.2151 goalAssist: 0.1496 -> 0.1446
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Granit Xhaka', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 0.9518, xG scale 1.4594, xA scale 1.5808, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.389` `squad_xg_sum=0.9518` `squad_xa_sum=0.5711` `team_expected_assists=0.9029` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0549 -> 0.0801 expected_assists: 0.087 -> 0.1375
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 13 by projected rating, fifa_overall 73.0)	`match_squad_rank=13` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.077` `goal_rate=0.1229` `scorer_weight=0.1001`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1285` `assist_rate=0.1446` `provider_weight=0.1719`	-

Matt Grimes M 88 0 0.05 0.01 359% 0 0.4 6.9 39.0 14 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9953, quality 0.465)	`market_value_eur=5200000` `quality_factor=0.4646` `boost_factor=0.9953` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0333)	`blend_weight=0.0333` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=0.9761` `prior_reliability=0.1333` `competition=Premier League`	totalCross: 2.0676 -> 2.1613 fouls: 0.7791 -> 0.7531 wonTackle: 0.6086 -> 0.5883 wasFouled: 0.4719 -> 0.4887 interceptionWon: 0.3916 -> 0.3785 onTargetScoringAttempt: 0.1526 -> 0.1475
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Noah Sadiki', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.284, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.8004, 'skill_edge': -0.1558, 'height_edge': 0.02, 'attacker_skill': 0.4759, 'defender_skill': 0.8221, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Noah Sadiki', 'market_value': 42000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8221}], 'xg_override': 0.036}	aerialWon: 0.7309 -> 1.0 totalShots: 0.5891 -> 0.4715 shotOffTarget: 0.44 -> 0.3522 onTargetScoringAttempt: 0.1475 -> 0.1181 goals: 0.0549 -> 0.0439 expectedGoals: 0.045 -> 0.036
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 0.9518, xG scale 1.4594, xA scale 1.5808, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.389` `squad_xg_sum=0.9518` `squad_xa_sum=0.5711` `team_expected_assists=0.9029` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.036 -> 0.0525 expected_assists: 0.068 -> 0.1075
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 76.0)	`match_squad_rank=6` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0511` `goal_rate=0.0439` `scorer_weight=0.0573`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1019` `assist_rate=0.0615` `provider_weight=0.1016`	-

Caleb Yirenkyi M 75 0 0.05 0.00 2008% 1 0.1 6.7 37.1 14 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0178, quality 0.660)	`market_value_eur=23000000` `quality_factor=0.6595` `boost_factor=1.0178` `reference_quality=0.505` `history_matches=8` `history_shrink=0.3846`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0250)	`blend_weight=0.025` `history_matches=8` `season_overlap_discount=1.0` `per90_to_match_factor=0.8281` `prior_reliability=0.1` `competition=Premier League`	interceptionWon: 1.0472 -> 1.021 wonTackle: 0.3542 -> 0.3453 fouls: 1.9683 -> 1.9605 totalCross: 0.2785 -> 0.2853
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Noah Sadiki', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.399, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.9686, 'skill_edge': -0.0573, 'height_edge': 0.036, 'attacker_skill': 0.6946, 'defender_skill': 0.8221, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Noah Sadiki', 'market_value': 42000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8221}], 'xg_override': 0.0355}	aerialWon: 0.8202 -> 1.1 totalShots: 0.5422 -> 0.5252 shotOffTarget: 0.3957 -> 0.3833
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 0.9518, xG scale 1.4594, xA scale 1.5808, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.389` `squad_xg_sum=0.9518` `squad_xa_sum=0.5711` `team_expected_assists=0.9029` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0355 -> 0.0518 expected_assists: 0.0364 -> 0.0575
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 12 by projected rating, fifa_overall 73.0)	`match_squad_rank=12` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0505` `goal_rate=0.0447` `scorer_weight=0.0574`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0559` `assist_rate=0.059` `provider_weight=0.0719`	-

Tatsuhiro Sakamoto F 81 0 0.14 0.01 1167% 1 0.2 6.8 16.4 25 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9914, quality 0.454)	`market_value_eur=4300000` `quality_factor=0.4535` `boost_factor=0.9914` `reference_quality=0.505` `history_matches=4` `history_shrink=0.5556`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0096)	`blend_weight=0.0096` `history_matches=4` `season_overlap_discount=1.0` `per90_to_match_factor=0.9018` `prior_reliability=0.0267` `competition=Premier League`	totalCross: 1.7243 -> 1.7729 wonTackle: 1.3155 -> 1.3463 totalShots: 1.1772 -> 1.1903 wasFouled: 1.0648 -> 1.0545 fouls: 0.8867 -> 0.8782 interceptionWon: 0.5359 -> 0.5307
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Trai Hume, Reinildo Mandava', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 0.8578, 'skill_edge': -0.0567, 'height_edge': -0.04, 'attacker_skill': 0.4535, 'defender_skill': 0.5797, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Trai Hume', 'market_value': 20000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.6914}, {'player': 'Reinildo Mandava', 'market_value': 4099999, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4679}], 'xg_override': 0.096}	totalShots: 1.1903 -> 1.021 shotOffTarget: 0.7915 -> 0.6789 onTargetScoringAttempt: 0.3819 -> 0.3276 goals: 0.1303 -> 0.1118 expectedGoals: 0.1119 -> 0.096
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 0.9518, xG scale 1.4594, xA scale 1.5808, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.389` `squad_xg_sum=0.9518` `squad_xa_sum=0.5711` `team_expected_assists=0.9029` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.096 -> 0.1401 expected_assists: 0.0395 -> 0.0624
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 74.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1307` `goal_rate=0.1118` `scorer_weight=0.1553`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0605` `assist_rate=0.0` `provider_weight=0.0499`	-

Team totals 1 1.39 10 6.8

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Daniel Bentley	G	4.4m	10.9%	Market-value substitute prior
Bobby Thomas	D	14.4m	23.3%	Market-value substitute prior
Aurèle Amenda	D	14.2m	22.9%	Market-value substitute prior
Victor Torp	M	23.9m	70.3%	Market-value substitute prior
Josh Eccles	M	22.3m	51.3%	Market-value substitute prior
Loum Tchaouna	F	26.0m	34.9%	Market-value substitute prior

Sunderland — projected line vs season baseline (4-2-3-1)

Player Pos Min Proj G Proj xG Season xG/g Proj Sh Season Sh/g Proj Rtg Season Rtg Rec Model / FBref Prior

Trai Hume D 87 0 0.04 0.05 ok 1 0.9 6.8 54.1 36 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0038, quality 0.631)	`market_value_eur=20000000` `quality_factor=0.6311` `boost_factor=1.0038` `reference_quality=0.505` `history_matches=45` `history_shrink=0.1`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0082)	`blend_weight=0.0082` `history_matches=45` `season_overlap_discount=1.0` `per90_to_match_factor=0.9672` `prior_reliability=0.1267` `competition=Premier League`	wonTackle: 1.0986 -> 1.1063 wasFouled: 0.6325 -> 0.6273
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Sidiki Cherif', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 1.3995, xG scale 0.9947, xA scale 1.1255, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.392` `squad_xg_sum=1.3995` `squad_xa_sum=0.8039` `team_expected_assists=0.9048` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.069 -> 0.0777
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 10 by projected rating, fifa_overall 78.0)	`match_squad_rank=10` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0394` `goal_rate=0.0703` `scorer_weight=0.0503`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0748` `assist_rate=0.001` `provider_weight=0.0622`	-

Luke O'Nien D 87 0 0.05 0.01 515% 1 0.2 6.8 22.9 38 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9949, quality 0.437)	`market_value_eur=660000` `quality_factor=0.4374` `boost_factor=0.9949` `reference_quality=0.505` `history_matches=15` `history_shrink=0.25`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0076)	`blend_weight=0.0076` `history_matches=15` `season_overlap_discount=1.0` `per90_to_match_factor=0.9627` `prior_reliability=0.0467` `competition=Premier League`	fouls: 0.7124 -> 0.7279 totalCross: 1.4 -> 1.4102 interceptionWon: 0.8092 -> 0.8031 wasFouled: 0.7525 -> 0.7468 wonTackle: 0.7108 -> 0.7054 totalShots: 0.6552 -> 0.6602
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Tatsuhiro Sakamoto', 'matchup_zone': 'Shape fallback'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 1.3995, xG scale 0.9947, xA scale 1.1255, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.392` `squad_xg_sum=1.3995` `squad_xa_sum=0.8039` `team_expected_assists=0.9048` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 73.0)	`match_squad_rank=7` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0493` `goal_rate=0.0` `scorer_weight=0.0405`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0188` `assist_rate=0.1575` `provider_weight=0.0238`	-

Nordi Mukiele D 83 0 0.09 0.07 42% 1 0.8 6.9 51.4 31 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0033, quality 0.602)	`market_value_eur=17100000` `quality_factor=0.6016` `boost_factor=1.0033` `reference_quality=0.505` `history_matches=39` `history_shrink=0.1136`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0049)	`blend_weight=0.0049` `history_matches=39` `season_overlap_discount=1.0` `per90_to_match_factor=0.9248` `prior_reliability=0.0667` `competition=Premier League`	totalCross: 2.6647 -> 2.6561 totalShots: 1.2723 -> 1.266 wonTackle: 1.23 -> 1.2239
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Jack Rudoni', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 1.3995, xG scale 0.9947, xA scale 1.1255, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.392` `squad_xg_sum=1.3995` `squad_xa_sum=0.8039` `team_expected_assists=0.9048` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0652 -> 0.0734
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 81.0)	`match_squad_rank=4` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.089` `goal_rate=0.0506` `scorer_weight=0.0746`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0708` `assist_rate=0.0255` `provider_weight=0.0587`	-

Reinildo Mandava D 86 0 0.00 0.00 49% 0 0.0 6.7 48.5 26 FBref Prior D_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost D_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 0.9978, quality 0.451)	`market_value_eur=4099999` `quality_factor=0.4511` `boost_factor=0.9978` `reference_quality=0.505` `history_matches=32` `history_shrink=0.1351`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0117)	`blend_weight=0.0117` `history_matches=32` `season_overlap_discount=1.0` `per90_to_match_factor=0.9611` `prior_reliability=0.1333` `competition=Premier League`	fouls: 1.2211 -> 1.2349 totalCross: 0.6665 -> 0.6756 interceptionWon: 1.1924 -> 1.2009
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Sidiki Cherif', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle'}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 1.3995, xG scale 0.9947, xA scale 1.1255, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.392` `squad_xg_sum=1.3995` `squad_xa_sum=0.8039` `team_expected_assists=0.9048` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 13 by projected rating, fifa_overall 80.0)	`match_squad_rank=13` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv_filled` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0013` `goal_rate=0.0` `scorer_weight=0.0013`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0346` `assist_rate=0.013` `provider_weight=0.0282`	-

Granit Xhaka M 90 0 0.05 0.02 130% 1 0.5 7.2 59.0 30 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0000, quality 0.505)	`market_value_eur=8500000` `quality_factor=0.5047` `boost_factor=1.0` `reference_quality=0.505` `history_matches=42` `history_shrink=0.1064`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0092)	`blend_weight=0.0092` `history_matches=42` `season_overlap_discount=1.0` `per90_to_match_factor=0.996` `prior_reliability=0.1333` `competition=Premier League`	wasFouled: 1.7737 -> 1.7895 totalShots: 0.681 -> 0.6747
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Jay DaSilva', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle', 'duel_win_rate': 0.499, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.0278, 'skill_edge': -0.0411, 'height_edge': 0.06, 'attacker_skill': 0.5047, 'defender_skill': 0.5961, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Jay DaSilva', 'market_value': 11600000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.5961}], 'xg_override': 0.0487}	aerialWon: 0.7194 -> 1.2 totalShots: 0.679 -> 0.6979 shotOffTarget: 0.4729 -> 0.486 onTargetScoringAttempt: 0.2061 -> 0.2118
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 1.3995, xG scale 0.9947, xA scale 1.1255, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.392` `squad_xg_sum=1.3995` `squad_xa_sum=0.8039` `team_expected_assists=0.9048` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.1624 -> 0.1828
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 85.0)	`match_squad_rank=1` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0472` `goal_rate=0.0464` `scorer_weight=0.0473`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=1` `assist_probability=0.1671` `assist_rate=0.099` `provider_weight=0.1462`	-

Noah Sadiki M 86 0 0.08 0.05 56% 1 0.6 6.8 54.6 31 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0093, quality 0.791)	`market_value_eur=42000000` `quality_factor=0.7909` `boost_factor=1.0093` `reference_quality=0.505` `history_matches=41` `history_shrink=0.1087`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0094)	`blend_weight=0.0094` `history_matches=41` `season_overlap_discount=1.0` `per90_to_match_factor=0.9541` `prior_reliability=0.1333` `competition=Premier League`	wonTackle: 0.6898 -> 0.6968 wasFouled: 1.1767 -> 1.1836
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Matt Grimes', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.556, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.1996, 'skill_edge': 0.1558, 'height_edge': -0.02, 'attacker_skill': 0.8221, 'defender_skill': 0.4759, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Matt Grimes', 'market_value': 5200000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.4759}], 'xg_override': 0.0819}	aerialWon: 0.5138 -> 0.9 totalShots: 0.8317 -> 0.9977 shotOffTarget: 0.4966 -> 0.5957 onTargetScoringAttempt: 0.3351 -> 0.402 bigChanceMissed: 0.1504 -> 0.1804 goals: 0.0765 -> 0.0918 expectedGoals: 0.0683 -> 0.0819 expectedGoalsOnTarget: 0.0301 -> 0.0361
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 1.3995, xG scale 0.9947, xA scale 1.1255, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.392` `squad_xg_sum=1.3995` `squad_xa_sum=0.8039` `team_expected_assists=0.9048` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0675 -> 0.076
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 80.0)	`match_squad_rank=5` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0783` `goal_rate=0.0918` `scorer_weight=0.0876`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0732` `assist_rate=0.1112` `provider_weight=0.095`	-

Chris Rigg M 73 0 0.11 0.02 539% 1 0.3 6.7 25.4 34 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0091, quality 0.669)	`market_value_eur=24000000` `quality_factor=0.6685` `boost_factor=1.0091` `reference_quality=0.505` `history_matches=22` `history_shrink=0.1852`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0016)	`blend_weight=0.0016` `history_matches=22` `season_overlap_discount=1.0` `per90_to_match_factor=0.8064` `prior_reliability=0.0133` `competition=Premier League`	-
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Matt Grimes', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.519, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.1451, 'skill_edge': 0.1027, 'height_edge': -0.004, 'attacker_skill': 0.704, 'defender_skill': 0.4759, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Matt Grimes', 'market_value': 5200000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.4759}], 'xg_override': 0.1145}	aerialWon: 0.28 -> 1.0 totalShots: 1.1327 -> 1.2971 shotOffTarget: 0.7085 -> 0.8113 onTargetScoringAttempt: 0.4242 -> 0.4858 expectedGoals: 0.1 -> 0.1145 expectedGoalsOnTarget: 0.0985 -> 0.1128 goals: 0.0935 -> 0.1071
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 1.3995, xG scale 0.9947, xA scale 1.1255, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.392` `squad_xg_sum=1.3995` `squad_xa_sum=0.8039` `team_expected_assists=0.9048` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.059 -> 0.0664
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 14 by projected rating, fifa_overall 74.0)	`match_squad_rank=14` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1077` `goal_rate=0.1071` `scorer_weight=0.1088`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0642` `assist_rate=0.0749` `provider_weight=0.0761`	-

Enzo Le Fée M 87 0 0.16 0.15 ok 2 1.1 6.9 57.3 33 FBref Prior M_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost M_service raw output over 8 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0060, quality 0.702)	`market_value_eur=28000000` `quality_factor=0.7022` `boost_factor=1.006` `reference_quality=0.505` `history_matches=44` `history_shrink=0.102`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0084)	`blend_weight=0.0084` `history_matches=44` `season_overlap_discount=1.0` `per90_to_match_factor=0.9634` `prior_reliability=0.1267` `competition=Premier League`	totalCross: 3.0227 -> 3.0143
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Matt Grimes, Caleb Yirenkyi', 'matchup_zone': 'Attacking Mid vs Double Pivot — Tight Space Battle', 'duel_win_rate': 0.27, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.0438, 'skill_edge': 0.0578, 'height_edge': -0.028, 'attacker_skill': 0.7136, 'defender_skill': 0.5852, 'attacker_role': 'CAM', 'attacker_mv_price_index': 0.95, 'defender_adjusted_parts': [{'player': 'Matt Grimes', 'market_value': 5200000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.4759}, {'player': 'Caleb Yirenkyi', 'market_value': 23000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.6946}], 'xg_override': 0.1606}	aerialWon: 0.4435 -> 0.9 totalShots: 1.5885 -> 1.6581 shotOffTarget: 1.0833 -> 1.1307 onTargetScoringAttempt: 0.5014 -> 0.5234 expectedGoalsOnTarget: 0.1656 -> 0.1729 goals: 0.1613 -> 0.1684 expectedGoals: 0.1539 -> 0.1606 bigChanceMissed: 0.1445 -> 0.1508
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 1.3995, xG scale 0.9947, xA scale 1.1255, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.392` `squad_xg_sum=1.3995` `squad_xa_sum=0.8039` `team_expected_assists=0.9048` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.1462 -> 0.1645
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 80.0)	`match_squad_rank=3` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1476` `goal_rate=0.1684` `scorer_weight=0.1656`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.1517` `assist_rate=0.108` `provider_weight=0.1325`	-

Nilson Angulo F 71 0 0.16 0.07 132% 2 0.6 6.8 41.0 32 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0090, quality 0.625)	`market_value_eur=19400000` `quality_factor=0.6251` `boost_factor=1.009` `reference_quality=0.505` `history_matches=15` `history_shrink=0.25`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0152)	`blend_weight=0.0152` `history_matches=15` `season_overlap_discount=1.0` `per90_to_match_factor=0.7931` `prior_reliability=0.0933` `competition=Premier League`	totalCross: 0.9321 -> 0.9523 wasFouled: 1.1039 -> 1.1215 interceptionWon: 1.1776 -> 1.1683 goals: 0.1787 -> 0.1846 totalShots: 1.5176 -> 1.5124 fouls: 1.4681 -> 1.463
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Ethan Pinnock, Stephen Mfuni', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.235, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.0511, 'skill_edge': 0.0807, 'height_edge': -0.046, 'attacker_skill': 0.6251, 'defender_skill': 0.4457, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Ethan Pinnock', 'market_value': 2800000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4465}, {'player': 'Stephen Mfuni', 'market_value': 2700000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4449}], 'xg_override': 0.1559}	totalShots: 1.5176 -> 1.5951 shotOffTarget: 0.9337 -> 0.9814 onTargetScoringAttempt: 0.5839 -> 0.6137 bigChanceMissed: 0.3327 -> 0.3497 goals: 0.1846 -> 0.194 expectedGoals: 0.1483 -> 0.1559
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 1.3995, xG scale 0.9947, xA scale 1.1255, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.392` `squad_xg_sum=1.3995` `squad_xa_sum=0.8039` `team_expected_assists=0.9048` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0704 -> 0.0792
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 73.0)	`match_squad_rank=9` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1437` `goal_rate=0.194` `scorer_weight=0.1838`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0761` `assist_rate=0.1255` `provider_weight=0.099`	-

Robin Roefs G 90 0 0.00 - n/a 0.0 6.7 53.8 31 FBref Prior G_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost G_service raw output over 3 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0069, quality 0.725)	`market_value_eur=31000000` `quality_factor=0.7249` `boost_factor=1.0069` `reference_quality=0.505` `history_matches=43` `history_shrink=0.1042`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0090)	`blend_weight=0.009` `history_matches=43` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	saves: 2.8496 -> 2.8284
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` `duel_context={'opponent_faced': 'Sidiki Cherif, Brandon Thomas-Asante', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.1997}`	-
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	-
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 1.3995, xG scale 0.9947, xA scale 1.1255, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.392` `squad_xg_sum=1.3995` `squad_xa_sum=0.8039` `team_expected_assists=0.9048` `team_xa_to_xg_ratio=0.65`	shots_faced: 11.4 -> 13.2 expected_saves: 2.79 -> 3.23 high_pressure_saves: 0.96 -> 1.11
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 11 by projected rating, fifa_overall 81.0)	`match_squad_rank=11` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=1.5422`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.0` `provider_weight=0.0` `squad_weight_total=0.8379`	-

Brian Brobbey F 79 1 0.52 0.27 91% 4 1.4 7.0 53.3 30 FBref Prior F_service

Step-by-step (10 steps)

Step	Description	Key params	Changes
`xgboost_raw`	XGBoost F_service raw output over 7 target models	-	-
`market_value_boost`	Market-value quality scaling (boost 1.0118, quality 0.834)	`market_value_eur=53000000` `quality_factor=0.834` `boost_factor=1.0118` `reference_quality=0.505` `history_matches=37` `history_shrink=0.119`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0072)	`blend_weight=0.0072` `history_matches=37` `season_overlap_discount=1.0` `per90_to_match_factor=0.8738` `prior_reliability=0.0933` `competition=Premier League`	wasFouled: 1.9205 -> 1.9111 totalOffside: 0.7753 -> 0.7832 totalShots: 2.6429 -> 2.6369 onTargetScoringAttempt: 1.3294 -> 1.3242
`final_cap`	Stats clamped to historical constraints and shot-component identity	`stats_clamped=0` `shot_identity_applied=false`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Ethan Pinnock, Stephen Mfuni', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.321, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.1775, 'skill_edge': 0.1747, 'height_edge': -0.054, 'attacker_skill': 0.834, 'defender_skill': 0.4457, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Ethan Pinnock', 'market_value': 2800000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4465}, {'player': 'Stephen Mfuni', 'market_value': 2700000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4449}], 'xg_override': 0.521}	totalShots: 2.6377 -> 3.1059 onTargetScoringAttempt: 1.3242 -> 1.5592 shotOffTarget: 1.3135 -> 1.5466 goals: 0.5666 -> 0.6672 expectedGoalsOnTarget: 0.5294 -> 0.6234 bigChanceMissed: 0.4833 -> 0.5691 expectedGoals: 0.4425 -> 0.521
`scoreline_reset`	Team scoreline allocation: goal/assist counters reset to 0; the team's 1 goals are handed out as whole numbers by integer_goal_projection / integer_assist_projection	`scoreline_reset=true` `team_goals=1`	goals: 1.0 -> 0
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 1.39 (squad xG sum 1.3995, xG scale 0.9947, xA scale 1.1255, team xA 0.90 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.392` `squad_xg_sum=1.3995` `squad_xa_sum=0.8039` `team_expected_assists=0.9048` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0446 -> 0.0502
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 78.0)	`match_squad_rank=2` `squad_role=starter` `squad_rank_basis=projected_rating` `affects_projection=false` `ranking_source=sofascore_plus_fifa_csv` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.4043` `goal_rate=0.6672` `scorer_weight=0.6476`	-
`integer_assist_projection`	Team scoreline allocation: 1 projected assists distributed by provider weight (xA x creation x appearance; nobody assists his own goal)	`allocation_applied=true` `team_assists=1` `assists_allocated=0` `assist_probability=0.049` `assist_rate=0.0` `provider_weight=0.0402`	-

Team totals 1 1.39 17 6.8

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Anthony Patterson	G	4.6m	11.9%	Market-value substitute prior
Daniel Ballard	D	14.6m	23.8%	Market-value substitute prior
Lutsharel Geertruida	D	14.4m	33.7%	Market-value substitute prior
Habib Diarra	M	18.7m	71.8%	Market-value substitute prior
Dayann Methalie	M	23.0m	54.5%	Market-value substitute prior
Malick Fofana	F	26.4m	60.8%	Market-value substitute prior

Antagonist / tactical summary

Midfield is evenly contested — Caleb Yirenkyi vs Noah Sadiki will be the pivotal duel. Total antagonist xG — Home: 0.73, Away: 1.08. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Midfield — possession 48.5% / 51.5%, dominance ratio 0.46, key duel: Caleb Yirenkyi vs Noah Sadiki

Jack Rudoni xG 0.17Sidiki Cherif xG 0.23Brandon Thomas-Asante xG 0.17Granit Xhaka xG 0.05Enzo Le Fée xG 0.16Brian Brobbey xG 0.52

Summary

Antagonist matchup analysis for Coventry vs Sunderland. Most likely scoreline: 1-1. Lambdas derived from 6 home and 6 away individual player duels. Head-to-head: Coventry 4W-5D-1L vs Sunderland over 10 meetings.

Tactical parity between 4-2-3-1 and 4-2-3-1. Match outcome driven primarily by individual player quality. Midfield is evenly contested — Caleb Yirenkyi vs Noah Sadiki will be the pivotal duel. Total antagonist xG — Home: 0.73, Away: 1.08. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Generated 2026-10-31 13:37:35 UTC · Soccer prediction reporting