Soccer prediction system deep audit

Slate 20261024 · immutable forensic review of 8 match prediction artifact(s)

Executive conclusion

In 2 match(es) the headline score is not the distribution's modal scoreline (a breach of the published-mode contract) — see the MOST_LIKELY_IS_MODE findings.

At a glance

8matches audited

0critical

0errors

7warnings

0grid non-reproducible

2headline != modal

0winner disagreements

0lambda inconsistencies

0with real result

-winner hit-rate

-decisive hit-rate

38%predicted draws

-actual draws

0confident misses

-mean Brier

Matches

Chelsea vs Tottenham

20261024_TOT@CHE · Premier League

WARNINGS risk 20 pred 1-0 (Chelsea)λ 1.79 / 1.04conf 0.75 warning 2 good 38

Problems (2)

Sev	Category	Rule	Message	Value	Expected
warning	model_behavior	`MOST_LIKELY_IS_MODE`	Headline score is not the modal scoreline (modal=1-1 (p=0.1105), shown=1-0 (p=0.1048)); the published exact score must be the grid mode, and the winning scenario belongs in the labelled most_likely_winning_score_home/away fields. Both lean Chelsea.	1-0	1-1
warning	provenance	`FBREF_PRIORS_ACTIVE`	FBref dataset files not found; team priors inactive.

Lambda evolution (audit trail)

Stage	Home λ	Away λ
Pre-competitive-floor	1.789	1.042
Pre-calibration	1.727	1.081
Final	1.789	1.042

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

Top scorelines	P
1-1	11.1%
1-0	10.5%
2-1	9.8%
2-0	9.4%
0-1	6.1%

Outcome — home 54.8% · draw 23.3% · away 21.9%

Markets — O2.5 53.8% · BTTS 54.0%

Passed checks (38)

Sev	Category	Rule	Message
good	integrity	`SCHEMA_KEYS_PRESENT`	All required sections present.
good	structure	`LINEUP_ELEVEN`	Chelsea: 11 starters.
good	structure	`FORMATION_VALID`	Chelsea: formation 3-4-2-1 valid.
good	structure	`LINEUP_ELEVEN`	Tottenham: 11 starters.
good	structure	`FORMATION_VALID`	Tottenham: formation 4-2-3-1 valid.
good	probability	`OUTCOME_SUMS_ONE`	outcome sums to 1.0000.
good	probability	`MARKETS_COMPLEMENTARY`	Over/Under and BTTS pairs each sum to 1.
good	probability	`SCORELINE_PROBS_NONNEG`	All scoreline probabilities within [0,1].
good	probability	`GRID_SUMS_ONE`	scoreline grid sums to 0.9973.
good	model_behavior	`GRID_REPRODUCIBLE`	Grid reproducible from lambdas (max cell diff 0.0007).
good	probability	`MLS_PROBABILITY_MATCHES`	Headline probability matches the grid.
good	model_behavior	`OUTCOME_MATCHES_GRID`	Outcome split matches the grid (max diff 0.0017).
good	model_behavior	`MARKETS_MATCH_GRID`	Markets match the grid (max diff 0.0026).
good	integrity	`WINNER_IS_VALID_TEAM`	predicted_winner 'Chelsea' is a participating side.
good	model_behavior	`WINNER_MATCHES_SCORE`	Winner agrees with headline score (Chelsea).
good	model_behavior	`WINNER_PROBABILITY_DIRECTION`	predicted_winner 'Chelsea' follows the published modal scoreline 1-0.
good	model_behavior	`LAMBDA_BOUNDS`	Lambdas in range (home 1.789, away 1.042).
good	provenance	`LAMBDA_FACTORS_PRESENT`	lambda_factors audit trail present.
good	model_behavior	`LAMBDA_CONSISTENCY`	Lambda identical across prediction / antagonist / factors.
good	model_behavior	`CALIBRATION_SCALE_BOUNDS`	Calibration scale 1.0 in range.
good	provenance	`XGBOOST_REASONING_PRESENT`	Match-level xgboost_reasoning present (XGBoost).
good	model_behavior	`LAMBDA_REASONING_CONSISTENCY`	Lambdas in xgboost_reasoning match final Poisson lambdas (1.789 / 1.042).
good	provenance	`XGBOOST_PLAYER_REASONING`	All 22 starters carry full XGBoost / FBref reasoning.
good	integrity	`TEAM_GOALS_SUM`	Chelsea: player goals reconcile to team total (1).
good	integrity	`TEAM_GOALS_SUM`	Tottenham: player goals reconcile to team total (0).
good	structure	`POSSESSION_SUMS_100`	Possession sums to 100.0%.
good	integrity	`PLAYER_STAT_REALISM`	All 22 projected player lines pass realism checks.
good	structure	`EXACTLY_ONE_GOALKEEPER`	Chelsea: exactly one goalkeeper (Mike Penders).
good	integrity	`SHOT_MATH_INTEGRITY`	All 11 projected lines have consistent shot components.
good	integrity	`TEAM_SHOTS_CAP`	Chelsea: 16 projected shots inside the realistic envelope.
good	structure	`EXACTLY_ONE_GOALKEEPER`	Tottenham: exactly one goalkeeper (Antonín Kinský).
good	integrity	`SHOT_MATH_INTEGRITY`	All 11 projected lines have consistent shot components.
good	integrity	`TEAM_SHOTS_CAP`	Tottenham: 15 projected shots inside the realistic envelope.
good	provenance	`STANDING_AVAILABLE`	League-standing context available for both sides.
good	provenance	`RANKING_PROVENANCE`	22/22 starters carry a ranking source.
good	provenance	`DATA_DEPTH`	21/22 starters have adequate recent history.
good	provenance	`PREGAME_DIRECTION_AGREES`	Pregame favorite agrees with the model winner (Chelsea).
good	provenance	`PREGAME_TOTAL_CONSISTENT`	Pregame total 2.83 ~ Poisson total 2.83.

Atlético Madrid vs Deportivo

20261024_COR@ATM · LaLiga

WARNINGS risk 10 pred 2-0 (Atlético Madrid)λ 2.90 / 0.66conf 0.84 warning 1 good 39

Problems (1)

Sev	Category	Rule	Message	Value	Expected
warning	provenance	`FBREF_PRIORS_ACTIVE`	FBref dataset files not found; team priors inactive.

Lambda evolution (audit trail)

Stage	Home λ	Away λ
Pre-mismatch	2.790	0.768
Pre-calibration	2.657	0.808
Final	2.903	0.655

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

Top scorelines	P
2-0	12.0%
3-0	11.6%
4-0	8.4%
1-0	8.3%
2-1	7.9%

Outcome — home 82.9% · draw 11.5% · away 5.6%

Markets — O2.5 69.0% · BTTS 45.4%

Passed checks (39)

Sev	Category	Rule	Message
good	integrity	`SCHEMA_KEYS_PRESENT`	All required sections present.
good	structure	`LINEUP_ELEVEN`	Atlético Madrid: 11 starters.
good	structure	`FORMATION_VALID`	Atlético Madrid: formation 4-2-3-1 valid.
good	structure	`LINEUP_ELEVEN`	Deportivo: 11 starters.
good	structure	`FORMATION_VALID`	Deportivo: formation 4-2-3-1 valid.
good	probability	`OUTCOME_SUMS_ONE`	outcome sums to 1.0000.
good	probability	`MARKETS_COMPLEMENTARY`	Over/Under and BTTS pairs each sum to 1.
good	probability	`SCORELINE_PROBS_NONNEG`	All scoreline probabilities within [0,1].
good	probability	`GRID_SUMS_ONE`	scoreline grid sums to 0.9975.
good	model_behavior	`GRID_REPRODUCIBLE`	Grid reproducible from lambdas (max cell diff 0.0002).
good	model_behavior	`MOST_LIKELY_IS_MODE`	Headline score 2-0 is the modal scoreline.
good	probability	`MLS_PROBABILITY_MATCHES`	Headline probability matches the grid.
good	model_behavior	`OUTCOME_MATCHES_GRID`	Outcome split matches the grid (max diff 0.0016).
good	model_behavior	`MARKETS_MATCH_GRID`	Markets match the grid (max diff 0.0025).
good	integrity	`WINNER_IS_VALID_TEAM`	predicted_winner 'Atlético Madrid' is a participating side.
good	model_behavior	`WINNER_MATCHES_SCORE`	Winner agrees with headline score (Atlético Madrid).
good	model_behavior	`WINNER_PROBABILITY_DIRECTION`	predicted_winner 'Atlético Madrid' follows the published modal scoreline 2-0.
good	model_behavior	`LAMBDA_BOUNDS`	Lambdas in range (home 2.903, away 0.655).
good	provenance	`LAMBDA_FACTORS_PRESENT`	lambda_factors audit trail present.
good	model_behavior	`LAMBDA_CONSISTENCY`	Lambda identical across prediction / antagonist / factors.
good	model_behavior	`CALIBRATION_SCALE_BOUNDS`	Calibration scale 1.0 in range.
good	provenance	`XGBOOST_REASONING_PRESENT`	Match-level xgboost_reasoning present (XGBoost).
good	model_behavior	`LAMBDA_REASONING_CONSISTENCY`	Lambdas in xgboost_reasoning match final Poisson lambdas (2.903 / 0.655).
good	provenance	`XGBOOST_PLAYER_REASONING`	All 22 starters carry full XGBoost / FBref reasoning.
good	integrity	`TEAM_GOALS_SUM`	Atlético Madrid: player goals reconcile to team total (2).
good	integrity	`TEAM_GOALS_SUM`	Deportivo: player goals reconcile to team total (0).
good	structure	`POSSESSION_SUMS_100`	Possession sums to 100.0%.
good	integrity	`PLAYER_STAT_REALISM`	All 22 projected player lines pass realism checks.
good	structure	`EXACTLY_ONE_GOALKEEPER`	Atlético Madrid: exactly one goalkeeper (Jan Oblak).
good	integrity	`SHOT_MATH_INTEGRITY`	All 11 projected lines have consistent shot components.
good	integrity	`TEAM_SHOTS_CAP`	Atlético Madrid: 21 projected shots inside the realistic envelope.
good	structure	`EXACTLY_ONE_GOALKEEPER`	Deportivo: exactly one goalkeeper (Leo Román).
good	integrity	`SHOT_MATH_INTEGRITY`	All 11 projected lines have consistent shot components.
good	integrity	`TEAM_SHOTS_CAP`	Deportivo: 10 projected shots inside the realistic envelope.
good	provenance	`STANDING_AVAILABLE`	League-standing context available for both sides.
good	provenance	`RANKING_PROVENANCE`	22/22 starters carry a ranking source.
good	provenance	`DATA_DEPTH`	22/22 starters have adequate recent history.
good	provenance	`PREGAME_DIRECTION_AGREES`	Pregame favorite agrees with the model winner (Atlético Madrid).
good	provenance	`PREGAME_TOTAL_CONSISTENT`	Pregame total 3.56 ~ Poisson total 3.56.

Racing Club vs Espanyol

20261024_ESP@SAN · LaLiga

WARNINGS risk 10 pred 1-1 (Draw)λ 1.31 / 1.46conf 0.7 warning 1 good 37

Problems (1)

Sev	Category	Rule	Message	Value	Expected
warning	provenance	`FBREF_PRIORS_ACTIVE`	FBref dataset files not found; team priors inactive.

Lambda evolution (audit trail)

Stage	Home λ	Away λ
Pre-competitive-floor	1.312	1.460
Pre-calibration	1.303	1.470
Final	1.312	1.460

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

Top scorelines	P
1-1	12.1%
0-1	9.0%
1-2	8.7%
1-0	8.1%
2-1	7.9%

Outcome — home 33.8% · draw 25.6% · away 40.6%

Markets — O2.5 52.4% · BTTS 56.2%

Passed checks (37)

Sev	Category	Rule	Message
good	integrity	`SCHEMA_KEYS_PRESENT`	All required sections present.
good	structure	`LINEUP_ELEVEN`	Racing Club: 11 starters.
good	structure	`FORMATION_VALID`	Racing Club: formation 4-2-3-1 valid.
good	structure	`LINEUP_ELEVEN`	Espanyol: 11 starters.
good	structure	`FORMATION_VALID`	Espanyol: formation 5-4-1 valid.
good	probability	`OUTCOME_SUMS_ONE`	outcome sums to 1.0000.
good	probability	`MARKETS_COMPLEMENTARY`	Over/Under and BTTS pairs each sum to 1.
good	probability	`SCORELINE_PROBS_NONNEG`	All scoreline probabilities within [0,1].
good	probability	`GRID_SUMS_ONE`	scoreline grid sums to 0.9962.
good	model_behavior	`GRID_REPRODUCIBLE`	Grid reproducible from lambdas (max cell diff 0.0012).
good	model_behavior	`MOST_LIKELY_IS_MODE`	Headline score 1-1 is the modal scoreline.
good	probability	`MLS_PROBABILITY_MATCHES`	Headline probability matches the grid.
good	model_behavior	`OUTCOME_MATCHES_GRID`	Outcome split matches the grid (max diff 0.0020).
good	model_behavior	`MARKETS_MATCH_GRID`	Markets match the grid (max diff 0.0038).
good	integrity	`WINNER_IS_VALID_TEAM`	predicted_winner 'Draw' is a participating side.
good	model_behavior	`WINNER_PROBABILITY_DIRECTION`	predicted_winner 'Draw' follows the published modal scoreline 1-1. It differs from the calibrated 1X2 favorite 'Espanyol', which is published separately and does not decide the call.
good	model_behavior	`LAMBDA_BOUNDS`	Lambdas in range (home 1.312, away 1.46).
good	provenance	`LAMBDA_FACTORS_PRESENT`	lambda_factors audit trail present.
good	model_behavior	`LAMBDA_CONSISTENCY`	Lambda identical across prediction / antagonist / factors.
good	model_behavior	`CALIBRATION_SCALE_BOUNDS`	Calibration scale 1.0 in range.
good	provenance	`XGBOOST_REASONING_PRESENT`	Match-level xgboost_reasoning present (XGBoost).
good	model_behavior	`LAMBDA_REASONING_CONSISTENCY`	Lambdas in xgboost_reasoning match final Poisson lambdas (1.312 / 1.46).
good	provenance	`XGBOOST_PLAYER_REASONING`	All 22 starters carry full XGBoost / FBref reasoning.
good	integrity	`TEAM_GOALS_SUM`	Racing Club: player goals reconcile to team total (1).
good	integrity	`TEAM_GOALS_SUM`	Espanyol: player goals reconcile to team total (1).
good	structure	`POSSESSION_SUMS_100`	Possession sums to 100.0%.
good	integrity	`PLAYER_STAT_REALISM`	All 22 projected player lines pass realism checks.
good	structure	`EXACTLY_ONE_GOALKEEPER`	Racing Club: exactly one goalkeeper (Julen Agirrezabala).
good	integrity	`SHOT_MATH_INTEGRITY`	All 11 projected lines have consistent shot components.
good	integrity	`TEAM_SHOTS_CAP`	Racing Club: 14 projected shots inside the realistic envelope.
good	structure	`EXACTLY_ONE_GOALKEEPER`	Espanyol: exactly one goalkeeper (Ángel Fortuño).
good	integrity	`SHOT_MATH_INTEGRITY`	All 11 projected lines have consistent shot components.
good	integrity	`TEAM_SHOTS_CAP`	Espanyol: 10 projected shots inside the realistic envelope.
good	provenance	`STANDING_AVAILABLE`	League-standing context available for both sides.
good	provenance	`RANKING_PROVENANCE`	22/22 starters carry a ranking source.
good	provenance	`DATA_DEPTH`	22/22 starters have adequate recent history.
good	provenance	`PREGAME_TOTAL_CONSISTENT`	Pregame total 2.77 ~ Poisson total 2.77.

Arsenal vs Everton

20261024_EVE@ARS · Premier League

WARNINGS risk 10 pred 1-0 (Arsenal)λ 1.85 / 1.02conf 0.76 warning 1 good 39

Problems (1)

Sev	Category	Rule	Message	Value	Expected
warning	model_behavior	`MOST_LIKELY_IS_MODE`	Headline score is not the modal scoreline (modal=1-1 (p=0.1074), shown=1-0 (p=0.1045)); the published exact score must be the grid mode, and the winning scenario belongs in the labelled most_likely_winning_score_home/away fields. Both lean Arsenal.	1-0	1-1

Lambda evolution (audit trail)

Stage	Home λ	Away λ
Pre-mismatch	1.785	1.085
Pre-competitive-floor	1.937	1.062
Pre-calibration	1.712	1.134
Final	1.854	1.016

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

Top scorelines	P
1-1	10.7%
1-0	10.4%
2-1	9.9%
2-0	9.7%
3-1	6.1%

Outcome — home 56.9% · draw 22.7% · away 20.4%

Markets — O2.5 54.7% · BTTS 53.9%

Passed checks (39)

Sev	Category	Rule	Message
good	integrity	`SCHEMA_KEYS_PRESENT`	All required sections present.
good	structure	`LINEUP_ELEVEN`	Arsenal: 11 starters.
good	structure	`FORMATION_VALID`	Arsenal: formation 4-2-3-1 valid.
good	structure	`LINEUP_ELEVEN`	Everton: 11 starters.
good	structure	`FORMATION_VALID`	Everton: formation 4-2-3-1 valid.
good	probability	`OUTCOME_SUMS_ONE`	outcome sums to 1.0000.
good	probability	`MARKETS_COMPLEMENTARY`	Over/Under and BTTS pairs each sum to 1.
good	probability	`SCORELINE_PROBS_NONNEG`	All scoreline probabilities within [0,1].
good	probability	`GRID_SUMS_ONE`	scoreline grid sums to 0.9970.
good	model_behavior	`GRID_REPRODUCIBLE`	Grid reproducible from lambdas (max cell diff 0.0006).
good	probability	`MLS_PROBABILITY_MATCHES`	Headline probability matches the grid.
good	model_behavior	`OUTCOME_MATCHES_GRID`	Outcome split matches the grid (max diff 0.0019).
good	model_behavior	`MARKETS_MATCH_GRID`	Markets match the grid (max diff 0.0030).
good	integrity	`WINNER_IS_VALID_TEAM`	predicted_winner 'Arsenal' is a participating side.
good	model_behavior	`WINNER_MATCHES_SCORE`	Winner agrees with headline score (Arsenal).
good	model_behavior	`WINNER_PROBABILITY_DIRECTION`	predicted_winner 'Arsenal' follows the published modal scoreline 1-0.
good	model_behavior	`LAMBDA_BOUNDS`	Lambdas in range (home 1.854, away 1.016).
good	provenance	`LAMBDA_FACTORS_PRESENT`	lambda_factors audit trail present.
good	model_behavior	`LAMBDA_CONSISTENCY`	Lambda identical across prediction / antagonist / factors.
good	model_behavior	`CALIBRATION_SCALE_BOUNDS`	Calibration scale 1.0 in range.
good	provenance	`XGBOOST_REASONING_PRESENT`	Match-level xgboost_reasoning present (XGBoost).
good	model_behavior	`LAMBDA_REASONING_CONSISTENCY`	Lambdas in xgboost_reasoning match final Poisson lambdas (1.854 / 1.016).
good	provenance	`FBREF_PRIORS_ACTIVE`	FBref team priors applied (blend weight 0.192, home E[lambda]=2.069, away E[lambda]=2.069).
good	provenance	`XGBOOST_PLAYER_REASONING`	All 22 starters carry full XGBoost / FBref reasoning.
good	integrity	`TEAM_GOALS_SUM`	Arsenal: player goals reconcile to team total (1).
good	integrity	`TEAM_GOALS_SUM`	Everton: player goals reconcile to team total (0).
good	structure	`POSSESSION_SUMS_100`	Possession sums to 100.0%.
good	integrity	`PLAYER_STAT_REALISM`	All 22 projected player lines pass realism checks.
good	structure	`EXACTLY_ONE_GOALKEEPER`	Arsenal: exactly one goalkeeper (David Raya).
good	integrity	`SHOT_MATH_INTEGRITY`	All 11 projected lines have consistent shot components.
good	integrity	`TEAM_SHOTS_CAP`	Arsenal: 13 projected shots inside the realistic envelope.
good	structure	`EXACTLY_ONE_GOALKEEPER`	Everton: exactly one goalkeeper (Jordan Pickford).
good	integrity	`SHOT_MATH_INTEGRITY`	All 11 projected lines have consistent shot components.
good	integrity	`TEAM_SHOTS_CAP`	Everton: 13 projected shots inside the realistic envelope.
good	provenance	`STANDING_AVAILABLE`	League-standing context available for both sides.
good	provenance	`RANKING_PROVENANCE`	22/22 starters carry a ranking source.
good	provenance	`DATA_DEPTH`	22/22 starters have adequate recent history.
good	provenance	`PREGAME_DIRECTION_AGREES`	Pregame favorite agrees with the model winner (Arsenal).
good	provenance	`PREGAME_TOTAL_CONSISTENT`	Pregame total 2.87 ~ Poisson total 2.87.

Coventry vs Fulham

20261024_FUL@COC · Premier League

WARNINGS risk 10 pred 1-1 (Draw)λ 1.20 / 1.30conf 0.7 warning 1 good 37

Problems (1)

Sev	Category	Rule	Message	Value	Expected
warning	provenance	`FBREF_PRIORS_ACTIVE`	FBref dataset files not found; team priors inactive.

Lambda evolution (audit trail)

Stage	Home λ	Away λ
Pre-competitive-floor	1.201	1.297
Pre-calibration	1.162	1.342
Final	1.201	1.297

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

Top scorelines	P
1-1	12.9%
0-1	10.5%
1-0	9.8%
0-0	8.3%
1-2	8.3%

Outcome — home 34.1% · draw 27.2% · away 38.7%

Markets — O2.5 45.6% · BTTS 50.9%

Passed checks (37)

Sev	Category	Rule	Message
good	integrity	`SCHEMA_KEYS_PRESENT`	All required sections present.
good	structure	`LINEUP_ELEVEN`	Coventry: 11 starters.
good	structure	`FORMATION_VALID`	Coventry: formation 4-2-3-1 valid.
good	structure	`LINEUP_ELEVEN`	Fulham: 11 starters.
good	structure	`FORMATION_VALID`	Fulham: formation 4-2-3-1 valid.
good	probability	`OUTCOME_SUMS_ONE`	outcome sums to 1.0000.
good	probability	`MARKETS_COMPLEMENTARY`	Over/Under and BTTS pairs each sum to 1.
good	probability	`SCORELINE_PROBS_NONNEG`	All scoreline probabilities within [0,1].
good	probability	`GRID_SUMS_ONE`	scoreline grid sums to 0.9966.
good	model_behavior	`GRID_REPRODUCIBLE`	Grid reproducible from lambdas (max cell diff 0.0013).
good	model_behavior	`MOST_LIKELY_IS_MODE`	Headline score 1-1 is the modal scoreline.
good	probability	`MLS_PROBABILITY_MATCHES`	Headline probability matches the grid.
good	model_behavior	`OUTCOME_MATCHES_GRID`	Outcome split matches the grid (max diff 0.0020).
good	model_behavior	`MARKETS_MATCH_GRID`	Markets match the grid (max diff 0.0034).
good	integrity	`WINNER_IS_VALID_TEAM`	predicted_winner 'Draw' is a participating side.
good	model_behavior	`WINNER_PROBABILITY_DIRECTION`	predicted_winner 'Draw' follows the published modal scoreline 1-1. It differs from the calibrated 1X2 favorite 'Fulham', which is published separately and does not decide the call.
good	model_behavior	`LAMBDA_BOUNDS`	Lambdas in range (home 1.201, away 1.297).
good	provenance	`LAMBDA_FACTORS_PRESENT`	lambda_factors audit trail present.
good	model_behavior	`LAMBDA_CONSISTENCY`	Lambda identical across prediction / antagonist / factors.
good	model_behavior	`CALIBRATION_SCALE_BOUNDS`	Calibration scale 1.0 in range.
good	provenance	`XGBOOST_REASONING_PRESENT`	Match-level xgboost_reasoning present (XGBoost).
good	model_behavior	`LAMBDA_REASONING_CONSISTENCY`	Lambdas in xgboost_reasoning match final Poisson lambdas (1.201 / 1.297).
good	provenance	`XGBOOST_PLAYER_REASONING`	All 22 starters carry full XGBoost / FBref reasoning.
good	integrity	`TEAM_GOALS_SUM`	Coventry: player goals reconcile to team total (1).
good	integrity	`TEAM_GOALS_SUM`	Fulham: player goals reconcile to team total (1).
good	structure	`POSSESSION_SUMS_100`	Possession sums to 100.0%.
good	integrity	`PLAYER_STAT_REALISM`	All 22 projected player lines pass realism checks.
good	structure	`EXACTLY_ONE_GOALKEEPER`	Coventry: exactly one goalkeeper (Carl Rushworth).
good	integrity	`SHOT_MATH_INTEGRITY`	All 11 projected lines have consistent shot components.
good	integrity	`TEAM_SHOTS_CAP`	Coventry: 10 projected shots inside the realistic envelope.
good	structure	`EXACTLY_ONE_GOALKEEPER`	Fulham: exactly one goalkeeper (Bernd Leno).
good	integrity	`SHOT_MATH_INTEGRITY`	All 11 projected lines have consistent shot components.
good	integrity	`TEAM_SHOTS_CAP`	Fulham: 16 projected shots inside the realistic envelope.
good	provenance	`STANDING_AVAILABLE`	League-standing context available for both sides.
good	provenance	`RANKING_PROVENANCE`	22/22 starters carry a ranking source.
good	provenance	`DATA_DEPTH`	22/22 starters have adequate recent history.
good	provenance	`PREGAME_TOTAL_CONSISTENT`	Pregame total 2.50 ~ Poisson total 2.50.

Aston Villa vs Man City

20261024_MCI@AVL · Premier League

WARNINGS risk 10 pred 0-2 (Man City)λ 0.89 / 2.62conf 0.8 warning 1 good 39

Problems (1)

Sev	Category	Rule	Message	Value	Expected
warning	provenance	`FBREF_PRIORS_ACTIVE`	FBref dataset files not found; team priors inactive.

Lambda evolution (audit trail)

Stage	Home λ	Away λ
Pre-mismatch	0.993	2.516
Pre-calibration	1.031	2.271
Final	0.890	2.619

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

Top scorelines	P
0-2	10.3%
1-2	9.1%
0-3	9.0%
1-3	8.0%
0-1	7.8%

Outcome — home 10.4% · draw 15.3% · away 74.3%

Markets — O2.5 68.1% · BTTS 54.7%

Passed checks (39)

Sev	Category	Rule	Message
good	integrity	`SCHEMA_KEYS_PRESENT`	All required sections present.
good	structure	`LINEUP_ELEVEN`	Aston Villa: 11 starters.
good	structure	`FORMATION_VALID`	Aston Villa: formation 4-2-3-1 valid.
good	structure	`LINEUP_ELEVEN`	Man City: 11 starters.
good	structure	`FORMATION_VALID`	Man City: formation 4-2-3-1 valid.
good	probability	`OUTCOME_SUMS_ONE`	outcome sums to 1.0000.
good	probability	`MARKETS_COMPLEMENTARY`	Over/Under and BTTS pairs each sum to 1.
good	probability	`SCORELINE_PROBS_NONNEG`	All scoreline probabilities within [0,1].
good	probability	`GRID_SUMS_ONE`	scoreline grid sums to 0.9959.
good	model_behavior	`GRID_REPRODUCIBLE`	Grid reproducible from lambdas (max cell diff 0.0003).
good	model_behavior	`MOST_LIKELY_IS_MODE`	Headline score 0-2 is the modal scoreline.
good	probability	`MLS_PROBABILITY_MATCHES`	Headline probability matches the grid.
good	model_behavior	`OUTCOME_MATCHES_GRID`	Outcome split matches the grid (max diff 0.0020).
good	model_behavior	`MARKETS_MATCH_GRID`	Markets match the grid (max diff 0.0042).
good	integrity	`WINNER_IS_VALID_TEAM`	predicted_winner 'Man City' is a participating side.
good	model_behavior	`WINNER_MATCHES_SCORE`	Winner agrees with headline score (Man City).
good	model_behavior	`WINNER_PROBABILITY_DIRECTION`	predicted_winner 'Man City' follows the published modal scoreline 0-2.
good	model_behavior	`LAMBDA_BOUNDS`	Lambdas in range (home 0.89, away 2.619).
good	provenance	`LAMBDA_FACTORS_PRESENT`	lambda_factors audit trail present.
good	model_behavior	`LAMBDA_CONSISTENCY`	Lambda identical across prediction / antagonist / factors.
good	model_behavior	`CALIBRATION_SCALE_BOUNDS`	Calibration scale 1.0 in range.
good	provenance	`XGBOOST_REASONING_PRESENT`	Match-level xgboost_reasoning present (XGBoost).
good	model_behavior	`LAMBDA_REASONING_CONSISTENCY`	Lambdas in xgboost_reasoning match final Poisson lambdas (0.89 / 2.619).
good	provenance	`XGBOOST_PLAYER_REASONING`	All 22 starters carry full XGBoost / FBref reasoning.
good	integrity	`TEAM_GOALS_SUM`	Aston Villa: player goals reconcile to team total (0).
good	integrity	`TEAM_GOALS_SUM`	Man City: player goals reconcile to team total (2).
good	structure	`POSSESSION_SUMS_100`	Possession sums to 100.0%.
good	integrity	`PLAYER_STAT_REALISM`	All 22 projected player lines pass realism checks.
good	structure	`EXACTLY_ONE_GOALKEEPER`	Aston Villa: exactly one goalkeeper (Zion Suzuki).
good	integrity	`SHOT_MATH_INTEGRITY`	All 11 projected lines have consistent shot components.
good	integrity	`TEAM_SHOTS_CAP`	Aston Villa: 14 projected shots inside the realistic envelope.
good	structure	`EXACTLY_ONE_GOALKEEPER`	Man City: exactly one goalkeeper (Gianluigi Donnarumma).
good	integrity	`SHOT_MATH_INTEGRITY`	All 11 projected lines have consistent shot components.
good	integrity	`TEAM_SHOTS_CAP`	Man City: 19 projected shots inside the realistic envelope.
good	provenance	`STANDING_AVAILABLE`	League-standing context available for both sides.
good	provenance	`RANKING_PROVENANCE`	22/22 starters carry a ranking source.
good	provenance	`DATA_DEPTH`	22/22 starters have adequate recent history.
good	provenance	`PREGAME_DIRECTION_AGREES`	Pregame favorite agrees with the model winner (Man City).
good	provenance	`PREGAME_TOTAL_CONSISTENT`	Pregame total 3.51 ~ Poisson total 3.51.

Rayo Vallecano vs Elche

20261024_ELC@RAY · LaLiga

CLEAN risk 0 pred 1-1 (Draw)λ 1.29 / 1.11conf 0.7 good 38

Problems (0)

none

Lambda evolution (audit trail)

Stage	Home λ	Away λ
Pre-competitive-floor	1.286	1.111
Pre-calibration	1.286	1.111
Final	1.286	1.111

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

Top scorelines	P
1-1	13.1%
1-0	11.6%
0-1	10.0%
0-0	9.2%
2-1	8.4%

Outcome — home 40.4% · draw 27.8% · away 31.9%

Markets — O2.5 43.0% · BTTS 48.7%

Passed checks (38)

Sev	Category	Rule	Message
good	integrity	`SCHEMA_KEYS_PRESENT`	All required sections present.
good	structure	`LINEUP_ELEVEN`	Rayo Vallecano: 11 starters.
good	structure	`FORMATION_VALID`	Rayo Vallecano: formation 4-2-3-1 valid.
good	structure	`LINEUP_ELEVEN`	Elche: 11 starters.
good	structure	`FORMATION_VALID`	Elche: formation 4-2-3-1 valid.
good	probability	`OUTCOME_SUMS_ONE`	outcome sums to 1.0000.
good	probability	`MARKETS_COMPLEMENTARY`	Over/Under and BTTS pairs each sum to 1.
good	probability	`SCORELINE_PROBS_NONNEG`	All scoreline probabilities within [0,1].
good	probability	`GRID_SUMS_ONE`	scoreline grid sums to 0.9974.
good	model_behavior	`GRID_REPRODUCIBLE`	Grid reproducible from lambdas (max cell diff 0.0012).
good	model_behavior	`MOST_LIKELY_IS_MODE`	Headline score 1-1 is the modal scoreline.
good	probability	`MLS_PROBABILITY_MATCHES`	Headline probability matches the grid.
good	model_behavior	`OUTCOME_MATCHES_GRID`	Outcome split matches the grid (max diff 0.0015).
good	model_behavior	`MARKETS_MATCH_GRID`	Markets match the grid (max diff 0.0026).
good	integrity	`WINNER_IS_VALID_TEAM`	predicted_winner 'Draw' is a participating side.
good	model_behavior	`WINNER_PROBABILITY_DIRECTION`	predicted_winner 'Draw' follows the published modal scoreline 1-1. It differs from the calibrated 1X2 favorite 'Rayo Vallecano', which is published separately and does not decide the call.
good	model_behavior	`LAMBDA_BOUNDS`	Lambdas in range (home 1.286, away 1.111).
good	provenance	`LAMBDA_FACTORS_PRESENT`	lambda_factors audit trail present.
good	model_behavior	`LAMBDA_CONSISTENCY`	Lambda identical across prediction / antagonist / factors.
good	model_behavior	`CALIBRATION_SCALE_BOUNDS`	Calibration scale 1.0 in range.
good	provenance	`XGBOOST_REASONING_PRESENT`	Match-level xgboost_reasoning present (XGBoost).
good	model_behavior	`LAMBDA_REASONING_CONSISTENCY`	Lambdas in xgboost_reasoning match final Poisson lambdas (1.286 / 1.111).
good	provenance	`FBREF_PRIORS_ACTIVE`	FBref team priors applied (blend weight 0.228, home E[lambda]=0.988, away E[lambda]=0.985).
good	provenance	`XGBOOST_PLAYER_REASONING`	All 22 starters carry full XGBoost / FBref reasoning.
good	integrity	`TEAM_GOALS_SUM`	Rayo Vallecano: player goals reconcile to team total (1).
good	integrity	`TEAM_GOALS_SUM`	Elche: player goals reconcile to team total (1).
good	structure	`POSSESSION_SUMS_100`	Possession sums to 100.0%.
good	integrity	`PLAYER_STAT_REALISM`	All 22 projected player lines pass realism checks.
good	structure	`EXACTLY_ONE_GOALKEEPER`	Rayo Vallecano: exactly one goalkeeper (Emil Audero).
good	integrity	`SHOT_MATH_INTEGRITY`	All 11 projected lines have consistent shot components.
good	integrity	`TEAM_SHOTS_CAP`	Rayo Vallecano: 13 projected shots inside the realistic envelope.
good	structure	`EXACTLY_ONE_GOALKEEPER`	Elche: exactly one goalkeeper (Matías Dituro).
good	integrity	`SHOT_MATH_INTEGRITY`	All 11 projected lines have consistent shot components.
good	integrity	`TEAM_SHOTS_CAP`	Elche: 13 projected shots inside the realistic envelope.
good	provenance	`STANDING_AVAILABLE`	League-standing context available for both sides.
good	provenance	`RANKING_PROVENANCE`	22/22 starters carry a ranking source.
good	provenance	`DATA_DEPTH`	22/22 starters have adequate recent history.
good	provenance	`PREGAME_TOTAL_CONSISTENT`	Pregame total 2.40 ~ Poisson total 2.40.

Valencia vs Villarreal

20261024_VIL@VAL · LaLiga

CLEAN risk 0 pred 0-1 (Villarreal)λ 0.90 / 1.45conf 0.74 good 40

Problems (0)

none

Lambda evolution (audit trail)

Stage	Home λ	Away λ
Pre-competitive-floor	0.899	1.451
Pre-calibration	0.919	1.421
Final	0.899	1.451

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

Top scorelines	P
0-1	13.7%
1-1	12.5%
0-2	10.0%
0-0	9.6%
1-2	9.0%

Outcome — home 23.2% · draw 26.9% · away 49.9%

Markets — O2.5 41.7% · BTTS 45.5%

Passed checks (40)

Sev	Category	Rule	Message
good	integrity	`SCHEMA_KEYS_PRESENT`	All required sections present.
good	structure	`LINEUP_ELEVEN`	Valencia: 11 starters.
good	structure	`FORMATION_VALID`	Valencia: formation 4-2-3-1 valid.
good	structure	`LINEUP_ELEVEN`	Villarreal: 11 starters.
good	structure	`FORMATION_VALID`	Villarreal: formation 4-4-2 valid.
good	probability	`OUTCOME_SUMS_ONE`	outcome sums to 0.9999.
good	probability	`MARKETS_COMPLEMENTARY`	Over/Under and BTTS pairs each sum to 1.
good	probability	`SCORELINE_PROBS_NONNEG`	All scoreline probabilities within [0,1].
good	probability	`GRID_SUMS_ONE`	scoreline grid sums to 0.9963.
good	model_behavior	`GRID_REPRODUCIBLE`	Grid reproducible from lambdas (max cell diff 0.0010).
good	model_behavior	`MOST_LIKELY_IS_MODE`	Headline score 0-1 is the modal scoreline.
good	probability	`MLS_PROBABILITY_MATCHES`	Headline probability matches the grid.
good	model_behavior	`OUTCOME_MATCHES_GRID`	Outcome split matches the grid (max diff 0.0016).
good	model_behavior	`MARKETS_MATCH_GRID`	Markets match the grid (max diff 0.0037).
good	integrity	`WINNER_IS_VALID_TEAM`	predicted_winner 'Villarreal' is a participating side.
good	model_behavior	`WINNER_MATCHES_SCORE`	Winner agrees with headline score (Villarreal).
good	model_behavior	`WINNER_PROBABILITY_DIRECTION`	predicted_winner 'Villarreal' follows the published modal scoreline 0-1.
good	model_behavior	`LAMBDA_BOUNDS`	Lambdas in range (home 0.899, away 1.451).
good	provenance	`LAMBDA_FACTORS_PRESENT`	lambda_factors audit trail present.
good	model_behavior	`LAMBDA_CONSISTENCY`	Lambda identical across prediction / antagonist / factors.
good	model_behavior	`CALIBRATION_SCALE_BOUNDS`	Calibration scale 1.0 in range.
good	provenance	`XGBOOST_REASONING_PRESENT`	Match-level xgboost_reasoning present (XGBoost).
good	model_behavior	`LAMBDA_REASONING_CONSISTENCY`	Lambdas in xgboost_reasoning match final Poisson lambdas (0.899 / 1.451).
good	provenance	`FBREF_PRIORS_ACTIVE`	FBref team priors applied (blend weight 0.228, home E[lambda]=0.328, away E[lambda]=0.326).
good	provenance	`XGBOOST_PLAYER_REASONING`	All 22 starters carry full XGBoost / FBref reasoning.
good	integrity	`TEAM_GOALS_SUM`	Valencia: player goals reconcile to team total (0).
good	integrity	`TEAM_GOALS_SUM`	Villarreal: player goals reconcile to team total (1).
good	structure	`POSSESSION_SUMS_100`	Possession sums to 100.0%.
good	integrity	`PLAYER_STAT_REALISM`	All 22 projected player lines pass realism checks.
good	structure	`EXACTLY_ONE_GOALKEEPER`	Valencia: exactly one goalkeeper (Stole Dimitrievski).
good	integrity	`SHOT_MATH_INTEGRITY`	All 11 projected lines have consistent shot components.
good	integrity	`TEAM_SHOTS_CAP`	Valencia: 14 projected shots inside the realistic envelope.
good	structure	`EXACTLY_ONE_GOALKEEPER`	Villarreal: exactly one goalkeeper (Péter Gulácsi).
good	integrity	`SHOT_MATH_INTEGRITY`	All 11 projected lines have consistent shot components.
good	integrity	`TEAM_SHOTS_CAP`	Villarreal: 16 projected shots inside the realistic envelope.
good	provenance	`STANDING_AVAILABLE`	League-standing context available for both sides.
good	provenance	`RANKING_PROVENANCE`	22/22 starters carry a ranking source.
good	provenance	`DATA_DEPTH`	22/22 starters have adequate recent history.
good	provenance	`PREGAME_DIRECTION_AGREES`	Pregame favorite agrees with the model winner (Villarreal).
good	provenance	`PREGAME_TOTAL_CONSISTENT`	Pregame total 2.35 ~ Poisson total 2.35.

Rule catalogue

Rule	Description	Triggered
`SCHEMA_KEYS_PRESENT`	All required top-level sections exist in the artifact.	ok
`LINEUP_ELEVEN`	Each side fields exactly 11 starters.	ok
`FORMATION_VALID`	Formation string parses and its outfield lines sum to 10.	ok
`POSSESSION_SUMS_100`	Estimated possession of both teams sums to ~100%.	ok
`TEAM_GOALS_SUM`	Players' projected integer goals sum to the team's headline score.	ok
`TEAM_TOTALS_SCORE_MATCH`	team_totals.projected_goals equals the headline score component.	ok
`TEAM_EG_LAMBDA_MATCH`	team_totals.expected_goals matches the team's Poisson lambda.	ok
`PLAYER_STAT_REALISM`	Projected player lines are internally consistent (xG<=shots, accurate<=total passes, minutes/rating in range).	ok
`EXACTLY_ONE_GOALKEEPER`	Each starting XI fields exactly one goalkeeper.	ok
`SHOT_MATH_INTEGRITY`	Shot components are physically consistent (on target <= total, on+off <= total; blocked shots cannot be negative).	ok
`PLAYER_MINUTES_LEAGUE_CAP`	No player projects more minutes than the competition clock allows (90 + stoppage, or extra time in cups).	ok
`TEAM_SHOTS_CAP`	Team projected shots stay inside the realistic historical envelope (~35).	ok
`OUTCOME_SUMS_ONE`	home_win + draw + away_win sums to 1.	ok
`MARKETS_COMPLEMENTARY`	Over/Under and BTTS pairs each sum to 1.	ok
`GRID_SUMS_ONE`	The scoreline probability grid sums to ~1.	ok
`SCORELINE_PROBS_NONNEG`	Every scoreline probability is a valid [0,1] value.	ok
`GRID_REPRODUCIBLE`	The reported scoreline grid is reproducible from the lambdas.	ok
`MOST_LIKELY_IS_MODE`	The headline most-likely score is the distribution's modal scoreline.	2 problems
`MLS_PROBABILITY_MATCHES`	most_likely_score_probability equals the grid value of that score.	ok
`OUTCOME_MATCHES_GRID`	The reported win/draw/loss split is derivable from the grid.	ok
`MARKETS_MATCH_GRID`	The reported Over/Under and BTTS are derivable from the grid.	ok
`WINNER_MATCHES_SCORE`	predicted_winner agrees with the headline score direction.	ok
`WINNER_PROBABILITY_DIRECTION`	predicted_winner derives from the published modal scoreline; diverging from the 1X2 favorite is expected.	ok
`WINNER_IS_VALID_TEAM`	predicted_winner names a participating team (or Draw).	ok
`MODAL_OUTCOME_IS_DRAW`	Draw is the single largest outcome bucket while a team is named.	ok
`LAMBDA_BOUNDS`	Both lambdas fall inside the engine's [0.55, 5.00] envelope.	ok
`LAMBDA_CONSISTENCY`	The lambda is identical across prediction / antagonist / lambda_factors.	ok
`LAMBDA_FACTORS_PRESENT`	The lambda_factors decomposition (audit trail) is present.	ok
`CALIBRATION_SCALE_BOUNDS`	competition_calibration_scale stays within [0.85, 1.35].	ok
`CALIBRATION_RECORDED`	When calibration ran, the pre-calibration lambdas are recorded.	ok
`STANDING_AVAILABLE`	League-standing context is available for a domestic fixture.	ok
`RANKING_PROVENANCE`	Player ranking/quality signals carry a recognised source.	ok
`DATA_DEPTH`	Starters have enough recent games to model rather than synthesise.	ok
`FBREF_PRIORS_ACTIVE`	FBref season-prior files exist so that branch is not silently inert.	5 problems
`XGBOOST_REASONING_PRESENT`	Match-level xgboost_reasoning payload is present with full tracing pipeline.	ok
`XGBOOST_PLAYER_REASONING`	Starters have xgboost_reasoning metadata tracking feature contributions and FBref priors.	ok
`LAMBDA_REASONING_CONSISTENCY`	Lambdas reported in xgboost_reasoning match the final Poisson lambdas.	ok
`PREGAME_DIRECTION_AGREES`	The pregame fair-line favorite agrees with the model winner.	ok
`PREGAME_TOTAL_CONSISTENT`	The pregame expected total is close to the sum of the lambdas.	ok
`RESULT_WINNER_HIT`	The predicted winner matched the real result.	ok
`RESULT_EXACT_SCORE_HIT`	The headline score matched the real scoreline exactly.	ok
`RESULT_OU_HIT`	The Over/Under 2.5 lean matched reality.	ok
`RESULT_BTTS_HIT`	The BTTS lean matched reality.	ok
`RESULT_CONFIDENT_MISS`	A high-confidence prediction did not reach the real outcome.	ok
`OUTCOME_BRIER`	Brier score of the win/draw/loss probabilities vs the real result.	ok

Generated 2026-10-24 13:38:11 UTC · Soccer prediction reporting