Soccer predictions: actual data vs predicted data

Slate 20261025 · 9 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
Celta vs Real Betis LaLiga	0-1	Real Betis	30.5% / 30.9% / 38.6%	32.4%	40.0%	0.71	0.92 / 1.08	-	pending
Liverpool vs Brighton Premier League	2-1	Liverpool	54.4% / 20.2% / 25.5%	72.6%	69.9%	0.74	2.26 / 1.51	-	pending
Man Utd vs Bournemouth Premier League	1-0	Man Utd	47.8% / 24.6% / 27.6%	53.5%	56.0%	0.73	1.64 / 1.19	-	pending
Hull vs Brentford Premier League	1-2	Brentford	21.2% / 22.3% / 56.5%	57.6%	56.5%	0.74	1.09 / 1.91	-	pending
Athletic Club vs Getafe LaLiga	0-0	Draw	38.1% / 32.2% / 29.7%	29.0%	37.0%	0.7	1.02 / 0.86	2-0 ()	MISS
Sunderland vs Leeds Premier League	1-1	Draw	34.3% / 25.6% / 40.1%	52.6%	56.4%	0.71	1.33 / 1.45	-	pending
Real Sociedad vs Levante LaLiga	2-1	Real Sociedad	58.4% / 21.9% / 19.8%	57.3%	55.5%	0.76	1.94 / 1.04	-	pending
Crystal Palace vs Newcastle Premier League	1-2	Newcastle	36.6% / 23.2% / 40.2%	63.8%	65.1%	0.7	1.60 / 1.69	-	pending
Barcelona vs Real Madrid LaLiga	2-1	Barcelona	61.9% / 17.6% / 20.4%	79.0%	72.9%	0.78	2.67 / 1.53	-	pending

Match detail

Celta vs Real Betis

LaLiga · 20261025 · 3-4-2-1 vs 4-2-3-1

Predicted 0-1Winner: Real BetisWinning H 1-0 (12.3%)Winning A 0-1 (14.4%)Confidence 0.71λ home/away 0.92 / 1.08Most-likely P 14.4%

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

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Celta vs Real Betis. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 3-4-2-1 vs 4-2-3-1 (duel lambdas: Home 0.898, Away 1.112). Final calibrated Poisson lambdas: Home 0.924, Away 1.080 (most likely score 0-1).

Home Starters 11/11 XGBoost (11 FBref) Away Starters 11/11 XGBoost (10 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 0.90 / 1.11 Final λ (H/A) 0.92 / 1.08

FBref Team Priors: Celta Vigo (Gls/90: 0.25, GA/90: 0.25 → E[λ]: 0.246) vs Real Betis (Gls/90: 1.3333333333333333, GA/90: 1.33 → E[λ]: 0.247) | 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	Celta	Real Betis
League position	8.00	4.00
Points per game	1.28	1.81
Attack index	0.93	1.02
Defense (concede) index	0.96	0.83
Form multiplier	0.97	1.03
Standing λ multiplier	0.95	0.94

Lambda decomposition (how the goal rates were built)

Stage	Celta λ	Real Betis λ
FBref Squad Prior E[λ] (weight 0.228)	0.246	0.247
Pre-competitive-floor	0.924	1.080
Pre-calibration (Micro-duels)	0.898	1.112
Final λ	0.924	1.080

competition calibration scale 1.0competition observed total goals 0.0quality gap applied Truemismatch score pressure applied Falsecompetitive underdog floor applied Falsestage discount 1.0

Scoreline distribution & markets

Top 5 scorelines	P
0-1	14.4%
0-0	13.6%
1-1	13.6%
1-0	12.3%
0-2	7.9%

Home win	30.5%
Draw	30.9%
Away win	38.6%
Over 2.5	32.4%
BTTS yes	40.0%

Pregame fair line — favorite Real Betis (38.6%), fair total 2.00, fair spread -0.16

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.

Celta — 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

Ferrán Jutglà F 73 0 0.16 0.18 ok 2 1.1 6.8 59.3 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 1.0009, quality 0.531)	`market_value_eur=10700000` `quality_factor=0.5306` `boost_factor=1.0009` `reference_quality=0.505` `history_matches=37` `history_shrink=0.119`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0134)	`blend_weight=0.0134` `history_matches=37` `season_overlap_discount=1.0` `per90_to_match_factor=0.8119` `prior_reliability=0.1733` `competition=La Liga`	totalShots: 1.446 -> 1.4309 totalOffside: 0.6429 -> 0.6343 onTargetScoringAttempt: 0.5843 -> 0.5765 wasFouled: 1.6484 -> 1.6556
`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': 'Pablo Fornals, Natan', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle', 'duel_win_rate': 0.181, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 0.869, 'skill_edge': -0.0571, 'height_edge': -0.032, 'attacker_skill': 0.5306, 'defender_skill': 0.6576, 'attacker_role': 'SS', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Pablo Fornals', 'market_value': 8400000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.5211}, {'player': 'Natan', 'market_value': 32000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7941}], 'xg_override': 0.1638}	totalShots: 1.4382 -> 1.2498 shotOffTarget: 0.8617 -> 0.7488 onTargetScoringAttempt: 0.5765 -> 0.501 expectedGoalsOnTarget: 0.3532 -> 0.3069 goals: 0.2086 -> 0.1813 expectedGoals: 0.1885 -> 0.1638 bigChanceMissed: 0.1424 -> 0.1237
`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 0.92 (squad xG sum 0.9410, xG scale 0.9819, xA scale 0.9819, team xA 0.69 = 0.75 x lambda)	`scoreline_reset=true` `team_lambda=0.924` `squad_xg_sum=0.941` `squad_xa_sum=0.7025` `team_expected_assists=0.6898` `team_xa_to_xg_ratio=0.7465`	-
`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 skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.1485` `goal_rate=0.1813` `scorer_weight=0.174`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0556` `assist_rate=0.0858` `provider_weight=0.0715`	-

Williot Swedberg F 66 0 0.19 0.09 102% 2 0.6 6.8 51.9 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 1.0042, quality 0.618)	`market_value_eur=18700000` `quality_factor=0.6181` `boost_factor=1.0042` `reference_quality=0.505` `history_matches=35` `history_shrink=0.125`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0130)	`blend_weight=0.013` `history_matches=35` `season_overlap_discount=1.0` `per90_to_match_factor=0.7327` `prior_reliability=0.16` `competition=La Liga`	wasFouled: 1.7832 -> 1.8037 totalShots: 1.3764 -> 1.3704
`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': 'Natan, Diego Llorente', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.194, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 0.9908, 'skill_edge': -0.0003, 'height_edge': -0.006, 'attacker_skill': 0.6181, 'defender_skill': 0.6187, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Natan', 'market_value': 32000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7941}, {'player': 'Diego Llorente', 'market_value': 2600000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4432}], 'xg_override': 0.1892}	totalShots: 1.3768 -> 1.3642 shotOffTarget: 0.7956 -> 0.7883 onTargetScoringAttempt: 0.5812 -> 0.5759
`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 0.92 (squad xG sum 0.9410, xG scale 0.9819, xA scale 0.9819, team xA 0.69 = 0.75 x lambda)	`scoreline_reset=true` `team_lambda=0.924` `squad_xg_sum=0.941` `squad_xa_sum=0.7025` `team_expected_assists=0.6898` `team_xa_to_xg_ratio=0.7465`	-
`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=0.1696` `goal_rate=0.1656` `scorer_weight=0.1675`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0724` `assist_rate=0.103` `provider_weight=0.0909`	-

Pablo Durán F 67 0 0.19 0.11 82% 2 0.9 6.8 51.4 28 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 0.9980, quality 0.451)	`market_value_eur=4099999` `quality_factor=0.4511` `boost_factor=0.998` `reference_quality=0.505` `history_matches=35` `history_shrink=0.125`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0081)	`blend_weight=0.0081` `history_matches=35` `season_overlap_discount=1.0` `per90_to_match_factor=0.7463` `prior_reliability=0.1` `competition=La Liga`	totalShots: 1.8661 -> 1.8551
`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': 'Pablo Fornals, Natan', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 0.8222, 'skill_edge': -0.0929, 'height_edge': -0.028, 'attacker_skill': 0.4511, 'defender_skill': 0.6576, 'attacker_role': 'SS', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Pablo Fornals', 'market_value': 8400000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.5211}, {'player': 'Natan', 'market_value': 32000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7941}], 'xg_override': 0.1972}	totalShots: 1.8639 -> 1.5325 shotOffTarget: 1.0848 -> 0.8919 onTargetScoringAttempt: 0.7791 -> 0.6406 goals: 0.3041 -> 0.25 expectedGoals: 0.2399 -> 0.1972 bigChanceMissed: 0.2375 -> 0.1953 expectedGoalsOnTarget: 0.1192 -> 0.098
`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 0.92 (squad xG sum 0.9410, xG scale 0.9819, xA scale 0.9819, team xA 0.69 = 0.75 x lambda)	`scoreline_reset=true` `team_lambda=0.924` `squad_xg_sum=0.941` `squad_xa_sum=0.7025` `team_expected_assists=0.6898` `team_xa_to_xg_ratio=0.7465`	-
`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 skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.176` `goal_rate=0.25` `scorer_weight=0.235`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0563` `assist_rate=0.0` `provider_weight=0.0463`	-

Javi Galán D 78 0 0.01 0.01 ok 0 0.2 6.8 60.1 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 0.9972, quality 0.437)	`market_value_eur=1700000` `quality_factor=0.4374` `boost_factor=0.9972` `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=0.8666` `prior_reliability=0.1` `competition=La Liga`	totalCross: 4.347 -> 4.339 wasFouled: 1.5447 -> 1.5516
`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 0.92 (squad xG sum 0.9410, xG scale 0.9819, xA scale 0.9819, team xA 0.69 = 0.75 x lambda)	`scoreline_reset=true` `team_lambda=0.924` `squad_xg_sum=0.941` `squad_xa_sum=0.7025` `team_expected_assists=0.6898` `team_xa_to_xg_ratio=0.7465`	-
`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 skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0058` `goal_rate=0.0` `scorer_weight=0.0052`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0754` `assist_rate=0.0069` `provider_weight=0.0627`	-

Aleix Febas M 85 0 0.07 0.04 84% 1 0.5 6.9 59.4 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 0.9986, quality 0.465)	`market_value_eur=5200000` `quality_factor=0.4646` `boost_factor=0.9986` `reference_quality=0.505` `history_matches=38` `history_shrink=0.1163`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0091)	`blend_weight=0.0091` `history_matches=38` `season_overlap_discount=1.0` `per90_to_match_factor=0.9438` `prior_reliability=0.12` `competition=La Liga`	fouls: 1.4937 -> 1.5039
`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': 0.393, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9901, 'skill_edge': -0.0108, 'height_edge': 0.004, 'attacker_skill': 0.4918, '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': 0.0707}	aerialWon: 0.1283 -> 0.8 totalShots: 0.8247 -> 0.8165 shotOffTarget: 0.5263 -> 0.5211
`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 0.92 (squad xG sum 0.9410, xG scale 0.9819, xA scale 0.9819, team xA 0.69 = 0.75 x lambda)	`scoreline_reset=true` `team_lambda=0.924` `squad_xg_sum=0.941` `squad_xa_sum=0.7025` `team_expected_assists=0.6898` `team_xa_to_xg_ratio=0.7465`	-
`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 skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.067` `goal_rate=0.0628` `scorer_weight=0.0649`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0883` `assist_rate=0.0769` `provider_weight=0.0814`	-

Ilaix Moriba M 84 0 0.06 0.04 59% 1 0.5 6.9 63.6 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 1.0019, quality 0.565)	`market_value_eur=13700000` `quality_factor=0.5649` `boost_factor=1.0019` `reference_quality=0.505` `history_matches=43` `history_shrink=0.1042`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0122)	`blend_weight=0.0122` `history_matches=43` `season_overlap_discount=1.0` `per90_to_match_factor=0.9341` `prior_reliability=0.18` `competition=La Liga`	fouls: 1.1384 -> 1.1245 interceptionWon: 1.1325 -> 1.1271
`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 Fornals', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.497, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.0694, 'skill_edge': 0.0272, 'height_edge': 0.02, 'attacker_skill': 0.5816, 'defender_skill': 0.5211, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Pablo Fornals', 'market_value': 8400000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.5211}], 'xg_override': 0.0568}	aerialWon: 0.0694 -> 1.1 totalShots: 0.7927 -> 0.8477 shotOffTarget: 0.5715 -> 0.6112 onTargetScoringAttempt: 0.2196 -> 0.2348 expectedGoalsOnTarget: 0.1402 -> 0.1499
`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 0.92 (squad xG sum 0.9410, xG scale 0.9819, xA scale 0.9819, team xA 0.69 = 0.75 x lambda)	`scoreline_reset=true` `team_lambda=0.924` `squad_xg_sum=0.941` `squad_xa_sum=0.7025` `team_expected_assists=0.6898` `team_xa_to_xg_ratio=0.7465`	-
`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=0.0543` `goal_rate=0.0662` `scorer_weight=0.0607`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0557` `assist_rate=0.0522` `provider_weight=0.054`	-

Sergio Carreira M 90 0 0.08 0.05 73% 1 0.4 7.0 60.8 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 1.0011, quality 0.540)	`market_value_eur=11500000` `quality_factor=0.5399` `boost_factor=1.0011` `reference_quality=0.505` `history_matches=41` `history_shrink=0.1087`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0155)	`blend_weight=0.0155` `history_matches=41` `season_overlap_discount=1.0` `per90_to_match_factor=0.9951` `prior_reliability=0.22` `competition=La Liga`	wasFouled: 1.3308 -> 1.3429 wonTackle: 0.5157 -> 0.5077 totalCross: 0.7397 -> 0.7329 interceptionWon: 0.3554 -> 0.3499 fouls: 0.8582 -> 0.8636
`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 Fornals', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.433, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9751, 'skill_edge': 0.015, 'height_edge': -0.032, 'attacker_skill': 0.5545, 'defender_skill': 0.5211, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Pablo Fornals', 'market_value': 8400000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.5211}], 'xg_override': 0.0794}	aerialWon: 0.1473 -> 0.8 totalShots: 0.8129 -> 0.7927 shotOffTarget: 0.5162 -> 0.5033 onTargetScoringAttempt: 0.2906 -> 0.2834 bigChanceMissed: 0.2169 -> 0.2115
`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 0.92 (squad xG sum 0.9410, xG scale 0.9819, xA scale 0.9819, team xA 0.69 = 0.75 x lambda)	`scoreline_reset=true` `team_lambda=0.924` `squad_xg_sum=0.941` `squad_xa_sum=0.7025` `team_expected_assists=0.6898` `team_xa_to_xg_ratio=0.7465`	-
`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_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=0.075` `goal_rate=0.0981` `scorer_weight=0.0893`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0943` `assist_rate=0.0816` `provider_weight=0.0865`	-

Marcos Alonso D 86 0 0.04 0.08 46% 1 0.7 7.0 61.8 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 0.9976, quality 0.437)	`market_value_eur=1300000` `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 0.0134)	`blend_weight=0.0134` `history_matches=37` `season_overlap_discount=1.0` `per90_to_match_factor=0.9564` `prior_reliability=0.1733` `competition=La Liga`	interceptionWon: 1.1696 -> 1.1934 wonTackle: 0.8508 -> 0.8591 fouls: 0.987 -> 0.9935
`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 0.92 (squad xG sum 0.9410, xG scale 0.9819, xA scale 0.9819, team xA 0.69 = 0.75 x lambda)	`scoreline_reset=true` `team_lambda=0.924` `squad_xg_sum=0.941` `squad_xa_sum=0.7025` `team_expected_assists=0.6898` `team_xa_to_xg_ratio=0.7465`	-
`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=0.0419` `goal_rate=0.0061` `scorer_weight=0.0342`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0451` `assist_rate=0.0` `provider_weight=0.037`	-

Yoel Lago D 89 0 0.02 0.00 238% 0 0.1 6.9 40.9 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 0.9989, quality 0.487)	`market_value_eur=7000000` `quality_factor=0.4866` `boost_factor=0.9989` `reference_quality=0.505` `history_matches=19` `history_shrink=0.2083`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0316)	`blend_weight=0.0316` `history_matches=19` `season_overlap_discount=1.0` `per90_to_match_factor=0.9897` `prior_reliability=0.2333` `competition=La Liga`	totalShots: 0.3764 -> 0.3645 totalCross: 0.0 -> 0.0089 wonTackle: 0.8126 -> 0.8048 interceptionWon: 1.2271 -> 1.233
`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 0.92 (squad xG sum 0.9410, xG scale 0.9819, xA scale 0.9819, team xA 0.69 = 0.75 x lambda)	`scoreline_reset=true` `team_lambda=0.924` `squad_xg_sum=0.941` `squad_xa_sum=0.7025` `team_expected_assists=0.6898` `team_xa_to_xg_ratio=0.7465`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 71.0)	`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=0.0158` `goal_rate=0.0` `scorer_weight=0.0127`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0121` `assist_rate=0.0` `provider_weight=0.0098`	-

Javi Rodríguez D 82 0 0.01 0.02 ok 0 0.3 6.7 53.5 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 1.0036, quality 0.606)	`market_value_eur=17500000` `quality_factor=0.6058` `boost_factor=1.0036` `reference_quality=0.505` `history_matches=37` `history_shrink=0.119`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0134)	`blend_weight=0.0134` `history_matches=37` `season_overlap_discount=1.0` `per90_to_match_factor=0.9116` `prior_reliability=0.1733` `competition=La Liga`	wasFouled: 0.7947 -> 0.7887 wonTackle: 0.7688 -> 0.7632
`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': '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 0.92 (squad xG sum 0.9410, xG scale 0.9819, xA scale 0.9819, team xA 0.69 = 0.75 x lambda)	`scoreline_reset=true` `team_lambda=0.924` `squad_xg_sum=0.941` `squad_xa_sum=0.7025` `team_expected_assists=0.6898` `team_xa_to_xg_ratio=0.7465`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 15 by projected rating, fifa_overall 76.0)	`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=0.0122` `goal_rate=0.0` `scorer_weight=0.0098`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0216` `assist_rate=0.0251` `provider_weight=0.0227`	-

Ionuț Radu G 90 0 0.00 - n/a 0.0 7.0 64.7 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 1.0004, quality 0.519)	`market_value_eur=9700000` `quality_factor=0.5189` `boost_factor=1.0004` `reference_quality=0.505` `history_matches=45` `history_shrink=0.1`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0173)	`blend_weight=0.0173` `history_matches=45` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.2667` `competition=La Liga`	saves: 2.8356 -> 2.8471
`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': 1.1121}`	-
`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 0.92 (squad xG sum 0.9410, xG scale 0.9819, xA scale 0.9819, team xA 0.69 = 0.75 x lambda)	`scoreline_reset=true` `team_lambda=0.924` `squad_xg_sum=0.941` `squad_xa_sum=0.7025` `team_expected_assists=0.6898` `team_xa_to_xg_ratio=0.7465`	shots_faced: 10.6 -> 10.3 expected_saves: 2.6 -> 2.52 high_pressure_saves: 0.89 -> 0.86
`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=0.0` `scorer_weight=0.0` `squad_weight_total=0.9627`	-
`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=0.6489`	-

Team totals 0 0.92 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
Jones El-Abdellaoui	F	25.8m	53.2%	Market-value substitute prior

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

Troy Parrott F 73 0 0.24 0.16 54% 2 0.6 6.7 33.2 8 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 1.0183, quality 0.615)	`market_value_eur=18400000` `quality_factor=0.615` `boost_factor=1.0183` `reference_quality=0.505` `history_matches=4` `history_shrink=0.5556`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0192)	`blend_weight=0.0192` `history_matches=4` `season_overlap_discount=1.0` `per90_to_match_factor=0.8061` `prior_reliability=0.0533` `competition=La Liga`	wonTackle: 0.658 -> 0.7035 totalShots: 1.6967 -> 1.7191 fouls: 0.2308 -> 0.2458 totalOffside: 0.2308 -> 0.2458 onTargetScoringAttempt: 0.716 -> 0.7022 wasFouled: 0.5912 -> 0.5798 goalAssist: 0.3195 -> 0.3134 goals: 0.2864 -> 0.2809
`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 Alonso, Yoel Lago', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.257, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.0831, 'skill_edge': 0.0626, 'height_edge': -0.006, 'attacker_skill': 0.615, 'defender_skill': 0.476, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Marcos Alonso', 'market_value': 1300000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'Yoel Lago', 'market_value': 7000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.5146}], 'xg_override': 0.2891}	totalShots: 1.7191 -> 1.862 shotOffTarget: 0.9806 -> 1.0621 onTargetScoringAttempt: 0.7022 -> 0.7606 bigChanceMissed: 0.4271 -> 0.4626 goals: 0.2809 -> 0.3042 expectedGoals: 0.2669 -> 0.2891 expectedGoalsOnTarget: 0.219 -> 0.2372
`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 1.08 (squad xG sum 1.2996, xG scale 0.8311, xA scale 0.8311, team xA 0.74 = 0.69 x lambda)	`scoreline_reset=true` `team_lambda=1.08` `squad_xg_sum=1.2996` `squad_xa_sum=0.8938` `team_expected_assists=0.7428` `team_xa_to_xg_ratio=0.6878`	expected_goals: 0.2891 -> 0.2403 expected_assists: 0.0553 -> 0.046
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 16 by projected rating, fifa_overall 78.0)	`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=0.2136` `goal_rate=0.3042` `scorer_weight=0.251`	-
`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=0.045` `assist_rate=0.3134` `provider_weight=0.0575`	-

Abdessamad Ezzalzouli F 81 1 0.28 0.24 ok 3 2.0 7.3 62.8 28 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 1.0115, 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`	-
`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 Alonso, Yoel Lago', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.306, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.1556, 'skill_edge': 0.1438, 'height_edge': -0.038, 'attacker_skill': 0.7957, 'defender_skill': 0.476, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Marcos Alonso', 'market_value': 1300000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'Yoel Lago', 'market_value': 7000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.5146}], 'xg_override': 0.34}	totalShots: 2.2517 -> 2.6021 shotOffTarget: 1.2962 -> 1.4979 onTargetScoringAttempt: 0.9555 -> 1.1042 expectedGoalsOnTarget: 0.4333 -> 0.5007 goals: 0.395 -> 0.4565 expectedGoals: 0.2942 -> 0.34 bigChanceMissed: 0.2394 -> 0.2767
`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 1.08 (squad xG sum 1.2996, xG scale 0.8311, xA scale 0.8311, team xA 0.74 = 0.69 x lambda)	`scoreline_reset=true` `team_lambda=1.08` `squad_xg_sum=1.2996` `squad_xa_sum=0.8938` `team_expected_assists=0.7428` `team_xa_to_xg_ratio=0.6878`	expected_goals: 0.34 -> 0.2825 expected_assists: 0.2117 -> 0.1758
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 86.0)	`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: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.2461` `goal_rate=0.4565` `scorer_weight=0.3531`	-
`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=0.1612` `assist_rate=0.1556` `provider_weight=0.1406`	-

Isco M 66 0 0.09 0.11 ok 1 1.0 6.9 59.1 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 0.9955, quality 0.445)	`market_value_eur=3600000` `quality_factor=0.4449` `boost_factor=0.9955` `reference_quality=0.505` `history_matches=15` `history_shrink=0.25`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0141)	`blend_weight=0.0141` `history_matches=15` `season_overlap_discount=1.0` `per90_to_match_factor=0.7352` `prior_reliability=0.0867` `competition=La Liga`	fouls: 0.1402 -> 0.1701 totalCross: 2.2419 -> 2.2661 onTargetScoringAttempt: 0.4183 -> 0.4355
`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 Alonso', 'matchup_zone': 'LW vs CB', 'duel_win_rate': 0.375, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9344, 'skill_edge': 0.0034, 'height_edge': -0.048, 'attacker_skill': 0.4449, 'defender_skill': 0.4374, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Marcos Alonso', 'market_value': 1300000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.1125}	aerialWon: 0.0 -> 0.9 totalShots: 1.4409 -> 1.3464 shotOffTarget: 1.0054 -> 0.9394 onTargetScoringAttempt: 0.4355 -> 0.4069 bigChanceMissed: 0.2684 -> 0.2508 expectedGoals: 0.1204 -> 0.1125 goals: 0.0886 -> 0.0828
`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 1.08 (squad xG sum 1.2996, xG scale 0.8311, xA scale 0.8311, team xA 0.74 = 0.69 x lambda)	`scoreline_reset=true` `team_lambda=1.08` `squad_xg_sum=1.2996` `squad_xa_sum=0.8938` `team_expected_assists=0.7428` `team_xa_to_xg_ratio=0.6878`	expected_goals: 0.1125 -> 0.0935 expected_assists: 0.1189 -> 0.0988
`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: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0893` `goal_rate=0.0828` `scorer_weight=0.0764`	-
`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=0.1156` `provider_weight=0.0969`	-

Antony F 80 0 0.19 0.23 ok 3 2.9 7.1 61.4 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 1.0088, quality 0.764)	`market_value_eur=37000000` `quality_factor=0.7641` `boost_factor=1.0088` `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=0.8851` `prior_reliability=0.2` `competition=La Liga`	totalShots: 2.6634 -> 2.6894 totalCross: 1.8778 -> 1.8893 onTargetScoringAttempt: 0.9878 -> 0.995 wasFouled: 1.3383 -> 1.3447
`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 Alonso, Yoel Lago', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.272, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.1053, 'skill_edge': 0.1297, 'height_edge': -0.058, 'attacker_skill': 0.7641, 'defender_skill': 0.476, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Marcos Alonso', 'market_value': 1300000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'Yoel Lago', 'market_value': 7000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.5146}], 'xg_override': 0.2268}	totalShots: 2.6894 -> 2.9726 shotOffTarget: 1.6756 -> 1.852 onTargetScoringAttempt: 0.995 -> 1.0998 bigChanceMissed: 0.4035 -> 0.446 expectedGoalsOnTarget: 0.2843 -> 0.3142 expectedGoals: 0.2052 -> 0.2268 goals: 0.1774 -> 0.1961
`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 1.08 (squad xG sum 1.2996, xG scale 0.8311, xA scale 0.8311, team xA 0.74 = 0.69 x lambda)	`scoreline_reset=true` `team_lambda=1.08` `squad_xg_sum=1.2996` `squad_xa_sum=0.8938` `team_expected_assists=0.7428` `team_xa_to_xg_ratio=0.6878`	expected_goals: 0.2268 -> 0.1885 expected_assists: 0.1398 -> 0.1162
`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=0.1718` `goal_rate=0.1961` `scorer_weight=0.1676`	-
`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=0.1097` `assist_rate=0.0524` `provider_weight=0.093`	-

Pablo Fornals M 81 0 0.06 0.07 ok 1 0.9 7.0 60.9 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 0.503)	`market_value_eur=8400000` `quality_factor=0.5035` `boost_factor=1.0` `reference_quality=0.505` `history_matches=43` `history_shrink=0.1042`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0099)	`blend_weight=0.0099` `history_matches=43` `season_overlap_discount=1.0` `per90_to_match_factor=0.9012` `prior_reliability=0.1467` `competition=La Liga`	totalShots: 1.1832 -> 1.1953 wasFouled: 0.9362 -> 0.9269 interceptionWon: 0.7654 -> 0.7741 totalCross: 3.4418 -> 3.4483 wonTackle: 0.2473 -> 0.253
`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, Ilaix Moriba', '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 1.08 (squad xG sum 1.2996, xG scale 0.8311, xA scale 0.8311, team xA 0.74 = 0.69 x lambda)	`scoreline_reset=true` `team_lambda=1.08` `squad_xg_sum=1.2996` `squad_xa_sum=0.8938` `team_expected_assists=0.7428` `team_xa_to_xg_ratio=0.6878`	expected_goals: 0.0756 -> 0.0628 expected_assists: 0.1525 -> 0.1267
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 85.0)	`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=0.0609` `goal_rate=0.0849` `scorer_weight=0.0674`	-
`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=0.119` `assist_rate=0.1199` `provider_weight=0.1063`	-

Marc Roca M 84 0 0.03 0.02 32% 1 0.5 7.0 50.9 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 0.9980, quality 0.452)	`market_value_eur=4200000` `quality_factor=0.4523` `boost_factor=0.998` `reference_quality=0.505` `history_matches=34` `history_shrink=0.1282`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0094)	`blend_weight=0.0094` `history_matches=34` `season_overlap_discount=1.0` `per90_to_match_factor=0.933` `prior_reliability=0.1133` `competition=La Liga`	totalShots: 0.5812 -> 0.591 wonTackle: 1.0442 -> 1.0499
`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 Carreira', '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 1.08 (squad xG sum 1.2996, xG scale 0.8311, xA scale 0.8311, team xA 0.74 = 0.69 x lambda)	`scoreline_reset=true` `team_lambda=1.08` `squad_xg_sum=1.2996` `squad_xa_sum=0.8938` `team_expected_assists=0.7428` `team_xa_to_xg_ratio=0.6878`	expected_goals: 0.038 -> 0.0316 expected_assists: 0.0551 -> 0.0458
`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=0.0311` `goal_rate=0.0369` `scorer_weight=0.0312`	-
`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=0.0448` `assist_rate=0.0613` `provider_weight=0.0485`	-

Fran García D 88 0 0.01 0.02 52% 0 0.1 6.8 30.9 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 0.9988, quality 0.485)	`market_value_eur=6900000` `quality_factor=0.4854` `boost_factor=0.9988` `reference_quality=0.505` `history_matches=19` `history_shrink=0.2083`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0217)	`blend_weight=0.0217` `history_matches=19` `season_overlap_discount=1.0` `per90_to_match_factor=0.9807` `prior_reliability=0.16` `competition=La Liga`	totalCross: 2.6108 -> 2.6516 totalShots: 0.3947 -> 0.3861 goalAssist: 0.0381 -> 0.0461 wasFouled: 1.1463 -> 1.1392 interceptionWon: 0.9635 -> 0.9692
`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': '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 1.08 (squad xG sum 1.2996, xG scale 0.8311, xA scale 0.8311, team xA 0.74 = 0.69 x lambda)	`scoreline_reset=true` `team_lambda=1.08` `squad_xg_sum=1.2996` `squad_xa_sum=0.8938` `team_expected_assists=0.7428` `team_xa_to_xg_ratio=0.6878`	expected_assists: 0.0481 -> 0.04
`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=0.0104` `goal_rate=0.0` `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=0.0392` `assist_rate=0.0461` `provider_weight=0.0392`	-

Natan D 89 0 0.04 0.02 125% 1 0.2 6.9 51.7 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 1.0077, quality 0.732)	`market_value_eur=32000000` `quality_factor=0.732` `boost_factor=1.0077` `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=0.9886` `prior_reliability=0.1333` `competition=La Liga`	wonTackle: 0.7865 -> 0.7982 interceptionWon: 0.9101 -> 0.9011
`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': 'Williot Swedberg', '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 1.08 (squad xG sum 1.2996, xG scale 0.8311, xA scale 0.8311, team xA 0.74 = 0.69 x lambda)	`scoreline_reset=true` `team_lambda=1.08` `squad_xg_sum=1.2996` `squad_xa_sum=0.8938` `team_expected_assists=0.7428` `team_xa_to_xg_ratio=0.6878`	expected_goals: 0.0478 -> 0.0397
`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=0.0389` `goal_rate=0.1088` `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=0.0103` `assist_rate=0.0` `provider_weight=0.0083`	-

Diego Llorente D 80 0 0.01 0.01 30% 0 0.1 6.9 50.1 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 0.9959, quality 0.437)	`market_value_eur=2600000` `quality_factor=0.4374` `boost_factor=0.9959` `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=0.8862` `prior_reliability=0.0667` `competition=La Liga`	interceptionWon: 1.1267 -> 1.1554 wonTackle: 0.7905 -> 0.7836
`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': 'Williot Swedberg', '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 1.08 (squad xG sum 1.2996, xG scale 0.8311, xA scale 0.8311, team xA 0.74 = 0.69 x lambda)	`scoreline_reset=true` `team_lambda=1.08` `squad_xg_sum=1.2996` `squad_xa_sum=0.8938` `team_expected_assists=0.7428` `team_xa_to_xg_ratio=0.6878`	-
`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=0.0145` `goal_rate=0.0` `scorer_weight=0.0117`	-
`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=0.0032` `assist_rate=0.0` `provider_weight=0.003`	-

Angel Ortiz D 70 0 0.01 0.00 112% 0 0.1 6.8 29.2 21 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 0.9983, quality 0.482)	`market_value_eur=6600000` `quality_factor=0.4818` `boost_factor=0.9983` `reference_quality=0.505` `history_matches=15` `history_shrink=0.25`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0087)	`blend_weight=0.0087` `history_matches=15` `season_overlap_discount=1.0` `per90_to_match_factor=0.7776` `prior_reliability=0.0533` `competition=La Liga`	totalCross: 1.6515 -> 1.6372 wasFouled: 0.8381 -> 0.8308 interceptionWon: 0.7306 -> 0.7243
`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 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 1.08 (squad xG sum 1.2996, xG scale 0.8311, xA scale 0.8311, team xA 0.74 = 0.69 x lambda)	`scoreline_reset=true` `team_lambda=1.08` `squad_xg_sum=1.2996` `squad_xa_sum=0.8938` `team_expected_assists=0.7428` `team_xa_to_xg_ratio=0.6878`	expected_assists: 0.0541 -> 0.045
`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: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0051` `goal_rate=0.0` `scorer_weight=0.0045`	-
`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=0.044` `assist_rate=0.0` `provider_weight=0.036`	-

Álvaro Valles G 90 0 0.00 - n/a 0.0 7.4 61.7 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 1.0003, quality 0.514)	`market_value_eur=9300000` `quality_factor=0.5142` `boost_factor=1.0003` `reference_quality=0.505` `history_matches=35` `history_shrink=0.125`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0163)	`blend_weight=0.0163` `history_matches=35` `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': 'Pablo Durán, Williot Swedberg', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 0.8976}`	-
`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 1.08 (squad xG sum 1.2996, xG scale 0.8311, xA scale 0.8311, team xA 0.74 = 0.69 x lambda)	`scoreline_reset=true` `team_lambda=1.08` `squad_xg_sum=1.2996` `squad_xa_sum=0.8938` `team_expected_assists=0.7428` `team_xa_to_xg_ratio=0.6878`	shots_faced: 8.5 -> 8.8 expected_saves: 2.08 -> 2.16 high_pressure_saves: 0.72 -> 0.74
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 77.0)	`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=1.129`	-
`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=0.6648`	-

Team totals 1 1.08 14 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

Antagonist / tactical summary

Midfield is evenly contested — Ilaix Moriba vs Pablo Fornals will be the pivotal duel. Total antagonist xG — Home: 0.76, Away: 0.97. Home danger zone: Shadow Striker vs CDM + CB — Zone 14 Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Midfield — possession 51.2% / 48.8%, dominance ratio 0.53, key duel: Ilaix Moriba vs Pablo Fornals

Williot Swedberg xG 0.19Ferrán Jutglà xG 0.16Pablo Durán xG 0.20Isco xG 0.11Abdessamad Ezzalzouli xG 0.34Troy Parrott xG 0.29

Summary

Antagonist matchup analysis for Celta vs Real Betis. Most likely scoreline: 0-1. Lambdas derived from 6 home and 4 away individual player duels. Head-to-head: Celta 5W-4D-1L vs Real Betis over 10 meetings.

Tactical parity between 3-4-2-1 and 4-2-3-1. Match outcome driven primarily by individual player quality. Midfield is evenly contested — Ilaix Moriba vs Pablo Fornals will be the pivotal duel. Total antagonist xG — Home: 0.76, Away: 0.97. Home danger zone: Shadow Striker vs CDM + CB — Zone 14 Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Liverpool vs Brighton

Premier League · 20261025 · 4-2-3-1 vs 4-2-3-1

Predicted 2-1Winner: LiverpoolWinning H 2-1 (8.9%)Winning A 1-2 (6.0%)Confidence 0.74λ home/away 2.26 / 1.51Most-likely P 8.9%

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

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Liverpool 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 2.036, Away 1.572). Final calibrated Poisson lambdas: Home 2.256, Away 1.514 (most likely score 2-1).

Home Starters 11/11 XGBoost (10 FBref) Away Starters 11/11 XGBoost (9 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 2.04 / 1.57 Final λ (H/A) 2.26 / 1.51

FBref Team Priors: Liverpool FC (Gls/90: 2.0, GA/90: 2.0 → E[λ]: 3.5) vs Brighton & Hove Albion (Gls/90: 2.5, GA/90: 3.5 → E[λ]: 3.448) | 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	Liverpool	Brighton
League position	5.00	3.00
Points per game	1.55	1.86
Attack index	1.01	1.39
Defense (concede) index	0.77	0.97
Form multiplier	0.98	1.02
Standing λ multiplier	0.90	1.14

Lambda decomposition (how the goal rates were built)

Stage	Liverpool λ	Brighton λ
FBref Squad Prior E[λ] (weight 0.192)	3.500	3.448
Pre-competitive-floor	2.256	1.514
Pre-calibration (Micro-duels)	2.036	1.572
Final λ	2.256	1.514

Scoreline distribution & markets

Top 5 scorelines	P
2-1	8.9%
1-1	7.9%
2-2	6.7%
3-1	6.7%
1-2	6.0%

Home win	54.4%
Draw	20.2%
Away win	25.5%
Over 2.5	72.6%
BTTS yes	69.9%

Pregame fair line — favorite Liverpool (54.4%), fair total 3.77, fair spread 0.74

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.

Liverpool — 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

Alexander Isak F 80 2 0.63 0.27 132% 3 1.1 6.8 42.1 15 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 1.0241, quality 0.891)	`market_value_eur=93000000` `quality_factor=0.8905` `boost_factor=1.0241` `reference_quality=0.505` `history_matches=19` `history_shrink=0.2083`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0181)	`blend_weight=0.0181` `history_matches=19` `season_overlap_discount=1.0` `per90_to_match_factor=0.8884` `prior_reliability=0.1333` `competition=Premier League`	totalShots: 2.5724 -> 2.5821 wasFouled: 0.4085 -> 0.4172
`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': 'Ferdi Kadıoğlu, Luka Vušković', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.25, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 1.0731, 'skill_edge': 0.0117, 'height_edge': 0.038, 'attacker_skill': 0.8905, 'defender_skill': 0.8645, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Ferdi Kadıoğlu', 'market_value': 41000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.839}, {'player': 'Luka Vušković', 'market_value': 69000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.89}], 'xg_override': 0.4523}	totalShots: 2.5821 -> 2.7709 shotOffTarget: 1.4511 -> 1.5572 onTargetScoringAttempt: 1.125 -> 1.2072 expectedGoalsOnTarget: 0.6124 -> 0.6572 bigChanceMissed: 0.5611 -> 0.6021 goals: 0.5059 -> 0.5429 expectedGoals: 0.4215 -> 0.4523
`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 2.26 (squad xG sum 1.6098, xG scale 1.4014, xA scale 1.9439, team xA 1.47 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.256` `squad_xg_sum=1.6098` `squad_xa_sum=0.7544` `team_expected_assists=1.4664` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.4523 -> 0.6338 expected_assists: 0.0384 -> 0.0746
`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=2` `goal_probability=0.4694` `goal_rate=0.5429` `scorer_weight=0.7481`	-
`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=0.0719` `assist_rate=0.0288` `provider_weight=0.0665`	-

Cody Gakpo F 84 0 0.27 0.14 89% 3 1.8 6.9 51.8 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 1.0120, quality 0.873)	`market_value_eur=72000000` `quality_factor=0.8734` `boost_factor=1.012` `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=0.9363` `prior_reliability=0.12` `competition=Premier League`	totalCross: 1.4358 -> 1.4589 totalShots: 2.103 -> 2.112 interceptionWon: 1.0494 -> 1.0581 fouls: 1.2633 -> 1.2702
`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': 'Ferdi Kadıoğlu, Luka Vušković', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.242, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 1.0617, 'skill_edge': 0.004, 'height_edge': 0.038, 'attacker_skill': 0.8734, 'defender_skill': 0.8645, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Ferdi Kadıoğlu', 'market_value': 41000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.839}, {'player': 'Luka Vušković', 'market_value': 69000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.89}], 'xg_override': 0.193}	totalShots: 2.112 -> 2.2423 shotOffTarget: 1.5296 -> 1.624 onTargetScoringAttempt: 0.573 -> 0.6084 bigChanceMissed: 0.249 -> 0.2644 expectedGoals: 0.1818 -> 0.193 goals: 0.175 -> 0.1858 expectedGoalsOnTarget: 0.0976 -> 0.1036
`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 2.26 (squad xG sum 1.6098, xG scale 1.4014, xA scale 1.9439, team xA 1.47 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.256` `squad_xg_sum=1.6098` `squad_xa_sum=0.7544` `team_expected_assists=1.4664` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.193 -> 0.2705 expected_assists: 0.1234 -> 0.2399
`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.237` `goal_rate=0.1858` `scorer_weight=0.2625`	-
`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=0.2133` `assist_rate=0.4386` `provider_weight=0.2999`	-

Florian Wirtz M 82 0 0.34 0.19 79% 2 1.6 7.0 55.3 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 1.0132, quality 0.883)	`market_value_eur=81000000` `quality_factor=0.8829` `boost_factor=1.0132` `reference_quality=0.505` `history_matches=38` `history_shrink=0.1163`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0086)	`blend_weight=0.0086` `history_matches=38` `season_overlap_discount=1.0` `per90_to_match_factor=0.9163` `prior_reliability=0.1133` `competition=Premier League`	totalShots: 1.8826 -> 1.8897
`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': 'Olivier Boscagli', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle', 'duel_win_rate': 0.543, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.0927, 'skill_edge': 0.0871, 'height_edge': -0.024, 'attacker_skill': 0.8829, 'defender_skill': 0.6894, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Olivier Boscagli', 'market_value': 18500000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.6894}], 'xg_override': 0.2412}	aerialWon: 0.5114 -> 0.9 totalShots: 1.8897 -> 2.0649 shotOffTarget: 1.2044 -> 1.316 onTargetScoringAttempt: 0.6817 -> 0.7449 expectedGoals: 0.2207 -> 0.2412 goals: 0.1941 -> 0.2121 expectedGoalsOnTarget: 0.1444 -> 0.1578 bigChanceMissed: 0.1094 -> 0.1195
`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 2.26 (squad xG sum 1.6098, xG scale 1.4014, xA scale 1.9439, team xA 1.47 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.256` `squad_xg_sum=1.6098` `squad_xa_sum=0.7544` `team_expected_assists=1.4664` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.2412 -> 0.338 expected_assists: 0.1319 -> 0.2564
`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=0.2868` `goal_rate=0.2121` `scorer_weight=0.3042`	-
`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=0.2262` `assist_rate=0.1411` `provider_weight=0.2694`	-

Bradley Barcola F 76 0 0.24 0.05 378% 2 0.4 6.7 32.8 10 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 1.0726, quality 0.892)	`market_value_eur=97000000` `quality_factor=0.8922` `boost_factor=1.0726` `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': 'Ferdi Kadıoğlu, Luka Vušković', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.206, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 1.0095, 'skill_edge': 0.0125, 'height_edge': -0.006, 'attacker_skill': 0.8922, 'defender_skill': 0.8645, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Ferdi Kadıoğlu', 'market_value': 41000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.839}, {'player': 'Luka Vušković', 'market_value': 69000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.89}], 'xg_override': 0.1741}	totalShots: 1.4086 -> 1.422 shotOffTarget: 0.7979 -> 0.8055 onTargetScoringAttempt: 0.6107 -> 0.6165
`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 2.26 (squad xG sum 1.6098, xG scale 1.4014, xA scale 1.9439, team xA 1.47 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.256` `squad_xg_sum=1.6098` `squad_xa_sum=0.7544` `team_expected_assists=1.4664` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1741 -> 0.244 expected_assists: 0.0784 -> 0.1524
`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` `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=0.2165` `goal_rate=0.1823` `scorer_weight=0.2529`	-
`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=0.1414` `assist_rate=0.0` `provider_weight=0.1219`	-

Alexis Mac Allister M 87 0 0.13 0.08 73% 1 1.0 6.9 56.7 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 1.0119, quality 0.879)	`market_value_eur=77000000` `quality_factor=0.8792` `boost_factor=1.0119` `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=0.9673` `prior_reliability=0.08` `competition=Premier League`	wasFouled: 0.9677 -> 0.9757 totalShots: 1.1797 -> 1.1732
`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': 'Yasin Ayari', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.468, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 1.0709, 'skill_edge': 0.0322, 'height_edge': 0.016, 'attacker_skill': 0.8893, 'defender_skill': 0.8177, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Yasin Ayari', 'market_value': 41000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8177}], 'xg_override': 0.0947}	totalShots: 1.1776 -> 1.2611 aerialWon: 0.8315 -> 0.9 shotOffTarget: 0.8164 -> 0.8743 onTargetScoringAttempt: 0.3612 -> 0.3868 goals: 0.1017 -> 0.1089 expectedGoals: 0.0884 -> 0.0947
`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 2.26 (squad xG sum 1.6098, xG scale 1.4014, xA scale 1.9439, team xA 1.47 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.256` `squad_xg_sum=1.6098` `squad_xa_sum=0.7544` `team_expected_assists=1.4664` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0947 -> 0.1327 expected_assists: 0.0766 -> 0.1489
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 84.0)	`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=0.1243` `goal_rate=0.1089` `scorer_weight=0.1457`	-
`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=0.1383` `assist_rate=0.0749` `provider_weight=0.1469`	-

Ryan Gravenberch M 79 0 0.12 0.03 244% 1 0.8 6.8 57.2 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 1.0123, quality 0.883)	`market_value_eur=81000000` `quality_factor=0.8829` `boost_factor=1.0123` `reference_quality=0.505` `history_matches=41` `history_shrink=0.1087`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0052)	`blend_weight=0.0052` `history_matches=41` `season_overlap_discount=1.0` `per90_to_match_factor=0.8831` `prior_reliability=0.0733` `competition=Premier League`	wasFouled: 1.3412 -> 1.3509
`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': 'Yasin Ayari', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.513, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 1.137, 'skill_edge': 0.0332, 'height_edge': 0.06, 'attacker_skill': 0.8915, 'defender_skill': 0.8177, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Yasin Ayari', 'market_value': 41000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8177}], 'xg_override': 0.0856}	aerialWon: 0.7048 -> 1.3 totalShots: 0.9977 -> 1.1344 shotOffTarget: 0.6734 -> 0.7657 onTargetScoringAttempt: 0.3192 -> 0.3629 expectedGoals: 0.0753 -> 0.0856 goals: 0.074 -> 0.0841 expectedGoalsOnTarget: 0.0629 -> 0.0715
`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 2.26 (squad xG sum 1.6098, xG scale 1.4014, xA scale 1.9439, team xA 1.47 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.256` `squad_xg_sum=1.6098` `squad_xa_sum=0.7544` `team_expected_assists=1.4664` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0856 -> 0.12 expected_assists: 0.0439 -> 0.0853
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 12 by projected rating, fifa_overall 85.0)	`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1131` `goal_rate=0.0841` `scorer_weight=0.1187`	-
`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=0.0818` `assist_rate=0.0691` `provider_weight=0.1066`	-

Miloš Kerkez D 84 0 0.06 0.02 191% 1 0.2 6.7 48.3 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 1.0086, quality 0.758)	`market_value_eur=36000000` `quality_factor=0.7581` `boost_factor=1.0086` `reference_quality=0.505` `history_matches=39` `history_shrink=0.1136`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0089)	`blend_weight=0.0089` `history_matches=39` `season_overlap_discount=1.0` `per90_to_match_factor=0.937` `prior_reliability=0.12` `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': '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 2.26 (squad xG sum 1.6098, xG scale 1.4014, xA scale 1.9439, team xA 1.47 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.256` `squad_xg_sum=1.6098` `squad_xa_sum=0.7544` `team_expected_assists=1.4664` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0437 -> 0.0612 expected_assists: 0.0172 -> 0.0334
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 17 by projected rating, fifa_overall 81.0)	`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=0.0594` `goal_rate=0.062` `scorer_weight=0.0732`	-
`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=0.0328` `assist_rate=0.0253` `provider_weight=0.0374`	-

Virgil van Dijk D 90 0 0.13 0.11 ok 1 0.7 7.3 60.3 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 1.0025, quality 0.585)	`market_value_eur=15500000` `quality_factor=0.5846` `boost_factor=1.0025` `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`	interceptionWon: 1.1681 -> 1.1621
`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': '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 2.26 (squad xG sum 1.6098, xG scale 1.4014, xA scale 1.9439, team xA 1.47 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.256` `squad_xg_sum=1.6098` `squad_xa_sum=0.7544` `team_expected_assists=1.4664` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0912 -> 0.1278 expected_assists: 0.032 -> 0.0622
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 1 by projected rating, fifa_overall 88.0)	`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=0.12` `goal_rate=0.1966` `scorer_weight=0.1598`	-
`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=0.0603` `assist_rate=0.0` `provider_weight=0.0498`	-

Jérémy Jacquet D 86 0 0.00 - n/a 0 0.0 7.2 38.6 10 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 1.0542, quality 0.866)	`market_value_eur=67000000` `quality_factor=0.8661` `boost_factor=1.0542` `reference_quality=0.505` `history_matches=5` `history_shrink=0.5`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0347)	`blend_weight=0.0347` `history_matches=5` `season_overlap_discount=1.0` `per90_to_match_factor=0.9606` `prior_reliability=0.1067` `competition=Premier League`	interceptionWon: 1.2066 -> 1.248 wasFouled: 0.9446 -> 0.9743 fouls: 1.085 -> 1.0682 totalShots: 0.442 -> 0.4267
`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 2.26 (squad xG sum 1.6098, xG scale 1.4014, xA scale 1.9439, team xA 1.47 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.256` `squad_xg_sum=1.6098` `squad_xa_sum=0.7544` `team_expected_assists=1.4664` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0273 -> 0.0531
`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: 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` `goal_rate=0.0` `scorer_weight=0.001`	-
`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=0.0517` `assist_rate=0.0` `provider_weight=0.0425`	-

Ronald Araújo D 79 0 0.04 0.04 ok 0 0.2 6.9 41.0 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 1.0083, quality 0.694)	`market_value_eur=27000000` `quality_factor=0.6941` `boost_factor=1.0083` `reference_quality=0.505` `history_matches=29` `history_shrink=0.1471`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0026)	`blend_weight=0.0026` `history_matches=29` `season_overlap_discount=1.0` `per90_to_match_factor=0.874` `prior_reliability=0.0267` `competition=Premier League`	fouls: 0.5287 -> 0.5385
`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 2.26 (squad xG sum 1.6098, xG scale 1.4014, xA scale 1.9439, team xA 1.47 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.256` `squad_xg_sum=1.6098` `squad_xa_sum=0.7544` `team_expected_assists=1.4664` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0251 -> 0.0352 expected_assists: 0.0321 -> 0.0624
`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=0.0346` `goal_rate=0.1067` `scorer_weight=0.044`	-
`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`	-

Alisson G 90 0 0.00 - n/a 0.0 7.1 54.0 23 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 1.0038, quality 0.595)	`market_value_eur=16500000` `quality_factor=0.5953` `boost_factor=1.0038` `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=1.0` `prior_reliability=0.1333` `competition=Premier League`	saves: 3.0016 -> 2.9895
`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, Pascal Groß', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.5724}`	-
`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 2.26 (squad xG sum 1.6098, xG scale 1.4014, xA scale 1.9439, team xA 1.47 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.256` `squad_xg_sum=1.6098` `squad_xa_sum=0.7544` `team_expected_assists=1.4664` `team_xa_to_xg_ratio=0.65`	shots_faced: 14.9 -> 14.4 expected_saves: 3.64 -> 3.53 high_pressure_saves: 1.26 -> 1.21
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 87.0)	`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=0.0` `scorer_weight=0.0` `squad_weight_total=2.4352`	-
`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=1.4623`	-

Team totals 2 2.26 17 6.9

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Giorgi Mamardashvili	G	5.5m	15.8%	Market-value substitute prior
Jeremie Frimpong	D	14.8m	35.6%	Market-value substitute prior
Curtis Jones	D	14.8m	35.3%	Market-value substitute prior
Dominik Szoboszlai	M	23.8m	58.0%	Market-value substitute prior
Mohamed Salah	M	22.8m	53.5%	Market-value substitute prior
Hugo Ekitiké	F	27.0m	63.8%	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 58.2 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 1.0081, quality 0.764)	`market_value_eur=37000000` `quality_factor=0.7641` `boost_factor=1.0081` `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: 2.9034 -> 2.8817
`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': 'Alexander Isak, Florian Wirtz', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 2.0359}`	-
`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 1.51 (squad xG sum 1.2450, xG scale 1.2161, xA scale 1.2161, team xA 1.16 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.514` `squad_xg_sum=1.245` `squad_xa_sum=0.955` `team_expected_assists=1.1614` `team_xa_to_xg_ratio=0.7671`	shots_faced: 19.3 -> 21.4 expected_saves: 4.72 -> 5.23 high_pressure_saves: 1.63 -> 1.8
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 81.0)	`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=0.0` `scorer_weight=0.0` `squad_weight_total=1.6849`	-
`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=1.246`	-

Ferdi Kadıoğlu D 87 0 0.06 0.05 ok 1 0.8 6.8 55.9 28 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 1.0092, quality 0.786)	`market_value_eur=41000000` `quality_factor=0.7859` `boost_factor=1.0092` `reference_quality=0.505` `history_matches=41` `history_shrink=0.1087`	-
`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': 'Alexander Isak', '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 1.51 (squad xG sum 1.2450, xG scale 1.2161, xA scale 1.2161, team xA 1.16 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.514` `squad_xg_sum=1.245` `squad_xa_sum=0.955` `team_expected_assists=1.1614` `team_xa_to_xg_ratio=0.7671`	expected_goals: 0.0453 -> 0.0551 expected_assists: 0.0816 -> 0.0992
`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: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0536` `goal_rate=0.042` `scorer_weight=0.0532`	-
`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=0.0944` `assist_rate=0.0` `provider_weight=0.0794`	-

Lewis Dunk D 90 0 0.06 0.04 37% 1 0.6 7.3 52.7 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 0.9979, quality 0.445)	`market_value_eur=3600000` `quality_factor=0.4449` `boost_factor=0.9979` `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=1.0` `prior_reliability=0.1333` `competition=Premier League`	fouls: 0.6454 -> 0.6542 totalCross: 0.2445 -> 0.2523 totalShots: 0.5983 -> 0.5921
`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': 'Bradley Barcola', '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 1.51 (squad xG sum 1.2450, xG scale 1.2161, xA scale 1.2161, team xA 1.16 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.514` `squad_xg_sum=1.245` `squad_xa_sum=0.955` `team_expected_assists=1.1614` `team_xa_to_xg_ratio=0.7671`	expected_goals: 0.0486 -> 0.0591
`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=0.0574` `goal_rate=0.2107` `scorer_weight=0.0739`	-
`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=0.0245` `assist_rate=0.0` `provider_weight=0.0198`	-

Charalampos Kostoulas F 81 1 0.31 0.11 187% 2 0.9 6.9 41.8 26 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 1.0141, quality 0.796)	`market_value_eur=43000000` `quality_factor=0.7957` `boost_factor=1.0141` `reference_quality=0.505` `history_matches=26` `history_shrink=0.1613`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0084)	`blend_weight=0.0084` `history_matches=26` `season_overlap_discount=1.0` `per90_to_match_factor=0.8962` `prior_reliability=0.08` `competition=Premier League`	wasFouled: 1.4367 -> 1.4247
`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': 'Jérémy Jacquet, Ronald Araújo', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.9464, 'skill_edge': -0.0125, 'height_edge': -0.024, 'attacker_skill': 0.7957, 'defender_skill': 0.8234, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Jérémy Jacquet', 'market_value': 67000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8886}, {'player': 'Ronald Araújo', 'market_value': 27000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7581}], 'xg_override': 0.2553}	totalShots: 2.2449 -> 2.1246 shotOffTarget: 1.4144 -> 1.3386 onTargetScoringAttempt: 0.8296 -> 0.7851 bigChanceMissed: 0.6424 -> 0.608 expectedGoalsOnTarget: 0.3567 -> 0.3376 goals: 0.328 -> 0.3104 expectedGoals: 0.2698 -> 0.2553
`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 1.51 (squad xG sum 1.2450, xG scale 1.2161, xA scale 1.2161, team xA 1.16 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.514` `squad_xg_sum=1.245` `squad_xa_sum=0.955` `team_expected_assists=1.1614` `team_xa_to_xg_ratio=0.7671`	expected_goals: 0.2553 -> 0.3106 expected_assists: 0.0671 -> 0.0816
`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=1` `goal_probability=0.267` `goal_rate=0.3104` `scorer_weight=0.3667`	-
`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=0.0784` `assist_rate=0.3496` `provider_weight=0.102`	-

Maxim De Cuyper D 80 0 0.19 0.11 79% 1 0.6 6.8 48.7 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 1.0077, 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 0.0103)	`blend_weight=0.0103` `history_matches=35` `season_overlap_discount=1.0` `per90_to_match_factor=0.8906` `prior_reliability=0.1267` `competition=Premier League`	totalShots: 0.7885 -> 0.805 totalCross: 4.7796 -> 4.7931 wasFouled: 0.8781 -> 0.8691 fouls: 1.0719 -> 1.0802
`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': 'Cody Gakpo', '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 1.51 (squad xG sum 1.2450, xG scale 1.2161, xA scale 1.2161, team xA 1.16 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.514` `squad_xg_sum=1.245` `squad_xa_sum=0.955` `team_expected_assists=1.1614` `team_xa_to_xg_ratio=0.7671`	expected_goals: 0.1548 -> 0.1883 expected_assists: 0.2068 -> 0.2515
`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=0.1716` `goal_rate=0.0984` `scorer_weight=0.1506`	-
`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=0.2224` `assist_rate=0.3407` `provider_weight=0.3144`	-

Yasin Ayari M 86 0 0.12 0.05 157% 1 0.7 6.8 48.0 23 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 0.786)	`market_value_eur=41000000` `quality_factor=0.7859` `boost_factor=1.0111` `reference_quality=0.505` `history_matches=33` `history_shrink=0.1316`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0040)	`blend_weight=0.004` `history_matches=33` `season_overlap_discount=1.0` `per90_to_match_factor=0.9579` `prior_reliability=0.0467` `competition=Premier League`	fouls: 0.6821 -> 0.6903 totalShots: 1.3266 -> 1.3322
`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': 'Alexis Mac Allister', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.372, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.9291, 'skill_edge': -0.0322, 'height_edge': -0.016, 'attacker_skill': 0.8177, 'defender_skill': 0.8893, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Alexis Mac Allister', 'market_value': 77000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8893}], 'xg_override': 0.1013}	aerialWon: 0.3138 -> 0.8 totalShots: 1.3322 -> 1.2377 shotOffTarget: 0.9465 -> 0.8794 onTargetScoringAttempt: 0.3786 -> 0.3518 expectedGoalsOnTarget: 0.1718 -> 0.1596 goals: 0.1247 -> 0.1159 expectedGoals: 0.109 -> 0.1013
`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 1.51 (squad xG sum 1.2450, xG scale 1.2161, xA scale 1.2161, team xA 1.16 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.514` `squad_xg_sum=1.245` `squad_xa_sum=0.955` `team_expected_assists=1.1614` `team_xa_to_xg_ratio=0.7671`	expected_goals: 0.1013 -> 0.1232 expected_assists: 0.0529 -> 0.0643
`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=0.1159` `goal_rate=0.1159` `scorer_weight=0.1351`	-
`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=0.0623` `assist_rate=0.0557` `provider_weight=0.066`	-

Diego Gómez M 80 0 0.20 0.13 52% 2 1.3 6.9 54.2 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 1.0081, 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 0.0098)	`blend_weight=0.0098` `history_matches=37` `season_overlap_discount=1.0` `per90_to_match_factor=0.8926` `prior_reliability=0.1267` `competition=Premier League`	totalShots: 1.6376 -> 1.6575 onTargetScoringAttempt: 0.5919 -> 0.5996 fouls: 1.5332 -> 1.5274 wonTackle: 1.3479 -> 1.3531
`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': 'Alexis Mac Allister', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.401, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.9726, 'skill_edge': -0.0547, 'height_edge': 0.036, 'attacker_skill': 0.7679, 'defender_skill': 0.8893, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Alexis Mac Allister', 'market_value': 77000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8893}], 'xg_override': 0.1663}	aerialWon: 3.0058 -> 1.2 totalShots: 1.6575 -> 1.6121 shotOffTarget: 1.0457 -> 1.017 onTargetScoringAttempt: 0.5996 -> 0.5832 bigChanceMissed: 0.2625 -> 0.2553 goals: 0.2282 -> 0.2219
`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 1.51 (squad xG sum 1.2450, xG scale 1.2161, xA scale 1.2161, team xA 1.16 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.514` `squad_xg_sum=1.245` `squad_xa_sum=0.955` `team_expected_assists=1.1614` `team_xa_to_xg_ratio=0.7671`	expected_goals: 0.1663 -> 0.2022 expected_assists: 0.1025 -> 0.1247
`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=0.1831` `goal_rate=0.2219` `scorer_weight=0.2527`	-
`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=0.1172` `assist_rate=0.1297` `provider_weight=0.1469`	-

Chema Andrés M 86 0 0.14 0.06 120% 1 0.4 6.8 18.1 8 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 1.0331, quality 0.660)	`market_value_eur=23000000` `quality_factor=0.6595` `boost_factor=1.0331` `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': 'Alexis Mac Allister, Ryan Gravenberch', 'matchup_zone': 'Attacking Mid vs Double Pivot — Tight Space Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.8956, 'skill_edge': -0.099, 'height_edge': 0.028, 'attacker_skill': 0.6704, 'defender_skill': 0.8904, 'attacker_role': 'CAM', 'attacker_mv_price_index': 0.95, 'defender_adjusted_parts': [{'player': 'Alexis Mac Allister', 'market_value': 77000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8893}, {'player': 'Ryan Gravenberch', 'market_value': 81000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8915}], 'xg_override': 0.1174}	aerialWon: 3.2669 -> 1.3 totalShots: 1.5196 -> 1.361 shotOffTarget: 1.0872 -> 0.9737 expectedGoalsOnTarget: 0.5189 -> 0.4647 onTargetScoringAttempt: 0.4324 -> 0.3873 goals: 0.143 -> 0.1281 expectedGoals: 0.1311 -> 0.1174
`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 1.51 (squad xG sum 1.2450, xG scale 1.2161, xA scale 1.2161, team xA 1.16 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.514` `squad_xg_sum=1.245` `squad_xa_sum=0.955` `team_expected_assists=1.1614` `team_xa_to_xg_ratio=0.7671`	expected_goals: 0.1174 -> 0.1428 expected_assists: 0.0705 -> 0.0857
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 86.0)	`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: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1331` `goal_rate=0.1281` `scorer_weight=0.1519`	-
`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=0.0821` `assist_rate=0.0657` `provider_weight=0.0823`	-

Pascal Groß M 90 0 0.22 0.11 94% 2 1.0 7.3 59.2 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 0.9965, quality 0.437)	`market_value_eur=2300000` `quality_factor=0.4374` `boost_factor=0.9965` `reference_quality=0.505` `history_matches=24` `history_shrink=0.1724`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0149)	`blend_weight=0.0149` `history_matches=24` `season_overlap_discount=1.0` `per90_to_match_factor=0.9962` `prior_reliability=0.1333` `competition=Premier League`	totalCross: 6.5603 -> 6.5962 totalShots: 1.6641 -> 1.6467 interceptionWon: 0.7503 -> 0.7391
`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': 'Virgil van Dijk', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle', 'duel_win_rate': 0.328, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.776, 'skill_edge': -0.0964, 'height_edge': -0.056, 'attacker_skill': 0.4374, 'defender_skill': 0.6516, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Virgil van Dijk', 'market_value': 15500000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.6516}], 'xg_override': 0.1779}	aerialWon: 0.6905 -> 1.1 totalShots: 1.6617 -> 1.2895 shotOffTarget: 1.0085 -> 0.7826 onTargetScoringAttempt: 0.6532 -> 0.5069 goals: 0.289 -> 0.2243 expectedGoals: 0.2292 -> 0.1779 expectedGoalsOnTarget: 0.1744 -> 0.1353 bigChanceMissed: 0.116 -> 0.09
`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 1.51 (squad xG sum 1.2450, xG scale 1.2161, xA scale 1.2161, team xA 1.16 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.514` `squad_xg_sum=1.245` `squad_xa_sum=0.955` `team_expected_assists=1.1614` `team_xa_to_xg_ratio=0.7671`	expected_goals: 0.1779 -> 0.2163 expected_assists: 0.1881 -> 0.2287
`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=0.1945` `goal_rate=0.2243` `scorer_weight=0.2603`	-
`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=0.2044` `assist_rate=0.2354` `provider_weight=0.2741`	-

Olivier Boscagli D 88 0 0.01 0.01 94% 1 0.2 7.0 35.1 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 1.0076, quality 0.616)	`market_value_eur=18500000` `quality_factor=0.6161` `boost_factor=1.0076` `reference_quality=0.505` `history_matches=17` `history_shrink=0.2273`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0197)	`blend_weight=0.0197` `history_matches=17` `season_overlap_discount=1.0` `per90_to_match_factor=0.9729` `prior_reliability=0.1333` `competition=Premier League`	fouls: 1.3317 -> 1.3438 totalCross: 2.958 -> 2.9476 wonTackle: 0.9656 -> 0.9562 onTargetScoringAttempt: 0.1035 -> 0.111 interceptionWon: 1.1302 -> 1.1367
`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 Wirtz', '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 1.51 (squad xG sum 1.2450, xG scale 1.2161, xA scale 1.2161, team xA 1.16 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.514` `squad_xg_sum=1.245` `squad_xa_sum=0.955` `team_expected_assists=1.1614` `team_xa_to_xg_ratio=0.7671`	expected_assists: 0.0781 -> 0.095
`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=0.0102` `goal_rate=0.0` `scorer_weight=0.0088`	-
`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=0.0906` `assist_rate=0.0276` `provider_weight=0.076`	-

Luka Vušković D 88 0 0.11 0.05 139% 1 0.8 7.1 37.5 10 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 1.0546, quality 0.869)	`market_value_eur=69000000` `quality_factor=0.8692` `boost_factor=1.0546` `reference_quality=0.505` `history_matches=5` `history_shrink=0.5`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0433)	`blend_weight=0.0433` `history_matches=5` `season_overlap_discount=1.0` `per90_to_match_factor=0.9782` `prior_reliability=0.1333` `competition=Premier League`	wasFouled: 1.5058 -> 1.4406 interceptionWon: 1.0456 -> 1.0003 totalShots: 0.9106 -> 0.9559 fouls: 1.666 -> 1.6362 totalCross: 0.4109 -> 0.3931 wonTackle: 0.776 -> 0.7636 goals: 0.2054 -> 0.1965 onTargetScoringAttempt: 0.2054 -> 0.1965
`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': 'Alexander Isak', '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 1.51 (squad xG sum 1.2450, xG scale 1.2161, xA scale 1.2161, team xA 1.16 = 0.77 x lambda)	`scoreline_reset=true` `team_lambda=1.514` `squad_xg_sum=1.245` `squad_xa_sum=0.955` `team_expected_assists=1.1614` `team_xa_to_xg_ratio=0.7671`	expected_goals: 0.0899 -> 0.1093
`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=0.1035` `goal_rate=0.1965` `scorer_weight=0.1366`	-
`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=0.0162` `assist_rate=0.0` `provider_weight=0.013`	-

Team totals 1 1.51 15 7.0

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
Mats Wieffer	D	14.5m	23.6%	Market-value substitute prior
Pascal Struijk	D	14.4m	33.9%	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 ~51% possession expected (Florian Wirtz vs Yasin Ayari). Total antagonist xG — Home: 1.24, 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 51.4% / 48.6%, dominance ratio 0.56, key duel: Florian Wirtz vs Yasin Ayari

Florian Wirtz xG 0.24Alexander Isak xG 0.45Cody Gakpo xG 0.19Pascal Groß xG 0.18Chema Andrés xG 0.12Charalampos Kostoulas xG 0.26

Summary

Antagonist matchup analysis for Liverpool vs Brighton. Most likely scoreline: 2-1. Lambdas derived from 6 home and 5 away individual player duels. Head-to-head: Liverpool 5W-1D-4L 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 ~51% possession expected (Florian Wirtz vs Yasin Ayari). Total antagonist xG — Home: 1.24, 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.

Man Utd vs Bournemouth

Premier League · 20261025 · 4-2-3-1 vs 4-2-3-1

Predicted 1-0Winner: Man UtdWinning H 1-0 (9.7%)Winning A 0-1 (7.0%)Confidence 0.73λ home/away 1.64 / 1.19Most-likely P 9.7%

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

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Man Utd vs Bournemouth. 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 1.613, Away 1.202). Final calibrated Poisson lambdas: Home 1.636, Away 1.185 (most likely score 1-0).

Home Starters 11/11 XGBoost (9 FBref) Away Starters 11/11 XGBoost (10 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 1.61 / 1.20 Final λ (H/A) 1.64 / 1.19

FBref Team Priors: Manchester United (Gls/90: 2.0, GA/90: 2.5 → E[λ]: 1.379) vs Bournemouth (Gls/90: 1.0, GA/90: 1.0 → E[λ]: 1.724) | 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 Utd	Bournemouth
League position	13.00	15.00
Points per game	1.16	1.09
Attack index	1.05	0.91
Defense (concede) index	1.05	1.05
Form multiplier	1.00	0.99
Standing λ multiplier	1.04	0.98

Lambda decomposition (how the goal rates were built)

Stage	Man Utd λ	Bournemouth λ
FBref Squad Prior E[λ] (weight 0.192)	1.379	1.724
Pre-competitive-floor	1.636	1.185
Pre-calibration (Micro-duels)	1.613	1.202
Final λ	1.636	1.185

Scoreline distribution & markets

Top 5 scorelines	P
1-1	11.6%
1-0	9.7%
2-1	9.4%
2-0	8.0%
0-1	7.0%

Home win	47.8%
Draw	24.6%
Away win	27.6%
Over 2.5	53.5%
BTTS yes	56.0%

Pregame fair line — favorite Man Utd (47.8%), fair total 2.82, fair spread 0.45

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 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 85 0 0.44 0.40 ok 3 2.2 7.0 56.6 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 1.0129, quality 0.875)	`market_value_eur=73000000` `quality_factor=0.8747` `boost_factor=1.0129` `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=0.9416` `prior_reliability=0.1333` `competition=Premier League`	totalCross: 3.8663 -> 3.8985 totalShots: 2.375 -> 2.3938 interceptionWon: 0.5957 -> 0.5897 onTargetScoringAttempt: 0.9015 -> 0.9066
`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, James Hill', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.281, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.1183, 'skill_edge': 0.1165, 'height_edge': -0.036, 'attacker_skill': 0.8747, 'defender_skill': 0.6158, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Adam Smith', 'market_value': 310000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'James Hill', 'market_value': 32000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7941}], 'xg_override': 0.3851}	totalShots: 2.3938 -> 2.677 shotOffTarget: 1.4735 -> 1.6478 onTargetScoringAttempt: 0.9066 -> 1.0139 bigChanceMissed: 0.663 -> 0.7414 expectedGoalsOnTarget: 0.4235 -> 0.4736 expectedGoals: 0.3444 -> 0.3851 goals: 0.3109 -> 0.3477
`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 1.64 (squad xG sum 1.4417, xG scale 1.1348, xA scale 1.1348, team xA 1.12 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.636` `squad_xg_sum=1.4417` `squad_xa_sum=0.9865` `team_expected_assists=1.1195` `team_xa_to_xg_ratio=0.6843`	expected_goals: 0.3851 -> 0.4369 expected_assists: 0.1593 -> 0.1808
`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` `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=0.354` `goal_rate=0.3477` `scorer_weight=0.3993`	-
`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=0.1654` `assist_rate=0.0301` `provider_weight=0.1446`	-

Bruno Fernandes M 89 1 0.33 0.23 44% 3 2.1 7.4 62.7 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 1.0076, 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 0.0096)	`blend_weight=0.0096` `history_matches=40` `season_overlap_discount=1.0` `per90_to_match_factor=0.9943` `prior_reliability=0.1333` `competition=Premier League`	totalShots: 2.5926 -> 2.6203 onTargetScoringAttempt: 0.8035 -> 0.8149 goals: 0.3959 -> 0.4064
`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 Wing vs Right Back — Flank Battle', 'duel_win_rate': 0.455, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.9637, 'skill_edge': -0.0367, 'height_edge': 0.012, 'attacker_skill': 0.732, 'defender_skill': 0.8135, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Adrien Truffert', 'market_value': 33000000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.8135}], 'xg_override': 0.2888}	aerialWon: 0.3391 -> 1.0 totalShots: 2.6203 -> 2.5252 shotOffTarget: 1.7891 -> 1.7242 onTargetScoringAttempt: 0.8149 -> 0.7853 goals: 0.4064 -> 0.3916 expectedGoalsOnTarget: 0.3032 -> 0.2922 expectedGoals: 0.2997 -> 0.2888 bigChanceMissed: 0.1948 -> 0.1877
`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 1.64 (squad xG sum 1.4417, xG scale 1.1348, xA scale 1.1348, team xA 1.12 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.636` `squad_xg_sum=1.4417` `squad_xa_sum=0.9865` `team_expected_assists=1.1195` `team_xa_to_xg_ratio=0.6843`	expected_goals: 0.2888 -> 0.3277 expected_assists: 0.2536 -> 0.2877
`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=1` `goal_probability=0.2794` `goal_rate=0.3916` `scorer_weight=0.4096`	-
`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=0.25` `assist_rate=0.3681` `provider_weight=0.3596`	-

Youri Tielemans M 84 0 0.08 0.05 52% 1 0.8 7.2 56.3 22 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 1.0125, quality 0.796)	`market_value_eur=43000000` `quality_factor=0.7957` `boost_factor=1.0125` `reference_quality=0.505` `history_matches=30` `history_shrink=0.1429`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0105)	`blend_weight=0.0105` `history_matches=30` `season_overlap_discount=1.0` `per90_to_match_factor=0.9312` `prior_reliability=0.1133` `competition=Premier League`	wasFouled: 1.7241 -> 1.7405 totalShots: 1.005 -> 1.0171 totalCross: 1.4984 -> 1.4884 wonTackle: 1.3753 -> 1.3666 onTargetScoringAttempt: 0.3049 -> 0.313
`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 Christie, Tyler Adams', 'matchup_zone': 'Attacking Mid vs Double Pivot — Tight Space Battle', 'duel_win_rate': 0.316, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.111, 'skill_edge': 0.0735, 'height_edge': 0.002, 'attacker_skill': 0.8058, 'defender_skill': 0.6424, 'attacker_role': 'CAM', 'attacker_mv_price_index': 0.95, 'defender_adjusted_parts': [{'player': 'Ryan Christie', 'market_value': 8600000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.5238}, {'player': 'Tyler Adams', 'market_value': 31000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.761}], 'xg_override': 0.0681}	totalShots: 1.0171 -> 1.13 shotOffTarget: 0.7 -> 0.7777 onTargetScoringAttempt: 0.313 -> 0.3477 aerialWon: 0.9806 -> 1.0 bigChanceMissed: 0.124 -> 0.1378 goals: 0.088 -> 0.0978 expectedGoals: 0.0613 -> 0.0681 expectedGoalsOnTarget: 0.0511 -> 0.0568
`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 1.64 (squad xG sum 1.4417, xG scale 1.1348, xA scale 1.1348, team xA 1.12 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.636` `squad_xg_sum=1.4417` `squad_xa_sum=0.9865` `team_expected_assists=1.1195` `team_xa_to_xg_ratio=0.6843`	expected_goals: 0.0681 -> 0.0773 expected_assists: 0.1098 -> 0.1246
`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=0.0744` `goal_rate=0.0978` `scorer_weight=0.0966`	-
`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=0.1172` `assist_rate=0.0815` `provider_weight=0.1025`	-

Amad Diallo F 80 0 0.14 0.07 97% 2 0.9 6.9 42.3 27 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 1.0120, quality 0.800)	`market_value_eur=44000000` `quality_factor=0.8002` `boost_factor=1.012` `reference_quality=0.505` `history_matches=32` `history_shrink=0.1351`	-
`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, James Hill', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.247, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.0691, 'skill_edge': 0.083, 'height_edge': -0.036, 'attacker_skill': 0.8002, 'defender_skill': 0.6158, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Adam Smith', 'market_value': 310000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'James Hill', 'market_value': 32000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7941}], 'xg_override': 0.1208}	totalShots: 1.5386 -> 1.6449 shotOffTarget: 0.9659 -> 1.0326 onTargetScoringAttempt: 0.5727 -> 0.6123 bigChanceMissed: 0.2228 -> 0.2382 expectedGoalsOnTarget: 0.1822 -> 0.1948 goals: 0.1724 -> 0.1843 expectedGoals: 0.113 -> 0.1208
`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 1.64 (squad xG sum 1.4417, xG scale 1.1348, xA scale 1.1348, team xA 1.12 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.636` `squad_xg_sum=1.4417` `squad_xa_sum=0.9865` `team_expected_assists=1.1195` `team_xa_to_xg_ratio=0.6843`	expected_goals: 0.1208 -> 0.1371 expected_assists: 0.0973 -> 0.1104
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 80.0)	`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=0.1281` `goal_rate=0.1843` `scorer_weight=0.1714`	-
`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=0.1045` `assist_rate=0.167` `provider_weight=0.138`	-

Carlos Baleba M 74 0 0.10 0.02 286% 1 0.5 6.8 43.1 22 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 1.0147, 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`	-
`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 Christie', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.582, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.2388, 'skill_edge': 0.1585, 'height_edge': 0.004, 'attacker_skill': 0.876, 'defender_skill': 0.5238, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Ryan Christie', 'market_value': 8600000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.5238}], 'xg_override': 0.084}	aerialWon: 1.7016 -> 1.0 totalShots: 0.8999 -> 1.1148 shotOffTarget: 0.6192 -> 0.7671 onTargetScoringAttempt: 0.2807 -> 0.3477 bigChanceMissed: 0.088 -> 0.109 expectedGoals: 0.0678 -> 0.084 goals: 0.0665 -> 0.0824 expectedGoalsOnTarget: 0.0266 -> 0.033
`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 1.64 (squad xG sum 1.4417, xG scale 1.1348, xA scale 1.1348, team xA 1.12 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.636` `squad_xg_sum=1.4417` `squad_xa_sum=0.9865` `team_expected_assists=1.1195` `team_xa_to_xg_ratio=0.6843`	expected_goals: 0.084 -> 0.0953 expected_assists: 0.0468 -> 0.0531
`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=0.0909` `goal_rate=0.0824` `scorer_weight=0.0942`	-
`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=0.0517` `assist_rate=0.0436` `provider_weight=0.0513`	-

Andrey Santos M 75 0 0.07 0.04 83% 1 0.1 6.9 40.6 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 1.0127, quality 0.800)	`market_value_eur=44000000` `quality_factor=0.8002` `boost_factor=1.0127` `reference_quality=0.505` `history_matches=30` `history_shrink=0.1429`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0056)	`blend_weight=0.0056` `history_matches=30` `season_overlap_discount=1.0` `per90_to_match_factor=0.8348` `prior_reliability=0.06` `competition=Premier League`	fouls: 0.6114 -> 0.6183 interceptionWon: 0.9241 -> 0.919
`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 Christie', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.566, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.2144, 'skill_edge': 0.1378, 'height_edge': 0.008, 'attacker_skill': 0.8301, 'defender_skill': 0.5238, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Ryan Christie', 'market_value': 8600000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.5238}], 'xg_override': 0.0628}	aerialWon: 0.9116 -> 1.1 totalShots: 0.5239 -> 0.6362 shotOffTarget: 0.37 -> 0.4493 onTargetScoringAttempt: 0.1539 -> 0.1869 expectedGoalsOnTarget: 0.1175 -> 0.1427 goals: 0.0709 -> 0.0861 bigChanceMissed: 0.0633 -> 0.0769 expectedGoals: 0.0517 -> 0.0628
`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 1.64 (squad xG sum 1.4417, xG scale 1.1348, xA scale 1.1348, team xA 1.12 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.636` `squad_xg_sum=1.4417` `squad_xa_sum=0.9865` `team_expected_assists=1.1195` `team_xa_to_xg_ratio=0.6843`	expected_goals: 0.0628 -> 0.0713 expected_assists: 0.0472 -> 0.0536
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 80.0)	`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=0.0688` `goal_rate=0.0861` `scorer_weight=0.086`	-
`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=0.0522` `assist_rate=0.0422` `provider_weight=0.0508`	-

Diogo Dalot D 88 0 0.08 0.04 114% 1 0.8 6.8 53.2 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 1.0067, quality 0.702)	`market_value_eur=28000000` `quality_factor=0.7022` `boost_factor=1.0067` `reference_quality=0.505` `history_matches=39` `history_shrink=0.1136`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0044)	`blend_weight=0.0044` `history_matches=39` `season_overlap_discount=1.0` `per90_to_match_factor=0.9731` `prior_reliability=0.06` `competition=Premier League`	totalCross: 1.8429 -> 1.8587 totalShots: 0.9954 -> 1.0095 interceptionWon: 1.0008 -> 1.0059
`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': '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 1.64 (squad xG sum 1.4417, xG scale 1.1348, xA scale 1.1348, team xA 1.12 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.636` `squad_xg_sum=1.4417` `squad_xa_sum=0.9865` `team_expected_assists=1.1195` `team_xa_to_xg_ratio=0.6843`	expected_goals: 0.0695 -> 0.0789 expected_assists: 0.0658 -> 0.0747
`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=0.0759` `goal_rate=0.003` `scorer_weight=0.0631`	-
`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=0.072` `assist_rate=0.0229` `provider_weight=0.0598`	-

Harry Maguire D 89 0 0.05 0.04 ok 1 0.6 6.9 53.5 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 1.0000, quality 0.506)	`market_value_eur=8600000` `quality_factor=0.5059` `boost_factor=1.0` `reference_quality=0.505` `history_matches=28` `history_shrink=0.1515`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0131)	`blend_weight=0.0131` `history_matches=28` `season_overlap_discount=1.0` `per90_to_match_factor=0.9894` `prior_reliability=0.1333` `competition=Premier League`	totalShots: 0.7606 -> 0.7701 wonTackle: 0.7133 -> 0.7039 interceptionWon: 1.0207 -> 1.0268 totalCross: 0.5972 -> 0.6023
`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': 'Marcus Tavernier', '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 1.64 (squad xG sum 1.4417, xG scale 1.1348, xA scale 1.1348, team xA 1.12 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.636` `squad_xg_sum=1.4417` `squad_xa_sum=0.9865` `team_expected_assists=1.1195` `team_xa_to_xg_ratio=0.6843`	expected_goals: 0.0437 -> 0.0496 expected_assists: 0.0582 -> 0.066
`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: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0484` `goal_rate=0.0052` `scorer_weight=0.0397`	-
`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=0.0639` `assist_rate=0.0386` `provider_weight=0.054`	-

Ayden Heaven D 85 0 0.06 0.01 310% 1 0.1 6.8 19.5 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 1.0148, quality 0.732)	`market_value_eur=32000000` `quality_factor=0.732` `boost_factor=1.0148` `reference_quality=0.505` `history_matches=18` `history_shrink=0.2174`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0094)	`blend_weight=0.0094` `history_matches=18` `season_overlap_discount=1.0` `per90_to_match_factor=0.9459` `prior_reliability=0.0667` `competition=Premier League`	totalShots: 0.4959 -> 0.5091 wonTackle: 0.8473 -> 0.8393 interceptionWon: 0.8007 -> 0.7932 wasFouled: 1.1902 -> 1.1968
`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 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 1.64 (squad xG sum 1.4417, xG scale 1.1348, xA scale 1.1348, team xA 1.12 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.636` `squad_xg_sum=1.4417` `squad_xa_sum=0.9865` `team_expected_assists=1.1195` `team_xa_to_xg_ratio=0.6843`	expected_goals: 0.0505 -> 0.0573
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 14 by projected rating, fifa_overall 75.0)	`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=0.0557` `goal_rate=0.0` `scorer_weight=0.0458`	-
`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=0.0152` `assist_rate=0.0` `provider_weight=0.0122`	-

Leny Yoro D 85 0 0.12 0.04 222% 0 0.2 6.6 31.3 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 1.0110, quality 0.791)	`market_value_eur=42000000` `quality_factor=0.7909` `boost_factor=1.011` `reference_quality=0.505` `history_matches=34` `history_shrink=0.1282`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0011)	`blend_weight=0.0011` `history_matches=34` `season_overlap_discount=1.0` `per90_to_match_factor=0.9486` `prior_reliability=0.0133` `competition=Premier League`	interceptionWon: 1.24 -> 1.2491
`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 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 1.64 (squad xG sum 1.4417, xG scale 1.1348, xA scale 1.1348, team xA 1.12 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.636` `squad_xg_sum=1.4417` `squad_xa_sum=0.9865` `team_expected_assists=1.1195` `team_xa_to_xg_ratio=0.6843`	expected_goals: 0.1066 -> 0.121 expected_assists: 0.0539 -> 0.0612
`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=0.114` `goal_rate=0.0` `scorer_weight=0.0968`	-
`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=0.0594` `assist_rate=0.0` `provider_weight=0.049`	-

Senne Lammens G 90 0 0.00 - n/a 0.0 7.0 59.0 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 1.0095, quality 0.770)	`market_value_eur=38000000` `quality_factor=0.7699` `boost_factor=1.0095` `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=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, Marcus Tavernier', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.2021}`	-
`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 1.64 (squad xG sum 1.4417, xG scale 1.1348, xA scale 1.1348, team xA 1.12 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.636` `squad_xg_sum=1.4417` `squad_xa_sum=0.9865` `team_expected_assists=1.1195` `team_xa_to_xg_ratio=0.6843`	shots_faced: 11.4 -> 11.3 expected_saves: 2.79 -> 2.77 high_pressure_saves: 0.96 -> 0.95
`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` `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=1.7147`	-
`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=1.1211`	-

Team totals 1 1.64 16 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
Patrick Dorgu	M	16.5m	52.4%	Market-value substitute prior
Matheus Cunha	F	27.1m	37.0%	Market-value substitute prior

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

James Hill D 90 0 0.01 0.03 43% 0 0.3 7.0 58.6 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 1.0087, quality 0.732)	`market_value_eur=32000000` `quality_factor=0.732` `boost_factor=1.0087` `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=0.9991` `prior_reliability=0.1333` `competition=Premier League`	fouls: 0.9075 -> 0.9141 totalShots: 0.4807 -> 0.4754
`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 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 1.19 (squad xG sum 1.4966, xG scale 0.7918, xA scale 0.8381, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.185` `squad_xg_sum=1.4966` `squad_xa_sum=0.919` `team_expected_assists=0.7703` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0899 -> 0.0753
`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 skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0145` `goal_rate=0.0` `scorer_weight=0.0117`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0725` `assist_rate=0.0235` `provider_weight=0.0602`	-

Adam Smith D 84 0 0.00 0.00 153% 0 0.0 6.6 34.2 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 0.9968, quality 0.437)	`market_value_eur=310000` `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=0.93` `prior_reliability=0.1333` `competition=Premier League`	interceptionWon: 0.8047 -> 0.8127 fouls: 1.3826 -> 1.3765 wonTackle: 0.8922 -> 0.8864
`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 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 1.19 (squad xG sum 1.4966, xG scale 0.7918, xA scale 0.8381, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.185` `squad_xg_sum=1.4966` `squad_xa_sum=0.919` `team_expected_assists=0.7703` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 18 by projected rating, fifa_overall 73.0)	`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 skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0013` `goal_rate=0.0` `scorer_weight=0.0013`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0161` `assist_rate=0.0715` `provider_weight=0.0202`	-

Đorđe Petrović G 90 0 0.00 - n/a 0.0 6.9 57.7 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 1.0064, quality 0.710)	`market_value_eur=29000000` `quality_factor=0.71` `boost_factor=1.0064` `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: 2.8419 -> 2.8569
`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': 1.6126}`	-
`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 1.19 (squad xG sum 1.4966, xG scale 0.7918, xA scale 0.8381, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.185` `squad_xg_sum=1.4966` `squad_xa_sum=0.919` `team_expected_assists=0.7703` `team_xa_to_xg_ratio=0.65`	shots_faced: 15.3 -> 15.5 expected_saves: 3.74 -> 3.79 high_pressure_saves: 1.29 -> 1.31
`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 skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0` `scorer_weight=0.0` `squad_weight_total=1.1944`	-
`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=0.7322`	-

António Silva D 90 0 0.14 0.08 72% 1 0.4 7.0 34.9 10 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 1.0341, quality 0.732)	`market_value_eur=32000000` `quality_factor=0.732` `boost_factor=1.0341` `reference_quality=0.505` `history_matches=5` `history_shrink=0.5`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0433)	`blend_weight=0.0433` `history_matches=5` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	totalShots: 0.8009 -> 0.7879 interceptionWon: 1.2294 -> 1.2195 wonTackle: 0.7933 -> 0.8023 wasFouled: 0.2014 -> 0.1927 onTargetScoringAttempt: 0.2014 -> 0.1927 fouls: 0.6516 -> 0.645
`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 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 1.19 (squad xG sum 1.4966, xG scale 0.7918, xA scale 0.8381, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.185` `squad_xg_sum=1.4966` `squad_xa_sum=0.919` `team_expected_assists=0.7703` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1828 -> 0.1447 expected_assists: 0.065 -> 0.0545
`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 skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.1347` `goal_rate=0.0` `scorer_weight=0.1158`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.053` `assist_rate=0.0` `provider_weight=0.0436`	-

Adrien Truffert D 90 0 0.04 0.04 ok 1 0.4 6.9 57.3 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 1.0073, 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=0.9994` `prior_reliability=0.1333` `competition=Premier League`	totalCross: 3.9625 -> 3.9493 wonTackle: 1.5036 -> 1.4945 totalShots: 0.7461 -> 0.7394
`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 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 1.19 (squad xG sum 1.4966, xG scale 0.7918, xA scale 0.8381, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.185` `squad_xg_sum=1.4966` `squad_xa_sum=0.919` `team_expected_assists=0.7703` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0504 -> 0.0399 expected_assists: 0.0834 -> 0.0699
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 80.0)	`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=0.0391` `goal_rate=0.0309` `scorer_weight=0.0322`	-
`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=0.1542` `provider_weight=0.0874`	-

Ryan Christie M 69 0 0.07 0.06 ok 1 0.8 6.8 50.0 24 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=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': 'Carlos Baleba', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.258, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.7612, 'skill_edge': -0.1585, 'height_edge': -0.004, 'attacker_skill': 0.5238, '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': 0.0898}	totalShots: 1.144 -> 0.8708 shotOffTarget: 0.7543 -> 0.5742 aerialWon: 1.1091 -> 1.0 onTargetScoringAttempt: 0.3897 -> 0.2966 bigChanceMissed: 0.3217 -> 0.2449 expectedGoals: 0.118 -> 0.0898 goals: 0.0938 -> 0.0714 expectedGoalsOnTarget: 0.0752 -> 0.0572
`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 1.19 (squad xG sum 1.4966, xG scale 0.7918, xA scale 0.8381, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.185` `squad_xg_sum=1.4966` `squad_xa_sum=0.919` `team_expected_assists=0.7703` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0898 -> 0.0711 expected_assists: 0.0514 -> 0.0431
`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` `ranking_status=combined`	-
`integer_goal_projection`	Team scoreline allocation skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0686` `goal_rate=0.0714` `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=0.0422` `assist_rate=0.0734` `provider_weight=0.0525`	-

Evanilson F 83 0 0.16 0.24 33% 2 1.6 6.7 54.7 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 1.0076, quality 0.739)	`market_value_eur=33000000` `quality_factor=0.7389` `boost_factor=1.0076` `reference_quality=0.505` `history_matches=41` `history_shrink=0.1087`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0080)	`blend_weight=0.008` `history_matches=41` `season_overlap_discount=1.0` `per90_to_match_factor=0.919` `prior_reliability=0.1133` `competition=Premier League`	fouls: 1.3764 -> 1.387 totalOffside: 0.9788 -> 0.971 wasFouled: 1.7476 -> 1.7552 interceptionWon: 0.7176 -> 0.7248 goalAssist: 0.3263 -> 0.3323 onTargetScoringAttempt: 0.6987 -> 0.6931
`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': 0.8974, 'skill_edge': -0.0358, 'height_edge': -0.034, 'attacker_skill': 0.7389, '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': 0.204}	totalShots: 1.7227 -> 1.546 shotOffTarget: 1.0248 -> 0.9197 onTargetScoringAttempt: 0.6931 -> 0.622 bigChanceMissed: 0.5293 -> 0.475 expectedGoals: 0.2273 -> 0.204 goals: 0.2228 -> 0.1999 expectedGoalsOnTarget: 0.1545 -> 0.1386
`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 1.19 (squad xG sum 1.4966, xG scale 0.7918, xA scale 0.8381, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.185` `squad_xg_sum=1.4966` `squad_xa_sum=0.919` `team_expected_assists=0.7703` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.204 -> 0.1615 expected_assists: 0.0664 -> 0.0556
`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=0.1491` `goal_rate=0.1999` `scorer_weight=0.1589`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0541` `assist_rate=0.3323` `provider_weight=0.0695`	-

Marcus Tavernier F 85 0 0.21 0.24 ok 2 1.7 7.4 51.3 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 1.0064, quality 0.694)	`market_value_eur=27000000` `quality_factor=0.6941` `boost_factor=1.0064` `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=0.9468` `prior_reliability=0.1333` `competition=Premier League`	wasFouled: 1.2239 -> 1.2305 totalShots: 2.516 -> 2.5099 onTargetScoringAttempt: 1.0633 -> 1.0575 totalCross: 4.6905 -> 4.6956
`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': 0.8502, 'skill_edge': -0.0559, 'height_edge': -0.046, 'attacker_skill': 0.6941, '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': 0.2602}	totalShots: 2.5102 -> 2.1342 shotOffTarget: 1.4527 -> 1.2351 onTargetScoringAttempt: 1.0575 -> 0.8991 expectedGoalsOnTarget: 0.4082 -> 0.3471 goals: 0.3935 -> 0.3346 expectedGoals: 0.3061 -> 0.2602 bigChanceMissed: 0.0933 -> 0.0793
`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 1.19 (squad xG sum 1.4966, xG scale 0.7918, xA scale 0.8381, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.185` `squad_xg_sum=1.4966` `squad_xa_sum=0.919` `team_expected_assists=0.7703` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.2602 -> 0.206 expected_assists: 0.1856 -> 0.1557
`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=0.1862` `goal_rate=0.3346` `scorer_weight=0.2554`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.1442` `assist_rate=0.0281` `provider_weight=0.1246`	-

Justin Kluivert M 76 0 0.14 0.09 61% 2 1.0 6.8 35.9 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 1.0085, quality 0.669)	`market_value_eur=24000000` `quality_factor=0.6685` `boost_factor=1.0085` `reference_quality=0.505` `history_matches=24` `history_shrink=0.1724`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0127)	`blend_weight=0.0127` `history_matches=24` `season_overlap_discount=1.0` `per90_to_match_factor=0.8417` `prior_reliability=0.1133` `competition=Premier League`	onTargetScoringAttempt: 0.6687 -> 0.6789 totalShots: 1.7924 -> 1.8008 totalCross: 2.1475 -> 2.1391 wonTackle: 0.585 -> 0.5901
`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': 0.377, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v1', 'xg_multiplier': 0.8493, 'skill_edge': -0.0505, 'height_edge': -0.052, 'attacker_skill': 0.6685, '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': 0.1733}	aerialWon: 0.4353 -> 0.8 totalShots: 1.8026 -> 1.531 shotOffTarget: 1.1237 -> 0.9544 onTargetScoringAttempt: 0.6789 -> 0.5766 expectedGoalsOnTarget: 0.3555 -> 0.3019 goals: 0.2058 -> 0.1748 expectedGoals: 0.2041 -> 0.1733 bigChanceMissed: 0.1377 -> 0.1169
`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 1.19 (squad xG sum 1.4966, xG scale 0.7918, xA scale 0.8381, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.185` `squad_xg_sum=1.4966` `squad_xa_sum=0.919` `team_expected_assists=0.7703` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1733 -> 0.1372 expected_assists: 0.103 -> 0.0863
`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 skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.1282` `goal_rate=0.1748` `scorer_weight=0.1381`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0827` `assist_rate=0.1018` `provider_weight=0.0856`	-

Rayan F 84 0 0.21 0.13 56% 2 1.5 6.9 48.9 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 1.0206, quality 0.848)	`market_value_eur=58000000` `quality_factor=0.8478` `boost_factor=1.0206` `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=0.9346` `prior_reliability=0.1333` `competition=Premier League`	totalCross: 2.0412 -> 2.0058 wasFouled: 0.8843 -> 0.869 totalShots: 2.1451 -> 2.157 onTargetScoringAttempt: 0.774 -> 0.7852 wonTackle: 0.6434 -> 0.6323 fouls: 0.3989 -> 0.392
`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': 0.195, 'pressing_faced': 'HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.993, 'skill_edge': 0.0133, 'height_edge': -0.018, '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': 0.2612}	totalShots: 2.157 -> 2.1419 shotOffTarget: 1.3711 -> 1.3615 onTargetScoringAttempt: 0.7852 -> 0.7797
`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 1.19 (squad xG sum 1.4966, xG scale 0.7918, xA scale 0.8381, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.185` `squad_xg_sum=1.4966` `squad_xa_sum=0.919` `team_expected_assists=0.7703` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.2612 -> 0.2069 expected_assists: 0.0971 -> 0.0814
`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=0.1869` `goal_rate=0.2513` `scorer_weight=0.2003`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0782` `assist_rate=0.0369` `provider_weight=0.0651`	-

Tyler Adams M 80 0 0.02 0.00 269% 0 0.1 6.9 48.1 21 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 1.0094, quality 0.725)	`market_value_eur=31000000` `quality_factor=0.7249` `boost_factor=1.0094` `reference_quality=0.505` `history_matches=30` `history_shrink=0.1429`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0025)	`blend_weight=0.0025` `history_matches=30` `season_overlap_discount=1.0` `per90_to_match_factor=0.8861` `prior_reliability=0.0267` `competition=Premier League`	interceptionWon: 1.5516 -> 1.5587
`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': 'Carlos Baleba', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.352, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.9004, 'skill_edge': -0.0518, 'height_edge': -0.016, 'attacker_skill': 0.761, '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': 0.0221}	aerialWon: 1.1101 -> 0.9 totalShots: 0.5478 -> 0.4932 shotOffTarget: 0.4067 -> 0.3662 onTargetScoringAttempt: 0.1411 -> 0.127
`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 1.19 (squad xG sum 1.4966, xG scale 0.7918, xA scale 0.8381, team xA 0.77 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.185` `squad_xg_sum=1.4966` `squad_xa_sum=0.919` `team_expected_assists=0.7703` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0364 -> 0.0305
`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 skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0173` `goal_rate=0.0298` `scorer_weight=0.0194`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.03` `assist_rate=0.0531` `provider_weight=0.0364`	-

Team totals 0 1.19 15 6.9

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

Antagonist / tactical summary

Home midfield dominates with ~52% possession expected (Carlos Baleba vs Tyler Adams). Total antagonist xG — Home: 1.01, Away: 1.01. 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 0.56, key duel: Carlos Baleba vs Tyler Adams

Bruno Fernandes xG 0.29Youri Tielemans xG 0.07Bryan Mbeumo xG 0.39Justin Kluivert xG 0.17Rayan xG 0.26Evanilson xG 0.20

Summary

Antagonist matchup analysis for Man Utd vs Bournemouth. Most likely scoreline: 1-0. Lambdas derived from 6 home and 6 away individual player duels. Head-to-head: Man Utd 4W-4D-2L vs Bournemouth 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 (Carlos Baleba vs Tyler Adams). Total antagonist xG — Home: 1.01, Away: 1.01. 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 Brentford

Premier League · 20261025 · 5-4-1 vs 4-2-3-1

Predicted 1-2Winner: BrentfordWinning H 2-1 (5.6%)Winning A 1-2 (9.9%)Confidence 0.74λ home/away 1.09 / 1.91Most-likely P 9.9%

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

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Hull vs Brentford. 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 1.165, Away 1.836). Final calibrated Poisson lambdas: Home 1.087, Away 1.909 (most likely score 1-2).

Home Starters 11/11 XGBoost (8 FBref) Away Starters 11/11 XGBoost (10 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 1.17 / 1.84 Final λ (H/A) 1.09 / 1.91

FBref Team Priors: Hull City (Gls/90: 1.5, GA/90: 1.5 → E[λ]: 2.069) vs Brentford (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	Brentford
League position	9.00	6.00
Points per game	1.32	1.40
Attack index	0.97	1.05
Defense (concede) index	0.88	0.94
Form multiplier	0.99	1.00
Standing λ multiplier	0.94	0.99

Lambda decomposition (how the goal rates were built)

Stage	Hull λ	Brentford λ
FBref Squad Prior E[λ] (weight 0.192)	2.069	2.069
Pre-mismatch	1.174	1.822
Pre-competitive-floor	1.141	2.005
Pre-calibration (Micro-duels)	1.165	1.836
Final λ	1.087	1.909

competition calibration scale 1.0competition observed total goals 0.0quality gap applied Truemismatch score pressure applied Truecompetitive underdog floor applied Falsestage discount 1.0

Scoreline distribution & markets

Top 5 scorelines	P
1-1	10.4%
1-2	9.9%
0-1	9.5%
0-2	9.1%
1-3	6.3%

Home win	21.2%
Draw	22.3%
Away win	56.5%
Over 2.5	57.6%
BTTS yes	56.5%

Pregame fair line — favorite Brentford (56.5%), fair total 3.00, fair spread -0.82

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

Oliver McBurnie F 86 1 0.25 0.12 106% 2 1.0 6.9 33.7 10 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 0.9983, quality 0.494)	`market_value_eur=7600000` `quality_factor=0.4939` `boost_factor=0.9983` `reference_quality=0.505` `history_matches=5` `history_shrink=0.5`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0433)	`blend_weight=0.0433` `history_matches=5` `season_overlap_discount=1.0` `per90_to_match_factor=0.9537` `prior_reliability=0.1333` `competition=Premier League`	totalShots: 1.9596 -> 1.916 goals: 0.2861 -> 0.2737 interceptionWon: 0.2691 -> 0.2574 onTargetScoringAttempt: 0.7105 -> 0.721 goalAssist: 0.2288 -> 0.2189 wonTackle: 0.2288 -> 0.2189 wasFouled: 2.1963 -> 2.2044
`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': 'Sepp van den Berg, Kristoffer Ajer', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 0.776, 'skill_edge': -0.1284, 'height_edge': -0.024, 'attacker_skill': 0.4939, 'defender_skill': 0.7792, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Sepp van den Berg', 'market_value': 33000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8002}, {'player': 'Kristoffer Ajer', 'market_value': 27000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7581}], 'xg_override': 0.224}	totalShots: 1.9701 -> 1.5288 shotOffTarget: 1.2491 -> 0.9693 onTargetScoringAttempt: 0.721 -> 0.5595 bigChanceMissed: 0.5638 -> 0.4375 expectedGoals: 0.2886 -> 0.224 goals: 0.2737 -> 0.2124 expectedGoalsOnTarget: 0.1327 -> 0.103
`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 1.09 (squad xG sum 0.9598, xG scale 1.1326, xA scale 1.2942, team xA 0.71 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.087` `squad_xg_sum=0.9598` `squad_xa_sum=0.5459` `team_expected_assists=0.7066` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.224 -> 0.2538 expected_assists: 0.0398 -> 0.0515
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 75.0)	`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=1` `goal_probability=0.2242` `goal_rate=0.2124` `scorer_weight=0.2431`	-
`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=0.0502` `assist_rate=0.2189` `provider_weight=0.0644`	-

Sorba Thomas F 74 0 0.14 0.03 390% 1 0.4 6.8 17.4 14 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 1.0129, quality 0.574)	`market_value_eur=14500000` `quality_factor=0.5737` `boost_factor=1.0129` `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=true`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Sepp van den Berg, Kristoffer Ajer', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 0.8111, 'skill_edge': -0.0925, 'height_edge': -0.036, 'attacker_skill': 0.5737, 'defender_skill': 0.7792, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Sepp van den Berg', 'market_value': 33000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8002}, {'player': 'Kristoffer Ajer', 'market_value': 27000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7581}], 'xg_override': 0.1223}	totalShots: 1.6333 -> 1.3248 shotOffTarget: 1.098 -> 0.8906 onTargetScoringAttempt: 0.5353 -> 0.4342 goals: 0.2181 -> 0.1769 expectedGoals: 0.1508 -> 0.1223
`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 1.09 (squad xG sum 0.9598, xG scale 1.1326, xA scale 1.2942, team xA 0.71 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.087` `squad_xg_sum=0.9598` `squad_xa_sum=0.5459` `team_expected_assists=0.7066` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1223 -> 0.1385 expected_assists: 0.0723 -> 0.0936
`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=0.1293` `goal_rate=0.1769` `scorer_weight=0.1731`	-
`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=0.0894` `assist_rate=0.0` `provider_weight=0.0749`	-

Regan Slater M 87 0 0.07 0.02 247% 1 0.4 6.8 34.3 10 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 1.0012, quality 0.513)	`market_value_eur=9200000` `quality_factor=0.513` `boost_factor=1.0012` `reference_quality=0.505` `history_matches=5` `history_shrink=0.5`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0433)	`blend_weight=0.0433` `history_matches=5` `season_overlap_discount=1.0` `per90_to_match_factor=0.9617` `prior_reliability=0.1333` `competition=Premier League`	totalShots: 1.0603 -> 1.0144 interceptionWon: 0.8896 -> 0.8511 fouls: 0.6427 -> 0.6773 goalAssist: 0.0957 -> 0.1124 onTargetScoringAttempt: 0.3331 -> 0.3187 totalCross: 1.1309 -> 1.1236 wasFouled: 0.6309 -> 0.6244
`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 Mid vs Right Back — Wide Channel', 'duel_win_rate': 0.293, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.7551, 'skill_edge': -0.1426, 'height_edge': -0.024, 'attacker_skill': 0.5251, 'defender_skill': 0.8419, 'attacker_role': 'LM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Michael Kayode', 'market_value': 39000000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.8419}], 'xg_override': 0.0655}	aerialWon: 1.2504 -> 0.9 totalShots: 1.0459 -> 0.7898 shotOffTarget: 0.7272 -> 0.5491 onTargetScoringAttempt: 0.3187 -> 0.2407 expectedGoals: 0.0867 -> 0.0655 goals: 0.0687 -> 0.0519
`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 1.09 (squad xG sum 0.9598, xG scale 1.1326, xA scale 1.2942, team xA 0.71 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.087` `squad_xg_sum=0.9598` `squad_xa_sum=0.5459` `team_expected_assists=0.7066` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0655 -> 0.0742 expected_assists: 0.0943 -> 0.122
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 72.0)	`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=0.0715` `goal_rate=0.0519` `scorer_weight=0.0655`	-
`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=0.1149` `assist_rate=0.1124` `provider_weight=0.1373`	-

Tim Iroegbunam M 83 0 0.08 0.02 296% 1 0.4 6.8 40.4 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 1.0032, quality 0.585)	`market_value_eur=15500000` `quality_factor=0.5846` `boost_factor=1.0032` `reference_quality=0.505` `history_matches=32` `history_shrink=0.1351`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0006)	`blend_weight=0.0006` `history_matches=32` `season_overlap_discount=1.0` `per90_to_match_factor=0.9207` `prior_reliability=0.0067` `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': 'Mikkel Damsgaard', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.393, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v1', 'xg_multiplier': 0.9169, 'skill_edge': -0.0645, 'height_edge': 0.008, 'attacker_skill': 0.6029, 'defender_skill': 0.7463, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Mikkel Damsgaard', 'market_value': 29000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.7463}], 'xg_override': 0.0745}	aerialWon: 1.392 -> 1.1 totalShots: 0.8031 -> 0.7364 shotOffTarget: 0.6028 -> 0.5527 onTargetScoringAttempt: 0.2003 -> 0.1837 expectedGoals: 0.0812 -> 0.0745 goals: 0.0644 -> 0.059
`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 1.09 (squad xG sum 0.9598, xG scale 1.1326, xA scale 1.2942, team xA 0.71 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.087` `squad_xg_sum=0.9598` `squad_xa_sum=0.5459` `team_expected_assists=0.7066` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0745 -> 0.0844 expected_assists: 0.0306 -> 0.0396
`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: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0809` `goal_rate=0.059` `scorer_weight=0.0739`	-
`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=0.0388` `assist_rate=0.0545` `provider_weight=0.0495`	-

Mohamed Belloumi F 79 0 0.20 0.07 179% 2 1.3 6.9 36.3 20 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 1.0145, quality 0.602)	`market_value_eur=17100000` `quality_factor=0.6016` `boost_factor=1.0145` `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': 'Sepp van den Berg, Kristoffer Ajer', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 0.7943, 'skill_edge': -0.0799, 'height_edge': -0.06, 'attacker_skill': 0.6016, 'defender_skill': 0.7792, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Sepp van den Berg', 'market_value': 33000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8002}, {'player': 'Kristoffer Ajer', 'market_value': 27000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7581}], 'xg_override': 0.1752}	totalShots: 2.28 -> 1.811 shotOffTarget: 1.436 -> 1.1406 onTargetScoringAttempt: 0.844 -> 0.6704 expectedGoalsOnTarget: 0.3131 -> 0.2487 bigChanceMissed: 0.2735 -> 0.2172 goals: 0.2507 -> 0.1991 expectedGoals: 0.2206 -> 0.1752
`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 1.09 (squad xG sum 0.9598, xG scale 1.1326, xA scale 1.2942, team xA 0.71 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.087` `squad_xg_sum=0.9598` `squad_xa_sum=0.5459` `team_expected_assists=0.7066` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1752 -> 0.1984 expected_assists: 0.1067 -> 0.1382
`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=0.18` `goal_rate=0.1991` `scorer_weight=0.2195`	-
`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=0.1291` `assist_rate=0.4415` `provider_weight=0.1727`	-

Ryan Giles D 90 0 0.00 0.00 62% 0 0.1 6.7 34.2 10 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 0.574)	`market_value_eur=14500000` `quality_factor=0.5737` `boost_factor=1.0103` `reference_quality=0.505` `history_matches=5` `history_shrink=0.5`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0412)	`blend_weight=0.0412` `history_matches=5` `season_overlap_discount=1.0` `per90_to_match_factor=0.9944` `prior_reliability=0.1267` `competition=Premier League`	totalCross: 9.2559 -> 9.1549 interceptionWon: 0.8586 -> 0.8448 fouls: 0.2724 -> 0.2612 wasFouled: 0.2316 -> 0.2221 goalAssist: 0.2316 -> 0.2221 wonTackle: 0.8423 -> 0.8507
`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, Caoimhin Kelleher', '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 1.09 (squad xG sum 0.9598, xG scale 1.1326, xA scale 1.2942, team xA 0.71 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.087` `squad_xg_sum=0.9598` `squad_xa_sum=0.5459` `team_expected_assists=0.7066` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.093 -> 0.1204
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 14 by projected rating, fifa_overall 72.0)	`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=0.0034` `goal_rate=0.0` `scorer_weight=0.0032`	-
`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=0.1134` `assist_rate=0.2221` `provider_weight=0.1505`	-

Lucas Herrington D 90 0 0.01 0.00 209% 0 0.1 6.6 33.3 10 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 1.0134, quality 0.594)	`market_value_eur=16399999` `quality_factor=0.5942` `boost_factor=1.0134` `reference_quality=0.505` `history_matches=5` `history_shrink=0.5`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0152)	`blend_weight=0.0152` `history_matches=5` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.0467` `competition=Premier League`	wonTackle: 0.8495 -> 0.8366 interceptionWon: 0.7526 -> 0.7412 totalShots: 0.3655 -> 0.36
`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': 'Jaidon Anthony', '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 1.09 (squad xG sum 0.9598, xG scale 1.1326, xA scale 1.2942, team xA 0.71 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.087` `squad_xg_sum=0.9598` `squad_xa_sum=0.5459` `team_expected_assists=0.7066` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 16 by projected rating, fifa_overall 66.0)	`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=0.0072` `goal_rate=0.0` `scorer_weight=0.0064`	-
`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=0.0136` `assist_rate=0.0` `provider_weight=0.0113`	-

John Egan D 88 0 0.01 0.01 165% 0 0.3 6.7 34.8 10 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 0.9899, quality 0.437)	`market_value_eur=1500000` `quality_factor=0.4374` `boost_factor=0.9899` `reference_quality=0.505` `history_matches=5` `history_shrink=0.5`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0433)	`blend_weight=0.0433` `history_matches=5` `season_overlap_discount=1.0` `per90_to_match_factor=0.9743` `prior_reliability=0.1333` `competition=Premier League`	fouls: 0.6466 -> 0.6186 wonTackle: 1.0022 -> 0.9799 totalShots: 0.4958 -> 0.4743 wasFouled: 0.1639 -> 0.1779 onTargetScoringAttempt: 0.2669 -> 0.2553 interceptionWon: 1.1211 -> 1.1147
`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 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 1.09 (squad xG sum 0.9598, xG scale 1.1326, xA scale 1.2942, team xA 0.71 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.087` `squad_xg_sum=0.9598` `squad_xa_sum=0.5459` `team_expected_assists=0.7066` `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 72.0)	`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=0.0135` `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=0.0027` `assist_rate=0.0` `provider_weight=0.0026`	-

Semi Ajayi D 87 0 0.12 0.07 79% 1 0.3 6.7 34.1 10 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 0.9899, quality 0.437)	`market_value_eur=930000` `quality_factor=0.4374` `boost_factor=0.9899` `reference_quality=0.505` `history_matches=5` `history_shrink=0.5`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0368)	`blend_weight=0.0368` `history_matches=5` `season_overlap_discount=1.0` `per90_to_match_factor=0.971` `prior_reliability=0.1133` `competition=Premier League`	onTargetScoringAttempt: 0.1394 -> 0.1554 goals: 0.1394 -> 0.1553 wasFouled: 0.1639 -> 0.1789 wonTackle: 0.8926 -> 0.8808 interceptionWon: 0.901 -> 0.9099 totalCross: 0.3568 -> 0.3647 fouls: 0.1928 -> 0.1857
`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 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 1.09 (squad xG sum 0.9598, xG scale 1.1326, xA scale 1.2942, team xA 0.71 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.087` `squad_xg_sum=0.9598` `squad_xa_sum=0.5459` `team_expected_assists=0.7066` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.104 -> 0.1178
`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: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1111` `goal_rate=0.1553` `scorer_weight=0.1472`	-
`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=0.0049` `assist_rate=0.0` `provider_weight=0.0046`	-

Lewie Coyle D 78 0 0.03 0.01 157% 0 0.1 6.6 32.4 10 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 0.9899, 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`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0282)	`blend_weight=0.0282` `history_matches=5` `season_overlap_discount=1.0` `per90_to_match_factor=0.8638` `prior_reliability=0.0867` `competition=Premier League`	wasFouled: 0.6065 -> 0.6643 fouls: 0.3032 -> 0.3321 totalShots: 0.3431 -> 0.3334 totalCross: 1.658 -> 1.6487
`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, Keane Lewis-Potter', '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 1.09 (squad xG sum 0.9598, xG scale 1.1326, xA scale 1.2942, team xA 0.71 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.087` `squad_xg_sum=0.9598` `squad_xa_sum=0.5459` `team_expected_assists=0.7066` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 17 by projected rating, fifa_overall 70.0)	`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=0.0339` `goal_rate=0.0` `scorer_weight=0.0276`	-
`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=0.0191` `assist_rate=0.0` `provider_weight=0.0154`	-

Konstantinos Tzolakis G 90 0 0.00 - n/a 0.0 7.0 37.0 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 1.0204, quality 0.641)	`market_value_eur=21000000` `quality_factor=0.6408` `boost_factor=1.0204` `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': 'Igor Thiago, Kevin Schade', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.8358}`	-
`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 1.09 (squad xG sum 0.9598, xG scale 1.1326, xA scale 1.2942, team xA 0.71 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.087` `squad_xg_sum=0.9598` `squad_xa_sum=0.5459` `team_expected_assists=0.7066` `team_xa_to_xg_ratio=0.65`	shots_faced: 17.4 -> 18.1 expected_saves: 4.25 -> 4.43 high_pressure_saves: 1.47 -> 1.53
`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=0.0` `scorer_weight=0.0` `squad_weight_total=1.1221`	-
`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=0.7801`	-

Team totals 1 1.09 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
Nobel Mendy	D	14.4m	23.3%	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

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

Igor Thiago F 89 1 0.54 0.49 ok 3 2.0 6.7 56.4 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 1.0113, quality 0.868)	`market_value_eur=68000000` `quality_factor=0.8677` `boost_factor=1.0113` `reference_quality=0.505` `history_matches=43` `history_shrink=0.1042`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0086)	`blend_weight=0.0086` `history_matches=43` `season_overlap_discount=1.0` `per90_to_match_factor=0.9884` `prior_reliability=0.1267` `competition=Premier League`	fouls: 1.9364 -> 1.9243 totalShots: 2.236 -> 2.2479 wasFouled: 0.5779 -> 0.5729 wonTackle: 0.5799 -> 0.5749
`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 Egan, Semi Ajayi', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.402, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.2964, 'skill_edge': 0.1936, 'height_edge': 0.008, 'attacker_skill': 0.8677, 'defender_skill': 0.4374, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'John Egan', 'market_value': 1500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'Semi Ajayi', 'market_value': 930000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.5144}	totalShots: 2.2479 -> 2.9142 shotOffTarget: 1.4076 -> 1.8248 onTargetScoringAttempt: 0.8301 -> 1.0761 bigChanceMissed: 0.5996 -> 0.7773 expectedGoals: 0.3968 -> 0.5144 expectedGoalsOnTarget: 0.3649 -> 0.4731 goals: 0.3267 -> 0.4235
`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 1.91 (squad xG sum 1.8156, xG scale 1.0514, xA scale 1.3606, team xA 1.24 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.909` `squad_xg_sum=1.8156` `squad_xa_sum=0.912` `team_expected_assists=1.2409` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.5144 -> 0.5408 expected_assists: 0.05 -> 0.068
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 14 by projected rating, fifa_overall 82.0)	`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=1` `goal_probability=0.4177` `goal_rate=0.4235` `scorer_weight=0.4537`	-
`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=0.0657` `assist_rate=0.0026` `provider_weight=0.0544`	-

Kevin Schade F 86 1 0.39 0.22 77% 3 1.4 7.0 50.8 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 1.0109, quality 0.831)	`market_value_eur=52000000` `quality_factor=0.8309` `boost_factor=1.0109` `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=0.9571` `prior_reliability=0.1333` `competition=Premier League`	totalShots: 2.1658 -> 2.1825 wonTackle: 0.8146 -> 0.8298 onTargetScoringAttempt: 0.7641 -> 0.7708 wasFouled: 1.2555 -> 1.2618
`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 Egan, Semi Ajayi', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.353, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.225, 'skill_edge': 0.177, 'height_edge': -0.024, 'attacker_skill': 0.8309, 'defender_skill': 0.4374, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'John Egan', 'market_value': 1500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'Semi Ajayi', 'market_value': 930000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.3711}	totalShots: 2.1825 -> 2.6736 shotOffTarget: 1.4017 -> 1.7171 onTargetScoringAttempt: 0.7708 -> 0.9442 bigChanceMissed: 0.4335 -> 0.531 goals: 0.3874 -> 0.4746 expectedGoals: 0.3029 -> 0.3711 expectedGoalsOnTarget: 0.2677 -> 0.3279
`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 1.91 (squad xG sum 1.8156, xG scale 1.0514, xA scale 1.3606, team xA 1.24 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.909` `squad_xg_sum=1.8156` `squad_xa_sum=0.912` `team_expected_assists=1.2409` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.3711 -> 0.3902 expected_assists: 0.0506 -> 0.0688
`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.3231` `goal_rate=0.4746` `scorer_weight=0.4861`	-
`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=0.0665` `assist_rate=0.0161` `provider_weight=0.055`	-

Mikkel Damsgaard M 80 0 0.08 0.06 ok 1 0.8 7.1 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 1.0072, 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 0.0045)	`blend_weight=0.0045` `history_matches=38` `season_overlap_discount=1.0` `per90_to_match_factor=0.8942` `prior_reliability=0.06` `competition=Premier League`	totalCross: 3.3131 -> 3.3341
`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': '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 1.91 (squad xG sum 1.8156, xG scale 1.0514, xA scale 1.3606, team xA 1.24 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.909` `squad_xg_sum=1.8156` `squad_xa_sum=0.912` `team_expected_assists=1.2409` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.1752 -> 0.2384
`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=0.0761` `goal_rate=0.0838` `scorer_weight=0.0846`	-
`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=0.2121` `assist_rate=0.1071` `provider_weight=0.1907`	-

Jaidon Anthony F 80 0 0.23 0.13 78% 2 1.2 6.8 49.8 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 1.0038, quality 0.625)	`market_value_eur=19400000` `quality_factor=0.6251` `boost_factor=1.0038` `reference_quality=0.505` `history_matches=42` `history_shrink=0.1064`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0041)	`blend_weight=0.0041` `history_matches=42` `season_overlap_discount=1.0` `per90_to_match_factor=0.8846` `prior_reliability=0.06` `competition=Premier League`	fouls: 1.0986 -> 1.1063 totalCross: 2.3011 -> 2.3079
`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 Egan, Semi Ajayi', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.236, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.0536, 'skill_edge': 0.0845, 'height_edge': -0.048, 'attacker_skill': 0.6251, 'defender_skill': 0.4374, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'John Egan', 'market_value': 1500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'Semi Ajayi', 'market_value': 930000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.2229}	totalShots: 1.6712 -> 1.7608 shotOffTarget: 1.0886 -> 1.1469 onTargetScoringAttempt: 0.5753 -> 0.6061 expectedGoalsOnTarget: 0.2879 -> 0.3033 goals: 0.2516 -> 0.2651 expectedGoals: 0.2116 -> 0.2229
`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 1.91 (squad xG sum 1.8156, xG scale 1.0514, xA scale 1.3606, team xA 1.24 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.909` `squad_xg_sum=1.8156` `squad_xa_sum=0.912` `team_expected_assists=1.2409` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.2229 -> 0.2344 expected_assists: 0.096 -> 0.1306
`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.209` `goal_rate=0.2651` `scorer_weight=0.2706`	-
`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=0.1224` `assist_rate=0.0575` `provider_weight=0.1045`	-

Mathias Jensen M 82 0 0.09 0.07 31% 1 1.0 7.1 57.4 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 0.9999, quality 0.502)	`market_value_eur=8300000` `quality_factor=0.5023` `boost_factor=0.9999` `reference_quality=0.505` `history_matches=37` `history_shrink=0.119`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0036)	`blend_weight=0.0036` `history_matches=37` `season_overlap_discount=1.0` `per90_to_match_factor=0.9107` `prior_reliability=0.0467` `competition=Premier League`	fouls: 1.095 -> 1.1051 totalCross: 4.4209 -> 4.4284 totalShots: 1.1158 -> 1.1208
`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 Egan', 'matchup_zone': 'LW vs CB', 'duel_win_rate': 0.429, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.0135, 'skill_edge': 0.0292, 'height_edge': -0.02, 'attacker_skill': 0.5023, 'defender_skill': 0.4374, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'John Egan', 'market_value': 1500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.0824}`	aerialWon: 0.2414 -> 1.1 totalShots: 1.1208 -> 1.1359 shotOffTarget: 0.7737 -> 0.7841
`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 1.91 (squad xG sum 1.8156, xG scale 1.0514, xA scale 1.3606, team xA 1.24 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.909` `squad_xg_sum=1.8156` `squad_xa_sum=0.912` `team_expected_assists=1.2409` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.166 -> 0.2259
`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.083` `goal_rate=0.1012` `scorer_weight=0.0989`	-
`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=0.2022` `assist_rate=0.1186` `provider_weight=0.1807`	-

Mamadou Sangare M 79 0 0.08 0.04 82% 1 0.6 6.9 34.9 10 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 1.0473, quality 0.821)	`market_value_eur=49000000` `quality_factor=0.8206` `boost_factor=1.0473` `reference_quality=0.505` `history_matches=5` `history_shrink=0.5`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0390)	`blend_weight=0.039` `history_matches=5` `season_overlap_discount=1.0` `per90_to_match_factor=0.8763` `prior_reliability=0.12` `competition=Premier League`	interceptionWon: 1.3954 -> 1.4549 totalShots: 1.0837 -> 1.1159 fouls: 2.1721 -> 2.1443 wasFouled: 0.8849 -> 0.9074 totalCross: 0.4392 -> 0.46 wonTackle: 1.9777 -> 1.9575 goalAssist: 0.0663 -> 0.0827 onTargetScoringAttempt: 0.2423 -> 0.2516
`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': 'Regan Slater', '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 1.91 (squad xG sum 1.8156, xG scale 1.0514, xA scale 1.3606, team xA 1.24 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.909` `squad_xg_sum=1.8156` `squad_xa_sum=0.912` `team_expected_assists=1.2409` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.052 -> 0.0707
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 81.0)	`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=0.0724` `goal_rate=0.0878` `scorer_weight=0.0853`	-
`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=0.0683` `assist_rate=0.0827` `provider_weight=0.0884`	-

Keane Lewis-Potter D 83 0 0.18 0.11 59% 1 1.0 6.7 55.6 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 1.0067, quality 0.710)	`market_value_eur=29000000` `quality_factor=0.71` `boost_factor=1.0067` `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=0.9253` `prior_reliability=0.1333` `competition=Premier League`	wasFouled: 0.9636 -> 0.9545 totalCross: 3.9695 -> 3.9757
`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': 'Lewie Coyle', 'matchup_zone': 'RWB vs LW + LB'}`	-
`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 1.91 (squad xG sum 1.8156, xG scale 1.0514, xA scale 1.3606, team xA 1.24 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.909` `squad_xg_sum=1.8156` `squad_xa_sum=0.912` `team_expected_assists=1.2409` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1669 -> 0.1755 expected_assists: 0.1368 -> 0.1861
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 15 by projected rating, fifa_overall 77.0)	`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=0.161` `goal_rate=0.1149` `scorer_weight=0.1404`	-
`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=0.1698` `assist_rate=0.1569` `provider_weight=0.2061`	-

Sepp van den Berg D 90 0 0.09 0.06 52% 1 0.5 6.9 42.6 25 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 1.0095, quality 0.739)	`market_value_eur=33000000` `quality_factor=0.7389` `boost_factor=1.0095` `reference_quality=0.505` `history_matches=32` `history_shrink=0.1351`	-
`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': '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 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 1.91 (squad xG sum 1.8156, xG scale 1.0514, xA scale 1.3606, team xA 1.24 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.909` `squad_xg_sum=1.8156` `squad_xa_sum=0.912` `team_expected_assists=1.2409` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0453 -> 0.0616
`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=0.0876` `goal_rate=0.0` `scorer_weight=0.0734`	-
`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=0.0597` `assist_rate=0.0301` `provider_weight=0.0518`	-

Kristoffer Ajer D 88 0 0.02 0.01 180% 0 0.1 6.7 46.1 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 1.0077, quality 0.694)	`market_value_eur=27000000` `quality_factor=0.6941` `boost_factor=1.0077` `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=0.9742` `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': '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 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 1.91 (squad xG sum 1.8156, xG scale 1.0514, xA scale 1.3606, team xA 1.24 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.909` `squad_xg_sum=1.8156` `squad_xa_sum=0.912` `team_expected_assists=1.2409` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0186 -> 0.0253
`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0168` `goal_rate=0.0` `scorer_weight=0.0135`	-
`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=0.025` `assist_rate=0.0` `provider_weight=0.0202`	-

Michael Kayode D 80 0 0.09 0.05 66% 1 0.3 6.7 56.8 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 1.0086, quality 0.775)	`market_value_eur=39000000` `quality_factor=0.7754` `boost_factor=1.0086` `reference_quality=0.505` `history_matches=42` `history_shrink=0.1064`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0083)	`blend_weight=0.0083` `history_matches=42` `season_overlap_discount=1.0` `per90_to_match_factor=0.8879` `prior_reliability=0.12` `competition=Premier League`	totalShots: 0.6317 -> 0.6469 wasFouled: 1.1019 -> 1.1091 totalCross: 1.122 -> 1.1291
`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': 'Regan Slater', 'matchup_zone': 'Left Mid vs Right Back — Wide Channel'}`	-
`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 1.91 (squad xG sum 1.8156, xG scale 1.0514, xA scale 1.3606, team xA 1.24 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.909` `squad_xg_sum=1.8156` `squad_xa_sum=0.912` `team_expected_assists=1.2409` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0554 -> 0.0754
`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0847` `goal_rate=0.128` `scorer_weight=0.1106`	-
`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=0.0726` `assist_rate=0.0297` `provider_weight=0.0603`	-

Caoimhin Kelleher G 90 0 0.00 - n/a 0.0 7.0 58.2 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 1.0065, quality 0.710)	`market_value_eur=29000000` `quality_factor=0.71` `boost_factor=1.0065` `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: 3.1212 -> 3.1155
`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': 1.1651}`	-
`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 1.91 (squad xG sum 1.8156, xG scale 1.0514, xA scale 1.3606, team xA 1.24 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.909` `squad_xg_sum=1.8156` `squad_xa_sum=0.912` `team_expected_assists=1.2409` `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 4 by projected rating, fifa_overall 80.0)	`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=1.9598`	-
`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=1.1031`	-

Team totals 2 1.91 16 6.9

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
Vitaly Janelt	M	22.8m	32.3%	Market-value substitute prior
Dango Ouattara	F	28.4m	55.2%	Market-value substitute prior

Antagonist / tactical summary

Away midfield dictates play with ~53% possession expected (Tim Iroegbunam vs Mamadou Sangare). Total antagonist xG — Home: 0.69, Away: 1.19. 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 47.3% / 52.7%, dominance ratio 0.44, key duel: Tim Iroegbunam vs Mamadou Sangare

Mohamed Belloumi xG 0.18Oliver McBurnie xG 0.22Sorba Thomas xG 0.12Mathias Jensen xG 0.08Igor Thiago xG 0.51Kevin Schade xG 0.37

Summary

Antagonist matchup analysis for Hull vs Brentford. Most likely scoreline: 1-2. Lambdas derived from 7 home and 4 away individual player duels. Head-to-head: Hull 4W-2D-3L vs Brentford over 9 meetings.

Tactical parity between 5-4-1 and 4-2-3-1. Match outcome driven primarily by individual player quality. Away midfield dictates play with ~53% possession expected (Tim Iroegbunam vs Mamadou Sangare). Total antagonist xG — Home: 0.69, Away: 1.19. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Athletic Club vs Getafe

LaLiga · 20261025 · 4-2-3-1 vs 5-3-2

Predicted 0-0Winner: DrawWinning H 1-0 (15.4%)Winning A 0-1 (13.0%)Confidence 0.7λ home/away 1.02 / 0.86Most-likely P 15.5%

Actual vs predicted — predicted 0-0 (Draw), actual 2-0 ()
Winner MISSExact score MISSO/U HITBTTS HIT

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Athletic Club vs Getafe. 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 5-3-2 (duel lambdas: Home 1.016, Away 0.860). Final calibrated Poisson lambdas: Home 1.016, Away 0.860 (most likely score 0-0).

Home Starters 11/11 XGBoost (10 FBref) Away Starters 11/11 XGBoost (8 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 1.02 / 0.86 Final λ (H/A) 1.02 / 0.86

FBref Team Priors: Athletic Club (Gls/90: 1.0, GA/90: 1.0 → E[λ]: 0.244) vs Getafe (Gls/90: 0.3333333333333333, GA/90: 0.33 → E[λ]: 0.247) | 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	Athletic Club	Getafe
League position	11.00	14.00
Points per game	1.32	1.14
Attack index	0.92	0.69
Defense (concede) index	0.87	0.88
Form multiplier	1.01	0.99
Standing λ multiplier	0.92	0.82

Lambda decomposition (how the goal rates were built)

Stage	Athletic Club λ	Getafe λ
FBref Squad Prior E[λ] (weight 0.228)	0.244	0.247
Pre-competitive-floor	1.016	0.860
Pre-calibration (Micro-duels)	1.016	0.860
Final λ	1.016	0.860

competition calibration scale 1.0competition observed total goals 0.0quality gap applied Falsemismatch score pressure applied Falsecompetitive underdog floor applied Falsestage discount 1.0

Scoreline distribution & markets

Top 5 scorelines	P
0-0	15.5%
1-0	15.4%
1-1	13.5%
0-1	13.0%
2-0	7.9%

Home win	38.1%
Draw	32.2%
Away win	29.7%
Over 2.5	29.0%
BTTS yes	37.0%

Pregame fair line — favorite Athletic Club (38.1%), fair total 1.88, fair spread 0.16

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.

Athletic Club — 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

Álex Berenguer F 84 0 0.12 0.08 49% 1 1.1 6.9 54.4 28 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=38` `history_shrink=0.1163`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0116)	`blend_weight=0.0116` `history_matches=38` `season_overlap_discount=1.0` `per90_to_match_factor=0.9373` `prior_reliability=0.1533` `competition=La Liga`	totalShots: 1.2965 -> 1.3046 wasFouled: 0.9958 -> 0.989
`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 Ives Valou, Abdel Abqar', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 0.9268, 'skill_edge': 0.0022, 'height_edge': -0.052, 'attacker_skill': 0.4511, 'defender_skill': 0.4461, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Jean Ives Valou', 'market_value': 550000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'Abdel Abqar', 'market_value': 3300000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4548}], 'xg_override': 0.1211}	totalShots: 1.3046 -> 1.2091 shotOffTarget: 0.9033 -> 0.8372 onTargetScoringAttempt: 0.3979 -> 0.3688 expectedGoalsOnTarget: 0.1886 -> 0.1748 goals: 0.1432 -> 0.1327 expectedGoals: 0.1307 -> 0.1211 bigChanceMissed: 0.1263 -> 0.1171
`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 1.02 (squad xG sum 1.0289, xG scale 0.9875, xA scale 0.8593, team xA 0.81 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.016` `squad_xg_sum=1.0289` `squad_xa_sum=0.9459` `team_expected_assists=0.8128` `team_xa_to_xg_ratio=0.8`	expected_assists: 0.1263 -> 0.1085
`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 skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.1127` `goal_rate=0.1327` `scorer_weight=0.1277`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.1028` `assist_rate=0.0189` `provider_weight=0.0868`	-

Iñaki Williams F 70 0 0.34 0.21 65% 2 1.6 6.8 55.6 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 0.9995, quality 0.490)	`market_value_eur=7300000` `quality_factor=0.4903` `boost_factor=0.9995` `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=0.7789` `prior_reliability=0.1333` `competition=La Liga`	totalShots: 2.3579 -> 2.3496 onTargetScoringAttempt: 0.8869 -> 0.8938 wasFouled: 0.5681 -> 0.5743 fouls: 0.8753 -> 0.8703
`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': 'Jean Ives Valou, Abdel Abqar', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.212, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.0174, 'skill_edge': 0.0199, 'height_edge': -0.008, 'attacker_skill': 0.4903, 'defender_skill': 0.4461, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Jean Ives Valou', 'market_value': 550000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'Abdel Abqar', 'market_value': 3300000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4548}], 'xg_override': 0.3467}	totalShots: 2.3648 -> 2.406 shotOffTarget: 1.471 -> 1.4966 onTargetScoringAttempt: 0.8938 -> 0.9094 goals: 0.3842 -> 0.3909 bigChanceMissed: 0.349 -> 0.3551 expectedGoals: 0.3408 -> 0.3467
`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 1.02 (squad xG sum 1.0289, xG scale 0.9875, xA scale 0.8593, team xA 0.81 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.016` `squad_xg_sum=1.0289` `squad_xa_sum=0.9459` `team_expected_assists=0.8128` `team_xa_to_xg_ratio=0.8`	expected_assists: 0.063 -> 0.0541
`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=0.2898` `goal_rate=0.3909` `scorer_weight=0.3803`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0527` `assist_rate=0.3051` `provider_weight=0.0676`	-

Peio Canales M 88 0 0.09 0.03 186% 1 0.7 6.9 44.3 6 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 1.0021, quality 0.519)	`market_value_eur=9700000` `quality_factor=0.5189` `boost_factor=1.0021` `reference_quality=0.505` `history_matches=5` `history_shrink=0.5`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0498)	`blend_weight=0.0498` `history_matches=5` `season_overlap_discount=1.0` `per90_to_match_factor=0.9822` `prior_reliability=0.1533` `competition=La Liga`	onTargetScoringAttempt: 0.3642 -> 0.3886 wasFouled: 1.2339 -> 1.2575 totalCross: 2.0042 -> 1.9895 interceptionWon: 0.7206 -> 0.706 totalShots: 1.0396 -> 1.0519 fouls: 0.671 -> 0.6588 wonTackle: 1.0518 -> 1.042 goalAssist: 0.1231 -> 0.117
`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': 'Dani Carvajal, Jean Ives Valou', 'matchup_zone': 'LW vs RWB + CB', 'duel_win_rate': 0.31, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.014, 'skill_edge': 0.0315, 'height_edge': -0.022, 'attacker_skill': 0.5189, 'defender_skill': 0.4489, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Dani Carvajal', 'market_value': 3400000, 'role': 'RWB', 'price_index': 0.7, 'adjusted_quality': 0.4604}, {'player': 'Jean Ives Valou', 'market_value': 550000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.0958}	aerialWon: 0.0 -> 0.9 totalShots: 1.0639 -> 1.0788 shotOffTarget: 0.6753 -> 0.6848 onTargetScoringAttempt: 0.3886 -> 0.394
`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 1.02 (squad xG sum 1.0289, xG scale 0.9875, xA scale 0.8593, team xA 0.81 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.016` `squad_xg_sum=1.0289` `squad_xa_sum=0.9459` `team_expected_assists=0.8128` `team_xa_to_xg_ratio=0.8`	expected_assists: 0.1102 -> 0.0947
`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 skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0903` `goal_rate=0.1018` `scorer_weight=0.0985`	-
`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=0.117` `provider_weight=0.0987`	-

Robert Navarro F 76 0 0.20 0.13 55% 2 0.8 7.1 53.3 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 0.9996, quality 0.493)	`market_value_eur=7500000` `quality_factor=0.4927` `boost_factor=0.9996` `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': 'Jean Ives Valou, Abdel Abqar', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.181, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 0.972, 'skill_edge': 0.021, 'height_edge': -0.04, 'attacker_skill': 0.4927, 'defender_skill': 0.4461, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Jean Ives Valou', 'market_value': 550000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}, {'player': 'Abdel Abqar', 'market_value': 3300000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4548}], 'xg_override': 0.2031}	totalShots: 1.7457 -> 1.6968 shotOffTarget: 1.1216 -> 1.0902 onTargetScoringAttempt: 0.6241 -> 0.6066 expectedGoalsOnTarget: 0.2947 -> 0.2864 goals: 0.2726 -> 0.265 expectedGoals: 0.209 -> 0.2031
`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 1.02 (squad xG sum 1.0289, xG scale 0.9875, xA scale 0.8593, team xA 0.81 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.016` `squad_xg_sum=1.0289` `squad_xa_sum=0.9459` `team_expected_assists=0.8128` `team_xa_to_xg_ratio=0.8`	expected_assists: 0.1536 -> 0.132
`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 skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.1818` `goal_rate=0.265` `scorer_weight=0.2497`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.1237` `assist_rate=0.0` `provider_weight=0.1056`	-

Beñat Gerenabarrena M 79 0 0.03 0.00 592% 0 0.1 6.8 41.1 8 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 0.9945, quality 0.465)	`market_value_eur=5200000` `quality_factor=0.4646` `boost_factor=0.9945` `reference_quality=0.505` `history_matches=6` `history_shrink=0.4545`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0414)	`blend_weight=0.0414` `history_matches=6` `season_overlap_discount=1.0` `per90_to_match_factor=0.8832` `prior_reliability=0.14` `competition=La Liga`	totalCross: 0.9183 -> 0.9847 interceptionWon: 1.8922 -> 1.9357 fouls: 2.2092 -> 2.2396 wonTackle: 1.6329 -> 1.6523 wasFouled: 1.4026 -> 1.4142 onTargetScoringAttempt: 0.1566 -> 0.1673
`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': 'Nemanja Gudelj', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.381, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9431, 'skill_edge': 0.0173, 'height_edge': -0.056, 'attacker_skill': 0.4759, 'defender_skill': 0.4374, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Nemanja Gudelj', 'market_value': 1900000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.4374}], 'xg_override': 0.0273}	aerialWon: 0.6972 -> 0.9 totalShots: 0.5063 -> 0.4775 shotOffTarget: 0.339 -> 0.3197 onTargetScoringAttempt: 0.1673 -> 0.1578
`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 1.02 (squad xG sum 1.0289, xG scale 0.9875, xA scale 0.8593, team xA 0.81 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.016` `squad_xg_sum=1.0289` `squad_xa_sum=0.9459` `team_expected_assists=0.8128` `team_xa_to_xg_ratio=0.8`	expected_assists: 0.0436 -> 0.0375
`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 skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0266` `goal_rate=0.0339` `scorer_weight=0.0293`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0368` `assist_rate=0.06` `provider_weight=0.0441`	-

Iñigo Ruiz de Galarreta M 70 0 0.04 0.02 101% 1 0.6 7.0 59.6 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 0.9977, quality 0.437)	`market_value_eur=1800000` `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 0.0038)	`blend_weight=0.0038` `history_matches=40` `season_overlap_discount=1.0` `per90_to_match_factor=0.7826` `prior_reliability=0.0533` `competition=La Liga`	fouls: 1.9279 -> 1.9355
`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': 'Nemanja Gudelj', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.368, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9236, 'skill_edge': 0.0, 'height_edge': -0.052, 'attacker_skill': 0.4374, 'defender_skill': 0.4374, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Nemanja Gudelj', 'market_value': 1900000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.4374}], 'xg_override': 0.0436}	aerialWon: 1.1363 -> 0.9 totalShots: 0.9331 -> 0.8618 shotOffTarget: 0.6941 -> 0.6411 onTargetScoringAttempt: 0.2379 -> 0.2197
`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 1.02 (squad xG sum 1.0289, xG scale 0.9875, xA scale 0.8593, team xA 0.81 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.016` `squad_xg_sum=1.0289` `squad_xa_sum=0.9459` `team_expected_assists=0.8128` `team_xa_to_xg_ratio=0.8`	expected_assists: 0.1381 -> 0.1187
`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 skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0422` `goal_rate=0.0439` `scorer_weight=0.0432`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.1119` `assist_rate=0.0986` `provider_weight=0.095`	-

Yuri Berchiche D 82 0 0.04 0.02 79% 1 1.0 7.1 60.5 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 0.9976, quality 0.437)	`market_value_eur=910000` `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 0.0057)	`blend_weight=0.0057` `history_matches=37` `season_overlap_discount=1.0` `per90_to_match_factor=0.9063` `prior_reliability=0.0733` `competition=La Liga`	totalShots: 1.1492 -> 1.1566 wonTackle: 0.9959 -> 0.9903
`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': 'Dani Carvajal', 'matchup_zone': 'Right Wingback vs Left Back — Overlapping Run'}`	-
`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 1.02 (squad xG sum 1.0289, xG scale 0.9875, xA scale 0.8593, team xA 0.81 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.016` `squad_xg_sum=1.0289` `squad_xa_sum=0.9459` `team_expected_assists=0.8128` `team_xa_to_xg_ratio=0.8`	expected_assists: 0.1701 -> 0.1461
`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=0.0406` `goal_rate=0.0` `scorer_weight=0.0331`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.1359` `assist_rate=0.1756` `provider_weight=0.1498`	-

Aitor Paredes D 89 0 0.05 0.02 117% 1 0.3 7.0 34.2 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 0.9991, quality 0.488)	`market_value_eur=7100000` `quality_factor=0.4878` `boost_factor=0.9991` `reference_quality=0.505` `history_matches=24` `history_shrink=0.1724`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0134)	`blend_weight=0.0134` `history_matches=24` `season_overlap_discount=1.0` `per90_to_match_factor=0.9846` `prior_reliability=0.12` `competition=La Liga`	totalCross: 0.4022 -> 0.4115 interceptionWon: 1.1453 -> 1.1373 fouls: 1.0571 -> 1.065
`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': 'Ivan Azón', '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 1.02 (squad xG sum 1.0289, xG scale 0.9875, xA scale 0.8593, team xA 0.81 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.016` `squad_xg_sum=1.0289` `squad_xa_sum=0.9459` `team_expected_assists=0.8128` `team_xa_to_xg_ratio=0.8`	-
`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=0.0465` `goal_rate=0.2279` `scorer_weight=0.0595`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0228` `assist_rate=0.0` `provider_weight=0.0185`	-

Yeray Álvarez D 79 0 0.03 0.02 72% 1 0.2 6.7 42.9 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 0.9937, quality 0.437)	`market_value_eur=930000` `quality_factor=0.4374` `boost_factor=0.9937` `reference_quality=0.505` `history_matches=11` `history_shrink=0.3125`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0271)	`blend_weight=0.0271` `history_matches=11` `season_overlap_discount=1.0` `per90_to_match_factor=0.8828` `prior_reliability=0.1333` `competition=La Liga`	totalCross: 0.6179 -> 0.649 totalShots: 0.5305 -> 0.552 wonTackle: 0.656 -> 0.6382 wasFouled: 0.809 -> 0.799 goalAssist: 0.114 -> 0.1229 interceptionWon: 1.0561 -> 1.0634
`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': 'Ivan Azón', '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 1.02 (squad xG sum 1.0289, xG scale 0.9875, xA scale 0.8593, team xA 0.81 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.016` `squad_xg_sum=1.0289` `squad_xa_sum=0.9459` `team_expected_assists=0.8128` `team_xa_to_xg_ratio=0.8`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 14 by projected rating, fifa_overall 77.0)	`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=0.0273` `goal_rate=0.0` `scorer_weight=0.0222`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0141` `assist_rate=0.1229` `provider_weight=0.0178`	-

Jesús Areso D 79 0 0.01 0.01 70% 0 0.1 6.7 39.1 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 0.9985, quality 0.468)	`market_value_eur=5500000` `quality_factor=0.4683` `boost_factor=0.9985` `reference_quality=0.505` `history_matches=32` `history_shrink=0.1351`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0094)	`blend_weight=0.0094` `history_matches=32` `season_overlap_discount=1.0` `per90_to_match_factor=0.8791` `prior_reliability=0.1067` `competition=La Liga`	wasFouled: 0.5308 -> 0.5413 interceptionWon: 1.1346 -> 1.1291
`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 Mojica', 'matchup_zone': 'Left Wingback vs Right Back — Overlapping Run'}`	-
`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 1.02 (squad xG sum 1.0289, xG scale 0.9875, xA scale 0.8593, team xA 0.81 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.016` `squad_xg_sum=1.0289` `squad_xa_sum=0.9459` `team_expected_assists=0.8128` `team_xa_to_xg_ratio=0.8`	expected_assists: 0.0499 -> 0.0429
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 17 by projected rating, fifa_overall 80.0)	`match_squad_rank=17` `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=0.0109` `goal_rate=0.0` `scorer_weight=0.009`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.042` `assist_rate=0.0223` `provider_weight=0.0343`	-

Unai Simón G 90 0 0.00 - n/a 0.0 7.1 61.4 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 1.0042, quality 0.641)	`market_value_eur=21000000` `quality_factor=0.6408` `boost_factor=1.0042` `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=1.0` `prior_reliability=0.2` `competition=La Liga`	saves: 2.961 -> 2.9748
`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': 'Martín Satriano, Ivan Azón', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 0.8596}`	-
`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 1.02 (squad xG sum 1.0289, xG scale 0.9875, xA scale 0.8593, team xA 0.81 = 0.80 x lambda)	`scoreline_reset=true` `team_lambda=1.016` `squad_xg_sum=1.0289` `squad_xa_sum=0.9459` `team_expected_assists=0.8128` `team_xa_to_xg_ratio=0.8`	-
`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` `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=1.1264`	-
`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=0.7546`	-

Team totals 0 1.02 11 6.9

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Alex Padilla	G	4.4m	10.9%	Market-value substitute prior
Dani Vivian	D	14.3m	33.0%	Market-value substitute prior
Aymeric Laporte	D	24.4m	35.1%	Market-value substitute prior
Mikel Jauregizar	M	24.7m	52.3%	Market-value substitute prior
Oihan Sancet	M	22.9m	32.6%	Market-value substitute prior
Nico Williams	F	25.8m	62.0%	Market-value substitute prior

Getafe — projected line vs season baseline (5-3-2)

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

Ivan Azón F 79 0 0.16 0.13 ok 2 1.9 6.8 50.0 7 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 0.9985, quality 0.495)	`market_value_eur=7700000` `quality_factor=0.4951` `boost_factor=0.9985` `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': 'Aitor Paredes, Yeray Álvarez', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.194, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 0.9915, 'skill_edge': 0.0082, 'height_edge': -0.014, 'attacker_skill': 0.4951, 'defender_skill': 0.4768, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Aitor Paredes', 'market_value': 7100000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.5162}, {'player': 'Yeray Álvarez', 'market_value': 930000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.1933}	totalShots: 1.9652 -> 1.9485 shotOffTarget: 1.3207 -> 1.3095 onTargetScoringAttempt: 0.6445 -> 0.639
`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 0.86 (squad xG sum 1.0202, xG scale 0.8429, xA scale 0.8707, team xA 0.56 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=0.86` `squad_xg_sum=1.0202` `squad_xa_sum=0.642` `team_expected_assists=0.559` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1933 -> 0.1629 expected_assists: 0.0615 -> 0.0535
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 70.0)	`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=0.1503` `goal_rate=0.2161` `scorer_weight=0.1782`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0521` `assist_rate=0.1924` `provider_weight=0.0669`	-

Martín Satriano F 82 0 0.19 0.18 ok 2 1.8 6.8 63.0 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 1.0011, quality 0.529)	`market_value_eur=10600000` `quality_factor=0.5295` `boost_factor=1.0011` `reference_quality=0.505` `history_matches=27` `history_shrink=0.1562`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0183)	`blend_weight=0.0183` `history_matches=27` `season_overlap_discount=1.0` `per90_to_match_factor=0.9122` `prior_reliability=0.18` `competition=La Liga`	wasFouled: 1.9801 -> 1.9933 interceptionWon: 0.598 -> 0.5871 fouls: 1.6919 -> 1.6857 totalCross: 0.6509 -> 0.6452 onTargetScoringAttempt: 0.6457 -> 0.6401
`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': 'Aitor Paredes, Yeray Álvarez', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.234, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.0495, 'skill_edge': 0.0237, 'height_edge': 0.01, 'attacker_skill': 0.5295, 'defender_skill': 0.4768, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Aitor Paredes', 'market_value': 7100000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.5162}, {'player': 'Yeray Álvarez', 'market_value': 930000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.4374}], 'xg_override': 0.2236}	totalShots: 2.0264 -> 2.1267 shotOffTarget: 1.3811 -> 1.4495 onTargetScoringAttempt: 0.6401 -> 0.6718 expectedGoalsOnTarget: 0.3117 -> 0.3271 goals: 0.2234 -> 0.2345 expectedGoals: 0.2131 -> 0.2236
`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 0.86 (squad xG sum 1.0202, xG scale 0.8429, xA scale 0.8707, team xA 0.56 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=0.86` `squad_xg_sum=1.0202` `squad_xa_sum=0.642` `team_expected_assists=0.559` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.2236 -> 0.1885 expected_assists: 0.0673 -> 0.0586
`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 skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.1718` `goal_rate=0.2345` `scorer_weight=0.196`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0569` `assist_rate=0.1403` `provider_weight=0.0732`	-

Mario Martín M 76 0 0.10 0.09 ok 1 0.8 6.7 59.6 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 1.0008, quality 0.531)	`market_value_eur=10700000` `quality_factor=0.5306` `boost_factor=1.0008` `reference_quality=0.505` `history_matches=41` `history_shrink=0.1087`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0033)	`blend_weight=0.0033` `history_matches=41` `season_overlap_discount=1.0` `per90_to_match_factor=0.8479` `prior_reliability=0.0467` `competition=La Liga`	wasFouled: 1.4903 -> 1.5014
`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': 'Beñat Gerenabarrena', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.497, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.0688, 'skill_edge': 0.0308, 'height_edge': 0.016, 'attacker_skill': 0.5444, 'defender_skill': 0.4759, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Beñat Gerenabarrena', 'market_value': 5200000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.4759}], 'xg_override': 0.1212}	aerialWon: 1.6935 -> 1.0 totalShots: 1.213 -> 1.2965 shotOffTarget: 0.7886 -> 0.8429 onTargetScoringAttempt: 0.4244 -> 0.4536 bigChanceMissed: 0.1342 -> 0.1434 expectedGoals: 0.1134 -> 0.1212 goals: 0.0883 -> 0.0944
`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 0.86 (squad xG sum 1.0202, xG scale 0.8429, xA scale 0.8707, team xA 0.56 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=0.86` `squad_xg_sum=1.0202` `squad_xa_sum=0.642` `team_expected_assists=0.559` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1212 -> 0.1022 expected_assists: 0.0554 -> 0.0482
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 74.0)	`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=0.0972` `goal_rate=0.0944` `scorer_weight=0.0862`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0471` `assist_rate=0.0722` `provider_weight=0.0559`	-

Ramón Terrats M 75 0 0.15 0.08 94% 2 1.1 6.8 60.6 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 0.9977, quality 0.440)	`market_value_eur=3200000` `quality_factor=0.4399` `boost_factor=0.9977` `reference_quality=0.505` `history_matches=37` `history_shrink=0.119`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0155)	`blend_weight=0.0155` `history_matches=37` `season_overlap_discount=1.0` `per90_to_match_factor=0.8334` `prior_reliability=0.2` `competition=La Liga`	totalCross: 1.7465 -> 1.7754 totalShots: 1.6548 -> 1.6724 wasFouled: 2.0963 -> 2.1069 interceptionWon: 0.6786 -> 0.6724 fouls: 1.0543 -> 1.0595
`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': 'Beñat Gerenabarrena', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.46, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.0144, 'skill_edge': -0.0142, 'height_edge': 0.024, 'attacker_skill': 0.4443, 'defender_skill': 0.4759, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Beñat Gerenabarrena', 'market_value': 5200000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.4759}], 'xg_override': 0.1776}	aerialWon: 0.8458 -> 1.0 totalShots: 1.6724 -> 1.6965 shotOffTarget: 1.095 -> 1.1108 onTargetScoringAttempt: 0.5598 -> 0.5679
`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 0.86 (squad xG sum 1.0202, xG scale 0.8429, xA scale 0.8707, team xA 0.56 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=0.86` `squad_xg_sum=1.0202` `squad_xa_sum=0.642` `team_expected_assists=0.559` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1776 -> 0.1497 expected_assists: 0.1222 -> 0.1063
`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_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=0.139` `goal_rate=0.1704` `scorer_weight=0.145`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.1008` `assist_rate=0.1204` `provider_weight=0.1052`	-

Nemanja Gudelj M 87 0 0.06 0.03 78% 1 0.9 6.7 61.3 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 0.9978, 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 0.0009)	`blend_weight=0.0009` `history_matches=41` `season_overlap_discount=1.0` `per90_to_match_factor=0.9634` `prior_reliability=0.0133` `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': 'Beñat Gerenabarrena', 'matchup_zone': 'CDM vs CDM', 'duel_win_rate': 0.459, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.0569, 'skill_edge': -0.0173, 'height_edge': 0.056, 'attacker_skill': 0.4374, 'defender_skill': 0.4759, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Beñat Gerenabarrena', 'market_value': 5200000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.4759}], 'xg_override': 0.0727}	aerialWon: 1.4295 -> 1.2 totalShots: 0.8625 -> 0.9116 shotOffTarget: 0.6512 -> 0.6883 onTargetScoringAttempt: 0.2065 -> 0.2182
`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 0.86 (squad xG sum 1.0202, xG scale 0.8429, xA scale 0.8707, team xA 0.56 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=0.86` `squad_xg_sum=1.0202` `squad_xa_sum=0.642` `team_expected_assists=0.559` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0727 -> 0.0613
`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 skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0595` `goal_rate=0.0619` `scorer_weight=0.0559`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0294` `assist_rate=0.0473` `provider_weight=0.0344`	-

Johan Mojica D 85 0 0.02 0.02 ok 1 0.4 6.7 59.5 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 0.9979, quality 0.437)	`market_value_eur=1400000` `quality_factor=0.4374` `boost_factor=0.9979` `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=0.9484` `prior_reliability=0.1333` `competition=La Liga`	totalCross: 7.0273 -> 7.0495 wasFouled: 1.2052 -> 1.22
`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': 'Jesús Areso', 'matchup_zone': 'Left Wingback vs Right Back — Overlapping Run'}`	-
`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 0.86 (squad xG sum 1.0202, xG scale 0.8429, xA scale 0.8707, team xA 0.56 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=0.86` `squad_xg_sum=1.0202` `squad_xa_sum=0.642` `team_expected_assists=0.559` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0999 -> 0.087
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 76.0)	`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=0.0197` `goal_rate=0.0` `scorer_weight=0.0159`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0833` `assist_rate=0.0372` `provider_weight=0.0696`	-

Jean Ives Valou D 90 0 0.03 - n/a 0 0.0 6.6 5.8 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 0.9855, quality 0.437)	`market_value_eur=550000` `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': 'Peio Canales', 'matchup_zone': 'LW vs RWB + 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 0.86 (squad xG sum 1.0202, xG scale 0.8429, xA scale 0.8707, team xA 0.56 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=0.86` `squad_xg_sum=1.0202` `squad_xa_sum=0.642` `team_expected_assists=0.559` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0354 -> 0.0298
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 15 by projected rating, fifa_overall 66.0)	`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=0.0294` `goal_rate=0.0` `scorer_weight=0.0238`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0028` `assist_rate=0.0` `provider_weight=0.0026`	-

Abdel Abqar D 80 0 0.00 0.01 81% 0 0.1 6.6 50.4 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 0.9970, quality 0.441)	`market_value_eur=3300000` `quality_factor=0.4412` `boost_factor=0.997` `reference_quality=0.505` `history_matches=27` `history_shrink=0.1562`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0068)	`blend_weight=0.0068` `history_matches=27` `season_overlap_discount=1.0` `per90_to_match_factor=0.8872` `prior_reliability=0.0667` `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': 'Iñaki Williams', '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 0.86 (squad xG sum 1.0202, xG scale 0.8429, xA scale 0.8707, team xA 0.56 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=0.86` `squad_xg_sum=1.0202` `squad_xa_sum=0.642` `team_expected_assists=0.559` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 14 by projected rating, fifa_overall 79.0)	`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=0.0013` `goal_rate=0.0` `scorer_weight=0.0013`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0075` `assist_rate=0.0664` `provider_weight=0.0094`	-

Dani Carvajal D 82 0 0.02 0.01 95% 0 0.2 6.8 37.5 18 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 0.9959, quality 0.442)	`market_value_eur=3400000` `quality_factor=0.4424` `boost_factor=0.9959` `reference_quality=0.505` `history_matches=18` `history_shrink=0.2174`	-
`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': 'Peio Canales', 'matchup_zone': 'LW vs RWB + 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 0.86 (squad xG sum 1.0202, xG scale 0.8429, xA scale 0.8707, team xA 0.56 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=0.86` `squad_xg_sum=1.0202` `squad_xa_sum=0.642` `team_expected_assists=0.559` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.104 -> 0.0906
`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_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=0.0219` `goal_rate=0.0` `scorer_weight=0.0177`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0866` `assist_rate=0.1525` `provider_weight=0.1133`	-

Kiko Femenía D 64 0 0.02 0.03 37% 0 0.2 6.6 55.7 22 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=525000` `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=0.7122` `prior_reliability=0.1` `competition=La Liga`	interceptionWon: 0.7823 -> 0.7756
`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': 'Álex Berenguer', '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 0.86 (squad xG sum 1.0202, xG scale 0.8429, xA scale 0.8707, team xA 0.56 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=0.86` `squad_xg_sum=1.0202` `squad_xa_sum=0.642` `team_expected_assists=0.559` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0545 -> 0.0475
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 18 by projected rating, fifa_overall 74.0)	`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 skipped: 0 projected goals	`team_goals=0` `allocation_applied=false` `goals_allocated=0` `goal_probability=0.0175` `goal_rate=0.0316` `scorer_weight=0.0203`	-
`integer_assist_projection`	Team scoreline allocation skipped: 0 projected assists	`allocation_applied=false` `team_assists=0` `assists_allocated=0` `assist_probability=0.0464` `assist_rate=0.0042` `provider_weight=0.038`	-

David Soria G 90 0 0.00 - n/a 0.0 7.1 68.6 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 0.9981, quality 0.437)	`market_value_eur=2700000` `quality_factor=0.4374` `boost_factor=0.9981` `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=1.0` `prior_reliability=0.2` `competition=La Liga`	saves: 3.1365 -> 3.1306 wasFouled: 0.2497 -> 0.2549
`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': 'Iñaki Williams, Robert Navarro', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.0156}`	-
`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 0.86 (squad xG sum 1.0202, xG scale 0.8429, xA scale 0.8707, team xA 0.56 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=0.86` `squad_xg_sum=1.0202` `squad_xa_sum=0.642` `team_expected_assists=0.559` `team_xa_to_xg_ratio=0.65`	shots_faced: 9.6 -> 9.7 expected_saves: 2.34 -> 2.38
`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=0.0` `scorer_weight=0.0` `squad_weight_total=0.8413`	-
`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=0.597`	-

Team totals 0 0.86 11 6.8

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Jiří Letáček	G	4.4m	10.9%	Market-value substitute prior
Andrés García	D	13.8m	22.2%	Market-value substitute prior
Zaid Romero	D	13.7m	21.9%	Market-value substitute prior
Orel Mangala	M	22.2m	31.2%	Market-value substitute prior
Mauro Arambarri	M	22.2m	51.0%	Market-value substitute prior
Christantus Uche	F	25.8m	58.1%	Market-value substitute prior

Antagonist / tactical summary

Midfield is evenly contested — Peio Canales vs Mario Martín will be the pivotal duel. Total antagonist xG — Home: 0.84, Away: 0.84. 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 50.6% / 49.4%, dominance ratio 0.50, key duel: Peio Canales vs Mario Martín

Peio Canales xG 0.10Iñaki Williams xG 0.35Álex Berenguer xG 0.12Ivan Azón xG 0.19Martín Satriano xG 0.22Johan Mojica xG 0.02

Summary

Antagonist matchup analysis for Athletic Club vs Getafe. Most likely scoreline: 0-0. Lambdas derived from 6 home and 7 away individual player duels. Head-to-head: Athletic Club 2W-6D-2L vs Getafe over 10 meetings.

Tactical parity between 4-2-3-1 and 5-3-2. Match outcome driven primarily by individual player quality. Midfield is evenly contested — Peio Canales vs Mario Martín will be the pivotal duel. Total antagonist xG — Home: 0.84, Away: 0.84. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Sunderland vs Leeds

Premier League · 20261025 · 4-2-3-1 vs 3-4-2-1

Predicted 1-1Winner: DrawWinning H 1-0 (8.1%)Winning A 0-1 (8.9%)Confidence 0.71λ home/away 1.33 / 1.45Most-likely P 12.1%

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

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Sunderland 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 1.089, Away 1.166). Final calibrated Poisson lambdas: Home 1.327, Away 1.454 (most likely score 1-1).

Home Starters 11/11 XGBoost (11 FBref) Away Starters 11/11 XGBoost (10 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 1.09 / 1.17 Final λ (H/A) 1.33 / 1.45

FBref Team Priors: Sunderland (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	Sunderland	Leeds
League position	20.00	7.00
Points per game	0.93	1.40
Attack index	0.90	0.91
Defense (concede) index	0.81	1.20
Form multiplier	0.97	1.03
Standing λ multiplier	0.88	1.04

Lambda decomposition (how the goal rates were built)

Stage	Sunderland λ	Leeds λ
FBref Squad Prior E[λ] (weight 0.192)	0.690	0.690
Pre-competitive-floor	1.056	1.283
Pre-calibration (Micro-duels)	1.089	1.166
Final λ	1.327	1.454

competition calibration scale 1.189competition 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	12.1%
0-1	8.9%
1-2	8.7%
1-0	8.1%
2-1	7.9%

Home win	34.3%
Draw	25.6%
Away win	40.1%
Over 2.5	52.6%
BTTS yes	56.4%

Pregame fair line — favorite Leeds (40.1%), fair total 2.78, fair spread -0.13

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.

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

Luke O'Nien D 86 0 0.02 0.00 376% 1 0.1 6.7 21.3 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 0.9944, quality 0.437)	`market_value_eur=660000` `quality_factor=0.4374` `boost_factor=0.9944` `reference_quality=0.505` `history_matches=13` `history_shrink=0.2778`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0084)	`blend_weight=0.0084` `history_matches=13` `season_overlap_discount=1.0` `per90_to_match_factor=0.9539` `prior_reliability=0.0467` `competition=Premier League`	fouls: 0.9138 -> 0.9291 totalCross: 1.2324 -> 1.245 interceptionWon: 0.8246 -> 0.8176 wonTackle: 0.7698 -> 0.7633 totalShots: 0.5533 -> 0.5596 wasFouled: 0.6836 -> 0.6778
`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': '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 1.33 (squad xG sum 1.3328, xG scale 0.9956, xA scale 0.9956, team xA 0.90 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.327` `squad_xg_sum=1.3328` `squad_xa_sum=0.9052` `team_expected_assists=0.9012` `team_xa_to_xg_ratio=0.6792`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 10 by projected rating, fifa_overall 73.0)	`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=0.0222` `goal_rate=0.0` `scorer_weight=0.018`	-
`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=0.0199` `assist_rate=0.2019` `provider_weight=0.0251`	-

Nordi Mukiele D 88 0 0.14 0.07 89% 1 0.8 6.8 50.7 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 1.0035, quality 0.602)	`market_value_eur=17100000` `quality_factor=0.6016` `boost_factor=1.0035` `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=0.9759` `prior_reliability=0.0667` `competition=Premier League`	totalCross: 3.2164 -> 3.2046 fouls: 1.2957 -> 1.2888 wonTackle: 1.24 -> 1.2334 totalShots: 1.0834 -> 1.0777
`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': '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 1.33 (squad xG sum 1.3328, xG scale 0.9956, xA scale 0.9956, team xA 0.90 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.327` `squad_xg_sum=1.3328` `squad_xa_sum=0.9052` `team_expected_assists=0.9012` `team_xa_to_xg_ratio=0.6792`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 81.0)	`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=0.1277` `goal_rate=0.085` `scorer_weight=0.1093`	-
`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=0.0891` `assist_rate=0.0415` `provider_weight=0.0746`	-

Robin Roefs G 90 0 0.00 - n/a 0.0 7.0 54.0 27 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 1.0073, quality 0.725)	`market_value_eur=31000000` `quality_factor=0.7249` `boost_factor=1.0073` `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`	saves: 2.9394 -> 2.9159
`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, Jayden Bogle', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.1657}`	-
`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 1.33 (squad xG sum 1.3328, xG scale 0.9956, xA scale 0.9956, team xA 0.90 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.327` `squad_xg_sum=1.3328` `squad_xa_sum=0.9052` `team_expected_assists=0.9012` `team_xa_to_xg_ratio=0.6792`	shots_faced: 11.1 -> 13.8 expected_saves: 2.72 -> 3.38 high_pressure_saves: 0.93 -> 1.16
`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=0.0` `scorer_weight=0.0` `squad_weight_total=1.3703`	-
`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=0.7923`	-

Omar Alderete D 78 0 0.04 0.02 69% 0 0.4 6.8 44.3 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 1.0009, quality 0.529)	`market_value_eur=10600000` `quality_factor=0.5295` `boost_factor=1.0009` `reference_quality=0.505` `history_matches=34` `history_shrink=0.1282`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0017)	`blend_weight=0.0017` `history_matches=34` `season_overlap_discount=1.0` `per90_to_match_factor=0.8679` `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': 'Dominic Calvert-Lewin', '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 1.33 (squad xG sum 1.3328, xG scale 0.9956, xA scale 0.9956, team xA 0.90 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.327` `squad_xg_sum=1.3328` `squad_xa_sum=0.9052` `team_expected_assists=0.9012` `team_xa_to_xg_ratio=0.6792`	-
`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` `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=0.0355` `goal_rate=0.0022` `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=0.0133` `assist_rate=0.0` `provider_weight=0.0107`	-

Reinildo Mandava D 87 0 0.00 0.00 30% 0 0.0 6.7 48.3 22 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 0.9976, quality 0.451)	`market_value_eur=4099999` `quality_factor=0.4511` `boost_factor=0.9976` `reference_quality=0.505` `history_matches=29` `history_shrink=0.1471`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0127)	`blend_weight=0.0127` `history_matches=29` `season_overlap_discount=1.0` `per90_to_match_factor=0.9683` `prior_reliability=0.1333` `competition=Premier League`	fouls: 1.5412 -> 1.5524 interceptionWon: 1.1812 -> 1.1908 totalCross: 0.7024 -> 0.712 totalShots: 0.393 -> 0.388
`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': 'Lukas Nmecha', '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 1.33 (squad xG sum 1.3328, xG scale 0.9956, xA scale 0.9956, team xA 0.90 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.327` `squad_xg_sum=1.3328` `squad_xa_sum=0.9052` `team_expected_assists=0.9012` `team_xa_to_xg_ratio=0.6792`	-
`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: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0021` `goal_rate=0.0` `scorer_weight=0.002`	-
`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=0.0437` `assist_rate=0.0212` `provider_weight=0.0358`	-

Noah Sadiki M 83 0 0.09 0.06 51% 1 0.7 6.8 55.0 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 1.0100, 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 0.0101)	`blend_weight=0.0101` `history_matches=38` `season_overlap_discount=1.0` `per90_to_match_factor=0.9267` `prior_reliability=0.1333` `competition=Premier League`	wonTackle: 0.7677 -> 0.774 wasFouled: 1.3517 -> 1.3568
`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, Lukas Nmecha, Anton Stach', '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 1.33 (squad xG sum 1.3328, xG scale 0.9956, xA scale 0.9956, team xA 0.90 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.327` `squad_xg_sum=1.3328` `squad_xa_sum=0.9052` `team_expected_assists=0.9012` `team_xa_to_xg_ratio=0.6792`	-
`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=0.087` `goal_rate=0.0739` `scorer_weight=0.0797`	-
`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=0.0624` `assist_rate=0.0763` `provider_weight=0.0709`	-

Brian Brobbey F 81 1 0.43 0.28 54% 3 1.4 7.0 54.7 28 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 1.0120, quality 0.834)	`market_value_eur=53000000` `quality_factor=0.834` `boost_factor=1.012` `reference_quality=0.505` `history_matches=36` `history_shrink=0.122`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0074)	`blend_weight=0.0074` `history_matches=36` `season_overlap_discount=1.0` `per90_to_match_factor=0.8988` `prior_reliability=0.0933` `competition=Premier League`	totalOffside: 0.5035 -> 0.514 wasFouled: 1.9025 -> 1.8932 totalShots: 2.8469 -> 2.8397 onTargetScoringAttempt: 1.2978 -> 1.2928
`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': 'Tarik Muharemović, Nico Elvedi', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.213, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 1.0197, 'skill_edge': 0.0554, 'height_edge': -0.042, 'attacker_skill': 0.834, 'defender_skill': 0.7109, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Tarik Muharemović', 'market_value': 43000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8462}, {'player': 'Nico Elvedi', 'market_value': 11000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.5755}], 'xg_override': 0.4309}	totalShots: 2.8419 -> 2.8979 shotOffTarget: 1.5491 -> 1.5796 onTargetScoringAttempt: 1.2928 -> 1.3183 goals: 0.5353 -> 0.5458 expectedGoalsOnTarget: 0.5333 -> 0.5438 expectedGoals: 0.4226 -> 0.4309 bigChanceMissed: 0.3856 -> 0.3932
`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 1.33 (squad xG sum 1.3328, xG scale 0.9956, xA scale 0.9956, team xA 0.90 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.327` `squad_xg_sum=1.3328` `squad_xa_sum=0.9052` `team_expected_assists=0.9012` `team_xa_to_xg_ratio=0.6792`	-
`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=0.3489` `goal_rate=0.5458` `scorer_weight=0.5316`	-
`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=0.0409` `assist_rate=0.0` `provider_weight=0.0334`	-

Enzo Le Fée M 89 0 0.24 0.17 41% 2 1.1 6.9 58.4 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 1.0064, quality 0.702)	`market_value_eur=28000000` `quality_factor=0.7022` `boost_factor=1.0064` `reference_quality=0.505` `history_matches=41` `history_shrink=0.1087`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0090)	`blend_weight=0.009` `history_matches=41` `season_overlap_discount=1.0` `per90_to_match_factor=0.987` `prior_reliability=0.1267` `competition=Premier League`	totalCross: 3.0953 -> 3.0862
`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': '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 1.33 (squad xG sum 1.3328, xG scale 0.9956, xA scale 0.9956, team xA 0.90 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.327` `squad_xg_sum=1.3328` `squad_xa_sum=0.9052` `team_expected_assists=0.9012` `team_xa_to_xg_ratio=0.6792`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 80.0)	`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=0.2153` `goal_rate=0.2228` `scorer_weight=0.2253`	-
`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=0.16` `assist_rate=0.1379` `provider_weight=0.1455`	-

Habib Diarra M 84 0 0.13 0.12 ok 1 0.9 6.7 49.7 23 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 1.0144, quality 0.764)	`market_value_eur=37000000` `quality_factor=0.7641` `boost_factor=1.0144` `reference_quality=0.505` `history_matches=22` `history_shrink=0.1852`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0024)	`blend_weight=0.0024` `history_matches=22` `season_overlap_discount=1.0` `per90_to_match_factor=0.9331` `prior_reliability=0.02` `competition=Premier League`	goalAssist: 0.0985 -> 0.1058 wasFouled: 2.0981 -> 2.093
`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': 'Anton Stach', 'matchup_zone': 'CAM vs CM', 'duel_win_rate': 0.38, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.9413, 'skill_edge': 0.012, 'height_edge': -0.052, 'attacker_skill': 0.7752, 'defender_skill': 0.7484, 'attacker_role': 'CAM', 'attacker_mv_price_index': 0.95, 'defender_adjusted_parts': [{'player': 'Anton Stach', 'market_value': 31000000, 'role': 'CM', 'price_index': 0.9, 'adjusted_quality': 0.7484}], 'xg_override': 0.1337}`	aerialWon: 0.6625 -> 1.0 totalShots: 1.4992 -> 1.4112 shotOffTarget: 1.0042 -> 0.9453 onTargetScoringAttempt: 0.4904 -> 0.4616 bigChanceMissed: 0.2083 -> 0.1961 expectedGoalsOnTarget: 0.1621 -> 0.1526 expectedGoals: 0.142 -> 0.1337 goals: 0.12 -> 0.113
`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 1.33 (squad xG sum 1.3328, xG scale 0.9956, xA scale 0.9956, team xA 0.90 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.327` `squad_xg_sum=1.3328` `squad_xa_sum=0.9052` `team_expected_assists=0.9012` `team_xa_to_xg_ratio=0.6792`	-
`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=0.1246` `goal_rate=0.113` `scorer_weight=0.1181`	-
`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=0.0931` `assist_rate=0.1058` `provider_weight=0.1028`	-

Trai Hume D 85 0 0.06 0.06 ok 1 0.9 6.7 54.6 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 1.0040, quality 0.631)	`market_value_eur=20000000` `quality_factor=0.6311` `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=0.9468` `prior_reliability=0.1267` `competition=Premier League`	wasFouled: 0.8536 -> 0.8461 wonTackle: 1.1363 -> 1.1438
`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': '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 1.33 (squad xG sum 1.3328, xG scale 0.9956, xA scale 0.9956, team xA 0.90 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.327` `squad_xg_sum=1.3328` `squad_xa_sum=0.9052` `team_expected_assists=0.9012` `team_xa_to_xg_ratio=0.6792`	-
`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=0.0585` `goal_rate=0.119` `scorer_weight=0.0754`	-
`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=0.0016` `provider_weight=0.0801`	-

Granit Xhaka M 90 0 0.03 0.02 31% 1 0.5 7.3 59.5 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 1.0000, quality 0.505)	`market_value_eur=8500000` `quality_factor=0.5047` `boost_factor=1.0` `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=0.9947` `prior_reliability=0.1333` `competition=Premier League`	wasFouled: 1.5944 -> 1.613 totalCross: 4.2514 -> 4.2634 totalShots: 0.7339 -> 0.7267 wonTackle: 1.1969 -> 1.19
`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': 'Tarik Muharemović', 'matchup_zone': 'LW vs CB', 'duel_win_rate': 0.234, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.7271, 'skill_edge': -0.1537, 'height_edge': -0.032, 'attacker_skill': 0.5047, 'defender_skill': 0.8462, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Tarik Muharemović', 'market_value': 43000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8462}], 'xg_override': 0.0273}	aerialWon: 0.9277 -> 1.2 totalShots: 0.7322 -> 0.5324 shotOffTarget: 0.5618 -> 0.4085 onTargetScoringAttempt: 0.1704 -> 0.1239 expectedGoals: 0.0376 -> 0.0273 goals: 0.037 -> 0.0269
`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 1.33 (squad xG sum 1.3328, xG scale 0.9956, xA scale 0.9956, team xA 0.90 = 0.68 x lambda)	`scoreline_reset=true` `team_lambda=1.327` `squad_xg_sum=1.3328` `squad_xa_sum=0.9052` `team_expected_assists=0.9012` `team_xa_to_xg_ratio=0.6792`	-
`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=0.0268` `goal_rate=0.0269` `scorer_weight=0.0271`	-
`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=0.1619` `assist_rate=0.1114` `provider_weight=0.1413`	-

Team totals 1 1.33 14 6.9

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
Dayann Methalie	M	23.0m	54.5%	Market-value substitute prior
Chris Rigg	M	15.8m	71.1%	Market-value substitute prior
Malick Fofana	F	26.4m	60.8%	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

Dominic Calvert-Lewin F 85 1 0.44 0.42 ok 3 1.8 6.9 55.5 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 1.0063, quality 0.694)	`market_value_eur=27000000` `quality_factor=0.6941` `boost_factor=1.0063` `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=0.9414` `prior_reliability=0.1333` `competition=Premier League`	totalShots: 2.4763 -> 2.4706 totalOffside: 0.819 -> 0.8247 onTargetScoringAttempt: 1.0445 -> 1.039
`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, Omar Alderete', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle', 'duel_win_rate': 0.297, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 1.04, 'skill_edge': -0.0008, 'height_edge': 0.028, 'attacker_skill': 0.6941, 'defender_skill': 0.6959, 'attacker_role': 'SS', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Noah Sadiki', 'market_value': 42000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8221}, {'player': 'Omar Alderete', 'market_value': 10600000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.5697}], 'xg_override': 0.4292}	totalShots: 2.4708 -> 2.5697 shotOffTarget: 1.4318 -> 1.4891 onTargetScoringAttempt: 1.039 -> 1.0806 bigChanceMissed: 0.9268 -> 0.9639 expectedGoalsOnTarget: 0.624 -> 0.649 goals: 0.4387 -> 0.4562 expectedGoals: 0.4127 -> 0.4292
`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 1.45 (squad xG sum 1.4149, xG scale 1.0277, xA scale 1.1324, team xA 0.95 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.454` `squad_xg_sum=1.4149` `squad_xa_sum=0.8346` `team_expected_assists=0.9451` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.4292 -> 0.4411
`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: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.3567` `goal_rate=0.4562` `scorer_weight=0.4659`	-
`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=0.0353` `assist_rate=0.005` `provider_weight=0.0287`	-

Lukas Nmecha F 74 0 0.17 0.11 56% 2 0.7 6.7 43.9 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 1.0003, quality 0.512)	`market_value_eur=9100000` `quality_factor=0.5118` `boost_factor=1.0003` `reference_quality=0.505` `history_matches=35` `history_shrink=0.125`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0011)	`blend_weight=0.0011` `history_matches=35` `season_overlap_discount=1.0` `per90_to_match_factor=0.8189` `prior_reliability=0.0133` `competition=Premier League`	totalCross: 0.7038 -> 0.7119
`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, Omar Alderete', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle', 'duel_win_rate': 0.207, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 0.9077, 'skill_edge': -0.0828, 'height_edge': 0.02, 'attacker_skill': 0.5118, 'defender_skill': 0.6959, 'attacker_role': 'SS', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Noah Sadiki', 'market_value': 42000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8221}, {'player': 'Omar Alderete', 'market_value': 10600000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.5697}], 'xg_override': 0.1606}	totalShots: 1.5344 -> 1.3928 shotOffTarget: 0.9753 -> 0.8853 onTargetScoringAttempt: 0.5565 -> 0.5051 bigChanceMissed: 0.4044 -> 0.3671 expectedGoalsOnTarget: 0.325 -> 0.295 goals: 0.1911 -> 0.1735 expectedGoals: 0.1769 -> 0.1606
`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 1.45 (squad xG sum 1.4149, xG scale 1.0277, xA scale 1.1324, team xA 0.95 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.454` `squad_xg_sum=1.4149` `squad_xa_sum=0.8346` `team_expected_assists=0.9451` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.059 -> 0.0668
`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: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1521` `goal_rate=0.1735` `scorer_weight=0.1751`	-
`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=0.0646` `assist_rate=0.119` `provider_weight=0.0835`	-

Anton Stach M 87 0 0.16 0.09 80% 2 1.3 7.0 56.9 24 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 1.0085, quality 0.725)	`market_value_eur=31000000` `quality_factor=0.7249` `boost_factor=1.0085` `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=0.9654` `prior_reliability=0.1333` `competition=Premier League`	totalCross: 6.6895 -> 6.6581 wasFouled: 1.2915 -> 1.3093 interceptionWon: 1.5057 -> 1.5211 fouls: 1.7226 -> 1.7142 totalShots: 1.4495 -> 1.4548
`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': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.477, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 1.0395, 'skill_edge': -0.0331, 'height_edge': 0.06, 'attacker_skill': 0.7484, 'defender_skill': 0.8221, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Noah Sadiki', 'market_value': 42000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8221}], 'xg_override': 0.1569}	aerialWon: 2.2793 -> 1.4 totalShots: 1.4548 -> 1.5123 shotOffTarget: 0.9905 -> 1.0296 onTargetScoringAttempt: 0.4592 -> 0.4773 bigChanceMissed: 0.2995 -> 0.3113 goals: 0.1951 -> 0.2028 expectedGoals: 0.1509 -> 0.1569
`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 1.45 (squad xG sum 1.4149, xG scale 1.0277, xA scale 1.1324, team xA 0.95 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.454` `squad_xg_sum=1.4149` `squad_xa_sum=0.8346` `team_expected_assists=0.9451` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.1103 -> 0.1249
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 80.0)	`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=0.1489` `goal_rate=0.2028` `scorer_weight=0.197`	-
`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=0.1174` `assist_rate=0.0952` `provider_weight=0.1131`	-

Gabriel Gudmundsson M 83 0 0.04 0.01 427% 1 0.2 6.9 38.5 30 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 1.0068, quality 0.686)	`market_value_eur=26000000` `quality_factor=0.6858` `boost_factor=1.0068` `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': 'Nordi Mukiele', 'matchup_zone': 'Left Wingback vs Right Back — Overlapping Run', 'duel_win_rate': 0.426, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.0378, 'skill_edge': 0.0417, 'height_edge': -0.016, 'attacker_skill': 0.765, 'defender_skill': 0.6723, 'attacker_role': 'LWB', 'attacker_mv_price_index': 0.7, 'defender_adjusted_parts': [{'player': 'Nordi Mukiele', 'market_value': 17100000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.6723}], 'xg_override': 0.0388}	aerialWon: 0.8318 -> 1.1 totalShots: 0.6562 -> 0.681 shotOffTarget: 0.4924 -> 0.511 onTargetScoringAttempt: 0.1638 -> 0.17
`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 1.45 (squad xG sum 1.4149, xG scale 1.0277, xA scale 1.1324, team xA 0.95 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.454` `squad_xg_sum=1.4149` `squad_xa_sum=0.8346` `team_expected_assists=0.9451` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0857 -> 0.097
`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=0.0391` `goal_rate=0.0409` `scorer_weight=0.0408`	-
`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=0.0771` `provider_weight=0.0909`	-

Ao Tanaka M 75 0 0.07 0.05 33% 1 0.4 6.9 34.4 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 1.0035, quality 0.592)	`market_value_eur=16200000` `quality_factor=0.5921` `boost_factor=1.0035` `reference_quality=0.505` `history_matches=32` `history_shrink=0.1351`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0029)	`blend_weight=0.0029` `history_matches=32` `season_overlap_discount=1.0` `per90_to_match_factor=0.8367` `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': 'Noah Sadiki', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.383, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.9015, 'skill_edge': -0.095, 'height_edge': 0.028, 'attacker_skill': 0.6109, 'defender_skill': 0.8221, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Noah Sadiki', 'market_value': 42000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8221}], 'xg_override': 0.065}	aerialWon: 0.3851 -> 1.1 totalShots: 0.9412 -> 0.8485 shotOffTarget: 0.7152 -> 0.6448 onTargetScoringAttempt: 0.226 -> 0.2037 bigChanceMissed: 0.1632 -> 0.1471 expectedGoals: 0.0721 -> 0.065 goals: 0.0669 -> 0.0603
`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 1.45 (squad xG sum 1.4149, xG scale 1.0277, xA scale 1.1324, team xA 0.95 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.454` `squad_xg_sum=1.4149` `squad_xa_sum=0.8346` `team_expected_assists=0.9451` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0651 -> 0.0737
`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=0.0646` `goal_rate=0.0603` `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=0.071` `assist_rate=0.0701` `provider_weight=0.0769`	-

Ethan Ampadu M 90 0 0.04 0.05 ok 1 0.6 7.0 59.6 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 1.0078, quality 0.739)	`market_value_eur=33000000` `quality_factor=0.7389` `boost_factor=1.0078` `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=0.9977` `prior_reliability=0.1333` `competition=Premier League`	wonTackle: 1.2072 -> 1.2004
`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': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.459, 'pressing_faced': 'HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 1.0133, 'skill_edge': -0.027, 'height_edge': 0.036, 'attacker_skill': 0.7622, 'defender_skill': 0.8221, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Noah Sadiki', 'market_value': 42000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8221}], 'xg_override': 0.0422}	aerialWon: 1.2611 -> 1.1 totalShots: 0.706 -> 0.7154 shotOffTarget: 0.5387 -> 0.5459
`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 1.45 (squad xG sum 1.4149, xG scale 1.0277, xA scale 1.1324, team xA 0.95 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.454` `squad_xg_sum=1.4149` `squad_xa_sum=0.8346` `team_expected_assists=0.9451` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0553 -> 0.0626
`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=0.0425` `goal_rate=0.0523` `scorer_weight=0.0482`	-
`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=0.0607` `assist_rate=0.0626` `provider_weight=0.0669`	-

Jayden Bogle M 77 0 0.17 0.08 103% 1 0.5 6.8 55.2 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 1.0040, quality 0.623)	`market_value_eur=19200000` `quality_factor=0.6231` `boost_factor=1.004` `reference_quality=0.505` `history_matches=39` `history_shrink=0.1136`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0084)	`blend_weight=0.0084` `history_matches=39` `season_overlap_discount=1.0` `per90_to_match_factor=0.8586` `prior_reliability=0.1133` `competition=Premier League`	wasFouled: 1.1738 -> 1.1851 totalShots: 1.1481 -> 1.1385
`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': "Luke O'Nien", 'matchup_zone': 'Right Wingback vs Left Back — Overlapping Run', 'duel_win_rate': 0.533, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.1956, 'skill_edge': 0.1171, 'height_edge': 0.016, 'attacker_skill': 0.6976, 'defender_skill': 0.4374, 'attacker_role': 'RWB', 'attacker_mv_price_index': 0.7, 'defender_adjusted_parts': [{'player': "Luke O'Nien", 'market_value': 660000, 'role': 'LB', 'price_index': 0.7, 'adjusted_quality': 0.4374}], 'xg_override': 0.1616}	totalShots: 1.1447 -> 1.3687 aerialWon: 0.8434 -> 1.0 shotOffTarget: 0.7426 -> 0.8879 onTargetScoringAttempt: 0.4021 -> 0.4808 expectedGoalsOnTarget: 0.226 -> 0.2702 expectedGoals: 0.1352 -> 0.1616 goals: 0.1205 -> 0.1441
`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 1.45 (squad xG sum 1.4149, xG scale 1.0277, xA scale 1.1324, team xA 0.95 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.454` `squad_xg_sum=1.4149` `squad_xa_sum=0.8346` `team_expected_assists=0.9451` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0872 -> 0.0987
`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: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.153` `goal_rate=0.1441` `scorer_weight=0.1524`	-
`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=0.094` `assist_rate=0.092` `provider_weight=0.1022`	-

Tarik Muharemović D 90 0 0.13 0.05 163% 1 0.5 7.0 35.8 10 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 1.0436, quality 0.796)	`market_value_eur=43000000` `quality_factor=0.7957` `boost_factor=1.0436` `reference_quality=0.505` `history_matches=5` `history_shrink=0.5`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0433)	`blend_weight=0.0433` `history_matches=5` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	interceptionWon: 1.2227 -> 1.2564 wasFouled: 1.0578 -> 1.0336 totalShots: 1.0447 -> 1.0211 onTargetScoringAttempt: 0.1469 -> 0.1622 totalCross: 0.1728 -> 0.187 wonTackle: 0.7442 -> 0.7336 goalAssist: 0.2392 -> 0.2288
`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': '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 1.45 (squad xG sum 1.4149, xG scale 1.0277, xA scale 1.1324, team xA 0.95 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.454` `squad_xg_sum=1.4149` `squad_xa_sum=0.8346` `team_expected_assists=0.9451` `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 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: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1209` `goal_rate=0.0` `scorer_weight=0.1031`	-
`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=0.0111` `assist_rate=0.2288` `provider_weight=0.014`	-

Nico Elvedi D 90 0 0.04 0.01 275% 1 0.2 6.8 27.5 10 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 1.0049, quality 0.534)	`market_value_eur=11000000` `quality_factor=0.5341` `boost_factor=1.0049` `reference_quality=0.505` `history_matches=4` `history_shrink=0.5556`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0144)	`blend_weight=0.0144` `history_matches=4` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.04` `competition=Premier League`	wasFouled: 0.0 -> 0.0481 wonTackle: 0.7687 -> 0.7576 totalShots: 0.6424 -> 0.6331 interceptionWon: 1.0194 -> 1.0287 onTargetScoringAttempt: 0.3906 -> 0.385
`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 1.45 (squad xG sum 1.4149, xG scale 1.0277, xA scale 1.1324, team xA 0.95 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.454` `squad_xg_sum=1.4149` `squad_xa_sum=0.8346` `team_expected_assists=0.9451` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.1662 -> 0.1884
`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: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0431` `goal_rate=0.0` `scorer_weight=0.0353`	-
`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=0.1717` `assist_rate=0.0` `provider_weight=0.1507`	-

James Justin D 89 0 0.06 0.06 ok 1 0.9 6.9 59.8 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 1.0035, quality 0.595)	`market_value_eur=16500000` `quality_factor=0.5953` `boost_factor=1.0035` `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=0.9932` `prior_reliability=0.1333` `competition=Premier League`	wonTackle: 1.3488 -> 1.3614
`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 1.45 (squad xG sum 1.4149, xG scale 1.0277, xA scale 1.1324, team xA 0.95 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.454` `squad_xg_sum=1.4149` `squad_xa_sum=0.8346` `team_expected_assists=0.9451` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0741 -> 0.0839
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 76.0)	`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=0.0628` `goal_rate=0.0443` `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=0.0805` `assist_rate=0.1366` `provider_weight=0.1049`	-

James Trafford G 90 0 0.00 - n/a 0.0 7.0 14.2 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 1.0284, quality 0.770)	`market_value_eur=38000000` `quality_factor=0.7699` `boost_factor=1.0284` `reference_quality=0.505` `history_matches=9` `history_shrink=0.3571`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0309)	`blend_weight=0.0309` `history_matches=9` `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': 'Brian Brobbey, Habib Diarra', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.0895}`	-
`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 1.45 (squad xG sum 1.4149, xG scale 1.0277, xA scale 1.1324, team xA 0.95 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.454` `squad_xg_sum=1.4149` `squad_xa_sum=0.8346` `team_expected_assists=0.9451` `team_xa_to_xg_ratio=0.65`	shots_faced: 10.4 -> 12.6 expected_saves: 2.55 -> 3.08 high_pressure_saves: 0.87 -> 1.06
`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=0.0` `scorer_weight=0.0` `squad_weight_total=1.4817`	-
`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=0.9401`	-

Team totals 1 1.45 16 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
Sean Longstaff	M	15.1m	70.3%	Market-value substitute prior
Ilia Gruev	M	22.4m	51.6%	Market-value substitute prior
Noah Okafor	F	25.7m	55.2%	Market-value substitute prior

Antagonist / tactical summary

Midfield is evenly contested — Noah Sadiki vs Ethan Ampadu will be the pivotal duel. Total antagonist xG — Home: 0.59, Away: 1.05. 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 0.50, key duel: Noah Sadiki vs Ethan Ampadu

Granit Xhaka xG 0.03Habib Diarra xG 0.13Brian Brobbey xG 0.43Dominic Calvert-Lewin xG 0.43Lukas Nmecha xG 0.16Gabriel Gudmundsson xG 0.04

Summary

Antagonist matchup analysis for Sunderland vs Leeds. Most likely scoreline: 1-1. Lambdas derived from 3 home and 7 away individual player duels. Head-to-head: Sunderland 3W-4D-3L vs Leeds 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 — Noah Sadiki vs Ethan Ampadu will be the pivotal duel. Total antagonist xG — Home: 0.59, Away: 1.05. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Shadow Striker vs CDM + CB — Zone 14 Battle.

Real Sociedad vs Levante

LaLiga · 20261025 · 4-4-2 vs 4-3-3

Predicted 2-1Winner: Real SociedadWinning H 2-1 (9.9%)Winning A 1-2 (5.3%)Confidence 0.76λ home/away 1.94 / 1.04Most-likely P 9.9%

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

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Real Sociedad vs Levante. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 4-4-2 vs 4-3-3 (duel lambdas: Home 2.009, Away 1.137). Final calibrated Poisson lambdas: Home 1.944, Away 1.040 (most likely score 2-1).

Home Starters 11/11 XGBoost (8 FBref) Away Starters 11/11 XGBoost (6 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 2.01 / 1.14 Final λ (H/A) 1.94 / 1.04

FBref Team Priors: Real Sociedad (Gls/90: 0.75, GA/90: 0.75 → E[λ]: 0.928) vs Levante UD (Gls/90: 1.6666666666666667, GA/90: 1.67 → E[λ]: 0.926) | 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 Sociedad	Levante
League position	5.00	17.00
Points per game	1.61	1.08
Attack index	1.15	0.99
Defense (concede) index	1.12	1.15
Form multiplier	1.03	0.97
Standing λ multiplier	1.12	1.05

Lambda decomposition (how the goal rates were built)

Stage	Real Sociedad λ	Levante λ
FBref Squad Prior E[λ] (weight 0.228)	0.928	0.926
Pre-mismatch	2.023	1.129
Pre-competitive-floor	2.207	1.101
Pre-calibration (Micro-duels)	2.009	1.137
Final λ	1.944	1.040

competition calibration scale 0.947competition 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
1-1	10.3%
2-1	9.9%
1-0	9.8%
2-0	9.6%
3-1	6.4%

Home win	58.4%
Draw	21.9%
Away win	19.8%
Over 2.5	57.3%
BTTS yes	55.5%

Pregame fair line — favorite Real Sociedad (58.4%), fair total 2.98, fair spread 0.90

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 Sociedad — 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

Jon Martin D 90 0 0.09 0.06 53% 0 0.4 6.8 53.0 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 1.0068, quality 0.669)	`market_value_eur=24000000` `quality_factor=0.6685` `boost_factor=1.0068` `reference_quality=0.505` `history_matches=31` `history_shrink=0.1389`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0241)	`blend_weight=0.0241` `history_matches=31` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.2667` `competition=La Liga`	wonTackle: 0.706 -> 0.7251 interceptionWon: 0.9525 -> 0.9356
`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': 'Víctor García', '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 1.94 (squad xG sum 1.5225, xG scale 1.2768, xA scale 1.5962, team xA 1.26 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.944` `squad_xg_sum=1.5225` `squad_xa_sum=0.7916` `team_expected_assists=1.2636` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0706 -> 0.0901 expected_assists: 0.0098 -> 0.0156
`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=0.0862` `goal_rate=0.0961` `scorer_weight=0.1091`	-
`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=0.0155` `assist_rate=0.0` `provider_weight=0.013`	-

Héctor Fort D 63 0 0.08 0.01 1486% 1 0.2 6.6 30.8 13 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 1.0052, quality 0.575)	`market_value_eur=14600000` `quality_factor=0.5748` `boost_factor=1.0052` `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': 'Roger Brugué', '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 1.94 (squad xG sum 1.5225, xG scale 1.2768, xA scale 1.5962, team xA 1.26 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.944` `squad_xg_sum=1.5225` `squad_xa_sum=0.7916` `team_expected_assists=1.2636` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.066 -> 0.0843 expected_assists: 0.0172 -> 0.0275
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 17 by projected rating, fifa_overall 72.0)	`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=0.0808` `goal_rate=0.3823` `scorer_weight=0.1054`	-
`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=0.0271` `assist_rate=0.2726` `provider_weight=0.0344`	-

Álex Remiro G 90 0 0.00 - n/a 0.0 7.1 63.6 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 1.0004, quality 0.518)	`market_value_eur=9600000` `quality_factor=0.5177` `boost_factor=1.0004` `reference_quality=0.505` `history_matches=43` `history_shrink=0.1042`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0181)	`blend_weight=0.0181` `history_matches=43` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.2667` `competition=La Liga`	saves: 2.9463 -> 2.9698
`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, Roger Brugué', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.1367}`	-
`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 1.94 (squad xG sum 1.5225, xG scale 1.2768, xA scale 1.5962, team xA 1.26 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.944` `squad_xg_sum=1.5225` `squad_xa_sum=0.7916` `team_expected_assists=1.2636` `team_xa_to_xg_ratio=0.65`	shots_faced: 10.8 -> 9.9 expected_saves: 2.64 -> 2.42 high_pressure_saves: 0.91 -> 0.83
`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=0.0` `scorer_weight=0.0` `squad_weight_total=2.1049`	-
`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=1.3089`	-

Mamadou Sarr D 83 0 0.16 0.03 524% 0 0.1 6.6 19.2 17 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 1.0245, quality 0.669)	`market_value_eur=24000000` `quality_factor=0.6685` `boost_factor=1.0245` `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': 'Roger Brugué', '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 1.94 (squad xG sum 1.5225, xG scale 1.2768, xA scale 1.5962, team xA 1.26 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.944` `squad_xg_sum=1.5225` `squad_xa_sum=0.7916` `team_expected_assists=1.2636` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1254 -> 0.1601 expected_assists: 0.0091 -> 0.0145
`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=0.1479` `goal_rate=0.0` `scorer_weight=0.1281`	-
`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=0.0144` `assist_rate=0.0` `provider_weight=0.0123`	-

Sergio Gómez D 85 0 0.02 0.02 ok 1 0.4 6.8 60.6 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 1.0022, quality 0.575)	`market_value_eur=14600000` `quality_factor=0.5748` `boost_factor=1.0022` `reference_quality=0.505` `history_matches=42` `history_shrink=0.1064`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0175)	`blend_weight=0.0175` `history_matches=42` `season_overlap_discount=1.0` `per90_to_match_factor=0.9426` `prior_reliability=0.2533` `competition=La Liga`	totalCross: 5.2217 -> 5.2302
`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': '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 1.94 (squad xG sum 1.5225, xG scale 1.2768, xA scale 1.5962, team xA 1.26 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.944` `squad_xg_sum=1.5225` `squad_xa_sum=0.7916` `team_expected_assists=1.2636` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0186 -> 0.0237 expected_assists: 0.1323 -> 0.2112
`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0234` `goal_rate=0.0` `scorer_weight=0.019`	-
`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=0.1904` `assist_rate=0.1606` `provider_weight=0.244`	-

Beñat Turrientes M 79 0 0.08 0.02 235% 1 0.6 6.9 54.0 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 1.0010, quality 0.529)	`market_value_eur=10600000` `quality_factor=0.5295` `boost_factor=1.001` `reference_quality=0.505` `history_matches=31` `history_shrink=0.1389`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0175)	`blend_weight=0.0175` `history_matches=31` `season_overlap_discount=1.0` `per90_to_match_factor=0.8728` `prior_reliability=0.1933` `competition=La Liga`	fouls: 0.7861 -> 0.7724 onTargetScoringAttempt: 0.3033 -> 0.314 wasFouled: 0.9006 -> 0.8901 interceptionWon: 0.5154 -> 0.5064
`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, Daniel Requena', 'matchup_zone': 'CM vs CDM + CM', 'duel_win_rate': 0.342, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.0611, 'skill_edge': 0.0476, 'height_edge': -0.006, 'attacker_skill': 0.5431, 'defender_skill': 0.4374, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Oriol Rey', 'market_value': 1700000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.4374}, {'player': 'Daniel Requena', 'market_value': 1500000, 'role': 'CM', 'price_index': 0.9, 'adjusted_quality': 0.4374}], 'xg_override': 0.0596}	aerialWon: 0.3334 -> 1.0 totalShots: 1.0552 -> 1.1197 shotOffTarget: 0.7412 -> 0.7865 onTargetScoringAttempt: 0.314 -> 0.3332
`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 1.94 (squad xG sum 1.5225, xG scale 1.2768, xA scale 1.5962, team xA 1.26 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.944` `squad_xg_sum=1.5225` `squad_xa_sum=0.7916` `team_expected_assists=1.2636` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0596 -> 0.0761 expected_assists: 0.0559 -> 0.0892
`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_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=0.0733` `goal_rate=0.0624` `scorer_weight=0.078`	-
`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=0.0853` `assist_rate=0.0562` `provider_weight=0.0894`	-

Orri Óskarsson F 78 1 0.53 0.29 84% 3 1.2 6.8 55.6 27 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 1.0066, 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`	-
`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': 0.264, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.094, 'skill_edge': 0.088, 'height_edge': -0.024, 'attacker_skill': 0.6408, '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': 0.4168}	totalShots: 2.3207 -> 2.5388 shotOffTarget: 1.3108 -> 1.434 onTargetScoringAttempt: 1.0099 -> 1.1048 bigChanceMissed: 0.9177 -> 1.004 expectedGoalsOnTarget: 0.7178 -> 0.7853 goals: 0.473 -> 0.5175 expectedGoals: 0.381 -> 0.4168
`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 1.94 (squad xG sum 1.5225, xG scale 1.2768, xA scale 1.5962, team xA 1.26 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.944` `squad_xg_sum=1.5225` `squad_xa_sum=0.7916` `team_expected_assists=1.2636` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.4168 -> 0.5323 expected_assists: 0.0351 -> 0.056
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 10 by projected rating, fifa_overall 74.0)	`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=1` `goal_probability=0.4127` `goal_rate=0.5175` `scorer_weight=0.6471`	-
`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=0.0545` `assist_rate=0.0` `provider_weight=0.0448`	-

Mikel Oyarzabal F 79 1 0.30 0.32 ok 2 1.5 6.9 63.1 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 1.0062, quality 0.694)	`market_value_eur=27000000` `quality_factor=0.6941` `boost_factor=1.0062` `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=0.8763` `prior_reliability=0.2` `competition=La Liga`	wasFouled: 1.1754 -> 1.1877 wonTackle: 0.8096 -> 0.7982 onTargetScoringAttempt: 0.6804 -> 0.6874 goalAssist: 0.1648 -> 0.1707
`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': 0.268, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.0999, 'skill_edge': 0.112, 'height_edge': -0.044, 'attacker_skill': 0.6941, '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': 0.2332}	totalShots: 1.9195 -> 2.1113 shotOffTarget: 1.2321 -> 1.3552 onTargetScoringAttempt: 0.6874 -> 0.7561 bigChanceMissed: 0.6713 -> 0.7384 expectedGoalsOnTarget: 0.2492 -> 0.2741 expectedGoals: 0.212 -> 0.2332 goals: 0.2112 -> 0.2323
`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 1.94 (squad xG sum 1.5225, xG scale 1.2768, xA scale 1.5962, team xA 1.26 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.944` `squad_xg_sum=1.5225` `squad_xa_sum=0.7916` `team_expected_assists=1.2636` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.2332 -> 0.2978 expected_assists: 0.112 -> 0.1788
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 90.0)	`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.2576` `goal_rate=0.2323` `scorer_weight=0.2968`	-
`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=0.1637` `assist_rate=0.1707` `provider_weight=0.2235`	-

Ander Barrenetxea F 67 0 0.21 0.06 224% 2 0.9 6.9 56.3 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 1.0036, quality 0.603)	`market_value_eur=17200000` `quality_factor=0.6027` `boost_factor=1.0036` `reference_quality=0.505` `history_matches=36` `history_shrink=0.122`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0127)	`blend_weight=0.0127` `history_matches=36` `season_overlap_discount=1.0` `per90_to_match_factor=0.7428` `prior_reliability=0.16` `competition=La Liga`	interceptionWon: 0.4537 -> 0.4479 totalCross: 2.2798 -> 2.2744
`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': 0.211, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.0159, 'skill_edge': 0.0708, 'height_edge': -0.06, 'attacker_skill': 0.6027, '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': 0.1612}	totalShots: 1.6295 -> 1.6554 shotOffTarget: 1.0875 -> 1.1048 onTargetScoringAttempt: 0.542 -> 0.5506
`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 1.94 (squad xG sum 1.5225, xG scale 1.2768, xA scale 1.5962, team xA 1.26 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.944` `squad_xg_sum=1.5225` `squad_xa_sum=0.7916` `team_expected_assists=1.2636` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1612 -> 0.2058 expected_assists: 0.0951 -> 0.1518
`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_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=0.186` `goal_rate=0.2224` `scorer_weight=0.2573`	-
`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=0.1408` `assist_rate=0.1071` `provider_weight=0.1644`	-

Carlos Soler M 86 0 0.15 0.09 60% 1 0.9 6.9 63.0 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 0.9992, quality 0.482)	`market_value_eur=6600000` `quality_factor=0.4818` `boost_factor=0.9992` `reference_quality=0.505` `history_matches=40` `history_shrink=0.1111`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0111)	`blend_weight=0.0111` `history_matches=40` `season_overlap_discount=1.0` `per90_to_match_factor=0.9526` `prior_reliability=0.1533` `competition=La Liga`	wonTackle: 0.3931 -> 0.4025 totalShots: 1.1542 -> 1.146 fouls: 0.6296 -> 0.6226
`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': 'Jeremy Toljan', 'matchup_zone': 'Left Mid vs Right Back — Wide Channel', 'duel_win_rate': 0.476, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.0234, 'skill_edge': 0.024, 'height_edge': -0.008, 'attacker_skill': 0.4907, 'defender_skill': 0.4374, 'attacker_role': 'LM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Jeremy Toljan', 'market_value': 1900000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.4374}], 'xg_override': 0.1157}	aerialWon: 0.1694 -> 1.1 totalShots: 1.1502 -> 1.1771 shotOffTarget: 0.7956 -> 0.8142 onTargetScoringAttempt: 0.3546 -> 0.3629
`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 1.94 (squad xG sum 1.5225, xG scale 1.2768, xA scale 1.5962, team xA 1.26 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.944` `squad_xg_sum=1.5225` `squad_xa_sum=0.7916` `team_expected_assists=1.2636` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1157 -> 0.1477 expected_assists: 0.1401 -> 0.2237
`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=0.1373` `goal_rate=0.1106` `scorer_weight=0.1432`	-
`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=0.2004` `assist_rate=0.1133` `provider_weight=0.1922`	-

Takefusa Kubo F 70 0 0.18 0.05 227% 1 0.4 6.8 44.7 26 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 0.641)	`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 0.0147)	`blend_weight=0.0147` `history_matches=29` `season_overlap_discount=1.0` `per90_to_match_factor=0.7751` `prior_reliability=0.1533` `competition=La Liga`	wasFouled: 2.2211 -> 2.2132 totalShots: 1.1417 -> 1.1351 onTargetScoringAttempt: 0.4022 -> 0.3963
`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': 0.228, 'pressing_faced': 'LOW', 'numerical_situation': '1v2', 'xg_multiplier': 1.0411, 'skill_edge': 0.088, 'height_edge': -0.06, 'attacker_skill': 0.6408, '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': 0.1398}	totalShots: 1.1357 -> 1.1824 shotOffTarget: 0.7394 -> 0.7698 onTargetScoringAttempt: 0.3963 -> 0.4126 bigChanceMissed: 0.1989 -> 0.2071 goals: 0.1373 -> 0.1429 expectedGoals: 0.1343 -> 0.1398
`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 1.94 (squad xG sum 1.5225, xG scale 1.2768, xA scale 1.5962, team xA 1.26 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.944` `squad_xg_sum=1.5225` `squad_xa_sum=0.7916` `team_expected_assists=1.2636` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1398 -> 0.1785 expected_assists: 0.0984 -> 0.1571
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 80.0)	`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=0.1635` `goal_rate=0.1429` `scorer_weight=0.1814`	-
`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=0.1454` `assist_rate=0.0967` `provider_weight=0.1553`	-

Team totals 2 1.94 14 6.8

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Unai Marrero	G	4.4m	10.9%	Market-value substitute prior
Jon Aramburu	D	14.1m	22.7%	Market-value substitute prior
Igor Zubeldia	D	24.9m	31.6%	Market-value substitute prior
Jon Gorrotxategi	M	19.1m	71.3%	Market-value substitute prior
Luka Sučić	M	26.7m	41.6%	Market-value substitute prior
Gonçalo Guedes	F	20.8m	72.2%	Market-value substitute prior

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

Roger Brugué F 82 0 0.25 0.36 30% 2 1.9 6.9 49.8 18 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 0.9965, quality 0.437)	`market_value_eur=2800000` `quality_factor=0.4374` `boost_factor=0.9965` `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': 'Héctor Fort, Mamadou Sarr', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 0.7521, 'skill_edge': -0.1087, 'height_edge': -0.06, 'attacker_skill': 0.4374, 'defender_skill': 0.6789, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Héctor Fort', 'market_value': 14600000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.6258}, {'player': 'Mamadou Sarr', 'market_value': 24000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.732}], 'xg_override': 0.2675}	totalShots: 2.1875 -> 1.6452 shotOffTarget: 1.2253 -> 0.9215 onTargetScoringAttempt: 0.9622 -> 0.7237 bigChanceMissed: 0.3782 -> 0.2844 goals: 0.3687 -> 0.2773 expectedGoals: 0.3557 -> 0.2675 expectedGoalsOnTarget: 0.3369 -> 0.2534
`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 1.04 (squad xG sum 1.0996, xG scale 0.9458, xA scale 1.1262, team xA 0.68 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.04` `squad_xg_sum=1.0996` `squad_xa_sum=0.6002` `team_expected_assists=0.676` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.2675 -> 0.253 expected_assists: 0.0599 -> 0.0675
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 88.0)	`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=0.2235` `goal_rate=0.2773` `scorer_weight=0.2608`	-
`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=0.0653` `assist_rate=0.0323` `provider_weight=0.054`	-

Iván Romero F 73 1 0.31 0.26 ok 2 1.8 7.0 61.3 28 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 0.9995, quality 0.491)	`market_value_eur=7400000` `quality_factor=0.4915` `boost_factor=0.9995` `reference_quality=0.505` `history_matches=38` `history_shrink=0.1163`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0126)	`blend_weight=0.0126` `history_matches=38` `season_overlap_discount=1.0` `per90_to_match_factor=0.8069` `prior_reliability=0.1667` `competition=La Liga`	totalShots: 2.3124 -> 2.3202 wasFouled: 1.8472 -> 1.8402
`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 Gómez', 'matchup_zone': 'Left Wing vs Right Back — Flank Battle', 'duel_win_rate': 0.417, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.908, 'skill_edge': -0.0666, 'height_edge': 0.004, 'attacker_skill': 0.4915, 'defender_skill': 0.6394, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Sergio Gómez', 'market_value': 14600000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.6394}], 'xg_override': 0.3228}	totalShots: 2.3202 -> 2.1067 shotOffTarget: 1.4011 -> 1.2722 onTargetScoringAttempt: 0.9081 -> 0.8246 bigChanceMissed: 0.5186 -> 0.4709 expectedGoals: 0.3555 -> 0.3228 goals: 0.3511 -> 0.3188 expectedGoalsOnTarget: 0.218 -> 0.1979
`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 1.04 (squad xG sum 1.0996, xG scale 0.9458, xA scale 1.1262, team xA 0.68 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.04` `squad_xg_sum=1.0996` `squad_xa_sum=0.6002` `team_expected_assists=0.676` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.3228 -> 0.3054 expected_assists: 0.0585 -> 0.0659
`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=1` `goal_probability=0.2632` `goal_rate=0.3188` `scorer_weight=0.3021`	-
`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=0.0638` `assist_rate=0.0057` `provider_weight=0.0527`	-

Víctor García F 75 0 0.10 0.07 40% 1 0.6 6.7 48.2 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 0.9967, quality 0.437)	`market_value_eur=1500000` `quality_factor=0.4374` `boost_factor=0.9967` `reference_quality=0.505` `history_matches=26` `history_shrink=0.1613`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0084)	`blend_weight=0.0084` `history_matches=26` `season_overlap_discount=1.0` `per90_to_match_factor=0.8342` `prior_reliability=0.08` `competition=La Liga`	totalShots: 1.25 -> 1.2395 totalCross: 4.415 -> 4.4071
`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': 'Jon Martin', 'matchup_zone': 'Right Wing vs Left Back — Flank Battle', 'duel_win_rate': 0.309, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v1', 'xg_multiplier': 0.7479, 'skill_edge': -0.1395, 'height_edge': -0.032, 'attacker_skill': 0.4374, 'defender_skill': 0.7474, 'attacker_role': 'RW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Jon Martin', 'market_value': 24000000, 'role': 'LB', 'price_index': 0.7, 'adjusted_quality': 0.7474}], 'xg_override': 0.1072}	totalShots: 1.2462 -> 0.9321 shotOffTarget: 0.8078 -> 0.6042 onTargetScoringAttempt: 0.4384 -> 0.3279 expectedGoalsOnTarget: 0.2811 -> 0.2102 bigChanceMissed: 0.2739 -> 0.2048 expectedGoals: 0.1433 -> 0.1072 goals: 0.1295 -> 0.0969
`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 1.04 (squad xG sum 1.0996, xG scale 0.9458, xA scale 1.1262, team xA 0.68 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.04` `squad_xg_sum=1.0996` `squad_xa_sum=0.6002` `team_expected_assists=0.676` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1072 -> 0.1014 expected_assists: 0.0813 -> 0.0916
`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: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0964` `goal_rate=0.0969` `scorer_weight=0.0948`	-
`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=0.0875` `assist_rate=0.0` `provider_weight=0.0733`	-

Daniel Requena M 87 0 0.05 0.01 469% 1 0.2 6.7 32.6 8 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 0.9873, quality 0.437)	`market_value_eur=1500000` `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=true`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Beñat Turrientes', 'matchup_zone': 'Central Mid vs Central Mid — Box-to-Box Midfield Contest', 'duel_win_rate': 0.4, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9713, 'skill_edge': -0.0476, 'height_edge': 0.028, 'attacker_skill': 0.4374, 'defender_skill': 0.5431, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Beñat Turrientes', 'market_value': 10600000, 'role': 'CM', 'price_index': 0.9, 'adjusted_quality': 0.5431}], 'xg_override': 0.0487}	aerialWon: 0.0 -> 1.2 totalShots: 0.693 -> 0.6731 shotOffTarget: 0.506 -> 0.4915 onTargetScoringAttempt: 0.187 -> 0.1816
`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 1.04 (squad xG sum 1.0996, xG scale 0.9458, xA scale 1.1262, team xA 0.68 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.04` `squad_xg_sum=1.0996` `squad_xa_sum=0.6002` `team_expected_assists=0.676` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 14 by projected rating, fifa_overall 84.0)	`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=0.0451` `goal_rate=0.0481` `scorer_weight=0.0458`	-
`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=0.0341` `assist_rate=0.0296` `provider_weight=0.0342`	-

Oriol Rey M 86 0 0.03 0.01 366% 1 0.2 6.8 34.8 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 0.9967, quality 0.437)	`market_value_eur=1700000` `quality_factor=0.4374` `boost_factor=0.9967` `reference_quality=0.505` `history_matches=26` `history_shrink=0.1613`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0210)	`blend_weight=0.021` `history_matches=26` `season_overlap_discount=1.0` `per90_to_match_factor=0.9543` `prior_reliability=0.2` `competition=La Liga`	totalShots: 0.6064 -> 0.5937 wasFouled: 0.7716 -> 0.7621 totalCross: 0.2453 -> 0.2402
`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': 'Beñat Turrientes', 'matchup_zone': 'Holding Mid vs Central Mid — Midfield Anchor Duel', 'duel_win_rate': 0.336, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9066, 'skill_edge': -0.0476, 'height_edge': -0.016, 'attacker_skill': 0.4374, 'defender_skill': 0.5431, 'attacker_role': 'CDM', 'attacker_mv_price_index': 0.85, 'defender_adjusted_parts': [{'player': 'Beñat Turrientes', 'market_value': 10600000, 'role': 'CM', 'price_index': 0.9, 'adjusted_quality': 0.5431}], 'xg_override': 0.0352}	aerialWon: 1.1859 -> 0.9 totalShots: 0.6032 -> 0.5469 shotOffTarget: 0.4547 -> 0.4122 onTargetScoringAttempt: 0.1485 -> 0.1346
`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 1.04 (squad xG sum 1.0996, xG scale 0.9458, xA scale 1.1262, team xA 0.68 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.04` `squad_xg_sum=1.0996` `squad_xa_sum=0.6002` `team_expected_assists=0.676` `team_xa_to_xg_ratio=0.65`	-
`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=0.0328` `goal_rate=0.0348` `scorer_weight=0.0331`	-
`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=0.0421` `assist_rate=0.0479` `provider_weight=0.0477`	-

Jon Ander Olasagasti M 84 0 0.11 0.10 ok 1 1.0 6.9 64.0 29 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 0.9977, quality 0.437)	`market_value_eur=2700000` `quality_factor=0.4374` `boost_factor=0.9977` `reference_quality=0.505` `history_matches=39` `history_shrink=0.1136`	-
`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': 'Beñat Turrientes', 'matchup_zone': 'Central Mid vs Central Mid — Box-to-Box Midfield Contest', 'duel_win_rate': 0.36, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9125, 'skill_edge': -0.0476, 'height_edge': -0.012, 'attacker_skill': 0.4374, 'defender_skill': 0.5431, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Beñat Turrientes', 'market_value': 10600000, 'role': 'CM', 'price_index': 0.9, 'adjusted_quality': 0.5431}], 'xg_override': 0.1172}	aerialWon: 0.7264 -> 0.9 totalShots: 1.2602 -> 1.1499 shotOffTarget: 0.794 -> 0.7245 onTargetScoringAttempt: 0.4662 -> 0.4254 expectedGoalsOnTarget: 0.1944 -> 0.1774 goals: 0.1533 -> 0.1399 expectedGoals: 0.1284 -> 0.1172 bigChanceMissed: 0.0953 -> 0.087
`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 1.04 (squad xG sum 1.0996, xG scale 0.9458, xA scale 1.1262, team xA 0.68 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.04` `squad_xg_sum=1.0996` `squad_xa_sum=0.6002` `team_expected_assists=0.676` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1172 -> 0.1108 expected_assists: 0.13 -> 0.1463
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 84.0)	`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=0.1049` `goal_rate=0.1399` `scorer_weight=0.1258`	-
`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=0.1361` `assist_rate=0.1509` `provider_weight=0.1633`	-

Manuel Sánchez D 88 0 0.00 0.01 41% 0 0.2 6.7 54.9 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 0.9991, quality 0.479)	`market_value_eur=6400000` `quality_factor=0.4793` `boost_factor=0.9991` `reference_quality=0.505` `history_matches=39` `history_shrink=0.1136`	-
`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': 'Takefusa Kubo', '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 1.04 (squad xG sum 1.0996, xG scale 0.9458, xA scale 1.1262, team xA 0.68 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.04` `squad_xg_sum=1.0996` `squad_xa_sum=0.6002` `team_expected_assists=0.676` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0594 -> 0.0669
`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_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=0.0047` `goal_rate=0.0028` `scorer_weight=0.0045`	-
`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=0.0263` `provider_weight=0.0535`	-

Aïssa Mandi D 85 0 0.00 0.01 ok 0 0.1 6.8 51.9 8 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 0.9908, quality 0.437)	`market_value_eur=1100000` `quality_factor=0.4374` `boost_factor=0.9908` `reference_quality=0.505` `history_matches=6` `history_shrink=0.4545`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0591)	`blend_weight=0.0591` `history_matches=6` `season_overlap_discount=1.0` `per90_to_match_factor=0.9458` `prior_reliability=0.2` `competition=La Liga`	wonTackle: 0.756 -> 0.8045 fouls: 0.4414 -> 0.4153 interceptionWon: 1.0925 -> 1.0838 wasFouled: 0.7703 -> 0.762
`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': 'Orri Óskarsson', '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 1.04 (squad xG sum 1.0996, xG scale 0.9458, xA scale 1.1262, team xA 0.68 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.04` `squad_xg_sum=1.0996` `squad_xa_sum=0.6002` `team_expected_assists=0.676` `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 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=0.0047` `goal_rate=0.0` `scorer_weight=0.0043`	-
`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=0.0082` `assist_rate=0.0` `provider_weight=0.0072`	-

Ifeanyi Ndukwe D 90 0 0.00 - n/a 0 0.0 6.7 18.4 7 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 0.9861, quality 0.440)	`market_value_eur=3200000` `quality_factor=0.4399` `boost_factor=0.9861` `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': 'Orri Óskarsson', '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 1.04 (squad xG sum 1.0996, xG scale 0.9458, xA scale 1.1262, team xA 0.68 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.04` `squad_xg_sum=1.0996` `squad_xa_sum=0.6002` `team_expected_assists=0.676` `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 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=0.0` `goal_rate=0.0` `scorer_weight=0.001`	-
`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=0.0007` `assist_rate=0.0` `provider_weight=0.001`	-

Jeremy Toljan D 84 0 0.03 0.01 124% 0 0.2 6.6 61.0 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 0.9979, quality 0.437)	`market_value_eur=1900000` `quality_factor=0.4374` `boost_factor=0.9979` `reference_quality=0.505` `history_matches=43` `history_shrink=0.1042`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0036)	`blend_weight=0.0036` `history_matches=43` `season_overlap_discount=1.0` `per90_to_match_factor=0.9328` `prior_reliability=0.0533` `competition=La Liga`	totalCross: 2.4008 -> 2.409
`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 Soler', 'matchup_zone': 'Left Mid vs Right Back — Wide Channel'}`	-
`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 1.04 (squad xG sum 1.0996, xG scale 0.9458, xA scale 1.1262, team xA 0.68 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.04` `squad_xg_sum=1.0996` `squad_xa_sum=0.6002` `team_expected_assists=0.676` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0726 -> 0.0818
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 18 by projected rating, fifa_overall 74.0)	`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=0.0271` `goal_rate=0.0003` `scorer_weight=0.022`	-
`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=0.0785` `assist_rate=0.2351` `provider_weight=0.1022`	-

Mathew Ryan G 90 0 0.00 0.01 100% 0.1 7.0 67.9 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 0.9978, quality 0.437)	`market_value_eur=1700000` `quality_factor=0.4374` `boost_factor=0.9978` `reference_quality=0.505` `history_matches=42` `history_shrink=0.1064`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0138)	`blend_weight=0.0138` `history_matches=42` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.2` `competition=La Liga`	saves: 3.0151 -> 3.0379
`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': 'Orri Óskarsson, Mikel Oyarzabal', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 2.0092}`	-
`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 1.04 (squad xG sum 1.0996, xG scale 0.9458, xA scale 1.1262, team xA 0.68 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.04` `squad_xg_sum=1.0996` `squad_xa_sum=0.6002` `team_expected_assists=0.676` `team_xa_to_xg_ratio=0.65`	shots_faced: 19.1 -> 18.5 expected_saves: 4.68 -> 4.53 high_pressure_saves: 1.61 -> 1.56
`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=1.0647`	-
`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=0.6667`	-

Team totals 1 1.04 11 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

Antagonist / tactical summary

Midfield is evenly contested — Beñat Turrientes vs Jon Ander Olasagasti will be the pivotal duel. Total antagonist xG — Home: 1.13, Away: 0.90. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Left Wing vs Right Back — Flank Battle.

Midfield — possession 51.0% / 49.0%, dominance ratio 0.52, key duel: Beñat Turrientes vs Jon Ander Olasagasti

Orri Óskarsson xG 0.42Mikel Oyarzabal xG 0.23Ander Barrenetxea xG 0.16Iván Romero xG 0.32Víctor García xG 0.11Roger Brugué xG 0.27

Summary

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

Tactical parity between 4-4-2 and 4-3-3. Match outcome driven primarily by individual player quality. Midfield is evenly contested — Beñat Turrientes vs Jon Ander Olasagasti will be the pivotal duel. Total antagonist xG — Home: 1.13, Away: 0.90. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Left Wing vs Right Back — Flank Battle.

Crystal Palace vs Newcastle

Premier League · 20261025 · 3-4-2-1 vs 4-2-3-1

Predicted 1-2Winner: NewcastleWinning H 2-1 (8.1%)Winning A 1-2 (8.5%)Confidence 0.7λ home/away 1.60 / 1.69Most-likely P 8.5%

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

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Crystal Palace vs Newcastle. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 3-4-2-1 vs 4-2-3-1 (duel lambdas: Home 1.364, Away 1.619). Final calibrated Poisson lambdas: Home 1.602, Away 1.686 (most likely score 1-2).

Home Starters 11/11 XGBoost (9 FBref) Away Starters 11/11 XGBoost (8 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 1.36 / 1.62 Final λ (H/A) 1.60 / 1.69

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	Crystal Palace	Newcastle
League position	14.00	11.00
Points per game	1.16	1.24
Attack index	1.01	0.94
Defense (concede) index	1.08	1.32
Form multiplier	0.99	1.00
Standing λ multiplier	1.04	1.09

Lambda decomposition (how the goal rates were built)

Stage	Crystal Palace λ	Newcastle λ
Pre-competitive-floor	1.602	1.686
Pre-calibration (Micro-duels)	1.363	1.619
Final λ	1.602	1.686

Scoreline distribution & markets

Top 5 scorelines	P
1-1	10.2%
1-2	8.5%
2-1	8.1%
2-2	6.8%
0-1	6.2%

Home win	36.6%
Draw	23.2%
Away win	40.2%
Over 2.5	63.8%
BTTS yes	65.1%

Pregame fair line — favorite Newcastle (40.2%), fair total 3.29, fair spread -0.08

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.

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

Tyrick Mitchell M 85 0 0.09 0.06 42% 1 0.6 6.8 57.2 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 1.0048, quality 0.660)	`market_value_eur=23000000` `quality_factor=0.6595` `boost_factor=1.0048` `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=0.9399` `prior_reliability=0.1333` `competition=Premier League`	wonTackle: 1.7126 -> 1.7056
`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': 'Lukáš Horníček', 'matchup_zone': 'Left Wingback vs Right Back — Overlapping Run', 'duel_win_rate': 0.329, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v1', 'xg_multiplier': 0.8664, 'skill_edge': -0.0309, 'height_edge': -0.06, 'attacker_skill': 0.7379, 'defender_skill': 0.8066, 'attacker_role': 'LWB', 'attacker_mv_price_index': 0.7, 'defender_adjusted_parts': [{'player': 'Lukáš Horníček', 'market_value': 25000000, 'role': 'GK', 'price_index': 0.55, 'adjusted_quality': 0.8066}], 'xg_override': 0.0773}	totalShots: 1.1183 -> 0.9689 shotOffTarget: 0.8342 -> 0.7228 aerialWon: 0.8185 -> 0.9 onTargetScoringAttempt: 0.2831 -> 0.2453 bigChanceMissed: 0.1668 -> 0.1445 expectedGoalsOnTarget: 0.1424 -> 0.1234 goals: 0.1006 -> 0.0872 expectedGoals: 0.0892 -> 0.0773
`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 1.60 (squad xG sum 1.3541, xG scale 1.1831, xA scale 1.3335, team xA 1.04 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.602` `squad_xg_sum=1.3541` `squad_xa_sum=0.7809` `team_expected_assists=1.0413` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0773 -> 0.0915 expected_assists: 0.0899 -> 0.1199
`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: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0874` `goal_rate=0.0872` `scorer_weight=0.0986`	-
`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=0.113` `assist_rate=0.0818` `provider_weight=0.113`	-

Daichi Kamada M 83 0 0.07 0.06 ok 1 0.5 6.8 55.4 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 1.0008, quality 0.526)	`market_value_eur=10300000` `quality_factor=0.526` `boost_factor=1.0008` `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=0.9228` `prior_reliability=0.1333` `competition=Premier League`	wonTackle: 1.5315 -> 1.5509 totalCross: 1.1368 -> 1.1449 interceptionWon: 0.4584 -> 0.4637
`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': 'Lewis Hall', 'matchup_zone': 'Right Wingback vs Left Back — Overlapping Run', 'duel_win_rate': 0.285, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.8309, 'skill_edge': -0.1351, 'height_edge': 0.02, 'attacker_skill': 0.5761, 'defender_skill': 0.8762, 'attacker_role': 'RWB', 'attacker_mv_price_index': 0.7, 'defender_adjusted_parts': [{'player': 'Lewis Hall', 'market_value': 52000000, 'role': 'LB', 'price_index': 0.7, 'adjusted_quality': 0.8762}], 'xg_override': 0.0586}	aerialWon: 0.3158 -> 1.1 totalShots: 0.8775 -> 0.7291 shotOffTarget: 0.5699 -> 0.4735 onTargetScoringAttempt: 0.3076 -> 0.2556 bigChanceMissed: 0.1789 -> 0.1486 goals: 0.0813 -> 0.0676 expectedGoals: 0.0705 -> 0.0586
`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 1.60 (squad xG sum 1.3541, xG scale 1.1831, xA scale 1.3335, team xA 1.04 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.602` `squad_xg_sum=1.3541` `squad_xa_sum=0.7809` `team_expected_assists=1.0413` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0586 -> 0.0693 expected_assists: 0.0947 -> 0.1263
`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=0.067` `goal_rate=0.0676` `scorer_weight=0.075`	-
`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=0.1186` `assist_rate=0.0917` `provider_weight=0.1237`	-

Adam Wharton M 86 0 0.09 0.03 204% 1 0.7 7.0 56.7 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 1.0128, quality 0.881)	`market_value_eur=79000000` `quality_factor=0.8811` `boost_factor=1.0128` `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=0.9597` `prior_reliability=0.1333` `competition=Premier League`	totalCross: 2.2635 -> 2.2459 totalShots: 1.1184 -> 1.1074 fouls: 0.9038 -> 0.8949 interceptionWon: 1.0643 -> 1.0727
`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': 'Nico González', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.447, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.9949, 'skill_edge': 0.0125, 'height_edge': -0.016, 'attacker_skill': 0.8877, 'defender_skill': 0.8599, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Nico González', 'market_value': 54000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8599}], 'xg_override': 0.0736}	aerialWon: 0.4829 -> 1.1 totalShots: 1.1154 -> 1.1097
`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 1.60 (squad xG sum 1.3541, xG scale 1.1831, xA scale 1.3335, team xA 1.04 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.602` `squad_xg_sum=1.3541` `squad_xa_sum=0.7809` `team_expected_assists=1.0413` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0736 -> 0.0871 expected_assists: 0.1485 -> 0.1979
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 3 by projected rating, fifa_overall 82.0)	`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=0.0834` `goal_rate=0.0739` `scorer_weight=0.0873`	-
`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=0.1795` `assist_rate=0.0974` `provider_weight=0.1583`	-

Anan Khalaili M 87 0 0.11 0.09 ok 1 0.6 6.9 39.8 11 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 1.0245, quality 0.669)	`market_value_eur=24000000` `quality_factor=0.6685` `boost_factor=1.0245` `reference_quality=0.505` `history_matches=5` `history_shrink=0.5`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0260)	`blend_weight=0.026` `history_matches=5` `season_overlap_discount=1.0` `per90_to_match_factor=0.9638` `prior_reliability=0.08` `competition=Premier League`	wonTackle: 2.1681 -> 2.2997 totalCross: 2.4018 -> 2.4855 interceptionWon: 1.1809 -> 1.2128 fouls: 0.7922 -> 0.8134 wasFouled: 0.1974 -> 0.2131 onTargetScoringAttempt: 0.3952 -> 0.3849 totalShots: 1.171 -> 1.1611
`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': 'Nico González', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.362, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 0.8709, 'skill_edge': -0.0758, 'height_edge': -0.012, 'attacker_skill': 0.6914, 'defender_skill': 0.8599, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Nico González', 'market_value': 54000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8599}], 'xg_override': 0.0913}	totalShots: 1.1611 -> 1.0112 shotOffTarget: 0.7758 -> 0.6756 aerialWon: 1.1675 -> 1.1 bigChanceMissed: 0.5056 -> 0.4403 onTargetScoringAttempt: 0.3849 -> 0.3352 goals: 0.1454 -> 0.1266 expectedGoals: 0.1048 -> 0.0913 expectedGoalsOnTarget: 0.0825 -> 0.0718
`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 1.60 (squad xG sum 1.3541, xG scale 1.1831, xA scale 1.3335, team xA 1.04 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.602` `squad_xg_sum=1.3541` `squad_xa_sum=0.7809` `team_expected_assists=1.0413` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0913 -> 0.108 expected_assists: 0.1038 -> 0.1384
`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=0.1024` `goal_rate=0.1266` `scorer_weight=0.135`	-
`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=0.1292` `assist_rate=0.0748` `provider_weight=0.1123`	-

Axel Disasi D 82 0 0.03 0.05 34% 0 0.2 6.7 51.8 22 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 1.0054, quality 0.585)	`market_value_eur=15500000` `quality_factor=0.5846` `boost_factor=1.0054` `reference_quality=0.505` `history_matches=17` `history_shrink=0.2273`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0020)	`blend_weight=0.002` `history_matches=17` `season_overlap_discount=1.0` `per90_to_match_factor=0.9062` `prior_reliability=0.0133` `competition=Premier League`	fouls: 0.4774 -> 0.4854
`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': 'Joelinton', '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 1.60 (squad xG sum 1.3541, xG scale 1.1831, xA scale 1.3335, team xA 1.04 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.602` `squad_xg_sum=1.3541` `squad_xa_sum=0.7809` `team_expected_assists=1.0413` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 14 by projected rating, fifa_overall 77.0)	`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=0.0299` `goal_rate=0.0` `scorer_weight=0.0243`	-
`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=0.0126` `assist_rate=0.0` `provider_weight=0.0107`	-

Chris Richards D 90 0 0.10 0.07 34% 1 0.4 6.8 52.9 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 1.0060, quality 0.677)	`market_value_eur=25000000` `quality_factor=0.6773` `boost_factor=1.006` `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=0.9998` `prior_reliability=0.1333` `competition=Premier League`	interceptionWon: 1.0899 -> 1.0991 wonTackle: 0.7908 -> 0.7979 totalCross: 0.5049 -> 0.4998
`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': 'William Osula', '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 1.60 (squad xG sum 1.3541, xG scale 1.1831, xA scale 1.3335, team xA 1.04 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.602` `squad_xg_sum=1.3541` `squad_xa_sum=0.7809` `team_expected_assists=1.0413` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0849 -> 0.1004 expected_assists: 0.0239 -> 0.0319
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 80.0)	`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=0.0955` `goal_rate=0.0123` `scorer_weight=0.0803`	-
`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=0.0314` `assist_rate=0.0` `provider_weight=0.0255`	-

Jaydee Canvot D 89 0 0.06 0.05 ok 0 0.2 6.8 45.8 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 1.0117, quality 0.732)	`market_value_eur=32000000` `quality_factor=0.732` `boost_factor=1.0117` `reference_quality=0.505` `history_matches=24` `history_shrink=0.1724`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0149)	`blend_weight=0.0149` `history_matches=24` `season_overlap_discount=1.0` `per90_to_match_factor=0.9939` `prior_reliability=0.1333` `competition=Premier League`	wonTackle: 1.101 -> 1.1142 interceptionWon: 1.128 -> 1.1186 fouls: 0.5775 -> 0.5689 totalCross: 0.3616 -> 0.3562
`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': 'William Osula', '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 1.60 (squad xG sum 1.3541, xG scale 1.1831, xA scale 1.3335, team xA 1.04 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.602` `squad_xg_sum=1.3541` `squad_xa_sum=0.7809` `team_expected_assists=1.0413` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0497 -> 0.0588 expected_assists: 0.0294 -> 0.0392
`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=0.0571` `goal_rate=0.0` `scorer_weight=0.047`	-
`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=0.0384` `assist_rate=0.0` `provider_weight=0.0314`	-

Walter Benítez G 90 0 0.00 - n/a 0.0 7.0 7.5 30 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 0.9890, quality 0.446)	`market_value_eur=3700000` `quality_factor=0.4461` `boost_factor=0.989` `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': 'William Osula, Yoane Wissa', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.6187}`	-
`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 1.60 (squad xG sum 1.3541, xG scale 1.1831, xA scale 1.3335, team xA 1.04 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.602` `squad_xg_sum=1.3541` `squad_xa_sum=0.7809` `team_expected_assists=1.0413` `team_xa_to_xg_ratio=0.65`	shots_faced: 15.4 -> 16.0 expected_saves: 3.77 -> 3.91 high_pressure_saves: 1.29 -> 1.35
`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=1.6514`	-
`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=0.917`	-

Ismaïla Sarr F 82 0 0.40 0.31 ok 2 1.4 6.7 52.3 22 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 1.0128, 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`	-
`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': 'Nico González, Sven Botman', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle', 'duel_win_rate': 0.237, 'pressing_faced': 'HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.9508, 'skill_edge': -0.0155, 'height_edge': -0.018, 'attacker_skill': 0.7957, 'defender_skill': 0.8301, 'attacker_role': 'SS', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Nico González', 'market_value': 54000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8599}, {'player': 'Sven Botman', 'market_value': 33000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8002}], 'xg_override': 0.3411}	totalShots: 2.1221 -> 2.0177 shotOffTarget: 1.2846 -> 1.2214 onTargetScoringAttempt: 0.8375 -> 0.7963 bigChanceMissed: 0.6197 -> 0.5892 expectedGoalsOnTarget: 0.5496 -> 0.5226 expectedGoals: 0.3588 -> 0.3411 goals: 0.3237 -> 0.3078
`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 1.60 (squad xG sum 1.3541, xG scale 1.1831, xA scale 1.3335, team xA 1.04 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.602` `squad_xg_sum=1.3541` `squad_xa_sum=0.7809` `team_expected_assists=1.0413` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.3411 -> 0.4037 expected_assists: 0.047 -> 0.0627
`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=0.3322` `goal_rate=0.3078` `scorer_weight=0.3693`	-
`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=0.0608` `assist_rate=0.0071` `provider_weight=0.0502`	-

Yéremy Pino F 68 0 0.13 0.07 88% 1 0.9 6.9 52.6 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 1.0062, quality 0.694)	`market_value_eur=27000000` `quality_factor=0.6941` `boost_factor=1.0062` `reference_quality=0.505` `history_matches=41` `history_shrink=0.1087`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0052)	`blend_weight=0.0052` `history_matches=41` `season_overlap_discount=1.0` `per90_to_match_factor=0.7558` `prior_reliability=0.0733` `competition=Premier League`	totalCross: 2.6662 -> 2.6844 fouls: 0.8563 -> 0.8625 totalOffside: 0.3032 -> 0.3087 wasFouled: 1.02 -> 1.0254
`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': 'Nico González, Sven Botman', 'matchup_zone': 'Shadow Striker vs CDM + CB — Zone 14 Battle', 'duel_win_rate': 0.18, 'pressing_faced': 'HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.8219, 'skill_edge': -0.0612, 'height_edge': -0.06, 'attacker_skill': 0.6941, 'defender_skill': 0.8301, 'attacker_role': 'SS', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Nico González', 'market_value': 54000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8599}, {'player': 'Sven Botman', 'market_value': 33000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8002}], 'xg_override': 0.1126}	totalShots: 1.2761 -> 1.0488 shotOffTarget: 0.7784 -> 0.6398 onTargetScoringAttempt: 0.4914 -> 0.4039 expectedGoalsOnTarget: 0.1694 -> 0.1392 goals: 0.1654 -> 0.1359 expectedGoals: 0.137 -> 0.1126 bigChanceMissed: 0.1365 -> 0.1122
`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 1.60 (squad xG sum 1.3541, xG scale 1.1831, xA scale 1.3335, team xA 1.04 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.602` `squad_xg_sum=1.3541` `squad_xa_sum=0.7809` `team_expected_assists=1.0413` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1126 -> 0.1332 expected_assists: 0.1097 -> 0.1463
`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=0.1247` `goal_rate=0.1359` `scorer_weight=0.1523`	-
`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=0.1361` `assist_rate=0.008` `provider_weight=0.117`	-

Jean-Philippe Mateta F 76 1 0.35 0.24 43% 2 0.8 6.7 50.1 20 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 1.0081, quality 0.710)	`market_value_eur=29000000` `quality_factor=0.71` `boost_factor=1.0081` `reference_quality=0.505` `history_matches=33` `history_shrink=0.1316`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0040)	`blend_weight=0.004` `history_matches=33` `season_overlap_discount=1.0` `per90_to_match_factor=0.8484` `prior_reliability=0.0467` `competition=Premier League`	fouls: 0.8452 -> 0.8515
`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': 'Sven Botman, Malick Thiaw', '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': -0.053, 'height_edge': -0.006, 'attacker_skill': 0.71, 'defender_skill': 0.8279, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Sven Botman', 'market_value': 33000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8002}, {'player': 'Malick Thiaw', 'market_value': 46000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8555}], 'xg_override': 0.2918}	totalShots: 1.7051 -> 1.5571 bigChanceMissed: 1.0224 -> 0.9337 shotOffTarget: 0.9286 -> 0.848 onTargetScoringAttempt: 0.7704 -> 0.7035 goals: 0.3596 -> 0.3284 expectedGoals: 0.3195 -> 0.2918 expectedGoalsOnTarget: 0.1923 -> 0.1756
`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 1.60 (squad xG sum 1.3541, xG scale 1.1831, xA scale 1.3335, team xA 1.04 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.602` `squad_xg_sum=1.3541` `squad_xa_sum=0.7809` `team_expected_assists=1.0413` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.2918 -> 0.3452 expected_assists: 0.0354 -> 0.0472
`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=1` `goal_probability=0.2919` `goal_rate=0.3284` `scorer_weight=0.3822`	-
`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=0.0461` `assist_rate=0.0` `provider_weight=0.0378`	-

Team totals 1 1.60 13 6.8

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Dean Henderson	G	5.5m	15.4%	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
Jørgen Strand Larsen	F	27.4m	55.5%	Market-value substitute prior

Newcastle — 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

Lukáš Horníček G 90 0 0.00 - n/a 0.0 6.9 37.0 10 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 1.0258, quality 0.677)	`market_value_eur=25000000` `quality_factor=0.6773` `boost_factor=1.0258` `reference_quality=0.505` `history_matches=5` `history_shrink=0.5`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0433)	`blend_weight=0.0433` `history_matches=5` `season_overlap_discount=1.0` `per90_to_match_factor=1.0` `prior_reliability=0.1333` `competition=Premier League`	saves: 3.1507 -> 3.2092
`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': 'Ismaïla Sarr, Jean-Philippe Mateta', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.3635}`	-
`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 1.69 (squad xG sum 1.3547, xG scale 1.2445, xA scale 1.5693, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.686` `squad_xg_sum=1.3547` `squad_xa_sum=0.6983` `team_expected_assists=1.0959` `team_xa_to_xg_ratio=0.65`	shots_faced: 13.0 -> 15.2 expected_saves: 3.19 -> 3.72 high_pressure_saves: 1.09 -> 1.28
`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: 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=1.813`	-
`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=1.1691`	-

Yoane Wissa F 83 1 0.33 0.20 65% 2 0.8 6.7 43.0 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 1.0080, quality 0.660)	`market_value_eur=23000000` `quality_factor=0.6595` `boost_factor=1.008` `reference_quality=0.505` `history_matches=24` `history_shrink=0.1724`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0142)	`blend_weight=0.0142` `history_matches=24` `season_overlap_discount=1.0` `per90_to_match_factor=0.9201` `prior_reliability=0.1267` `competition=Premier League`	wasFouled: 1.5917 -> 1.6241 totalShots: 1.7158 -> 1.733 totalOffside: 0.3222 -> 0.3314 goalAssist: 0.1027 -> 0.1081
`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': 'Axel Disasi, 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': 0.9038, 'skill_edge': -0.0134, 'height_edge': -0.052, 'attacker_skill': 0.6595, 'defender_skill': 0.6893, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Axel Disasi', 'market_value': 15500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.6376}, {'player': 'Chris Richards', 'market_value': 25000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7411}], 'xg_override': 0.267}	totalShots: 1.733 -> 1.5663 shotOffTarget: 0.8778 -> 0.7934 onTargetScoringAttempt: 0.8399 -> 0.7591 bigChanceMissed: 0.7641 -> 0.6906 expectedGoalsOnTarget: 0.3499 -> 0.3162 expectedGoals: 0.2954 -> 0.267 goals: 0.29 -> 0.2621
`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 1.69 (squad xG sum 1.3547, xG scale 1.2445, xA scale 1.5693, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.686` `squad_xg_sum=1.3547` `squad_xa_sum=0.6983` `team_expected_assists=1.0959` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.267 -> 0.3323 expected_assists: 0.0365 -> 0.0573
`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` `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=0.2827` `goal_rate=0.2621` `scorer_weight=0.3271`	-
`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=0.0557` `assist_rate=0.1081` `provider_weight=0.0716`	-

Harvey Barnes F 85 0 0.20 0.13 46% 2 1.5 6.9 58.4 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 1.0085, quality 0.770)	`market_value_eur=38000000` `quality_factor=0.7699` `boost_factor=1.0085` `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=0.9444` `prior_reliability=0.1333` `competition=Premier League`	totalShots: 1.836 -> 1.8279 totalCross: 2.2211 -> 2.2137 onTargetScoringAttempt: 0.5935 -> 0.588
`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': 'Axel Disasi, Chris Richards', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.204, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 1.0062, 'skill_edge': 0.0362, 'height_edge': -0.032, 'attacker_skill': 0.7699, 'defender_skill': 0.6893, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Axel Disasi', 'market_value': 15500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.6376}, {'player': 'Chris Richards', 'market_value': 25000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7411}], 'xg_override': 0.1586}	totalShots: 1.8305 -> 1.8418 shotOffTarget: 1.2425 -> 1.2502
`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 1.69 (squad xG sum 1.3547, xG scale 1.2445, xA scale 1.5693, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.686` `squad_xg_sum=1.3547` `squad_xa_sum=0.6983` `team_expected_assists=1.0959` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1586 -> 0.1974 expected_assists: 0.1149 -> 0.1803
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 8 by projected rating, fifa_overall 80.0)	`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=0.1791` `goal_rate=0.2292` `scorer_weight=0.2467`	-
`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=0.165` `assist_rate=0.3738` `provider_weight=0.2254`	-

Matías Fernández-Pardo F 73 0 0.17 0.01 2770% 2 0.1 6.7 24.4 8 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 1.0652, quality 0.853)	`market_value_eur=60000000` `quality_factor=0.8526` `boost_factor=1.0652` `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': 'Axel Disasi, Chris Richards', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.249, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 1.0727, 'skill_edge': 0.0735, 'height_edge': -0.024, 'attacker_skill': 0.8526, 'defender_skill': 0.6893, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Axel Disasi', 'market_value': 15500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.6376}, {'player': 'Chris Richards', 'market_value': 25000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7411}], 'xg_override': 0.1384}	totalShots: 1.4569 -> 1.5629 shotOffTarget: 0.9787 -> 1.0499 onTargetScoringAttempt: 0.4782 -> 0.513 expectedGoals: 0.129 -> 0.1384 goals: 0.1122 -> 0.1204
`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 1.69 (squad xG sum 1.3547, xG scale 1.2445, xA scale 1.5693, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.686` `squad_xg_sum=1.3547` `squad_xa_sum=0.6983` `team_expected_assists=1.0959` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.1384 -> 0.1722 expected_assists: 0.0508 -> 0.0797
`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1582` `goal_rate=0.1204` `scorer_weight=0.1558`	-
`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=0.0766` `assist_rate=0.0` `provider_weight=0.0638`	-

Joelinton M 85 0 0.10 0.08 ok 1 1.1 6.8 47.8 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 1.0064, quality 0.641)	`market_value_eur=21000000` `quality_factor=0.6408` `boost_factor=1.0064` `reference_quality=0.505` `history_matches=27` `history_shrink=0.1562`	-
`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': 'Axel Disasi', 'matchup_zone': 'LW vs CB', 'duel_win_rate': 0.405, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 0.9786, 'skill_edge': 0.0014, 'height_edge': -0.016, 'attacker_skill': 0.6408, 'defender_skill': 0.6376, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Axel Disasi', 'market_value': 15500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.6376}], 'xg_override': 0.0794}`	aerialWon: 2.7851 -> 1.2 totalShots: 1.2113 -> 1.1854 shotOffTarget: 0.9169 -> 0.8973 onTargetScoringAttempt: 0.2944 -> 0.2881
`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 1.69 (squad xG sum 1.3547, xG scale 1.2445, xA scale 1.5693, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.686` `squad_xg_sum=1.3547` `squad_xa_sum=0.6983` `team_expected_assists=1.0959` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0794 -> 0.0988 expected_assists: 0.0574 -> 0.0901
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 9 by projected rating, fifa_overall 80.0)	`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=0.0941` `goal_rate=0.0888` `scorer_weight=0.106`	-
`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=0.0862` `assist_rate=0.0677` `provider_weight=0.0987`	-

Nico González M 69 0 0.04 0.01 471% 1 0.1 6.9 30.9 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 1.0151, quality 0.837)	`market_value_eur=54000000` `quality_factor=0.837` `boost_factor=1.0151` `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': 'Ismaïla Sarr, 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 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 1.69 (squad xG sum 1.3547, xG scale 1.2445, xA scale 1.5693, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.686` `squad_xg_sum=1.3547` `squad_xa_sum=0.6983` `team_expected_assists=1.0959` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0332 -> 0.0413 expected_assists: 0.0493 -> 0.0774
`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=0` `goal_probability=0.0405` `goal_rate=0.0378` `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=0.0745` `assist_rate=0.056` `provider_weight=0.0826`	-

Lewis Hall D 88 0 0.07 0.03 138% 1 0.7 7.0 52.7 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 1.0122, quality 0.831)	`market_value_eur=52000000` `quality_factor=0.8309` `boost_factor=1.0122` `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=0.9776` `prior_reliability=0.1333` `competition=Premier League`	fouls: 0.701 -> 0.7093 wonTackle: 1.317 -> 1.3239
`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': 'Right Wingback vs Left 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 1.69 (squad xG sum 1.3547, xG scale 1.2445, xA scale 1.5693, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.686` `squad_xg_sum=1.3547` `squad_xa_sum=0.6983` `team_expected_assists=1.0959` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0549 -> 0.0683 expected_assists: 0.119 -> 0.1867
`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.066` `goal_rate=0.1693` `scorer_weight=0.0854`	-
`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=0.1703` `assist_rate=0.1118` `provider_weight=0.1787`	-

Sven Botman D 89 0 0.05 0.06 ok 1 0.5 6.9 47.7 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 1.0100, quality 0.739)	`market_value_eur=33000000` `quality_factor=0.7389` `boost_factor=1.01` `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=0.9878` `prior_reliability=0.1333` `competition=Premier League`	interceptionWon: 1.2285 -> 1.2133 wonTackle: 0.7581 -> 0.7487 fouls: 0.542 -> 0.5353 wasFouled: 1.0006 -> 0.9943
`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 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 1.69 (squad xG sum 1.3547, xG scale 1.2445, xA scale 1.5693, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.686` `squad_xg_sum=1.3547` `squad_xa_sum=0.6983` `team_expected_assists=1.0959` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0389 -> 0.0484 expected_assists: 0.0435 -> 0.0683
`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0472` `goal_rate=0.0134` `scorer_weight=0.0387`	-
`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=0.066` `assist_rate=0.0` `provider_weight=0.0546`	-

Malick Thiaw D 90 0 0.09 0.12 ok 0 0.5 6.8 56.9 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 1.0101, quality 0.809)	`market_value_eur=46000000` `quality_factor=0.8089` `boost_factor=1.0101` `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=0.9977` `prior_reliability=0.1333` `competition=Premier League`	wasFouled: 0.4585 -> 0.4685 interceptionWon: 0.915 -> 0.9062 wonTackle: 0.6165 -> 0.6106
`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 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 1.69 (squad xG sum 1.3547, xG scale 1.2445, xA scale 1.5693, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.686` `squad_xg_sum=1.3547` `squad_xa_sum=0.6983` `team_expected_assists=1.0959` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0721 -> 0.0897 expected_assists: 0.0184 -> 0.0289
`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0858` `goal_rate=0.0182` `scorer_weight=0.0718`	-
`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=0.0285` `assist_rate=0.0` `provider_weight=0.0231`	-

Lewis Miley M 84 0 0.05 0.04 48% 1 0.2 6.9 48.8 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 1.0103, quality 0.732)	`market_value_eur=32000000` `quality_factor=0.732` `boost_factor=1.0103` `reference_quality=0.505` `history_matches=28` `history_shrink=0.1515`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0118)	`blend_weight=0.0118` `history_matches=28` `season_overlap_discount=1.0` `per90_to_match_factor=0.9378` `prior_reliability=0.12` `competition=Premier League`	totalCross: 1.4546 -> 1.4374 wasFouled: 0.2527 -> 0.262 wonTackle: 1.042 -> 1.0358
`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 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 1.69 (squad xG sum 1.3547, xG scale 1.2445, xA scale 1.5693, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.686` `squad_xg_sum=1.3547` `squad_xa_sum=0.6983` `team_expected_assists=1.0959` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0424 -> 0.0528 expected_assists: 0.0743 -> 0.1166
`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=0.0514` `goal_rate=0.0422` `scorer_weight=0.0527`	-
`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=0.1101` `assist_rate=0.0755` `provider_weight=0.1177`	-

William Osula F 73 1 0.42 0.17 151% 2 0.8 6.9 47.5 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 1.0102, quality 0.710)	`market_value_eur=29000000` `quality_factor=0.71` `boost_factor=1.0102` `reference_quality=0.505` `history_matches=25` `history_shrink=0.1667`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0043)	`blend_weight=0.0043` `history_matches=25` `season_overlap_discount=1.0` `per90_to_match_factor=0.8134` `prior_reliability=0.04` `competition=Premier League`	fouls: 1.3521 -> 1.3462 wasFouled: 1.372 -> 1.3661 goalAssist: 0.1972 -> 0.2022
`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': 'Axel Disasi, Chris Richards', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.217, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v2', 'xg_multiplier': 1.0254, 'skill_edge': 0.0093, 'height_edge': 0.008, 'attacker_skill': 0.71, 'defender_skill': 0.6893, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Axel Disasi', 'market_value': 15500000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.6376}, {'player': 'Chris Richards', 'market_value': 25000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.7411}], 'xg_override': 0.3393}	totalShots: 2.1252 -> 2.1792 shotOffTarget: 1.2081 -> 1.2388 onTargetScoringAttempt: 0.9108 -> 0.9339 goals: 0.467 -> 0.4789 bigChanceMissed: 0.4493 -> 0.4607 expectedGoalsOnTarget: 0.4342 -> 0.4452 expectedGoals: 0.3309 -> 0.3393
`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 1.69 (squad xG sum 1.3547, xG scale 1.2445, xA scale 1.5693, team xA 1.10 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.686` `squad_xg_sum=1.3547` `squad_xa_sum=0.6983` `team_expected_assists=1.0959` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.3393 -> 0.4224 expected_assists: 0.0721 -> 0.1131
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 75.0)	`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=1` `goal_probability=0.3445` `goal_rate=0.4789` `scorer_weight=0.528`	-
`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=0.1069` `assist_rate=0.2022` `provider_weight=0.1414`	-

Team totals 2 1.69 14 6.9

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Ewen Jaouen	G	5.1m	14.2%	Market-value substitute prior
Tino Livramento	D	15.0m	36.4%	Market-value substitute prior
Amar Dedić	D	14.7m	23.9%	Market-value substitute prior
Aladji Bamba	M	18.5m	71.0%	Market-value substitute prior
Sean Steur	M	21.2m	51.6%	Market-value substitute prior
Nick Woltemade	F	23.1m	75.2%	Market-value substitute prior

Antagonist / tactical summary

Midfield is evenly contested — Adam Wharton vs Nico González will be the pivotal duel. Total antagonist xG — Home: 1.05, Away: 0.98. Home danger zone: Shadow Striker vs CDM + CB — Zone 14 Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Midfield — possession 50.0% / 50.0%, dominance ratio 0.49, key duel: Adam Wharton vs Nico González

Jean-Philippe Mateta xG 0.29Ismaïla Sarr xG 0.34Yéremy Pino xG 0.11Joelinton xG 0.08William Osula xG 0.34Yoane Wissa xG 0.27

Summary

Antagonist matchup analysis for Crystal Palace vs Newcastle. Most likely scoreline: 1-2. Lambdas derived from 7 home and 5 away individual player duels. Head-to-head: Crystal Palace 2W-3D-5L vs Newcastle over 10 meetings.

Tactical parity between 3-4-2-1 and 4-2-3-1. Match outcome driven primarily by individual player quality. Midfield is evenly contested — Adam Wharton vs Nico González will be the pivotal duel. Total antagonist xG — Home: 1.05, Away: 0.98. Home danger zone: Shadow Striker vs CDM + CB — Zone 14 Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.

Barcelona vs Real Madrid

LaLiga · 20261025 · 4-2-3-1 vs 3-4-1-2

Predicted 2-1Winner: BarcelonaWinning H 2-1 (8.2%)Winning A 1-2 (4.7%)Confidence 0.78λ home/away 2.67 / 1.53Most-likely P 8.2%

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

XGBoost Reasoning & FBref Dataset Tracing

Antagonist + XGBoost match analysis for Barcelona vs Real Madrid. 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 3-4-1-2 (duel lambdas: Home 2.835, Away 1.840). Final calibrated Poisson lambdas: Home 2.674, Away 1.527 (most likely score 2-1).

Home Starters 11/11 XGBoost (11 FBref) Away Starters 11/11 XGBoost (10 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 2.83 / 1.84 Final λ (H/A) 2.67 / 1.53

FBref Team Priors: FC Barcelona (Gls/90: 3.6666666666666665, GA/90: 4.0 → E[λ]: 3.5) vs Real Madrid (Gls/90: 3.0, GA/90: 3.33 → 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	Real Madrid
League position	1.00	3.00
Points per game	2.34	1.94
Attack index	1.87	1.23
Defense (concede) index	0.81	0.89
Form multiplier	1.02	0.98
Standing λ multiplier	1.19	1.03

Lambda decomposition (how the goal rates were built)

Stage	Barcelona λ	Real Madrid λ
FBref Squad Prior E[λ] (weight 0.228)	3.500	3.500
Pre-calibration (Micro-duels)	2.835	1.840
Final λ	2.674	1.527

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

Scoreline distribution & markets

Top 5 scorelines	P
2-1	8.2%
3-1	7.3%
2-2	6.2%
1-1	6.2%
3-2	5.6%

Home win	61.9%
Draw	17.6%
Away win	20.4%
Over 2.5	79.0%
BTTS yes	72.9%

Pregame fair line — favorite Barcelona (61.9%), fair total 4.20, fair spread 1.15

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 (4-2-3-1)

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

Raphinha F 78 1 0.60 0.57 ok 3 2.3 7.9 60.2 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 1.0156, quality 0.859)	`market_value_eur=63000000` `quality_factor=0.8589` `boost_factor=1.0156` `reference_quality=0.505` `history_matches=29` `history_shrink=0.1471`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0159)	`blend_weight=0.0159` `history_matches=29` `season_overlap_discount=1.0` `per90_to_match_factor=0.8613` `prior_reliability=0.1667` `competition=La Liga`	goals: 0.9518 -> 0.9641 totalCross: 2.7355 -> 2.7468 fouls: 0.3374 -> 0.3485 onTargetScoringAttempt: 1.8397 -> 1.8487 totalOffside: 0.5008 -> 0.5093 wonTackle: 0.4262 -> 0.4194 wasFouled: 0.7008 -> 0.6951
`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 Cucurella, Dean Huijsen', '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': 0.9291, 'skill_edge': -0.0122, 'height_edge': -0.036, 'attacker_skill': 0.8589, 'defender_skill': 0.886, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Marc Cucurella', 'market_value': 62000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8842}, {'player': 'Dean Huijsen', 'market_value': 66000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8879}], 'xg_override': 0.5149}	totalShots: 3.2273 -> 2.9985 onTargetScoringAttempt: 1.8487 -> 1.7176 shotOffTarget: 1.3786 -> 1.2809 expectedGoalsOnTarget: 1.116 -> 1.0369 goals: 0.9641 -> 0.8957 bigChanceMissed: 0.7397 -> 0.6873 expectedGoals: 0.5542 -> 0.5149
`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 2.67 (squad xG sum 2.2996, xG scale 1.1628, xA scale 1.2424, team xA 1.74 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.674` `squad_xg_sum=2.2996` `squad_xa_sum=1.399` `team_expected_assists=1.7381` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.5149 -> 0.5987 expected_assists: 0.1826 -> 0.2269
`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.4505` `goal_rate=0.8957` `scorer_weight=0.7484`	-
`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=0.203` `assist_rate=0.299` `provider_weight=0.2836`	-

Anthony Gordon F 77 0 0.38 0.27 44% 2 1.5 6.9 51.8 26 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 1.0154, quality 0.884)	`market_value_eur=82000000` `quality_factor=0.8837` `boost_factor=1.0154` `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=0.8582` `prior_reliability=0.1333` `competition=La Liga`	goalAssist: 0.2381 -> 0.2504 totalShots: 2.1326 -> 2.1426 totalCross: 1.2496 -> 1.24 onTargetScoringAttempt: 0.7739 -> 0.7798
`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 Cucurella, Dean Huijsen', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.191, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.9867, 'skill_edge': -0.0011, 'height_edge': -0.008, 'attacker_skill': 0.8837, 'defender_skill': 0.886, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Marc Cucurella', 'market_value': 62000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8842}, {'player': 'Dean Huijsen', 'market_value': 66000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8879}], 'xg_override': 0.3306}	totalShots: 2.1426 -> 2.1141 shotOffTarget: 1.3587 -> 1.3406 onTargetScoringAttempt: 0.7798 -> 0.7694 bigChanceMissed: 0.3846 -> 0.3795
`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 2.67 (squad xG sum 2.2996, xG scale 1.1628, xA scale 1.2424, team xA 1.74 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.674` `squad_xg_sum=2.2996` `squad_xa_sum=1.399` `team_expected_assists=1.7381` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.3306 -> 0.3844 expected_assists: 0.1171 -> 0.1455
`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.3191` `goal_rate=0.2618` `scorer_weight=0.3149`	-
`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=0.1354` `assist_rate=0.2504` `provider_weight=0.1819`	-

Fermín López M 80 0 0.28 0.28 ok 2 1.9 7.1 63.5 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 1.0143, quality 0.896)	`market_value_eur=113000000` `quality_factor=0.8965` `boost_factor=1.0143` `reference_quality=0.505` `history_matches=36` `history_shrink=0.122`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0122)	`blend_weight=0.0122` `history_matches=36` `season_overlap_discount=1.0` `per90_to_match_factor=0.888` `prior_reliability=0.1533` `competition=La Liga`	onTargetScoringAttempt: 0.8838 -> 0.8965 goals: 0.2234 -> 0.2348 wasFouled: 0.9768 -> 0.9837 wonTackle: 0.6874 -> 0.6931
`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 Cucurella', 'matchup_zone': 'LW vs CB', 'duel_win_rate': 0.43, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 1.014, 'skill_edge': 0.0055, 'height_edge': 0.004, 'attacker_skill': 0.8965, 'defender_skill': 0.8842, 'attacker_role': 'LW', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Marc Cucurella', 'market_value': 62000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8842}], 'xg_override': 0.2419}	aerialWon: 0.5714 -> 0.9 totalShots: 2.1389 -> 2.1689 shotOffTarget: 1.2424 -> 1.2598 onTargetScoringAttempt: 0.8965 -> 0.9091 expectedGoalsOnTarget: 0.5662 -> 0.5741
`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 2.67 (squad xG sum 2.2996, xG scale 1.1628, xA scale 1.2424, team xA 1.74 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.674` `squad_xg_sum=2.2996` `squad_xa_sum=1.399` `team_expected_assists=1.7381` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.2419 -> 0.2813 expected_assists: 0.1781 -> 0.2213
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 90.0)	`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.2452` `goal_rate=0.2381` `scorer_weight=0.2776`	-
`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=0.1985` `assist_rate=0.167` `provider_weight=0.2105`	-

Lamine Yamal F 83 1 0.75 0.52 46% 5 3.6 8.4 67.3 28 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 1.0148, quality 0.900)	`market_value_eur=237000000` `quality_factor=0.9` `boost_factor=1.0148` `reference_quality=0.505` `history_matches=35` `history_shrink=0.125`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0146)	`blend_weight=0.0146` `history_matches=35` `season_overlap_discount=1.0` `per90_to_match_factor=0.9181` `prior_reliability=0.18` `competition=La Liga`	totalOffside: 0.9147 -> 0.9312 totalCross: 2.6881 -> 2.7035 goalAssist: 0.6874 -> 0.6773 fouls: 1.2032 -> 1.1955 interceptionWon: 0.2533 -> 0.2595
`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 Cucurella, Dean Huijsen', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.198, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.9975, 'skill_edge': 0.0063, 'height_edge': -0.008, 'attacker_skill': 0.9, 'defender_skill': 0.886, 'attacker_role': 'ST', 'attacker_mv_price_index': 1.0, 'defender_adjusted_parts': [{'player': 'Marc Cucurella', 'market_value': 62000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8842}, {'player': 'Dean Huijsen', 'market_value': 66000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8879}], 'xg_override': 0.6468}	totalShots: 5.081 -> 5.0683 shotOffTarget: 3.3763 -> 3.3679
`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 assists: 1.0 -> 0
`team_lambda_rescale`	Player expected_goals/expected_assists rescaled to unconditional contributions so squad sum equals team lambda 2.67 (squad xG sum 2.2996, xG scale 1.1628, xA scale 1.2424, team xA 1.74 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.674` `squad_xg_sum=2.2996` `squad_xa_sum=1.399` `team_expected_assists=1.7381` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.6468 -> 0.7522 expected_assists: 0.2476 -> 0.3076
`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.5287` `goal_rate=0.7925` `scorer_weight=0.9095`	-
`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=0.2648` `assist_rate=0.6773` `provider_weight=0.3845`	-

Pedri M 80 0 0.10 0.05 111% 1 0.6 7.2 64.6 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 1.0144, quality 0.900)	`market_value_eur=164000000` `quality_factor=0.8997` `boost_factor=1.0144` `reference_quality=0.505` `history_matches=36` `history_shrink=0.122`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0122)	`blend_weight=0.0122` `history_matches=36` `season_overlap_discount=1.0` `per90_to_match_factor=0.8863` `prior_reliability=0.1533` `competition=La Liga`	wonTackle: 0.9161 -> 0.9237
`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': 'Yan Diomande, Jude Bellingham, Aurélien Tchouaméni', 'matchup_zone': 'Attacking Mid vs Double Pivot — Tight Space 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 2.67 (squad xG sum 2.2996, xG scale 1.1628, xA scale 1.2424, team xA 1.74 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.674` `squad_xg_sum=2.2996` `squad_xa_sum=1.399` `team_expected_assists=1.7381` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0855 -> 0.0994 expected_assists: 0.1821 -> 0.2262
`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0946` `goal_rate=0.0927` `scorer_weight=0.1047`	-
`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=0.2024` `assist_rate=0.2025` `provider_weight=0.2461`	-

Rodri M 82 0 0.11 0.08 42% 1 0.7 7.5 68.9 23 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 1.0156, quality 0.837)	`market_value_eur=54000000` `quality_factor=0.837` `boost_factor=1.0156` `reference_quality=0.505` `history_matches=27` `history_shrink=0.1562`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0041)	`blend_weight=0.0041` `history_matches=27` `season_overlap_discount=1.0` `per90_to_match_factor=0.9159` `prior_reliability=0.04` `competition=La Liga`	wonTackle: 1.7173 -> 1.7103 interceptionWon: 1.3892 -> 1.3836
`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': 'Yan Diomande, Aurélien Tchouaméni', 'matchup_zone': 'Attacking Mid vs Double Pivot — Tight Space 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 2.67 (squad xG sum 2.2996, xG scale 1.1628, xA scale 1.2424, team xA 1.74 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.674` `squad_xg_sum=2.2996` `squad_xa_sum=1.399` `team_expected_assists=1.7381` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0957 -> 0.1113 expected_assists: 0.1041 -> 0.1293
`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.1053` `goal_rate=0.0726` `scorer_weight=0.0937`	-
`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=0.1213` `assist_rate=0.0662` `provider_weight=0.1034`	-

João Cancelo D 83 0 0.05 0.02 139% 1 0.6 6.9 55.2 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 0.9993, quality 0.493)	`market_value_eur=7500000` `quality_factor=0.4927` `boost_factor=0.9993` `reference_quality=0.505` `history_matches=22` `history_shrink=0.1852`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0024)	`blend_weight=0.0024` `history_matches=22` `season_overlap_discount=1.0` `per90_to_match_factor=0.9251` `prior_reliability=0.02` `competition=La Liga`	totalCross: 2.4228 -> 2.417
`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': 'Federico Valverde', 'matchup_zone': 'Right Wingback vs Left 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 2.67 (squad xG sum 2.2996, xG scale 1.1628, xA scale 1.2424, team xA 1.74 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.674` `squad_xg_sum=2.2996` `squad_xa_sum=1.399` `team_expected_assists=1.7381` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0471 -> 0.0548 expected_assists: 0.1527 -> 0.1897
`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` `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=0.0533` `goal_rate=0.3069` `scorer_weight=0.0685`	-
`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=0.1728` `assist_rate=0.0747` `provider_weight=0.1518`	-

Gerard Martín D 82 0 0.02 0.01 83% 0 0.2 7.1 56.0 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 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=0.9162` `prior_reliability=0.1333` `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': 'Kylian Mbappé', '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 2.67 (squad xG sum 2.2996, xG scale 1.1628, xA scale 1.2424, team xA 1.74 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.674` `squad_xg_sum=2.2996` `squad_xa_sum=1.399` `team_expected_assists=1.7381` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0229 -> 0.0285
`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=0.0165` `goal_rate=0.0` `scorer_weight=0.0133`	-
`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=0.0` `provider_weight=0.0228`	-

Pau Cubarsí D 85 0 0.05 0.02 105% 0 0.2 7.2 55.6 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 1.0141, quality 0.891)	`market_value_eur=94000000` `quality_factor=0.891` `boost_factor=1.0141` `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=0.9471` `prior_reliability=0.1333` `competition=La Liga`	wonTackle: 0.7334 -> 0.7407 interceptionWon: 1.1487 -> 1.1416 totalShots: 0.4521 -> 0.4573
`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': 'Kylian Mbappé', '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 2.67 (squad xG sum 2.2996, xG scale 1.1628, xA scale 1.2424, team xA 1.74 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.674` `squad_xg_sum=2.2996` `squad_xa_sum=1.399` `team_expected_assists=1.7381` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.0427 -> 0.0497
`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=0.0485` `goal_rate=0.0113` `scorer_weight=0.0398`	-
`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=0.0248` `assist_rate=0.0` `provider_weight=0.0201`	-

Joan García G 84 0 0.00 - n/a 0.0 7.1 62.2 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 1.0111, quality 0.809)	`market_value_eur=46000000` `quality_factor=0.8089` `boost_factor=1.0111` `reference_quality=0.505` `history_matches=36` `history_shrink=0.122`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0159)	`blend_weight=0.0159` `history_matches=36` `season_overlap_discount=1.0` `per90_to_match_factor=0.9288` `prior_reliability=0.2` `competition=La Liga`	saves: 2.7355 -> 2.7167
`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': 'Kylian Mbappé, Jude Bellingham', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 1.8402}`	-
`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 2.67 (squad xG sum 2.2996, xG scale 1.1628, xA scale 1.2424, team xA 1.74 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.674` `squad_xg_sum=2.2996` `squad_xa_sum=1.399` `team_expected_assists=1.7381` `team_xa_to_xg_ratio=0.65`	shots_faced: 17.5 -> 14.5 expected_saves: 4.28 -> 3.55 high_pressure_saves: 1.47 -> 1.22
`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: 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=2.8912`	-
`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=1.8872`	-

Xavi Espart D 88 0 0.03 0.01 320% 1 0.2 6.9 26.0 21 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 1.0194, quality 0.686)	`market_value_eur=26000000` `quality_factor=0.6858` `boost_factor=1.0194` `reference_quality=0.505` `history_matches=9` `history_shrink=0.3571`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0464)	`blend_weight=0.0464` `history_matches=9` `season_overlap_discount=1.0` `per90_to_match_factor=0.9799` `prior_reliability=0.2` `competition=La Liga`	wasFouled: 0.9951 -> 0.9641 totalCross: 0.6563 -> 0.6865 interceptionWon: 0.9864 -> 1.0164 fouls: 1.5835 -> 1.5555 goalAssist: 0.3055 -> 0.3216 goals: 0.2455 -> 0.2341
`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': 'Eduardo Camavinga', '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 2.67 (squad xG sum 2.2996, xG scale 1.1628, xA scale 1.2424, team xA 1.74 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=2.674` `squad_xg_sum=2.2996` `squad_xa_sum=1.399` `team_expected_assists=1.7381` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.069 -> 0.0857
`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: 2 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=2` `allocation_applied=true` `goals_allocated=0` `goal_probability=0.0338` `goal_rate=0.2341` `scorer_weight=0.043`	-
`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=0.0821` `assist_rate=0.3216` `provider_weight=0.1071`	-

Team totals 2 2.67 19 7.3

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
Karim Adeyemi	F	19.1m	61.9%	Market-value substitute prior

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

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

Marc Cucurella D 88 0 0.03 0.07 63% 0 0.3 6.8 57.6 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 1.0115, quality 0.857)	`market_value_eur=62000000` `quality_factor=0.8569` `boost_factor=1.0115` `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=0.9801` `prior_reliability=0.12` `competition=La Liga`	totalCross: 1.5102 -> 1.4974 wasFouled: 0.7829 -> 0.7763 fouls: 0.7318 -> 0.7256
`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': 'Fermín López', '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 1.53 (squad xG sum 1.5869, xG scale 0.9623, xA scale 1.1580, team xA 0.99 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.527` `squad_xg_sum=1.5869` `squad_xa_sum=0.8571` `team_expected_assists=0.9925` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.037 -> 0.0428
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 17 by projected rating, fifa_overall 85.0)	`match_squad_rank=17` `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=0.026` `goal_rate=0.0` `scorer_weight=0.021`	-
`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=0.066` `provider_weight=0.0535`	-

Eduardo Camavinga M 78 0 0.03 0.01 149% 0 0.3 7.0 59.4 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 1.0117, quality 0.817)	`market_value_eur=48000000` `quality_factor=0.8168` `boost_factor=1.0117` `reference_quality=0.505` `history_matches=35` `history_shrink=0.125`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0070)	`blend_weight=0.007` `history_matches=35` `season_overlap_discount=1.0` `per90_to_match_factor=0.8658` `prior_reliability=0.0867` `competition=La Liga`	interceptionWon: 0.7623 -> 0.771 wasFouled: 1.1843 -> 1.176
`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': 'Xavi Espart', 'matchup_zone': 'Left Wingback vs Right Back — Overlapping Run', 'duel_win_rate': 0.475, 'pressing_faced': 'MEDIUM', 'numerical_situation': '1v1', 'xg_multiplier': 1.1097, 'skill_edge': 0.0466, 'height_edge': 0.028, 'attacker_skill': 0.8686, 'defender_skill': 0.765, 'attacker_role': 'LWB', 'attacker_mv_price_index': 0.7, 'defender_adjusted_parts': [{'player': 'Xavi Espart', 'market_value': 26000000, 'role': 'RB', 'price_index': 0.7, 'adjusted_quality': 0.765}], 'xg_override': 0.0326}	aerialWon: 0.4488 -> 1.1 totalShots: 0.413 -> 0.4583 shotOffTarget: 0.2963 -> 0.3288 onTargetScoringAttempt: 0.1167 -> 0.1295
`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 1.5869, xG scale 0.9623, xA scale 1.1580, team xA 0.99 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.527` `squad_xg_sum=1.5869` `squad_xa_sum=0.8571` `team_expected_assists=0.9925` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0457 -> 0.0529
`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` `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=0.0428` `scorer_weight=0.0353`	-
`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=0.0515` `assist_rate=0.0359` `provider_weight=0.0475`	-

Aurélien Tchouaméni M 79 0 0.05 0.07 ok 1 1.3 7.0 62.7 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 1.0135, quality 0.875)	`market_value_eur=73000000` `quality_factor=0.8747` `boost_factor=1.0135` `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=true`	-
`enriched_model_output`	Model output after duel-context enrichment (pre router post-processing)	`duel_context_applied=true` duel_context={'opponent_faced': 'Pedri', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.498, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 1.0711, 'skill_edge': -0.0076, 'height_edge': 0.056, 'attacker_skill': 0.883, 'defender_skill': 0.8999, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Pedri', 'market_value': 164000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8999}], 'xg_override': 0.0536}	aerialWon: 1.4642 -> 1.2 totalShots: 0.9994 -> 1.0705 shotOffTarget: 0.7871 -> 0.8431 onTargetScoringAttempt: 0.2123 -> 0.2274
`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 1.5869, xG scale 0.9623, xA scale 1.1580, team xA 0.99 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.527` `squad_xg_sum=1.5869` `squad_xa_sum=0.8571` `team_expected_assists=0.9925` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0549 -> 0.0636
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 6 by projected rating, fifa_overall 84.0)	`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=0.0503` `goal_rate=0.0549` `scorer_weight=0.0522`	-
`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=0.0616` `assist_rate=0.0571` `provider_weight=0.0649`	-

Federico Valverde M 82 0 0.09 0.08 ok 1 1.3 7.0 68.4 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 1.0127, quality 0.887)	`market_value_eur=87000000` `quality_factor=0.8873` `boost_factor=1.0127` `reference_quality=0.505` `history_matches=40` `history_shrink=0.1111`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0135)	`blend_weight=0.0135` `history_matches=40` `season_overlap_discount=1.0` `per90_to_match_factor=0.9144` `prior_reliability=0.1867` `competition=La Liga`	wonTackle: 0.6717 -> 0.6847 wasFouled: 0.6002 -> 0.5921
`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 Cancelo', 'matchup_zone': 'Right Wingback vs Left Back — Overlapping Run', 'duel_win_rate': 0.561, 'pressing_faced': 'LOW', 'numerical_situation': '1v1', 'xg_multiplier': 1.237, 'skill_edge': 0.1652, 'height_edge': -0.004, 'attacker_skill': 0.898, 'defender_skill': 0.5308, 'attacker_role': 'RWB', 'attacker_mv_price_index': 0.7, 'defender_adjusted_parts': [{'player': 'João Cancelo', 'market_value': 7500000, 'role': 'LB', 'price_index': 0.7, 'adjusted_quality': 0.5308}], 'xg_override': 0.0931}	aerialWon: 0.7368 -> 1.1 totalShots: 0.9499 -> 1.175 shotOffTarget: 0.6587 -> 0.8148 onTargetScoringAttempt: 0.2912 -> 0.3602 goals: 0.0908 -> 0.1123 expectedGoalsOnTarget: 0.0829 -> 0.1025 expectedGoals: 0.0753 -> 0.0931
`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 1.5869, xG scale 0.9623, xA scale 1.1580, team xA 0.99 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.527` `squad_xg_sum=1.5869` `squad_xa_sum=0.8571` `team_expected_assists=0.9925` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0693 -> 0.0802
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 4 by projected rating, fifa_overall 87.0)	`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=0.0857` `goal_rate=0.1123` `scorer_weight=0.1016`	-
`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=0.0771` `assist_rate=0.0648` `provider_weight=0.0772`	-

Dean Huijsen D 82 0 0.09 0.06 56% 1 0.5 7.0 60.0 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 1.0138, quality 0.864)	`market_value_eur=66000000` `quality_factor=0.8644` `boost_factor=1.0138` `reference_quality=0.505` `history_matches=34` `history_shrink=0.1282`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0167)	`blend_weight=0.0167` `history_matches=34` `season_overlap_discount=1.0` `per90_to_match_factor=0.9133` `prior_reliability=0.2` `competition=La Liga`	totalShots: 0.5436 -> 0.5498 fouls: 0.9422 -> 0.9366 onTargetScoringAttempt: 0.3172 -> 0.3119
`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 1.53 (squad xG sum 1.5869, xG scale 0.9623, xA scale 1.1580, team xA 0.99 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.527` `squad_xg_sum=1.5869` `squad_xa_sum=0.8571` `team_expected_assists=0.9925` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.0363 -> 0.042
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 5 by projected rating, fifa_overall 81.0)	`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=0.0825` `goal_rate=0.0285` `scorer_weight=0.0689`	-
`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=0.0411` `assist_rate=0.0173` `provider_weight=0.0336`	-

Ibrahima Konaté D 89 0 0.05 0.03 81% 1 0.3 6.8 60.9 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 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=0.9877` `prior_reliability=0.1333` `competition=La Liga`	wonTackle: 0.6798 -> 0.6917 fouls: 1.1435 -> 1.1375
`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': '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 1.53 (squad xG sum 1.5869, xG scale 0.9623, xA scale 1.1580, team xA 0.99 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.527` `squad_xg_sum=1.5869` `squad_xa_sum=0.8571` `team_expected_assists=0.9925` `team_xa_to_xg_ratio=0.65`	-
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 16 by projected rating, fifa_overall 84.0)	`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=0.0472` `goal_rate=0.0069` `scorer_weight=0.0387`	-
`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=0.0143` `assist_rate=0.0` `provider_weight=0.0115`	-

Antonio Rüdiger D 88 0 0.05 0.04 ok 1 0.5 6.9 41.8 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 0.477)	`market_value_eur=6200000` `quality_factor=0.4769` `boost_factor=0.9984` `reference_quality=0.505` `history_matches=21` `history_shrink=0.1923`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0083)	`blend_weight=0.0083` `history_matches=21` `season_overlap_discount=1.0` `per90_to_match_factor=0.9793` `prior_reliability=0.0667` `competition=La Liga`	interceptionWon: 0.7771 -> 0.7706 wasFouled: 0.2651 -> 0.2711
`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': '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 1.53 (squad xG sum 1.5869, xG scale 0.9623, xA scale 1.1580, team xA 0.99 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.527` `squad_xg_sum=1.5869` `squad_xa_sum=0.8571` `team_expected_assists=0.9925` `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 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=0.0481` `goal_rate=0.0427` `scorer_weight=0.0452`	-
`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=0.0148` `assist_rate=0.0` `provider_weight=0.0119`	-

Thibaut Courtois G 90 0 0.00 - n/a 0.0 7.0 63.0 23 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 1.0023, quality 0.574)	`market_value_eur=14500000` `quality_factor=0.5737` `boost_factor=1.0023` `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=1.0` `prior_reliability=0.2` `competition=La Liga`	saves: 2.7104 -> 2.695
`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, Raphinha', 'matchup_zone': 'Goalkeeper shot-stopping context', 'opponent_lambda': 2.8353}`	-
`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 1.5869, xG scale 0.9623, xA scale 1.1580, team xA 0.99 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.527` `squad_xg_sum=1.5869` `squad_xa_sum=0.8571` `team_expected_assists=0.9925` `team_xa_to_xg_ratio=0.65`	shots_faced: 26.9 -> 25.4 expected_saves: 6.58 -> 6.22 high_pressure_saves: 2.27 -> 2.14
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 7 by projected rating, fifa_overall 90.0)	`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=0.0` `scorer_weight=0.0` `squad_weight_total=1.5727`	-
`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=1.0718`	-

Jude Bellingham M 83 0 0.20 0.17 ok 2 1.6 7.1 66.8 23 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 1.0148, quality 0.899)	`market_value_eur=151000000` `quality_factor=0.8995` `boost_factor=1.0148` `reference_quality=0.505` `history_matches=35` `history_shrink=0.125`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0146)	`blend_weight=0.0146` `history_matches=35` `season_overlap_discount=1.0` `per90_to_match_factor=0.9238` `prior_reliability=0.18` `competition=La Liga`	totalShots: 2.0296 -> 2.0402 interceptionWon: 0.5849 -> 0.5763 onTargetScoringAttempt: 0.7994 -> 0.8078 goalAssist: 0.1397 -> 0.1477 fouls: 1.1525 -> 1.1457 goals: 0.2175 -> 0.2243 wasFouled: 2.8032 -> 2.7972
`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': 'Pedri', 'matchup_zone': 'Central Mid vs Defensive Mid — Pressing/Creation Duel', 'duel_win_rate': 0.506, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v1', 'xg_multiplier': 1.0822, 'skill_edge': -0.0001, 'height_edge': 0.056, 'attacker_skill': 0.8998, 'defender_skill': 0.8999, 'attacker_role': 'CM', 'attacker_mv_price_index': 0.9, 'defender_adjusted_parts': [{'player': 'Pedri', 'market_value': 164000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8999}], 'xg_override': 0.2085}	aerialWon: 0.7921 -> 1.2 totalShots: 2.0402 -> 2.2079 shotOffTarget: 1.2301 -> 1.3312 onTargetScoringAttempt: 0.8078 -> 0.8742 expectedGoalsOnTarget: 0.317 -> 0.3431 goals: 0.2243 -> 0.2427 expectedGoals: 0.1927 -> 0.2085 bigChanceMissed: 0.183 -> 0.198
`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 1.5869, xG scale 0.9623, xA scale 1.1580, team xA 0.99 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.527` `squad_xg_sum=1.5869` `squad_xa_sum=0.8571` `team_expected_assists=0.9925` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.2085 -> 0.2006 expected_assists: 0.1453 -> 0.1683
`rankings_attached`	Display-only provenance: no published number changes. Global ranking attached (match squad rank 2 by projected rating, fifa_overall 90.0)	`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=0.1818` `goal_rate=0.2427` `scorer_weight=0.2271`	-
`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=0.1549` `assist_rate=0.1477` `provider_weight=0.1704`	-

Yan Diomande F 81 0 0.10 0.03 203% 1 0.5 6.9 57.3 11 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 1.0492, quality 0.899)	`market_value_eur=132000000` `quality_factor=0.8986` `boost_factor=1.0492` `reference_quality=0.505` `history_matches=7` `history_shrink=0.4167`	-
`fbref_prior_blend`	FBref 2026-27 prior blended (weight 0.0126)	`blend_weight=0.0126` `history_matches=7` `season_overlap_discount=1.0` `per90_to_match_factor=0.8954` `prior_reliability=0.0467` `competition=La Liga`	totalCross: 1.9738 -> 2.0136 wasFouled: 2.251 -> 2.2225 totalShots: 1.1221 -> 1.1079 wonTackle: 0.8958 -> 0.8845 fouls: 0.7034 -> 0.6945 interceptionWon: 0.5316 -> 0.5249 onTargetScoringAttempt: 0.4204 -> 0.4151
`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': 'Pedri, Rodri', 'matchup_zone': 'Attacking Mid vs Double Pivot — Tight Space Battle', 'duel_win_rate': 0.239, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.9979, 'skill_edge': 0.0086, 'height_edge': -0.01, 'attacker_skill': 0.899, 'defender_skill': 0.8799, 'attacker_role': 'CAM', 'attacker_mv_price_index': 0.95, 'defender_adjusted_parts': [{'player': 'Pedri', 'market_value': 164000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8999}, {'player': 'Rodri', 'market_value': 54000000, 'role': 'CDM', 'price_index': 0.85, 'adjusted_quality': 0.8599}], 'xg_override': 0.0988}	-
`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 1.5869, xG scale 0.9623, xA scale 1.1580, team xA 0.99 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.527` `squad_xg_sum=1.5869` `squad_xa_sum=0.8571` `team_expected_assists=0.9925` `team_xa_to_xg_ratio=0.65`	expected_assists: 0.1326 -> 0.1536
`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` `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=0.0907` `goal_rate=0.1077` `scorer_weight=0.1008`	-
`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=0.1424` `assist_rate=0.1944` `provider_weight=0.192`	-

Kylian Mbappé F 89 1 0.68 0.70 ok 5 4.3 7.9 64.2 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 1.0138, quality 0.900)	`market_value_eur=212000000` `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 0.0151)	`blend_weight=0.0151` `history_matches=38` `season_overlap_discount=1.0` `per90_to_match_factor=0.9897` `prior_reliability=0.2` `competition=La Liga`	totalShots: 4.8434 -> 4.8899 onTargetScoringAttempt: 2.2079 -> 2.2444 totalOffside: 1.4266 -> 1.41 totalCross: 1.2446 -> 1.2358 goals: 0.746 -> 0.7547
`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, Gerard Martín', 'matchup_zone': 'Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle', 'duel_win_rate': 0.199, 'pressing_faced': 'ULTRA_HIGH', 'numerical_situation': '1v2', 'xg_multiplier': 0.9987, 'skill_edge': 0.0191, 'height_edge': -0.02, 'attacker_skill': 0.9, 'defender_skill': 0.8575, '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': 'Gerard Martín', 'market_value': 36000000, 'role': 'CB', 'price_index': 0.75, 'adjusted_quality': 0.8168}], 'xg_override': 0.7032}	totalShots: 4.8899 -> 4.8835
`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 1.53 (squad xG sum 1.5869, xG scale 0.9623, xA scale 1.1580, team xA 0.99 = 0.65 x lambda)	`scoreline_reset=true` `team_lambda=1.527` `squad_xg_sum=1.5869` `squad_xa_sum=0.8571` `team_expected_assists=0.9925` `team_xa_to_xg_ratio=0.65`	expected_goals: 0.7032 -> 0.6765 expected_assists: 0.1718 -> 0.199
`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: 1 projected goals distributed by scorer weight (xG x finishing x appearance)	`team_goals=1` `allocation_applied=true` `goals_allocated=1` `goal_probability=0.4916` `goal_rate=0.7537` `scorer_weight=0.7218`	-
`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=0.1805` `assist_rate=0.2471` `provider_weight=0.2488`	-

Team totals 1 1.53 17 7.0

Bench Candidates (7 modeled substitutes with empirical playing time)

Player	Pos	Exp Min	App Prob	Substitute Modeling Details
Andriy Lunin	G	4.8m	12.7%	Market-value substitute prior
Trent Alexander-Arnold	D	12.9m	44.3%	Market-value substitute prior
Álvaro Carreras	D	15.1m	24.9%	Market-value substitute prior
Arda Güler	M	12.9m	47.2%	Market-value substitute prior
Franco Mastantuono	M	23.4m	56.4%	Market-value substitute prior
Vinícius Júnior	F	27.1m	37.1%	Market-value substitute prior

Antagonist / tactical summary

Midfield is evenly contested — Pedri vs Jude Bellingham will be the pivotal duel. Total antagonist xG — Home: 1.73, Away: 1.19. 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 49.5% / 50.5%, dominance ratio 0.49, key duel: Pedri vs Jude Bellingham

Fermín López xG 0.24Lamine Yamal xG 0.65Raphinha xG 0.51Kylian Mbappé xG 0.70Yan Diomande xG 0.10Eduardo Camavinga xG 0.03

Summary

Antagonist matchup analysis for Barcelona vs Real Madrid. Most likely scoreline: 2-1. Lambdas derived from 4 home and 6 away individual player duels. Head-to-head: Barcelona 7W-0D-3L vs Real Madrid over 10 meetings.

Tactical parity between 4-2-3-1 and 3-4-1-2. Match outcome driven primarily by individual player quality. Midfield is evenly contested — Pedri vs Jude Bellingham will be the pivotal duel. Total antagonist xG — Home: 1.73, Away: 1.19. 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 2026-10-25 15:17:36 UTC · Soccer prediction reporting