Deep slate report 20260901

Which reports actually fed the diagnosis. A source shown as missing was not silently assumed.

High-level causal issue register

Source	Status	Evidence supplied	Path	Bytes
prediction review	loaded	official outcomes and scoring errors	/app/output/reports/prediction_review_20260901.json	n/a
forensic deep audit	loaded	probability, provenance, starter and decision contracts	/app/output/reports/PREDICTION_SYSTEM_DEEP_AUDIT_20260901.json	n/a
continuous learning	loaded	pipeline-stage quality, calibration, probability lineage	/app/output/learning/continuous_learning_daily_20260901.json	n/a
game artifacts	loaded	per-game model trace and structural audit source	/app/output/games	n/a

Deterministic findings across data integrity, model validation, calibration, score distribution, pipeline stages, simulation, and coverage. Critical is reserved for broken evidence or an unexplained contract violation; high marks material model/process risk. Each row pairs the action to take (Do this) with the wrong conclusion to avoid (Don't conclude).

Cross-slate trends (persistent-pattern memory)

Escalated persistent patterns

A finding that has repeated at or beyond the escalation streak is no longer single-slate noise. It carries its target registered lever and an action-oriented guardrail.

Finding streaks across the window

Metric direction across the window

Pipeline stage regression

Every stored stage is scored against the same official finals. A later recovery remains visible; it does not erase an unstable intermediate flip.

Probability lineage and calibration

All columns are pick-side probabilities, so away favorites are not accidentally compared using home-side semantics.

Score distribution and margin compression

Narrow predicted margins next to wide actual margins is the signature the score-variance tail is too tight.

Per-game evidence coverage

Every game, in depth (15)

Severity	Domain	Finding and evidence	Affected	Action & guardrail
HIGH	artifact_integrity	Canonical artifact warning requires reconciliation structural audit totals: {'critical': 0, 'error': 0, 'warning': 1, 'info': 108}	slate-wide	Do thisFix the flagged stat or presentation contract before treating this slate as clean evidence. Don't concludeA warning flags a display or contract issue; it does not prove the model picked the wrong side.
HIGH	score_distribution	Published margins are compressed relative to realized margins 13/15 predictions were ≤2 runs; 9/15 actual margins were ≥4; mean predicted margin 1.67 vs actual 4.20	CHW@HOU, DET@MIN, MIL@CHC, NYM@TB, PHI@ARI, STL@LAD, TOR@CLE	Do thisCheck whether predicted margins are too narrow across several recent out-of-sample slates (measuring tail coverage and margin MAE) before widening the score spread. Don't concludeA single blowout game is not a reason to widen every score prediction.
HIGH	calibration	Published confidence buckets were poorly aligned on this slate weighted absolute calibration error 0.2476 across 15 games	ATL@WAS, BAL@COL, CHW@HOU, DET@MIN, MIA@KC, MIL@CHC, NYM@TB, NYY@LAA, OAK@TEX, PHI@ARI, SD@CIN, SEA@BOS, SF@PIT, STL@LAD, TOR@CLE	Do thisWatch the same confidence buckets over the full rolling window before changing any calibration weight. Don't concludeA single slate has far too few games to establish a stable calibration curve.
HIGH	model_validation	Validated run models explain very little holdout variance home/away/total R² = 0.0361/0.0227/0.0280	ATL@WAS, BAL@COL, CHW@HOU, DET@MIN, MIA@KC, MIL@CHC, NYM@TB, NYY@LAA, OAK@TEX, PHI@ARI, SD@CIN, SEA@BOS, SF@PIT, STL@LAD, TOR@CLE	Do thisTreat exact run scores as low-confidence estimates, and look for stronger features or a better target before tuning the score tails. Don't concludeA small positive R² only means the model beats guessing the average; it does not make the exact scores reliable.
MEDIUM	calibration	Most high-raw picks lost before/after shrinkage 0/4 games with raw pick probability ≥0.65 won	BAL@COL, MIL@CHC, SD@CIN, STL@LAD	Do thisTrack how often high-raw-probability picks lose across multiple slates, and keep the raw to calibrated to published lineage. Don't concludeThis one slate is not enough to justify a new per-bucket shrinkage rule.
INFO	coverage	Every settled game received structural QA 15/15 games structurally audited	ATL@WAS, BAL@COL, CHW@HOU, DET@MIN, MIA@KC, MIL@CHC, NYM@TB, NYY@LAA, OAK@TEX, PHI@ARI, SD@CIN, SEA@BOS, SF@PIT, STL@LAD, TOR@CLE	Do thisKeep the per-game audit receipt in every report. Don't concludeFull audit coverage does not cancel the findings raised inside the audited games.
INFO	probability_lineage	Material calibrated-to-published movements have recorded lineage 9 material movement(s), all explained by named score-coherence and/or recency transformations	ATL@WAS, CHW@HOU, DET@MIN, MIA@KC, MIL@CHC, PHI@ARI, SD@CIN, SEA@BOS, STL@LAD	Do thisKeep showing every probability layer and the named transformation between them. Don't concludeA published probability above the calibrated one is not a governor bypass when a recorded transformation explains it.
INFO	simulation	Simulation produced no winner-direction intervention on this slate flip candidates 0; published flips 0	ATL@WAS, BAL@COL, CHW@HOU, DET@MIN, MIA@KC, MIL@CHC, NYM@TB, NYY@LAA, OAK@TEX, PHI@ARI, SD@CIN, SEA@BOS, SF@PIT, STL@LAD, TOR@CLE	Do thisMeasure the simulation's contribution across rolling slates, separating score realization from winner flipping. Don't concludeZero winner flips is not a defect and does not mean the simulation did no score-reconciliation work.

Finding	Persistence	Target lever(s)	Direction	Guardrail
simulation_inert_on_slate	5 consecutive · 5 in window	simulation.allow_score	confirm the simulation module is running, not silently no-oping	simulation_inert_on_slate has persisted 5 consecutive slates; this is no longer single-slate noise. Treat it as an actionable pattern and raise the code/diagnostic action named; do not force a parameter move.
weak_run_model_signal	5 consecutive · 5 in window	training.xgb.n_estimators, training.xgb.max_depth, training.xgb.learning_rate	investigate features/target via isolated training backtest (manual)	weak_run_model_signal has persisted 5 consecutive slates; this is no longer single-slate noise. Treat it as an actionable pattern and raise the code/diagnostic action named; do not force a parameter move.

Finding	Consecutive slates	Incidence	Present now	Fix type
high_raw_probability_cluster	1	2/5	yes	param
pipeline_winner_instability	0	2/5	no	code/diagnostic
score_margin_compression	2	2/5	yes	param
simulation_inert_on_slate	5	5/5	yes	code/diagnostic
single_slate_calibration_instability	2	4/5	yes	param
weak_run_model_signal	5	5/5	yes	code/diagnostic

Metric	Direction	Series (old → new)
winner_accuracy	regressing	[0.4667, 0.9412, 0.5, 0.6667, 0.4667]
weighted_calibration_error	regressing	[0.2179, 0.3961, 0.0693, 0.1661, 0.2476]
margin_bias	regressing	[-1.333, -1.882, -2.786, -3.5, -2.533]
team_run_error_on_losses	regressing	[5.0, 3.0, 7.0, 8.25, 7.125]

Stage	Winner hits	Accuracy	Team-run MAE	Total-run MAE	Margin MAE
historical_ml	7/15	46.7%	3.098	4.689	4.177
variance_applied	6/15	40%	3.067	4.667	4.133
enhanced_prediction	6/15	40%	3.267	4.667	4.933
embedded_savant_run_environment	6/15	40%	3.167	4.467	4.867
fundamental_edge_target_review	7/15	46.7%	3.133	4.533	4.8
capped_target	7/15	46.7%	3.133	4.533	4.8
simulation_result	7/15	46.7%	3.133	4.533	4.8
published_final	7/15	46.7%	3.133	4.533	4.8

Transition	Δ accuracy	Δ hits	Δ team-run MAE	Δ margin MAE	Games whose winner/tie changed
historical_ml → variance_applied	-0.0667	-1	-0.031	-0.0436	ATL@WAS: WAS → TIE; MIA@KC: KC → TIE; SD@CIN: SD → TIE; STL@LAD: LAD → TIE
variance_applied → enhanced_prediction	0	0	0.2	0.8	ATL@WAS: TIE → ATL; MIA@KC: TIE → KC; NYM@TB: NYM → TB; SD@CIN: TIE → SD; SEA@BOS: SEA → BOS; STL@LAD: TIE → LAD
enhanced_prediction → embedded_savant_run_environment	0	0	-0.1	-0.0666	TOR@CLE: TOR → TIE
embedded_savant_run_environment → fundamental_edge_target_review	0.0667	1	-0.0334	-0.0667	TOR@CLE: TIE → CLE
fundamental_edge_target_review → capped_target	0	0	0	0
capped_target → simulation_result	0	0	0	0
simulation_result → published_final	0	0	0	0

Game	Outcome	Raw model	Calibrated	Governor	Score coherence	Published	Explanation
ATL@WAS	miss	0.5085	0.5082	0.5082	0.5466	0.5317	score coherence followed by matchup-recency guardrail
BAL@COL	miss	0.7515	0.7373	0.7373	0.7461	0.7461	score_coherence aligned probability to the published score margin
CHW@HOU	hit	0.5085	0.5082	0.5082	0.5466	0.53	score coherence followed by matchup-recency guardrail
DET@MIN	miss	0.5032	0.505	0.505	0.545	0.53	score coherence followed by matchup-recency guardrail
MIA@KC	miss	0.6275	0.6227	0.6227	0.6464	0.6464	score_coherence aligned probability to the published score margin
MIL@CHC	miss	0.7934	0.6442	0.6442	0.6986	0.6986	score_coherence aligned probability to the published score margin
NYM@TB	hit	0.5085	0.5082	0.5082	0.5466	0.5269	score coherence followed by matchup-recency guardrail
NYY@LAA	hit	0.534	0.5332	0.5332	0.5591	0.54	score coherence followed by matchup-recency guardrail
OAK@TEX	hit	0.551	0.5499	0.5499	0.5674	0.5487	published probability matches the calibrated/governed value
PHI@ARI	hit	0.585	0.5816	0.5816	0.5833	0.5602	score coherence followed by matchup-recency guardrail
SD@CIN	miss	0.689	0.6778	0.6778	0.6739	0.6	score coherence followed by matchup-recency guardrail
SEA@BOS	miss	0.6328	0.6273	0.6273	0.6486	0.53	score coherence followed by matchup-recency guardrail
SF@PIT	hit	0.5615	0.56	0.56	0.5725	0.5445	score coherence followed by matchup-recency guardrail
STL@LAD	miss	0.6793	0.6696	0.6696	0.6698	0.58	score coherence followed by matchup-recency guardrail
TOR@CLE	hit	0.546	0.545	0.545	0.565	0.5552	no recorded transformation explains the publication delta

Sides analyzed	Broke on pitching	Broke on offense	Both	Variance	Mean starter ERA error	Mean lineup runs vs proj
30	9	3	2	16	1.48	1.07

Slate	Games	Winner acc	Mean starter ERA error	Mean lineup runs vs proj	Sides broke on pitching	Sides broke on offense
20260905	15	0.533	-1.17	-0.07	2	6
20260904	16	0.625	-0.23	-0.44	5	12
20260903	9	0.556	0.24	-0.83	4	6
20260902	15	0.333	0.72	0.30	7	6
20260901	15	0.467	—	—
20260831	12	0.667	—	—
20260830	14	0.500	—	—
20260829	17	0.941	—	—
20260828	15	0.467	—	—

Game	Pred margin	Actual margin	Pred − actual
ATL@WAS	2	4	-2
BAL@COL	3	2	1
CHW@HOU	1	4	-3
DET@MIN	1	13	-12
MIA@KC	2	3	-1
MIL@CHC	2	5	-3
NYM@TB	1	4	-3
NYY@LAA	2	4	-2
OAK@TEX	1	3	-2
PHI@ARI	1	6	-5
SD@CIN	3	1	2
SEA@BOS	2	3	-1
SF@PIT	1	1	0
STL@LAD	2	5	-3
TOR@CLE	1	5	-4

Team	Starter	Projected	Actual	ER err
ATL	AJ Smith-Shawver	3 ER / 13 outs	5 ER / 12 outs	2
WAS	Jake Irvin	2 ER / 15 outs	2 ER / 12 outs	0

Team	Pitcher	Outs	ERA	ER	H	HR	BB	SO	Context (as-of)
ATL	AJ Smith-Shawver (Starter)	12 / 13	11.25 / 6.23 (+5.02)	5 / 3	6 / 6	1 / 0	3 / 2	5 / 2	venue away: 6.23 vs 5.4 ERA (ok) · opp run factor 1 (strong) · thin sample: 3 prior appearances this season (high uncertainty)
WAS	Jake Irvin (Starter)	12 / 15	4.5 / 3.6	2 / 2	5 / 5	1 / 0	1 / 1	5 / 7	venue home: 7.04 vs 4.76 ERA (worse here) · opp run factor 1.05 (strong) · 17 prior appearances this season
WAS	Clayton Beeter (Closer)	3 / 3	27 / 0 (+27.00)	3 / 0	3 / 2	1 / 0	0 / 0	0 / 2
WAS	Orlando Ribalta (Setup Man)	3 / 3	0 / 9 (-9.00)	0 / 1	0 / 1	0 / 1	0 / 0	2 / 1

Team	Batter	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
WAS	Yohandy Morales (DH)	5 / 3	3 / 1	1 / 0	0 / 0	1 / 0	1 / 1	2 / 0	0 / 0	2 / 1	2.18	0.398 elite	projection too low
WAS	Daylen Lile (LF)	3 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	2 / 1	2 / 1	-1.07	0.329 above average	bad day
ATL	Ha-Seong Kim (SS)	3 / 3	3 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0.94	0.216 weak	good day
WAS	Harry Ford (C)	4 / 3	2 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	1 / 1	0.94	0.264 weak	good day
ATL	Ozzie Albies (DH)	4 / 4	1 / 1	0 / 0	0 / 0	1 / 0	1 / 0	1 / 0	0 / 0	0 / 1	0.93	0.284 below average	good day
ATL	Ronald Acuña Jr. (RF)	4 / 5	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	2 / 1	-0.8	0.362 elite	bad day
ATL	Austin Riley (3B)	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	2 / 1	-0.8	0.298 below average	projection too high
ATL	Michael Harris II (CF)	4 / 5	1 / 2	0 / 1	0 / 0	1 / 1	2 / 1	1 / 1	0 / 0	1 / 1	-0.78	0.341 elite	bad day
WAS	Brady House (3B)	3 / 4	1 / 1	1 / 0	0 / 0	0 / 0	3 / 1	0 / 0	1 / 0	1 / 1	0.64	0.336 above average	projection too low
WAS	CJ Abrams (SS)	4 / 4	2 / 1	0 / 1	0 / 0	0 / 0	2 / 1	1 / 1	1 / 0	1 / 1	0.49	0.328 above average	—
ATL	Mauricio Dubón (2B)	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	1 / 0	-0.47	0.302 below average	—
WAS	Abimelec Ortiz (1B)	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	-0.47	0.294 below average	—
WAS	Jacob Young (CF)	4 / 3	2 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0 / 0	1 / 1	0.47	0.319 average	—
WAS	Dylan Crews (RF)	4 / 5	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	1 / 1	1 / 0	2 / 1	-0.45	0.307 average	—
ATL	Drake Baldwin (C)	4 / 5	1 / 1	1 / 0	0 / 0	0 / 0	1 / 1	0 / 1	0 / 0	1 / 1	0.31	0.367 elite	—
ATL	Brewer Hicklen (LF)	3 / 3	1 / 1	0 / 1	0 / 0	0 / 0	0 / 1	1 / 1	1 / 1	2 / 1	-0.31	0.368 elite	—
ATL	Matt Olson (1B)	4 / 4	1 / 1	1 / 0	0 / 0	0 / 0	0 / 1	1 / 1	0 / 0	1 / 1	0.31	0.355 elite	—
WAS	Jorbit Vivas (2B)	3 / 3	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0 / 0	0.31	0.298 below average	—

Signal	Home-away edge	Meaning
Season quality	-0.602	positive favors home; negative favors away
Recent form	-0.318	positive favors home; negative favors away
Starting-pitcher matchup	-0.043	positive favors home; negative favors away
Composite	-0.865	positive favors home; negative favors away

Stage	Away-home runs	Confidence	What changed
historical_ml	4.77-4.83	n/a	Base ML historical model prediction
variance_applied	5-5	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	6-5	0.532	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by winner_starter_is_below_average, matchup_recency_guardrail)
embedded_savant_run_environment	7-5	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	7-5	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	7-5	n/a	Capped at max 12 realistic runs
simulation_result	7-5	n/a	Final inning-by-inning simulation result

Stage	Before	After	Toward final pick	What changed
variance_applied	4.77-4.83	5-5	0.062	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-5	6-5	1	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by winner_starter_is_below_average, matchup_recency_guardrail)
embedded_savant_run_environment	6-5	7-5	1	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	7-5	7-5	-0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	7-5	7-5	-0	Capped at max 12 realistic runs
simulation_result	7-5	7-5	-0	Final inning-by-inning simulation result

Team	Player	Role	Net effect	Largest change
ATL	Drake Baldwin	batter / C	better	predicted_RBI 2.24
ATL	Ronald Acuña Jr.	batter / RF	better	predicted_H 1.82
WAS	CJ Abrams	batter / SS	deboosted	predicted_2B 1.802
WAS	Abimelec Ortiz	batter / 1B	worse	predicted_AB 1.778
WAS	Yohandy Morales	batter / DH	worse	predicted_AB 1.742
WAS	Harry Ford	batter / C	worse	predicted_BB 1.591
ATL	Matt Olson	batter / 1B	better	predicted_RBI 1.42
WAS	Brady House	batter / 3B	better	predicted_RBI 1.34

Severity	Rule	Finding	Subject
info	CURRENT_PITCHER_DUPLICATION	ATL pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	ATL
info	CURRENT_PITCHER_DUPLICATION	WAS pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	WAS
info	CURRENT_FINALIZED_MATCH_BOX	ATL model-prediction batting totals differ from final simulation box score (AB: current=37, sim=38; H: current=10, sim=11; 2B: current=2, sim=0; HR: current=1, sim=3; RBI: current=5, sim=7; R: current=5, sim=7; BB: current=3, sim=2; SO: current=8, sim=11).	ATL
info	CURRENT_PLAYER_MATCH_BOX	ATL current finalized hitter lines are model predictions and differ from the simulation box score: Drake Baldwin: H current=1 box=2, HR current=0 box=1, RBI current=1 box=3; Mauricio Dubón: H current=1 box=0, RBI current=1 box=0, R current=0 box=1, SO current=0 box=1; Ozzie Albies: H current=1 box=0; Michael Harris II: 2B current=1 box=0, HR current=1 box=0, RBI current=1 box=0, SO current=1 box=2; Brewer Hicklen: 2B current=1 box=0, RBI current=1 box=0, SO current=1 box=2; Matt Olson: HR current=0 box=1, RBI current=1 box=2, BB current=0 box=1, SO current=1 box=2; Ronald Acuña Jr.: H current=1 box=3, HR current=0 box=1, RBI current=0 box=1, BB current=1 box=0; Ha-Seong Kim: AB current=3 box=4, H current=1 box=0, SO current=1 box=2	ATL
info	CURRENT_FINALIZED_MATCH_BOX	WAS model-prediction batting totals differ from final simulation box score (AB: current=32, sim=38; H: current=8, sim=11; 2B: current=2, sim=5; 3B: current=0, sim=2; HR: current=1, sim=0; R: current=4, sim=5; BB: current=3, sim=4; SO: current=8, sim=4).	WAS
info	CURRENT_PLAYER_MATCH_BOX	WAS current finalized hitter lines are model predictions and differ from the simulation box score: Brady House: H current=1 box=2, 2B current=0 box=1, RBI current=1 box=2, R current=0 box=1; Yohandy Morales: AB current=3 box=5, H current=1 box=0, RBI current=1 box=0, SO current=1 box=2; Daylen Lile: AB current=4 box=5, H current=1 box=2, 2B current=0 box=1, 3B current=0 box=1; Dylan Crews: H current=1 box=2, 2B current=1 box=0; CJ Abrams: AB current=4 box=5, H current=1 box=2, 2B current=1 box=2, RBI current=1 box=0; Jorbit Vivas: AB current=3 box=4, 3B current=0 box=1, BB current=1 box=0; Abimelec Ortiz: AB current=3 box=5, 2B current=0 box=1, SO current=1 box=0; Harry Ford: AB current=3 box=2, BB current=1 box=2, SO current=1 box=0	WAS
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

Team	Starter	Projected	Actual	ER err
BAL	Kyle Bradish	3 ER / 16 outs	2 ER / 15 outs	-1
COL	Gabriel Hughes	5 ER / 13 outs	1 ER / 14 outs	-4

Team	Pitcher	Outs	ERA	ER	H	HR	BB	SO	Context (as-of)
BAL	Kyle Bradish (Starter)	15 / 16	3.6 / 5.06	2 / 3	6 / 6	1 / 1	2 / 2	4 / 2	venue away: 4.5 vs 3.46 ERA (worse here) · opp run factor 1.04 (average) · 26 prior appearances this season
COL	Gabriel Hughes (Starter)	14 / 13	1.93 / 10.38 (-8.45)	1 / 5	4 / 7	0 / 1	2 / 0	4 / 9	venue home: 4.79 vs 9 ERA (ok) · opp run factor 1.08 (average) · 10 prior appearances this season
COL	Brennan Bernardino (Setup Man)	1 / 4	0 / 0	0 / 0	0 / 2	0 / 0	0 / 0	0 / 3
COL	Jaden Hill (Middle Reliever (High Leverage))	3 / 3	9 / 9	1 / 1	1 / 1	1 / 1	0 / 0	1 / 1
COL	Jordan Romano (Closer)	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 2

Team	Batter	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
BAL	Jackson Holliday (2B)	5 / 5	3 / 1	0 / 0	0 / 0	1 / 0	1 / 0	1 / 1	0 / 0	1 / 1	1.87	0.303 below average	good day
COL	Hunter Goodman (DH)	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	3 / 1	-1.73	0.324 above average	bad day
COL	Zac Veen (LF)	1 / 3	1 / 0	0 / 0	1 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 1	1.42	0.299 below average	good day
COL	Cole Carrigg (SS)	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	-1.11	0.323 above average	bad day
COL	Brett Sullivan (C)	3 / 3	2 / 1	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	1.11	0.286 below average	good day
BAL	Pete Alonso (DH)	3 / 5	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	1 / 1	1 / 2	-0.93	0.392 elite	bad day
COL	TJ Rumfield (1B)	3 / 4	1 / 1	0 / 0	0 / 0	1 / 0	2 / 0	1 / 0	1 / 1	1 / 1	0.93	0.334 above average	projection too low
BAL	Gunnar Henderson (SS)	4 / 5	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	-0.8	0.316 average	no signal
BAL	Carlos Narváez (C)	1 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 1	0.8	0.265 weak	good day
BAL	Leody Taveras (RF)	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	3 / 1	-0.47	0.273 weak	—
BAL	Colton Cowser (CF)	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	2 / 1	-0.47	0.321 above average	—
COL	Jake McCarthy (CF)	5 / 4	2 / 1	1 / 1	0 / 0	0 / 0	1 / 1	0 / 1	0 / 0	0 / 1	0.47	0.319 average	—
COL	Connor Norby (2B)	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	-0.47	0.301 below average	—
BAL	Dylan Beavers (LF)	4 / 3	1 / 1	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	-0.33	0.304 below average	—
BAL	Samuel Basallo (1B)	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 0	1 / 1	-0.14	0.322 above average	—
BAL	Christian Encarnacion-Strand (3B)	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0	0.343 elite	—
COL	Mickey Moniak (RF)	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	0 / 0	0 / 1	0	0.281 weak	—

Signal	Home-away edge	Meaning
Season quality	-1.087	positive favors home; negative favors away
Recent form	-1.418	positive favors home; negative favors away
Starting-pitcher matchup	-0.04	positive favors home; negative favors away
Composite	-1.581	positive favors home; negative favors away

Stage	Away-home runs	Confidence	What changed
historical_ml	5.44-4.39	n/a	Base ML historical model prediction
variance_applied	5-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	7-4	0.746	Enhanced ML with recent form, H2H, low-score analysis
embedded_savant_run_environment	7-4	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	7-4	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	7-4	n/a	Capped at max 12 realistic runs
simulation_result	7-4	n/a	Final inning-by-inning simulation result

Stage	Before	After	Toward final pick	What changed
variance_applied	5.44-4.39	5-4	-0.054	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-4	7-4	2	Enhanced ML with recent form, H2H, low-score analysis
embedded_savant_run_environment	7-4	7-4	-0	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	7-4	7-4	-0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	7-4	7-4	-0	Capped at max 12 realistic runs
simulation_result	7-4	7-4	-0	Final inning-by-inning simulation result

Team	Player	Role	Net effect	Largest change
BAL	Gunnar Henderson	batter / SS	better	predicted_H 1.95
BAL	Christian Encarnacion-Strand	batter / 3B	worse	predicted_AB 1.629
COL	Brett Sullivan	batter / C	better	predicted_H 1.28
BAL	Jackson Holliday	batter / 2B	boosted	predicted_H 1.27
BAL	Pete Alonso	batter / DH	boosted	predicted_AB 1.244
BAL	Colton Cowser	batter / CF	better	predicted_AB 1.24
COL	Connor Norby	batter / 2B	better	predicted_H 1.21
COL	Jake McCarthy	batter / CF	worse	predicted_AB 1.154

Severity	Rule	Finding	Subject
info	CURRENT_PITCHER_DUPLICATION	BAL pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	BAL
info	CURRENT_PITCHER_DUPLICATION	COL pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	COL
info	CURRENT_FINALIZED_MATCH_BOX	BAL model-prediction batting totals differ from final simulation box score (AB: current=33, sim=38; H: current=8, sim=11; 2B: current=1, sim=4; HR: current=1, sim=2; RBI: current=5, sim=7; R: current=5, sim=7; BB: current=3, sim=2; SO: current=10, sim=16).	BAL
info	CURRENT_PLAYER_MATCH_BOX	BAL current finalized hitter lines are model predictions and differ from the simulation box score: Pete Alonso: 2B current=0 box=1, HR current=1 box=0, BB current=1 box=0; Dylan Beavers: H current=1 box=0, 2B current=1 box=0, R current=1 box=0, SO current=1 box=2; Samuel Basallo: AB current=3 box=4, RBI current=1 box=0, R current=0 box=1, SO current=1 box=2; Leody Taveras: 2B current=0 box=1, BB current=0 box=1; Jackson Holliday: H current=1 box=2, RBI current=0 box=1, SO current=1 box=2; Christian Encarnacion-Strand: AB current=3 box=5, HR current=0 box=1, RBI current=0 box=1, R current=0 box=1; Colton Cowser: AB current=3 box=4, R current=0 box=1, SO current=1 box=2; Gunnar Henderson: H current=1 box=3, 2B current=0 box=2, HR current=0 box=1, RBI current=1 box=2	BAL
info	CURRENT_FINALIZED_MATCH_BOX	COL model-prediction batting totals differ from final simulation box score (AB: current=33, sim=34; 2B: current=2, sim=3; BB: current=3, sim=2; SO: current=9, sim=5; SF: current=0, sim=1).	COL
info	CURRENT_PLAYER_MATCH_BOX	COL current finalized hitter lines are model predictions and differ from the simulation box score: Jake McCarthy: AB current=4 box=5, 2B current=1 box=0, HR current=0 box=1; Mickey Moniak: H current=1 box=0, RBI current=1 box=0, SO current=1 box=0; Hunter Goodman: HR current=1 box=0, RBI current=1 box=0, R current=1 box=0, BB current=1 box=0; Cole Carrigg: H current=1 box=2, RBI current=1 box=0, R current=1 box=0, BB current=1 box=0; Willi Castro: AB current=4 box=3, H current=1 box=0, RBI current=0 box=1, SF current=0 box=1; Connor Norby: AB current=3 box=4, H current=1 box=2, 2B current=0 box=1, RBI current=0 box=1; Brett Sullivan: H current=1 box=2, 2B current=0 box=1, RBI current=0 box=1, R current=0 box=1; Zac Veen: BB current=0 box=1, SO current=1 box=0	COL
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

Team	Starter	Projected	Actual	ER err
CHW	Sean Burke	4 ER / 15 outs	0 ER / 17 outs	-4
HOU	Ronel Blanco	4 ER / 13 outs	0 ER / 12 outs	-4

Team	Pitcher	Outs	ERA	ER	H	HR	BB	SO	Context (as-of)
CHW	Sean Burke (Starter)	17 / 15	0 / 7.2 (-7.20)	0 / 4	1 / 7	0 / 0	4 / 1	6 / 4	venue away: 3.04 vs 3.99 ERA (ok) · opp run factor 0.97 (average) · 27 prior appearances this season
CHW	Bryan Hudson (Middle Reliever (High Leverage))	4 / 1	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	2 / 0
HOU	Ronel Blanco (Starter)	12 / 13	0 / 8.31 (-8.31)	0 / 4	3 / 5	0 / 1	3 / 3	3 / 4	venue home: 6.75 vs 7.2 ERA (ok) · opp run factor 1.02 (average) · 5 prior appearances this season

Team	Batter	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
HOU	Taylor Trammell (CF)	3 / 3	2 / 0	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	1.58	0.285 below average	good day
CHW	Munetaka Murakami (1B)	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	1 / 1	3 / 2	-1.4	0.35 elite	bad day
HOU	Yordan Alvarez (DH)	2 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	2 / 1	2 / 1	-1.07	0.45 elite	bad day
HOU	LaMonte Wade Jr. (LF)	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	2 / 1	-0.8	0.278 weak	projection too high
CHW	Drew Romo (C)	2 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 0	0 / 1	0.66	0.262 weak	good day
HOU	Isaac Paredes (3B)	4 / 4	1 / 1	0 / 0	0 / 0	1 / 0	1 / 0	1 / 0	0 / 1	0 / 0	0.6	0.32 average	no signal
CHW	Miguel Vargas (3B)	4 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	1 / 1	0 / 1	0.47	0.386 elite	—
HOU	Yainer Diaz (C)	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	2 / 0	-0.47	0.304 below average	—
HOU	Christian Walker (1B)	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	3 / 1	-0.47	0.3 below average	—
HOU	Cam Smith (RF)	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	-0.47	0.319 average	—
HOU	Jeremy Peña (SS)	3 / 5	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 0	1 / 1	-0.45	0.326 above average	—
CHW	Colson Montgomery (SS)	4 / 4	2 / 1	1 / 0	0 / 0	0 / 1	2 / 1	1 / 1	1 / 1	2 / 1	-0.15	0.307 average	—
CHW	Sam Antonacci (LF)	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	1 / 0	1 / 0	-0.14	0.342 elite	—
CHW	Tristan Peters (CF)	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	-0.14	0.319 average	—
CHW	Chase Meidroth (2B)	4 / 4	2 / 1	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0.14	0.292 below average	—
HOU	Jose Altuve (2B)	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	1 / 1	-0.14	0.272 weak	—
CHW	Andrew Benintendi (DH)	5 / 3	1 / 1	0 / 0	0 / 0	0 / 0	1 / 1	1 / 0	0 / 0	1 / 1	0	0.31 average	—

Signal	Home-away edge	Meaning
Season quality	-0.363	positive favors home; negative favors away
Recent form	0.283	positive favors home; negative favors away
Starting-pitcher matchup	-0.048	positive favors home; negative favors away
Composite	-0.415	positive favors home; negative favors away

Stage	Away-home runs	Confidence	What changed
historical_ml	5.04-3.45	n/a	Base ML historical model prediction
variance_applied	5-3	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	6-5	0.53	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run, matchup_recency_guardrail)
embedded_savant_run_environment	6-5	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	6-5	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	6-5	n/a	Capped at max 12 realistic runs
simulation_result	6-5	n/a	Final inning-by-inning simulation result

Stage	Before	After	Toward final pick	What changed
variance_applied	5.04-3.45	5-3	0.417	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-3	6-5	-1	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run, matchup_recency_guardrail)
embedded_savant_run_environment	6-5	6-5	-0	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	6-5	6-5	-0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	6-5	6-5	-0	Capped at max 12 realistic runs
simulation_result	6-5	6-5	-0	Final inning-by-inning simulation result

Team	Player	Role	Net effect	Largest change
CHW	Andrew Benintendi	batter / DH	better	predicted_SO 3.2
CHW	Chase Meidroth	batter / 2B	better	predicted_SO 2.062
HOU	Jose Altuve	batter / 2B	better	predicted_H 2.01
HOU	Ronel Blanco	pitcher / Starter	worse	predicted_ER 2
HOU	Bryan Abreu	pitcher / Middle Reliever (High Leverage)	better	predicted_ER -2
HOU	LaMonte Wade Jr.	batter / LF	better	predicted_BB 1.579
CHW	Randal Grichuk	batter / RF	better	predicted_AB 1.5
CHW	Colson Montgomery	batter / SS	worse	predicted_SO -1.434

Severity	Rule	Finding	Subject
info	CURRENT_PITCHER_DUPLICATION	CHW pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	CHW
info	CURRENT_PITCHER_DUPLICATION	HOU pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	HOU
info	CURRENT_FINALIZED_MATCH_BOX	CHW model-prediction batting totals differ from final simulation box score (AB: current=31, sim=38; H: current=8, sim=11; RBI: current=5, sim=6; R: current=5, sim=6).	CHW
info	CURRENT_PLAYER_MATCH_BOX	CHW current finalized hitter lines are model predictions and differ from the simulation box score: Sam Antonacci: AB current=3 box=4, H current=1 box=2, BB current=0 box=1; Munetaka Murakami: H current=1 box=2, RBI current=1 box=2, SO current=2 box=1; Randal Grichuk: AB current=3 box=4, R current=0 box=1, BB current=0 box=1, SO current=1 box=0; Tristan Peters: AB current=3 box=4, H current=1 box=2, 2B current=0 box=1, RBI current=0 box=1; Miguel Vargas: AB current=4 box=5, RBI current=1 box=0, R current=1 box=0, BB current=1 box=0; Colson Montgomery: HR current=1 box=0, RBI current=1 box=0, R current=1 box=0, SO current=1 box=0; Andrew Benintendi: AB current=3 box=5, R current=0 box=1, SO current=1 box=4; Drew Romo: AB current=3 box=4	CHW
info	CURRENT_FINALIZED_MATCH_BOX	HOU model-prediction batting totals differ from final simulation box score (AB: current=34, sim=35; H: current=8, sim=9; 2B: current=1, sim=2; HR: current=1, sim=0; RBI: current=4, sim=5; R: current=4, sim=5; BB: current=3, sim=2; SF: current=0, sim=1).	HOU
info	CURRENT_PLAYER_MATCH_BOX	HOU current finalized hitter lines are model predictions and differ from the simulation box score: Yordan Alvarez: AB current=4 box=5, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0; Jeremy Peña: 2B current=1 box=0, RBI current=1 box=0, SO current=1 box=2; Jose Altuve: H current=1 box=3, 2B current=0 box=1, RBI current=0 box=2, SO current=1 box=0; Isaac Paredes: AB current=4 box=5, H current=1 box=0, BB current=1 box=0; LaMonte Wade Jr.: AB current=3 box=2, 2B current=0 box=1, RBI current=0 box=1, R current=0 box=1; Yainer Diaz: SO current=0 box=1; Christian Walker: AB current=4 box=3, R current=0 box=1, SO current=1 box=2, SF current=0 box=1; Taylor Trammell: SO current=1 box=2, HBP current=0 box=1	HOU
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

Team	Starter	Projected	Actual	ER err
DET	Troy Melton	2 ER / 18 outs	5 ER / 16 outs	3
MIN	Andrew Morris	1 ER / 14 outs	1 ER / 3 outs	0

Team	Pitcher	Outs	ERA	ER	H	HR	BB	SO	Context (as-of)
DET	Troy Melton (Starter)	16 / 18	8.44 / 3 (+5.44)	5 / 2	6 / 2	2 / 0	1 / 4	4 / 6	venue away: 1.4 vs 2.37 ERA (ok) · opp run factor 0.97 (average) · 16 prior appearances this season
DET	Kyle Finnegan (Middle Reliever (High Leverage))	1 / 3	135 / 0 (+135.00)	5 / 0	3 / 0	1 / 0	2 / 0	0 / 0
MIN	Andrew Morris (Starter)	3 / 14	9 / 1.93 (+7.07)	1 / 1	2 / 1	0 / 0	0 / 1	0 / 5	venue home: 9 vs 0 ERA (ok) · opp run factor 1 (average) · 53 prior appearances this season

Team	Batter	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
MIN	Royce Lewis (1B)	4 / 4	3 / 1	1 / 0	0 / 0	1 / 0	3 / 0	2 / 0	1 / 0	0 / 1	2.51	0.316 average	no signal
MIN	Brooks Lee (3B)	3 / 5	0 / 1	0 / 0	0 / 0	0 / 1	1 / 1	0 / 1	0 / 0	1 / 1	-1.4	0.274 weak	projection too high
MIN	Kody Clemens (2B)	3 / 4	1 / 1	0 / 0	0 / 0	1 / 0	1 / 1	2 / 1	1 / 0	1 / 1	1.26	0.341 elite	projection too low
DET	Riley Greene (LF)	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	1 / 0	0 / 1	0 / 1	2 / 1	-1.11	0.367 elite	bad day
DET	Dillon Dingler (C)	4 / 5	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	0 / 0	0 / 1	-0.93	0.344 elite	bad day
DET	Javier Báez (RF)	4 / 3	0 / 1	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 0	-0.78	0.258 weak	projection too high
DET	Spencer Torkelson (1B)	3 / 3	1 / 0	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0.78	0.31 average	no signal
MIN	Kaelen Culpepper (SS)	5 / 4	2 / 1	1 / 0	0 / 0	0 / 0	3 / 0	1 / 1	0 / 0	0 / 1	0.78	0.32 average	no signal
MIN	Walker Jenkins (CF)	4 / 3	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	2 / 0	1 / 1	0 / 0	-0.78	0.253 weak	projection too high
MIN	Ryan Jeffers (C)	4 / 4	2 / 1	0 / 0	1 / 0	1 / 1	3 / 1	1 / 1	0 / 1	0 / 1	0.76	0.356 elite	projection too low
MIN	Josh Bell (DH)	3 / 5	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	2 / 1	2 / 0	2 / 1	0.66	0.333 above average	projection too low
DET	Kevin McGonigle (DH)	3 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	0 / 1	0 / 1	-0.33	0.35 elite	—
MIN	Trevor Larnach (LF)	2 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	2 / 0	1 / 0	0 / 1	0.33	0.329 above average	—
DET	Hao-Yu Lee (3B)	4 / 3	1 / 1	1 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	2 / 1	0.31	0.301 below average	—
MIN	Luke Keaschall (RF)	5 / 4	2 / 1	1 / 1	0 / 0	0 / 0	2 / 0	1 / 1	0 / 1	0 / 0	0.14	0.315 average	—
DET	Gleyber Torres (2B)	2 / 5	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	1 / 1	0	0.327 above average	—
DET	Max Clark (CF)	3 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	2 / 1	0	0.321 above average	—
DET	John Peck (SS)	4 / 4	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0	0.379 elite	—

Signal	Home-away edge	Meaning
Season quality	-0.297	positive favors home; negative favors away
Recent form	0.534	positive favors home; negative favors away
Starting-pitcher matchup	-0.025	positive favors home; negative favors away
Composite	-0.265	positive favors home; negative favors away

Stage	Away-home runs	Confidence	What changed
historical_ml	5.07-3.72	n/a	Base ML historical model prediction
variance_applied	5-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-3	0.53	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run, matchup_recency_guardrail)
embedded_savant_run_environment	4-3	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	4-3	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	4-3	n/a	Capped at max 12 realistic runs
simulation_result	4-3	n/a	Final inning-by-inning simulation result

Stage	Before	After	Toward final pick	What changed
variance_applied	5.07-3.72	5-4	-0.351	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-4	4-3	-0	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run, matchup_recency_guardrail)
embedded_savant_run_environment	4-3	4-3	-0	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	4-3	4-3	-0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	4-3	4-3	-0	Capped at max 12 realistic runs
simulation_result	4-3	4-3	-0	Final inning-by-inning simulation result

Team	Player	Role	Net effect	Largest change
MIN	Kody Clemens	batter / 2B	worse	predicted_SO 2.075
DET	John Peck	batter / SS	deboosted	predicted_RBI 1.86
MIN	Walker Jenkins	batter / CF	worse	predicted_RBI 1.47
DET	Hao-Yu Lee	batter / 3B	worse	predicted_SO 1.302
MIN	Brooks Lee	batter / 3B	worse	predicted_H -1.09
DET	Dillon Dingler	batter / C	worse	predicted_H -1.02
DET	Troy Melton	pitcher / Starter	better	predicted_ER -1
DET	Kenley Jansen	pitcher / Closer	worse	predicted_ER 1

Severity	Rule	Finding	Subject
info	CURRENT_PITCHER_DUPLICATION	DET pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	DET
info	CURRENT_PITCHER_DUPLICATION	MIN pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	MIN
info	CURRENT_FINALIZED_MATCH_BOX	DET model-prediction batting totals differ from final simulation box score (AB: current=35, sim=32; H: current=7, sim=5; 2B: current=2, sim=1; 3B: current=0, sim=1; RBI: current=3, sim=4; R: current=5, sim=4; BB: current=3, sim=2; SO: current=8, sim=10).	DET
info	CURRENT_PLAYER_MATCH_BOX	DET current finalized hitter lines are model predictions and differ from the simulation box score: Kevin McGonigle: 2B current=0 box=1, BB current=1 box=0; Hao-Yu Lee: H current=1 box=0, BB current=0 box=1, SO current=1 box=2; Dillon Dingler: AB current=5 box=4, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0; Riley Greene: AB current=4 box=3, H current=1 box=0, 2B current=1 box=0, R current=1 box=0; Javier Báez: H current=1 box=0, 2B current=1 box=0, R current=1 box=0, SO current=0 box=1; Gleyber Torres: AB current=5 box=4, RBI current=0 box=1, BB current=1 box=0; Spencer Torkelson: AB current=3 box=4, SO current=1 box=2; Max Clark: RBI current=1 box=0, R current=0 box=1	DET
info	CURRENT_FINALIZED_MATCH_BOX	MIN model-prediction batting totals differ from final simulation box score (AB: current=37, sim=30; H: current=9, sim=3; 2B: current=2, sim=0; HR: current=2, sim=0; RBI: current=5, sim=3; R: current=6, sim=3; BB: current=3, sim=4; HBP: current=0, sim=1).	MIN
info	CURRENT_PLAYER_MATCH_BOX	MIN current finalized hitter lines are model predictions and differ from the simulation box score: Luke Keaschall: H current=1 box=0, 2B current=1 box=0, R current=1 box=0, BB current=1 box=0; Kody Clemens: AB current=4 box=3, H current=1 box=0, RBI current=1 box=0, R current=1 box=0; Ryan Jeffers: HR current=1 box=0, BB current=1 box=0, SO current=1 box=0; Trevor Larnach: AB current=4 box=3, H current=1 box=0, BB current=0 box=1; Josh Bell: AB current=5 box=4, RBI current=1 box=0, SO current=1 box=0; Brooks Lee: AB current=5 box=4, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0; Royce Lewis: AB current=4 box=3, H current=1 box=0, SO current=1 box=0, HBP current=0 box=1; Kaelen Culpepper: AB current=4 box=3, H current=1 box=0, R current=1 box=0, BB current=0 box=1	MIN
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

Team	Starter	Projected	Actual	ER err
MIA	Tyler Phillips	4 ER / 15 outs	0 ER / 15 outs	-4
KC	Randy Dobnak	1 ER / 17 outs	6 ER / 18 outs	5

Team	Pitcher	Outs	ERA	ER	H	HR	BB	SO	Context (as-of)
MIA	Tyler Phillips (Starter)	15 / 15	0 / 7.2 (-7.20)	0 / 4	1 / 6	0 / 1	2 / 2	4 / 4	venue away: 6.83 vs 2.72 ERA (worse here) · opp run factor 0.98 (weak) · 32 prior appearances this season
MIA	Calvin Faucher (Closer)	3 / 6	0 / 4.5 (-4.50)	0 / 1	1 / 1	0 / 0	0 / 2	0 / 1
MIA	Tyler Zuber (Setup Man)	3 / 3	9 / 0 (+9.00)	1 / 0	2 / 0	1 / 0	0 / 0	0 / 1
KC	Randy Dobnak (Starter)	18 / 17	9 / 1.59 (+7.41)	6 / 1	7 / 1	2 / 0	2 / 1	6 / 5	venue home: 2.94 vs 1.14 ERA (worse here) · opp run factor 0.93 (average) · 10 prior appearances this season

Team	Batter	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
MIA	Graham Pauley (3B)	4 / 3	2 / 0	0 / 0	1 / 0	1 / 0	4 / 0	2 / 0	0 / 0	1 / 1	2.49	0.225 weak	good day
MIA	Heriberto Hernández (LF)	3 / 3	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	1 / 1	2 / 1	-1.4	0.353 elite	bad day
KC	Nick Loftin (2B)	4 / 3	1 / 1	0 / 0	0 / 0	1 / 0	1 / 0	1 / 0	0 / 0	2 / 0	0.93	0.297 below average	good day
MIA	Jakob Marsee (CF)	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	-0.8	0.305 average	no signal
KC	Bobby Witt Jr. (SS)	3 / 5	1 / 1	0 / 1	0 / 0	0 / 0	1 / 0	0 / 1	0 / 1	1 / 1	-0.64	0.383 elite	bad day
KC	Carter Jensen (C)	4 / 5	2 / 1	1 / 0	0 / 0	0 / 1	1 / 1	0 / 1	0 / 1	0 / 1	-0.48	0.304 below average	—
KC	Salvador Perez (DH)	4 / 5	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0 / 1	-0.47	0.308 average	—
KC	Josh Rojas (3B)	4 / 3	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0.47	0.217 weak	—
KC	Vinnie Pasquantino (1B)	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	1 / 1	-0.47	0.325 above average	—
KC	Kyle Isbel (CF)	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	-0.47	0.264 weak	—
KC	Jac Caglianone (RF)	3 / 5	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 0	1 / 1	-0.45	0.357 elite	—
KC	Isaac Collins (LF)	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	1 / 1	0.33	0.296 below average	—
MIA	Kyle Stowers (DH)	4 / 5	1 / 1	0 / 1	0 / 0	1 / 0	1 / 1	1 / 1	0 / 1	2 / 2	0.29	0.335 above average	—
MIA	Javier Sanoja (2B)	4 / 3	2 / 1	0 / 1	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0 / 0	0.16	0.287 below average	—
MIA	Griffin Conine (1B)	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 0	1 / 1	-0.14	0.373 elite	—
MIA	Otto Lopez (SS)	3 / 5	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	0 / 0	0 / 1	0	0.324 above average	—
MIA	Owen Caissie (RF)	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 1	0	0.297 below average	—
MIA	Joe Mack (C)	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 1	0	0.289 below average	—

Signal	Home-away edge	Meaning
Season quality	-0.27	positive favors home; negative favors away
Recent form	0.648	positive favors home; negative favors away
Starting-pitcher matchup	0.052	positive favors home; negative favors away
Composite	0.488	positive favors home; negative favors away

Stage	Away-home runs	Confidence	What changed
historical_ml	4.84-5	n/a	Base ML historical model prediction
variance_applied	5-5	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	3-5	0.646	Enhanced ML with recent form, H2H, low-score analysis
embedded_savant_run_environment	3-5	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	3-5	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	3-5	n/a	Capped at max 12 realistic runs
simulation_result	3-5	n/a	Final inning-by-inning simulation result

Stage	Before	After	Toward final pick	What changed
variance_applied	4.84-5	5-5	-0.163	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-5	3-5	2	Enhanced ML with recent form, H2H, low-score analysis
embedded_savant_run_environment	3-5	3-5	0	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	3-5	3-5	0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	3-5	3-5	0	Capped at max 12 realistic runs
simulation_result	3-5	3-5	0	Final inning-by-inning simulation result

Team	Player	Role	Net effect	Largest change
MIA	Graham Pauley	batter / 3B	worse	predicted_SO 2.341
KC	Salvador Perez	batter / DH	worse	predicted_AB -1.89
KC	Carter Jensen	batter / C	worse	predicted_AB -1.73
KC	Bobby Witt Jr.	batter / SS	better	predicted_AB -1.65
KC	Jac Caglianone	batter / RF	worse	predicted_H -1.48
MIA	Heriberto Hernández	batter / LF	better	predicted_RBI 1.43
KC	Kyle Isbel	batter / CF	worse	predicted_AB -1.43
MIA	Kyle Stowers	batter / DH	worse	predicted_SO -1.337

SF@PITHIT

Predicted: PIT 5-6 (conf +0.09, lean)Actual: PIT 12-13Brier 0.207total bias -14

Root cause: Right winner, wrong score

Evidence-backed diagnosis

published system edge 0.09 (lean); this is not the half-scale distance from 50%
team-run absolute error 14; total bias -14
pregame home-minus-away evidence: season +0.440, recent -0.353, pitcher -0.172, composite -0.043
confidence pipeline reached 0.560 before later publication/recency controls; published home probability was 0.545
score-volatility actions: no_change
forecast-file integrity: clean (0 defects, 7 informational note(s)); the saved artifact is not why the pick missed

Quantified contributing factors

total runs under-projected by 14 (11 vs 25)
SF offense under-projected by 7
PIT offense under-projected by 7
Logan Webb (SF) allowed 7 more ER than projected (9 vs 2)
Paul Skenes (PIT) allowed 4 more ER than projected (5 vs 1)

Team	Starter	Projected	Actual	ER err
SF	Logan Webb	2 ER / 18 outs	9 ER / 6 outs	7
PIT	Paul Skenes	1 ER / 16 outs	5 ER / 14 outs	4

Decision internals: run factors home×0.924 / away×1.096; savant lineup adjustment applied

Player-level breakdown (projected vs actual)

Every value is actual / projected. Run impact approximates how far each hitter's real line moved run expectancy from its projection; it orders the movers and is not a published stat.

Top deviations that drove the game:

Brandon Lowe (PIT 2B): 4-for-6, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 2.34]
Drew Cavanaugh (SF C): 3-for-5, 0 HR, 2 RBI (projected 0.0 H, 0.0 HR) [run impact 2.03]
Jacob Gonzalez (PIT SS): 2-for-2, 0 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 1.11]
Nick Gonzales (PIT DH): 3-for-5, 0 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 0.94]
Henry Davis (PIT C): 1-for-3, 0 HR, 2 RBI (projected 0.0 H, 0.0 HR) [run impact 0.8]

Starter miss: Logan Webb (SF) posted 40.50 ERA vs projected 3.00 (9 ER in 6 outs); he is materially worse pitching away (4.17 ERA there vs 3.39 on the other side), a venue-role weakness the model under-weighted.

What to do: Enable/raise the venue-split weight (context.venue_adjustment_weight) and validate in shadow; the split is credible for this starter.

Pitcher versus batter: what actually broke this game

Per-pitch matchup fields are not populated in the current feeds, so this is reconstructed from the settled box: each starter's workload-adjusted ERA, the bullpen share behind him, and the opposing lineup's real production against its projection, with the hitters who beat that projection named.

AWAY pitching

Logan Webb (SF) threw 6 outs, 9 ER = 40.5 ERA against a projected 3 ERA (37.5 ERA error)

Bullpen behind him: 0 ER over 3 outs.

The HOME lineup he faced produced 13 runs on 19 hits and 1 HR, against a projected 4 runs / 8 hits (9 runs vs projection).

Who did the damage to him:

Brandon Lowe: 4-for-6, 1 HR, 1 RBI [+2.34 runs vs projection] season xwOBA 0.322, above average - real projection error
Jacob Gonzalez: 2-for-2, 0 HR, 1 RBI [+1.11 runs vs projection] season xwOBA 0.306, average - a good day, not a model error
Nick Gonzales: 3-for-5, 0 HR, 1 RBI [+0.94 runs vs projection] season xwOBA 0.327, above average - real projection error

Verdict: pitching: the starter gave up runs at a far worse rate than projected

HOME pitching

Paul Skenes (PIT) threw 14 outs, 5 ER = 9.64 ERA against a projected 1.69 ERA (7.95 ERA error)

Bullpen behind him: 2 ER over 7 outs.

The AWAY lineup he faced produced 10 runs on 13 hits and 1 HR, against a projected 3 runs / 6 hits (7 runs vs projection).

Who did the damage to him:

Drew Cavanaugh: 3-for-5, 0 HR, 2 RBI [+2.03 runs vs projection] season xwOBA 0.324, above average - real projection error
Turner Hill: 4-for-5, 0 HR, 1 RBI [+0.77 runs vs projection] season xwOBA 0.346, elite - real projection error
Osleivis Basabe: 1-for-3, 0 HR, 1 RBI [+0.66 runs vs projection] season xwOBA 0.283, weak - a good day, not a model error

Verdict: pitching: the starter gave up runs at a far worse rate than projected

Starting pitchers & relievers: actual / projected + context (ERA is workload-aware: ER over 9 innings, so a short blow-up is not read like a quality start; ERA error in parentheses when material)

Team	Pitcher	Outs	ERA	ER	H	HR	BB	SO	Context (as-of)
SF	Logan Webb (Starter)	6 / 18	40.5 / 3 (+37.50)	9 / 2	9 / 6	0 / 0	2 / 0	1 / 8	venue away: 4.17 vs 3.39 ERA (worse here) · opp run factor 0.92 (strong) · 24 prior appearances this season
SF	Trent Harris (Middle Reliever (High Leverage))	3 / 3	0 / 18 (-18.00)	0 / 2	1 / 2	0 / 2	1 / 0	2 / 0
PIT	Paul Skenes (Starter)	14 / 16	9.64 / 1.69 (+7.95)	5 / 1	8 / 2	1 / 0	3 / 0	4 / 6	venue home: 3.59 vs 4.01 ERA (ok) · opp run factor 1.1 (weak) · 27 prior appearances this season
PIT	Gregory Soto (Middle Reliever (High Leverage))	2 / 4	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0
PIT	Luke Weaver (Setup Man)	2 / 3	27 / 9 (+18.00)	2 / 1	2 / 1	0 / 0	1 / 0	0 / 1
PIT	Mason Montgomery (Closer)	3 / 4	0 / 20.25 (-20.25)	0 / 3	2 / 3	0 / 1	0 / 1	0 / 1

Hitters: actual / projected (sorted by run impact vs projection)

Team	Batter	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
PIT	Brandon Lowe (2B)	6 / 4	4 / 1	0 / 0	0 / 0	1 / 0	1 / 1	3 / 1	0 / 0	0 / 1	2.34	0.322 above average	projection too low
SF	Drew Cavanaugh (C)	5 / 3	3 / 0	2 / 0	0 / 0	0 / 0	2 / 0	1 / 0	0 / 0	1 / 1	2.03	0.324 above average	projection too low
PIT	Jacob Gonzalez (SS)	2 / 3	2 / 1	1 / 0	0 / 0	0 / 0	1 / 0	1 / 0	1 / 0	0 / 1	1.11	0.306 average	no signal
PIT	Nick Gonzales (DH)	5 / 4	3 / 1	0 / 0	0 / 0	0 / 0	1 / 1	1 / 0	0 / 0	0 / 1	0.94	0.327 above average	projection too low
PIT	Henry Davis (C)	3 / 3	1 / 0	0 / 0	0 / 0	0 / 0	2 / 0	1 / 0	1 / 0	1 / 1	0.8	0.286 below average	good day
SF	Turner Hill (LF)	5 / 3	4 / 1	0 / 1	0 / 0	0 / 0	1 / 1	2 / 0	0 / 1	1 / 1	0.77	0.346 elite	projection too low
SF	Osleivis Basabe (2B)	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	1 / 0	2 / 0	2 / 0	1 / 1	0.66	0.283 weak	good day
PIT	Bryan Reynolds (LF)	5 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 0	3 / 1	0 / 1	0 / 1	-0.64	0.36 elite	bad day
SF	Christian Koss (SS)	5 / 3	2 / 1	0 / 0	0 / 0	0 / 0	1 / 0	2 / 0	0 / 0	1 / 1	0.47	0.24 weak	—
SF	Shay Whitcomb (3B)	3 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0.47	0.251 weak	—
PIT	Spencer Horwitz (1B)	6 / 4	2 / 1	0 / 0	0 / 0	0 / 0	2 / 0	1 / 0	0 / 0	3 / 1	0.47	0.312 average	—
SF	Rafael Devers (1B)	4 / 4	1 / 1	0 / 0	0 / 0	1 / 1	3 / 1	1 / 1	2 / 1	1 / 1	0.33	0.34 elite	—
PIT	Esmerlyn Valdez (RF)	5 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	0 / 1	3 / 1	-0.33	0.331 above average	—
PIT	Jared Triolo (3B)	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	1 / 0	0 / 1	0.33	0.289 below average	—
PIT	Oneil Cruz (CF)	6 / 5	4 / 1	0 / 0	0 / 0	0 / 1	4 / 1	1 / 1	0 / 1	1 / 1	0.15	0.352 elite	—
SF	Jung Hoo Lee (RF)	5 / 4	0 / 1	0 / 0	0 / 0	0 / 0	1 / 1	1 / 1	1 / 0	0 / 0	-0.14	0.315 average	—
SF	Bryce Eldridge (DH)	4 / 5	1 / 1	0 / 1	0 / 0	0 / 0	0 / 1	1 / 1	2 / 1	1 / 1	0.02	0.357 elite	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	0.44	positive favors home; negative favors away
Recent form	-0.353	positive favors home; negative favors away
Starting-pitcher matchup	-0.172	positive favors home; negative favors away
Composite	-0.043	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.28-6.15	n/a	Base ML historical model prediction
variance_applied	4-6	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-6	0.545	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run, matchup_recency_guardrail, season_quality_guardrail)
embedded_savant_run_environment	5-6	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	5-6	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	5-6	n/a	Capped at max 12 realistic runs
simulation_result	5-6	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.28-6.15	4-6	0.128	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-6	5-6	-1	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run, matchup_recency_guardrail, season_quality_guardrail)
embedded_savant_run_environment	5-6	5-6	0	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	5-6	5-6	0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	5-6	5-6	0	Capped at max 12 realistic runs
simulation_result	5-6	5-6	0	Final inning-by-inning simulation result

Largest player-level adjustments from the ultra-detailed trace

These are pregame model adjustments, not claims about actual postgame player error.

Team	Player	Role	Net effect	Largest change
SF	Turner Hill	batter / LF	worse	predicted_SO 1.583
PIT	Spencer Horwitz	batter / 1B	better	predicted_RBI 1.57
SF	Bryce Eldridge	batter / DH	worse	predicted_H -1.54
PIT	Esmerlyn Valdez	batter / RF	worse	predicted_SO -1.468
SF	Drew Gilbert	batter / CF	boosted	predicted_SO 1.433
SF	Drew Cavanaugh	batter / C	worse	predicted_AB 1.385
SF	Rafael Devers	batter / 1B	better	predicted_RBI 1.36
PIT	Jared Triolo	batter / 3B	better	predicted_SO 1.326

Structural audit findings

Severity	Rule	Finding	Subject
info	CURRENT_PITCHER_DUPLICATION	PIT pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	PIT
info	CURRENT_PITCHER_DUPLICATION	SF pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	SF
info	CURRENT_FINALIZED_MATCH_BOX	PIT model-prediction batting totals differ from final simulation box score (AB: current=33, sim=36; H: current=8, sim=11; 2B: current=1, sim=3; HR: current=1, sim=2; RBI: current=4, sim=6; R: current=4, sim=6; BB: current=3, sim=1).	PIT
info	CURRENT_PLAYER_MATCH_BOX	PIT current finalized hitter lines are model predictions and differ from the simulation box score: Nick Gonzales: H current=1 box=2, 2B current=0 box=1, RBI current=1 box=0, R current=0 box=1; Oneil Cruz: AB current=5 box=4, H current=1 box=2, HR current=1 box=0, BB current=1 box=0; Esmerlyn Valdez: AB current=3 box=4, RBI current=1 box=0, R current=1 box=0, BB current=1 box=0; Spencer Horwitz: AB current=4 box=5, HR current=0 box=1, RBI current=0 box=2, R current=0 box=1; Jared Triolo: AB current=3 box=4, H current=1 box=2, HR current=0 box=1, RBI current=0 box=1; Brandon Lowe: AB current=4 box=3, H current=1 box=0, RBI current=1 box=0, R current=1 box=0; Bryan Reynolds: H current=1 box=2, RBI current=0 box=1, BB current=1 box=0, SO current=1 box=0; Henry Davis: AB current=3 box=4, H current=0 box=1, 2B current=0 box=1, RBI current=0 box=1	PIT
info	CURRENT_FINALIZED_MATCH_BOX	SF model-prediction batting totals differ from final simulation box score (AB: current=31, sim=32; H: current=7, sim=6; 3B: current=0, sim=1; R: current=4, sim=5; BB: current=3, sim=1; SF: current=0, sim=1).	SF
info	CURRENT_PLAYER_MATCH_BOX	SF current finalized hitter lines are model predictions and differ from the simulation box score: Turner Hill: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, SO current=1 box=2; Bryce Eldridge: AB current=5 box=4, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Osleivis Basabe: AB current=3 box=4, H current=1 box=2, 2B current=0 box=1, RBI current=0 box=1; Jung Hoo Lee: 2B current=0 box=1; Drew Gilbert: AB current=3 box=4, 3B current=0 box=1, SO current=1 box=2; Christian Koss: H current=1 box=0, SO current=1 box=0; Rafael Devers: AB current=4 box=3, RBI current=1 box=2, BB current=1 box=0, SO current=1 box=0; Drew Cavanaugh: AB current=3 box=4, H current=0 box=1, R current=0 box=1, SO current=1 box=2	SF
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

Cross-game lever tension

Each candidate fix, and exactly which games it would help versus hurt. A lever that repairs one game and breaks two is visible here before it is proposed.

Luna deep post-mortem (full)

> Coverage receipt: COMPLETE · 9/8 sections; 15/15 games; 8/8 causes; 6/6 sources

1. Evidence receipt and gaps

Sources received

All seven manifest entries were checked against the supplied evidence:

Source	Path	Size	Status	Use
Prediction review	`/app/output/reports/prediction_review_20260901.json`	373,178 bytes	Loaded	Official outcomes and scoring errors
Forensic deep audit	`/app/output/reports/PREDICTION_SYSTEM_DEEP_AUDIT_20260901.json`	240,806 bytes	Loaded	Probability, provenance, starter, and decision contracts
Continuous learning	`/app/output/learning/continuous_learning_daily_20260901.json`	1,939,278 bytes	Loaded	Stage quality, calibration, lineage, score distribution, starter comparisons
Detailed report	`/app/output/reports/detailed_20260901.html`	355,722 bytes	Loaded	Full game and simulation rendering
Ultra-detailed player audit	`/app/output/reports/20260901_ultra_detailed_player_audit.html`	6,288,497 bytes	Loaded	Player-level prediction trace
Structural audit	`/app/output/reports/audit_report_20260901.html`	383,197 bytes	Loaded	Invariant findings
Integrity revision audit	`/app/output/reports/audit_report_20260901_integrity_revision.html`	Not supplied	Missing	Post-remediation proof for immutable revision artifacts

The missing integrity-revision audit is the principal evidence gap. It does not, by itself, prove a prediction error, but it prevents a fully closed artifact-revision audit.

Slate coverage

The evidence reports:

Expected games: 15.
Games reviewed: 15.
Continuous-learning games: 15.
Forensic deep-audit games: 15.
Structural audits completed: 15.
Starter comparisons: 15.
Canonical artifacts loaded: 15.
Settled games: 15.
Hits: 7.
Misses: 8.
Learning eligibility: eligible because all game artifacts passed critical/error gates.

The 15 reviewed games were:

ATL@WAS, BAL@COL, CHW@HOU, DET@MIN, MIA@KC, MIL@CHC, NYM@TB, NYY@LAA, OAK@TEX, PHI@ARI, SD@CIN, SEA@BOS, SF@PIT, STL@LAD, and TOR@CLE.

Internal contradictions and gaps

Missing integrity-revision artifact. The manifest explicitly marks one source as missing.
One material probability-lineage defect. TOR@CLE states that the publication delta has “no recorded transformation.” The delta is published probability 0.555236 versus governed/calibrated probability 0.545, a +0.010236 movement. This is a lineage/contract defect even though it did not change the winner.
One structural warning. STL@LAD has a pitch-count mismatch for Michael McGreevy: summary BF 24 versus box-score BF 23.
Aggregate audit contradiction requiring care. The slate-level structural summary reports zero critical and zero error findings, one warning, and 108 informational findings. That is consistent with the per-game records supplied, but “all artifacts passed critical/error gates” does not mean the slate is artifact-perfect: the warning, unexplained TOR probability delta, and missing integrity-revision source remain.
Calibration-label inconsistency. The calibration object defines error as average published confidence minus observed hit rate, with positive meaning overconfidence. The 0.50–0.55 and 0.55–0.60 buckets are labeled “underconfident” despite negative errors, which is directionally consistent with the definition. The high buckets have zero hits and are clearly overconfident. However, the displayed bucket errors for the high buckets equal average confidence rather than a separately rounded confidence-minus-hit-rate field; this is mathematically expected when observed hit rate is zero, not necessarily a defect.
Starter and player data are retrospective comparisons. They support diagnosis but do not establish that a pregame input was wrong unless the relevant pregame signal and transformation are shown.

---

2. Executive diagnosis

Overall slate result

The system went 7–8, or 46.67% winner accuracy. The continuous-learning summary reports:

Winner accuracy: 0.4667, 7 hits.
Winner Brier score: 0.3364 is not supplied slate-wide, but per-game Brier values are supplied.
Historical walk-forward accuracy: 0.566181.
Historical walk-forward Brier: 0.243016.
Model-validation winner accuracy: 0.592233.
Model-validation winner Brier: 0.238561.
Weighted absolute calibration error on this slate: 0.2476.
Mean team-run error: materially poor, with the enhanced_prediction stage the worst measured stage.
Exact run-model holdout fit: weak, with home/away/total R² of 0.0361 / 0.0227 / 0.0280 over 2,062 rows.

A correct winner is not equivalent to a good prediction. The slate contains several correct winners with large score errors, including SF@PIT and CHW@HOU.

Winner accuracy

Winner accuracy was poor relative to the model’s historical validation:

7 of 15 winners correct.
8 misses.
Four high-raw-probability selections lost: BAL@COL, MIL@CHC, SD@CIN, STL@LAD.
The high-raw cluster was 0–4 despite raw probabilities of 0.6793–0.7934 and published pick probabilities of 0.58–0.7461.

The main winner-direction problem was not uniformly one thing:

Starter projection was listed as the primary driver in four losses: BAL@COL, MIA@KC, SD@CIN, and SEA@BOS.
Run-environment miss was primary in two: DET@MIN and STL@LAD.
Small-edge losing-side miss was primary in ATL@WAS.
Wrong conviction on the selected side was explicitly secondary or primary in MIA@KC, MIL@CHC, SD@CIN, and STL@LAD.

Calibration quality

Calibration was the clearest probability-quality failure.

Weighted absolute calibration error: 0.2476 across 15 games.
0.50–0.55 bucket: average confidence 0.5352, observed hit rate 0.625, error −0.0898.
0.55–0.60 bucket: average confidence 0.5651, observed hit rate 0.6667, error −0.1015.
0.60–0.65 bucket: average confidence 0.6232, 0 hits in 2 games.
0.65–0.70 bucket: average confidence 0.6986, 0 hits in 1 game.
0.70–0.80 bucket: average confidence 0.7461, 0 hits in 1 game.

The low-confidence buckets were underconfident on this sample, while the higher-confidence buckets were severely overconfident. The evidence supports a bounded shadow adjustment, not an immediate permanent recalibration, because the high-bucket sample is only four games and the causal register correctly warns that a single slate is too small to establish a stable calibration curve.

Score accuracy

Score accuracy was materially poor and structurally biased toward narrow margins:

13 of 15 predicted margins were 2 runs or fewer.
9 of 15 actual margins were 4 or more.
Mean predicted margin: 1.67.
Mean actual margin: 4.20.
Actual margins ranged from 1 to 13.
The largest miss was DET@MIN: predicted margin 1, actual margin 13.
The enhanced_prediction stage had the largest team-run regression versus historical ML:
Team-run MAE: 3.2667 versus 3.0977 historical ML.
Team-run MAE delta: +0.169.
Margin MAE: 4.9333 versus 4.1769.
Winner accuracy: 6/15, or 40%.

The score_margin_compression causal issue is therefore materially supported and has severity high, not critical. The do-not-infer guardrail remains important: one blowout is not enough to widen all scores. Here, however, the distributional evidence spans nine actual margins of at least four and two consecutive slates in the supplied trend record.

Starter accuracy

Starter projections were a significant source of score and winner error:

Starter errors were cited in four losses as the primary driver.
The starter comparison fields are present for all 15 games.
Some large errors were likely variance:
SEA@BOS: Jedixson Paez +6 ER over projection.
SF@PIT: Logan Webb +7 ER and Paul Skenes +4 ER.
MIA@KC: Randy Dobnak +5 ER and Tyler Phillips −4 ER.
Some evidence supports underweighted context rather than pure variance:
MIA@KC: Dobnak’s home ERA was 2.945 versus 1.141 on the other side, with a credible venue split.
MIL@CHC: Robert Gasser’s away ERA was 5.595 versus 4.05 on the other side, with a credible venue split.

The lever-tension record says the pitcher-weight lever “helped” 10 games and hurt none by the supplied attribution, but that is a descriptive slate count, not a causal estimate.

Artifact integrity

Artifact integrity was mostly sound but not clean:

Structural summary: 0 critical, 0 error, 1 warning, 108 info.
Fourteen games had structural-audit status PASS.
STL@LAD had status WARNING because of a pitch-count discrepancy.
TOR@CLE had an unexplained +0.010236 publication delta.
The integrity-revision audit was missing.
All games were nevertheless marked learning-eligible because no critical/error structural gate failed.

The high-severity artifact_integrity_warning issue is appropriately preserved: the warning is a presentation or contract risk, not proof that the model chose the wrong side.

Causal issue register reconciliation

Issue	Severity	Evidence and affected games	Action	Guardrail
Canonical artifact warning requires reconciliation	High	1 warning, 108 info; no critical/error. No specific game list.	Fix the flagged stat/presentation contract before considering the slate clean.	Does not prove a wrong-side pick.
Published margins compressed	High	13/15 predicted margins ≤2; 9/15 actual margins ≥4; mean margin 1.67 versus 4.20. Affected games listed as `CHW@HOU`, `DET@MIN`, `MIL@CHC`, `NYM@TB`, `PHI@ARI`, `STL@LAD`, `TOR@CLE`.	Check rolling-slate tail coverage and margin MAE before widening score spread.	Do not widen every score because of one blowout.
Single-slate calibration instability	High	WAC error 0.2476 across all 15 games.	Monitor the same buckets over the rolling window before changing calibration weights.	Fifteen games do not establish a stable calibration curve.
Weak run-model signal	High	Home/away/total R² 0.0361/0.0227/0.0280 over 2,062 rows.	Investigate features and target via isolated training backtest.	Positive R² does not make exact scores reliable.
High raw-probability cluster	Medium	0/4 high-raw selections won.	Track across multiple slates; preserve raw→calibrated→published lineage.	Do not create a new bucket-shrinkage rule from one slate.
Complete slate audit coverage	Info	15/15 structurally audited.	Keep per-game audit receipts.	Coverage does not cancel internal findings.
Publication delta explained	Info	9 material movements had named transformations.	Continue displaying every layer and transformation.	Published > calibrated is not a bypass when explained.
Simulation inert on slate	Info for current slate; escalated across slates	0 flip candidates and 0 published flips.	Measure simulation contribution over rolling slates.	No winner flips does not prove no score reconciliation occurred.

---

3. Every game

Probabilities below are published probability for the selected/predicted side. The signed confidence edge is home probability − away probability; therefore negative values indicate an away selection.

ATL@WAS — Miss; wrong winner and wrong score

Prediction: ATL 7, WAS 5; winner ATL.
Actual: ATL 5, WAS 9; winner WAS.
Published selected-side probability: 0.5317 for ATL. Home-minus-away edge: −0.0634.
Bucket: Lean.
Errors: Brier 0.282705; team-run absolute error 6; total bias −2; predicted margin 2 versus actual margin 4.
Starter comparison:
AJ Smith-Shawver: projected 3 ER/13 outs; actual 5 ER/12 outs; +2 ER, −1 out.
Jake Irvin: projected 2 ER/15 outs; actual 2 ER/12 outs; 0 ER, −3 outs.
Edges: Season −0.6015; recent −0.318; pitcher −0.043; composite −0.865 home-minus-away.
Controls: Governor applied; raw pick probability 0.4969; governed 0.5082; recency guardrail applied; score coherence plus matchup-recency guardrail; volatility action no change.
Score path: Historical ML 4.77–4.83; variance 5–5; enhanced 6–5; Savant 7–5; later stages unchanged.
Player adjustments: Material changes 24. Positive adjustments included Drake Baldwin RBI +2.24, Ronald Acuña Jr. hits +1.82, Matt Olson RBI +1.42, and Brady House RBI +1.34. Washington also had deboosts to CJ Abrams, Abimelec Ortiz, Yohandy Morales, and Harry Ford.
Audit: Structural PASS, risk 0, no defects, seven informational notes. Forecast-file integrity clean.
Primary cause: Lean-side miss.
Secondary cause: None formally assigned.
Forensics: WAS offense was under-projected by 4. Smith-Shawver’s +2 ER was classified as an outlier start/variance; no credible venue or opponent-strength explanation was found.

BAL@COL — Miss; wrong winner and wrong score

Prediction: BAL 7, COL 4; winner BAL.
Actual: BAL 2, COL 4; winner COL.
Published selected-side probability: 0.746135 for BAL. Home-minus-away edge: −0.4923.
Bucket: Strong.
Errors: Brier 0.556717; team-run error 5; total bias +5; predicted margin 3 versus actual margin 2.
Starters:
Kyle Bradish: projected 3 ER/16 outs; actual 2 ER/15 outs; −1 ER, −1 out.
Gabriel Hughes: projected 5 ER/13 outs; actual 1 ER/14 outs; −4 ER, +1 out.
Edges: Season −1.0874; recent −1.418; pitcher −0.0404; composite −1.5815.
Controls: Raw 0.736061; governed 0.7373; no recency guardrail; score-coherence movement to published 0.746135; supported blowout-margin expansion.
Score path: Historical 5.44–4.39; variance 5–4; enhanced 7–4; Savant and later stages 7–4.
Player adjustments: Material changes 24. Gunnar Henderson hits +1.95, Brett Sullivan hits +1.28, Jackson Holliday hits +1.27, Pete Alonso AB +1.244, and Colton Cowser AB +1.24 were among the largest.
Audit: Structural PASS, risk 0, forecast artifact clean.
Primary cause: Starter projection.
Secondary cause: Run environment over-projected.
Forensics: Hughes allowed four fewer ER than projected. The evidence suggests possible opponent-strength weighting relevance, but the supplied synthesis recommends validation rather than a permanent change.

CHW@HOU — Hit; correct winner, wrong score

Prediction: CHW 6, HOU 5; winner CHW.
Actual: CHW 5, HOU 1; winner CHW.
Published selected-side probability: 0.53 for CHW. Home-minus-away edge: −0.0600.
Bucket: Lean.
Errors: Brier 0.2209; team-run error 5; total bias +5; predicted margin 1 versus actual margin 4.
Starters:
Sean Burke: projected 4 ER/15 outs; actual 0 ER/17 outs; −4 ER, +2 outs.
Ronel Blanco: projected 4 ER/13 outs; actual 0 ER/12 outs; −4 ER, −1 out.
Edges: Season −0.363; recent +0.283; pitcher −0.0483; composite −0.4152.
Controls: Raw 0.53; governed 0.5082; recency guardrail applied; score coherence/recency path; volatility no change.
Score path: Historical 4.84–5.00; variance 5–5; enhanced 3–5; Savant and later stages 3–5, although the supplied predicted score is 6–5, indicating a stored score-path versus published-score inconsistency that should be reconciled.
Player adjustments: No player-level forensic record was available.
Audit: Structural PASS, risk 0.
Finding: Correct winner/wrong score.
Cause: No formal primary driver; both starters materially outperformed ER projections, and HOU’s offense was over-projected.

DET@MIN — Miss; wrong winner and largest score miss

Prediction: DET 4, MIN 3; winner DET.
Actual: DET 2, MIN 15; winner MIN.
Published selected-side probability: 0.53 for DET. Home-minus-away edge: −0.0600.
Bucket: Lean.
Errors: Brier 0.2809; team-run error 14; total bias −10; predicted margin 1 versus actual margin 13.
Starters:
Troy Melton: projected 2 ER/18 outs; actual 5 ER/16 outs; +3 ER, −2 outs.
Andrew Morris: projected 1 ER/14 outs; actual 1 ER/3 outs; 0 ER, −11 outs.
Edges: Season −0.2973; recent +0.534; pitcher −0.0249; composite −0.2648.
Controls: Raw 0.53; governed 0.505; recency guardrail; enhanced stage explicitly shrank confidence because of a one-run margin and recency controls; no volatility adjustment.
Score path: Historical 5.07–3.72; variance 5–4; enhanced 4–3; Savant and final stages 4–3.
Player adjustments: Material changes 25. Large adjustments included Kody Clemens strikeout +2.075, John Peck RBI deboost −1.86, Walker Jenkins RBI −1.47, and Troy Melton projected ER adjustment −1.0.
Audit: Structural PASS, risk 0.
Primary cause: Run environment under-projected.
Secondary cause: None formally assigned.
Forensics: MIN scored 12 more than projected as a team. Melton allowed +3 ER, while the five-run ninth-inning-type bullpen contribution was also important: Kyle Finnegan allowed 5 ER versus 0 projected. The model treated the realized explosion as variance rather than a demonstrated persistent input failure.

MIA@KC — Miss; wrong winner and wrong score

Prediction: MIA 3, KC 5; winner KC.
Actual: MIA 6, KC 3; winner MIA.
Published selected-side probability: 0.646375 for KC. Home-minus-away edge: +0.2928.
Bucket: Strong.
Errors: Brier 0.417801; team-run error 5; total bias −1; predicted margin 2 versus actual margin 3.
Starters:
Tyler Phillips: projected 4 ER/15 outs; actual 0 ER/15 outs; −4 ER.
Randy Dobnak: projected 1 ER/17 outs; actual 6 ER/18 outs; +5 ER, +1 out.
Edges: Season −0.2701; recent +0.648; pitcher +0.052; composite +0.4878.
Controls: Raw 0.642335; governed 0.6227; no recency guardrail; score coherence raised published probability to 0.646375; no volatility action.
Score path: Historical 4.84–5.00; variance 5–5; enhanced 3–5; Savant and final stages 3–5.
Player adjustments: Material changes 21. Graham Pauley strikeout +2.341; Salvador Perez AB −1.89; Carter Jensen AB −1.73; Bobby Witt Jr. AB −1.65; Heriberto Hernández RBI +1.43.
Audit: Structural PASS, risk 0.
Primary cause: Starter projection.
Secondary cause: Wrong conviction side.
Forensics: Dobnak’s credible venue split showed materially worse home performance: 2.945 ERA at home versus 1.141 on the other side. The supplied recommendation is to enable or raise context.venue_adjustment_weight in shadow testing.

MIL@CHC — Miss; wrong winner and wrong score

Prediction: MIL 6, CHC 8; winner CHC.
Actual: MIL 9, CHC 4; winner MIL.
Published selected-side probability: 0.698612 for CHC. Home-minus-away edge: +0.3972.
Bucket: Strong.
Errors: Brier 0.488059; team-run error 7; total bias +1; predicted margin 2 versus actual margin 5.
Starters:
Robert Gasser: projected 5 ER/15 outs; actual 2 ER/13 outs; −3 ER, −2 outs.
Matthew Boyd: projected 4 ER/18 outs; actual 3 ER/16 outs; −1 ER, −2 outs.
Edges: Season −0.0089; recent +0.482; pitcher −0.023; composite +0.425.
Controls: Raw 0.914; calibrated/governed 0.6442 due to sample size 10; no recency guardrail; score coherence yielded 0.698612; supported blowout-margin expansion.
Score path: Historical 4.47–4.63; variance 4–5; enhanced 5–8; Savant 6–8; final unchanged.
Player adjustments: Material changes 21. Tyrone Taylor RBI +3.37 and Jackson Chourio RBI +2.28 were large positive adjustments; Seiya Suzuki AB −2.24 and Christian Yelich strikeout +1.86 were also substantial.
Audit: Structural PASS, sparse historical bucket diagnostic flag.
Primary cause: Wrong conviction side.
Secondary cause: Small-edge/side error despite a strong published selection.
Forensics: Both starters outperformed ER projections, especially Gasser. His credible away split was materially worse than his other-side ERA, suggesting underweighted venue-role information. The game was also a high-raw-probability cluster loss.

NYM@TB — Hit; correct winner, wrong score

Prediction: NYM 5, TB 6; winner TB.
Actual: NYM 2, TB 6; winner TB.
Published selected-side probability: 0.526945 for TB. Home-minus-away edge: +0.0539.
Bucket: Lean.
Errors: Brier 0.223781; team-run error 3; total bias +3; predicted margin 1 versus actual margin 4.
Starters:
Sean Manaea: projected 2 ER/18 outs; actual 5 ER/17 outs; +3 ER, −1 out.
Freddy Peralta: projected 4 ER/16 outs; actual 1 ER/18 outs; −3 ER, +2 outs.
Edges: Season +0.7945; recent +0.086; pitcher +0.1335; composite +0.9322.
Controls: Raw 0.53; governed 0.5082; recency guardrail; score coherence and recency path; no volatility action.
Score path: Historical 4.83–5.00; variance 5–5; enhanced 5–5; Savant/final 5–6.
Player adjustments: No player-level forensic record was available.
Audit: Structural PASS.
Finding: Correct winner/wrong score.
Cause: No formal primary driver; opposing starter errors offset in winner direction while NYM offense was over-projected.

NYY@LAA — Hit; correct winner, wrong score

Prediction: NYY 6, LAA 4; winner NYY.
Actual: NYY 7, LAA 3; winner NYY.
Published selected-side probability: 0.54 for NYY. Home-minus-away edge: −0.0800.
Bucket: Lean.
Errors: Brier 0.2116; team-run error 2; total bias 0; predicted margin 2 versus actual margin 4.
Starters:
Gerrit Cole: projected 1 ER/18 outs; actual 3 ER/16 outs; +2 ER, −2 outs.
Grayson Rodriguez: projected 3 ER/15 outs; actual 4 ER/18 outs; +1 ER, +3 outs.
Edges: Season −1.2252; recent −0.490; pitcher −0.0343; composite −1.2318.
Controls: Raw 0.54; governed 0.5332; recency guardrail; score coherence and recency path; no volatility action.
Score path: Historical stage not numerically displayed beyond final trace; published score 6–4.
Player adjustments: No player-level forensic record was available.
Audit: Structural PASS.
Finding: Correct winner/wrong score, but relatively accurate runs.
Cause: No formal primary or secondary driver.

OAK@TEX — Hit; correct winner, wrong score

Prediction: OAK 4, TEX 5; winner TEX.
Actual: OAK 5, TEX 8; winner TEX.
Published selected-side probability: 0.548725 for TEX. Home-minus-away edge: +0.0974.
Bucket: Lean.
Errors: Brier 0.203649; team-run error 4; total bias −4; predicted margin 1 versus actual margin 3.
Starters:
Brady Basso: projected 4 ER/13 outs; actual 2 ER/7 outs; −2 ER, −6 outs.
MacKenzie Gore: projected 3 ER/16 outs; actual 1 ER/16 outs; −2 ER.
Edges: Season +0.9405; recent +0.242; pitcher +0.0026; composite +0.7635.
Controls: Raw 0.5517; governed 0.5499; recency guardrail; published probability essentially matched governed value; no volatility action.
Score path: Published 4–5; detailed intermediate numeric path is not separately supplied.
Player adjustments: No player-level forensic record was available.
Audit: Structural PASS.
Finding: Correct winner/wrong score.
Cause: TEX offense under-projected by 3; total under-projected by 4.

PHI@ARI — Hit; correct winner, wrong score

Prediction: PHI 4, ARI 3; winner PHI.
Actual: PHI 7, ARI 1; winner PHI.
Published selected-side probability: 0.560194 for PHI. Home-minus-away edge: −0.1204.
Bucket: Moderate.
Errors: Brier 0.193429; team-run error 5; total bias −1; predicted margin 1 versus actual margin 6.
Starters:
Jesús Luzardo: projected 1 ER/20 outs; actual 1 ER/20 outs; exact ER and outs.
Eduardo Rodríguez: projected 1 ER/18 outs; actual 3 ER/15 outs; +2 ER, −3 outs.
Edges: Season −0.3931; recent −0.648; pitcher −0.0855; composite −0.7463.
Controls: Raw 0.5884; governed 0.5816; recency guardrail; score coherence/recency path; no volatility action.
Score path: Published 4–3; intermediate numeric stages not separately supplied.
Player adjustments: No player-level forensic record was available.
Audit: Structural PASS.
Finding: Correct winner/wrong score.
Cause: PHI offense under-projected by 3; Rodríguez also allowed two more ER than projected.

SD@CIN — Miss; wrong winner and wrong score

Prediction: SD 7, CIN 4; winner SD.
Actual: SD 3, CIN 4; winner CIN.
Published selected-side probability: 0.60 for SD. Home-minus-away edge: −0.2000.
Bucket: Strong.
Errors: Brier 0.36; team-run error 4; total bias +4; predicted margin 3 versus actual margin 1.
Starters:
Randy Vásquez: projected 3 ER/15 outs; actual 2 ER/14 outs; −1 ER, −1 out.
Nick Lodolo: projected 6 ER/12 outs; actual 2 ER/18 outs; −4 ER, +6 outs.
Edges: Season −0.8428; recent −0.699; pitcher +0.0764; composite −1.0102.
Controls: Raw 0.60; governed 0.6778, based on a raw/governor path with recency guardrail; published 0.60 after score coherence and matchup-recency guardrail; supported blowout-margin expansion.
Score path: Historical 4.46–3.81; variance 4–4; enhanced 6–4; Savant/final 7–4.
Player adjustments: Material changes 23. Xander Bogaerts RBI +3.53, Héctor Rodríguez RBI +2.55, Ty France RBI +2.26, and several SD lineup adjustments raised the selected-side projection.
Audit: Structural PASS.
Primary cause: Starter projection.
Secondary cause: Wrong conviction side.
Forensics: Lodolo allowed four fewer ER and six more outs than projected. The retrospective synthesis calls this variance and does not recommend a model change.

SEA@BOS — Miss; wrong winner and wrong score

Prediction: SEA 3, BOS 5; winner BOS.
Actual: SEA 9, BOS 6; winner SEA.
Published selected-side probability: 0.53 for BOS. Home-minus-away edge: +0.0600.
Bucket: Lean.
Errors: Brier 0.2809; team-run error 7; total bias −7; predicted margin 2 versus actual margin 3.
Starters:
Bryan Woo: projected 1 ER/17 outs; actual 1 ER/18 outs; exact ER, +1 out.
Jedixson Paez: projected 2 ER/9 outs; actual 8 ER/11 outs; +6 ER, +2 outs.
Edges: Season +0.9535; recent +0.551; pitcher −0.0054; composite +1.0103.
Controls: Raw 0.53; governed 0.6273; recency guardrail and multiple confidence-shrink controls; published 0.53 after score coherence/recency; supported blowout-margin expansion.
Score path: Historical 6.19–0.95; variance 6–1; enhanced 3–5; Savant/final 3–5.
Player adjustments: Material changes 24. Lazaro Montes RBI +2.554, Caleb Durbin RBI +1.63, Mickey Gasper RBI +1.556, and multiple SEA lineup deboosts.
Audit: Structural PASS.
Primary cause: Starter projection.
Secondary cause: Run environment under-projected.
Forensics: Paez allowed six more ER than projected. The synthesis classifies this as an outlier start/variance; no credible venue or opponent-strength signal explains it.

SF@PIT — Hit; correct winner, very wrong score

Prediction: SF 5, PIT 6; winner PIT.
Actual: SF 12, PIT 13; winner PIT.
Published selected-side probability: 0.544522 for PIT. Home-minus-away edge: +0.0890.
Bucket: Lean.
Errors: Brier 0.20746; team-run error 14; total bias −14; predicted margin 1 versus actual margin 1.
Starters:
Logan Webb: projected 2 ER/18 outs; actual 9 ER/6 outs; +7 ER, −12 outs.
Paul Skenes: projected 1 ER/16 outs; actual 5 ER/14 outs; +4 ER, −2 outs.
Edges: Season +0.4402; recent −0.3527; pitcher −0.1715; composite −0.0434.
Controls: Raw 0.559024; governed 0.56; recency guardrail; score coherence/recency path; no volatility action.
Score path: Published 5–6; intermediate numeric path not separately supplied.
Player adjustments: No player-level forensic record was available.
Audit: Structural PASS.
Finding: Correct winner/wrong score.
Cause: Both offenses and both starters were far outside the score projection. The winner survived a severe run-environment miss.

STL@LAD — Miss; wrong winner and wrong score

Prediction: STL 5, LAD 7; winner LAD.
Actual: STL 13, LAD 8; winner STL.
Published selected-side probability: 0.58 for LAD. Home-minus-away edge: +0.1600.
Bucket: Moderate.
Errors: Brier 0.3364; team-run error 9; total bias −9; predicted margin 2 versus actual margin 5.
Starters:
Michael McGreevy: projected 3 ER/17 outs; actual 4 ER/15 outs; +1 ER, −2 outs.
Eric Lauer: projected 5 ER/16 outs; actual 7 ER/9 outs; +2 ER, −7 outs.
Edges: Season +1.1307; recent +0.5433; pitcher +0.0064; composite +1.1318.
Controls: Raw 0.58; governed 0.6696; recency guardrail; score coherence/recency guardrail yielded published 0.58; supported blowout-margin expansion.
Score path: Historical 4.21–4.42; variance 4–4; enhanced 4–6; Savant 5–7; final 5–7.
Player adjustments: Material changes 27. Tommy Edman RBI +4.554, José Fermín RBI −2.638, Alex Freeland AB +2.214, Kyle Tucker hits +2.18, and several LAD adjustments.
Audit: WARNING, risk 10. Michael McGreevy pitch-count row had summary BF 24 versus box BF 23. One structural defect/warning was reported in the game-level diagnostic.
Primary cause: Run environment under-projected.
Secondary cause: Wrong conviction side.
Forensics: STL offense was under-projected by 8. Lauer allowed +2 ER, and the player-level results included multiple unexpected STL home runs. The synthesis labels the starter miss variance.

TOR@CLE — Hit; correct winner, wrong score; lineage defect

Prediction: TOR 4, CLE 5; winner CLE.
Actual: TOR 1, CLE 6; winner CLE.
Published selected-side probability: 0.555236 for CLE. Home-minus-away edge: +0.1105.
Bucket: Moderate.
Errors: Brier 0.197815; team-run error 4; total bias +2; predicted margin 1 versus actual margin 5.
Starters:
Spencer Miles: projected 2 ER/15 outs; actual 0 ER/3 outs; −2 ER, −12 outs.
Gavin Williams: projected 3 ER/16 outs; actual 0 ER/21 outs; −3 ER, +5 outs.
Edges: Season +0.4158; recent +0.3953; pitcher −0.1167; composite +0.0708.
Controls: Raw 0.548941; governed 0.545; no recency guardrail; score coherence probability 0.564983; published 0.555236.
Score path: Enhanced-to-fundamental transition changed the side from tie to CLE and improved winner accuracy.
Player adjustments: No player-level forensic record was available.
Audit: Structural PASS at the game level, but probability lineage contains a defect.
Finding: Correct winner/wrong score.
Cause: TOR offense over-projected by 3 and Williams allowed three fewer ER than projected.
Lineage issue: The +0.010236 published-minus-calibrated movement has no recorded transformation. This is an artifact/contract defect, not evidence that the winner was wrong.

---

4. Miss reconstruction

Earliest wrong-direction stage for each loss

A later recovery cannot erase an intermediate regression. The stage audit must therefore distinguish final outcome from path quality.

ATL@WAS

Historical ML selected WAS by the underlying home-minus-away direction in the transition record (WAS → TIE after variance).
variance_applied moved the direction from WAS toward a tie and cost one winner hit overall; ATL@WAS is explicitly listed as regressed.
The published pick ultimately favored ATL only weakly.
Classification: Small-edge side miss; likely winner-direction instability plus realized WAS offense. Starter variance contributed but was not the formal primary cause.

BAL@COL

The earliest harmful information was already present in the historical/quality edge: strong BAL superiority.
The enhanced stage expanded BAL from roughly 5–4 to 7–4.
Classification: Bad weighting or incomplete context, especially Hughes/opponent-strength and venue-role response; then starter error and run-environment overprojection. The loss was not caused by a late winner flip.

DET@MIN

Historical ML was near-neutral and variance moved the game toward DET/TIE dynamics.
Enhanced prediction reduced the projected score to DET 4, MIN 3 and retained the wrong DET direction.
Classification: Run-environment miss began in the narrow score construction and was amplified by an extreme MIN offensive/bullpen outcome. Melton’s +3 ER and Finnegan’s +5 ER were important, but the 12-run MIN underprojection is primarily a run-tail failure.

MIA@KC

Historical ML was near 5–5; variance stayed neutral.
Enhanced prediction changed to MIA 3, KC 5 and produced the KC direction.
Classification: Bad weighting of pitcher/context inputs. Dobnak’s credible home split and Phillips’ expected weakness were not sufficiently represented. This was a starter projection and conviction failure, not merely a random one-run miss.

MIL@CHC

The historical score was nearly even; variance moved to MIL 4, CHC 5.
Enhanced prediction sharply moved CHC to 8 runs and generated a strong CHC direction.
Classification: Winner-direction regression at the enhanced/recent-form stage, compounded by high-raw overconfidence and starter/venue mismatch. The model overweighted recent CHC offensive evidence and underweighted Gasser’s credible away weakness.

SD@CIN

Historical ML favored SD slightly.
Variance applied moved the direction from SD to tie, an intermediate winner-direction regression.
Enhanced prediction then recovered SD and increased its score to 6–4, and Savant raised it to 7–4.
Classification: The final side was still wrong, but the pipeline had an explicit intermediate regression at historical_ml → variance_applied; later recovery does not erase it. The final loss was principally a starter projection error: Lodolo allowed four fewer ER and six more outs than projected, with conviction amplified by a 0.60 published pick probability.

SEA@BOS

Historical ML produced SEA 6, BOS 1, favoring SEA by score.
Variance retained SEA direction.
Enhanced prediction reversed the score to SEA 3, BOS 5 based on recent form, starter-quality, recency, and season-quality controls.
This reversal was not listed as a winner regression in the transition table, but it is the earliest substantive movement toward the wrong final side in the game trace.
Classification: Bad weighting of recent/season and starter-quality controls, followed by an extreme Paez starter miss. The home confidence was weak in publication, so this was not a high-confidence calibration failure.

STL@LAD

Historical ML was nearly neutral.
Variance stayed tied.
Enhanced prediction moved to STL 4, LAD 6; Savant increased LAD to 7 and supported a blowout expansion.
Fundamental edges strongly favored LAD, and the published side remained LAD.
Classification: Run-environment underdispersion and over-reliance on LAD quality; secondary wrong conviction. McGreevy’s pitch-count warning is an artifact/data-quality issue, but it does not establish that the pick was caused by the warning.

Error-type classification across losses

Bad input: No critical bad-input defect was established. The STL pitch-count inconsistency is a warning; the missing integrity-revision file is an artifact gap.
Bad weighting: Best supported in MIA@KC and MIL@CHC through credible venue-role splits that were underweighted; also plausible in SEA@BOS due to aggressive form/quality controls.
Calibration: Strongly implicated in the confidence level of BAL@COL, MIL@CHC, SD@CIN, and STL@LAD; not necessarily the cause of the underlying side selection.
Starter error: Primary in four losses; substantial in nearly every loss.
Offense/bullpen miss: Strong in DET@MIN, STL@LAD, SEA@BOS, and ATL@WAS.
Run-environment miss: Primary in DET@MIN and STL@LAD, and broadly visible in the slate margin compression.
Irreducible variance: Supported for the large outlier starts in ATL@WAS, DET@MIN, SD@CIN, SEA@BOS, and STL@LAD where no credible contextual signal explains the deviation.

Pipeline-stage quality across all stages

Stage	Winner accuracy	Winner hits	Margin MAE	Team-run MAE	Total-run MAE
Historical ML	46.67%	7	4.1769	3.0977	4.6893
Variance applied	40.00%	6	4.1333	3.0667	4.6667
Enhanced prediction	40.00%	6	4.9333	3.2667	4.6667
Embedded Savant/run environment	40.00%	6	4.8667	3.1667	4.4667
Fundamental edge target review	46.67%	7	4.8000	3.1333	4.5333
Capped target	46.67%	7	4.8000	3.1333	4.5333
Simulation result	46.67%	7	4.8000	3.1333	4.5333
Published final	46.67%	7	4.8000	3.1333	4.5333

Transition audit

historical_ml → variance_applied
Margin MAE improved by 0.0436.
Team-run MAE improved by 0.031.
Winner accuracy fell by 6.67 percentage points, from 7 to 6.
Changed directions: ATL@WAS WAS→TIE, MIA@KC KC→TIE, SD@CIN SD→TIE, STL@LAD LAD→TIE.
ATL@WAS was explicitly regressed.
variance_applied → enhanced_prediction
Margin MAE worsened by 0.8.
Team-run MAE worsened by 0.2.
Winner accuracy was flat.
NYM@TB improved; SEA@BOS regressed.
Direction changed in six games, including SEA@BOS BOS→SEA in the stored direction notation.
enhanced_prediction → embedded_savant_run_environment
Small margin improvement of 0.0666 and team-run improvement of 0.1.
TOR@CLE moved from TOR to tie; no winner accuracy change.
embedded_savant_run_environment → fundamental_edge_target_review
TOR@CLE moved tie to CLE, improving one winner.
Margin and team-run MAE improved slightly.
All later transitions were inert:
No changed directions.
No metric changes.
No simulation winner intervention.

The key conclusion is that the enhanced stage was the main score-quality regression versus historical ML, while fundamental-edge review recovered one winner. The recovery does not erase the earlier regressions.

---

5. Cross-game counterfactuals

Lever tension 1: Widen score distributions

Lever: score.variance_multiplier / score.variance_floor

Helps: BAL@COL, MIA@KC, MIL@CHC, SD@CIN, and STL@LAD.
Hurts: None listed.
Net evidence: 5 helps versus 0 hurts.
Interpretation: This is the strongest direct slate-level case for a bounded score-tail shadow experiment. It is consistent with:
13/15 predicted margins ≤2.
9/15 actual margins ≥4.
Persistent negative margin bias.
Caution: The lever should not be globally widened solely because DET@MIN was a 13-run margin. The broader distributional evidence, not that single blowout, justifies testing.

Lever tension 2: Pitcher form weights and prior games

Lever: player.pitcher.v4 form weights / player.pitcher_prior_games

Helps: Ten games by the supplied attribution:
BAL@COL, CHW@HOU, DET@MIN, MIA@KC, MIL@CHC, NYM@TB, SD@CIN, SEA@BOS, SF@PIT, TOR@CLE.
Hurts: None listed.
Net evidence: 10 helps versus 0 hurts.
Interpretation: Strong descriptive support for a shadow test, but the “help” labels are based on starter error and should not be treated as proof that more recent form always improves forecasts. Extreme starter outcomes create noisy retrospective attribution.

Lever tension 3: Recent/season run-mean weighting

Lever: recent.* / season.* run-mean weighting

Helps: 13 games:

ATL@WAS, BAL@COL, CHW@HOU, DET@MIN, MIA@KC, MIL@CHC, NYM@TB, OAK@TEX, PHI@ARI, SD@CIN, SEA@BOS, SF@PIT, STL@LAD, TOR@CLE — the JSON’s help list contains 13 named games and omits two listed slate games from the help count.

Hurts: NYY@LAA, where team runs were already close, with an explicit warning that reweighting could disturb it.
Net evidence: 13 helps versus 1 hurt.
Interpretation: The net count supports investigation, but this lever is risky because recent and season signals already drove the wrong side in SEA@BOS and contributed to high-confidence errors in MIL@CHC/STL@LAD.

Tempting one-game fixes that must not be made

Do not widen every score because of DET@MIN. The causal register explicitly prohibits this.
Do not permanently shrink all high probabilities because BAL@COL, MIL@CHC, SD@CIN, and STL@LAD lost. The four-game cluster is a warning, not a stable calibration curve.
Do not increase venue weighting solely from MIA@KC. MIL@CHC provides supporting evidence, but the venue split must be validated across the rolling window.
Do not remove recency controls because of SEA@BOS. NYM@TB, NYY@LAA, and PHI@ARI show that recency/quality information can coexist with correct winners.
Do not treat the simulation as a failed winner selector solely because it produced zero flips. Its persistent inertness requires a code/diagnostic check, not an arbitrary parameter move.
Do not treat the STL pitch-count warning as the cause of the STL loss. It is a data-contract issue, not evidence of wrong-side causation.

---

6. Data and contract defects

Critical findings

Critical count: 0.
No critical data or contract finding was supplied.

Error findings

Error count: 0.
No error-severity structural finding was supplied.

Warning findings

STL@LAD: pitch-count mismatch

Player: Michael McGreevy.
Rule: PITCH_COUNT_PITCHER_MATCH.
Summary BF: 24.
Box-score BF: 23.
Source: simulation.
Severity: warning.
Game risk score: 10.
This is an artifact/data-contract warning, not established prediction causation.

Canonical artifact warning

Slate structural totals: 0 critical, 0 error, 1 warning, 108 info.
The exact aggregate warning above is the only supplied warning-level finding.

Informational findings

The evidence supplies an aggregate count of 108 informational findings but does not enumerate their individual messages. The major named informational categories are:

Complete 15/15 audit coverage.
Nine material calibrated-to-published movements with named transformations.
Zero simulation flip candidates and zero published flips.
Per-game forecast-file informational notes, usually seven or eight per game.
No material integrity defects in most game artifacts.

Because individual info messages are not present in the supplied JSON, they cannot be exhaustively itemized without inventing data.

Prediction-model defects versus presentation/artifact defects

Prediction-model/process risks

Compressed predicted margins and weak score tails.
Weak holdout run-model signal.
High raw-probability cluster losing 0–4.
Poor current-slate calibration.
Persistent simulation inertness.
Persistent run-model weakness.
Potentially underweighted pitcher venue-role effects in MIA@KC and MIL@CHC.
Potentially harmful enhancement-stage score regression.

Presentation/artifact defects

Missing integrity-revision audit artifact.
STL@LAD pitch-count contract warning.
TOR@CLE unexplained publication delta.
Possible score-path versus published-score reconciliation issue in CHW@HOU: the stored score evolution ends at 3–5 while the published prediction is 6–5. This should be checked against the canonical artifact before calling it a defect; the supplied JSON does not provide enough information to determine whether this is a display-path omission or a real artifact mismatch.

Coverage reconciliation

The slate cannot be called completely clean despite full coverage:

Coverage is complete: 15/15.
Critical/error gates are clear: yes.
Warning-free: no.
Complete revision evidence: no, because one manifest artifact is missing.
Complete probability lineage: no, because TOR has an unexplained material delta.

Probability-lineage audit

The required chain is raw model → calibrated → governor → score coherence/recency → published.

Games with generally recorded lineage

The following games show named transformations and no unexplained material delta:

ATL@WAS, BAL@COL, CHW@HOU, DET@MIN, MIA@KC, MIL@CHC, NYM@TB, NYY@LAA, OAK@TEX, PHI@ARI, SD@CIN, SEA@BOS, SF@PIT, and STL@LAD.

The continuous-learning finding states that nine material movements were explained by named score-coherence and/or recency transformations. The high-raw cluster specifically records all four movements as explained.

Bypass/defect

TOR@CLE: calibrated/governed probability 0.545; published probability 0.555236; delta +0.010236; transformation text says “no recorded transformation explains the publication delta.”
This qualifies as a material lineage defect under the supplied contract, although it did not alter the selected winner.

---

7. Cross-slate persistence

Four prior slates were loaded, creating a five-slate window from 2026-08-28 through 2026-09-01.

Finding streaks

Finding	Consecutive-slate count	Present now?	Interpretation
`high_raw_probability_cluster`	1	Yes	Current slate has the cluster, but the streak itself is not persistent.
`pipeline_winner_instability`	0	No	Not present now; incidence in window is 2 but no current consecutive streak.
`score_margin_compression`	2	Yes	Current and prior consecutive slate; persistent emerging run-model issue.
`simulation_inert_on_slate`	5	Yes	Established failure across all five slates.
`single_slate_calibration_instability`	2	Yes	Current and prior consecutive slate; persistent emerging calibration issue.
`weak_run_model_signal`	5	Yes	Established failure across all five slates.

The escalated_findings entries must be treated as established failures, not single-slate noise:

Simulation inertness

Streak: 5 consecutive slates.
Target lever listed: simulation.allow_score.
Metric group: run model.
Metrics hurt: team_run_mae, total_run_mae.
Direction: confirm simulation is running and not silently no-oping.
Parameter actionability: false; this requires code/diagnostic action rather than forced tuning.

Weak run-model signal

Streak: 5 consecutive slates.
Target levers listed: training.xgb.n_estimators, training.xgb.max_depth, training.xgb.learning_rate.
Metrics hurt: team_run_mae, total_run_mae.
Direction: investigate features and target with an isolated training backtest.
Parameter actionability: false in the registered finding; do not force a hyperparameter change from this report.

Metric trends

All reported directions must be stated explicitly:

Margin bias: Regressing.
Mean: −2.4068.
Series: −1.333, −1.882, −2.786, −3.5, −2.533.
Interpretation: increasingly underpredicted margins, consistent with compression.
Mean team-run error on losses: Regressing.
Mean: 6.075.
Series: 5.0, 3.0, 7.0, 8.25, 7.125.
Weighted calibration error: Regressing.
Mean: 0.2194.
Series: 0.2179, 0.3961, 0.0693, 0.1661, 0.2476.
Winner accuracy: Regressing.
Mean: 0.6083.
Series: 0.4667, 0.9412, 0.5, 0.6667, 0.4667.

Top-confidence-pick streak

The top-confidence pick history is:

2026-08-28: hit.
2026-08-29: hit.
2026-08-30: miss.
2026-08-31: hit.
2026-09-01: miss.

Current loss streak: 1. The top pick is not on a multi-slate losing streak, but the current miss is consistent with the separate high-confidence cluster concern.

Raw structural blocks

Margin distribution: persistent negative margin bias and current compression are present.
Pipeline transitions: variance application improved run errors slightly but reduced winner accuracy; enhancement materially worsened margin MAE; later recovery came at fundamental review; simulation and publication were inert.
Calibration monitoring: current high-confidence buckets are severely overconfident, while low-confidence buckets are underconfident.
High-raw cluster: current 0–4; only one-slate consecutive streak but incidence in the five-slate window is 2.
Simulation control: five-slate persistence, diagnostic-only status, 0 flips in 15 current games.

---

8. Adjustments for today

Recommendations are ranked by evidence strength and persistence. These are bounded shadow or diagnostic actions, not promotions.

1. Diagnose the inert simulation path — highest persistence, code action

Pattern: simulation_inert_on_slate, escalated for 5 consecutive slates.
Registered lever: simulation.allow_score.
Action: Do not change the parameter blindly. Run a code/diagnostic verification that simulation.allow_score is actually read, that score-level simulation executes, and that its output can differ from the capped target.
Expected metric: If the module is currently a no-op, restoring execution should change team_run_mae and total_run_mae; direction should be improvement.
Helps: Potentially all score forecasts, especially compressed-margin games: DET@MIN, SEA@BOS, SF@PIT, STL@LAD.
Hurts/regression risk: Any game where simulation adds unjustified tail variance, including relatively close-score successes such as NYY@LAA and TOR@CLE.
Gate: Require observable nonzero simulation contribution on shadow output and rollback if rolling team-run or total-run MAE worsens.

2. Run an isolated run-model feature/target backtest — highest persistent model weakness

Pattern: weak_run_model_signal, escalated for 5 consecutive slates.
Registered targets: training.xgb.n_estimators, training.xgb.max_depth, training.xgb.learning_rate.
Action: Code/diagnostic action, not an immediate parameter move. Test the existing features and target definition in an isolated temporal backtest before selecting a bounded hyperparameter shadow.
Expected metric: Improve team-run MAE and total-run MAE; R² should improve without degrading winner Brier.
Helps: All run forecasts, especially DET@MIN, SEA@BOS, SF@PIT, STL@LAD, and PHI@ARI.
Hurts/regression risk: Increased tail flexibility could worsen close-score games such as NYY@LAA, NYM@TB, and TOR@CLE.
Gate: Do not promote unless improvement persists across rolling temporal folds and does not materially worsen winner accuracy or Brier.

3. Bounded score-tail shadow experiment — persistent compression

Pattern: score_margin_compression, present for 2 consecutive slates; margin bias is regressing.
Exact lever: score.variance_multiplier — increase slightly, for example +0.05 in shadow only. Keep score.variance_floor unchanged initially to isolate the effect.
Expected metric: Reduce negative margin bias and improve team_run_mae/total_run_mae; expected direction is improvement.
Helps: The lever-tension list identifies BAL@COL, MIA@KC, MIL@CHC, SD@CIN, and STL@LAD.
Hurts: No hurt games were listed in the direct tension record, but regression risk exists for close or already-accurate games, especially NYY@LAA, NYM@TB, and TOR@CLE.
Gate: Shadow against the unchanged multiplier; rollback if margin MAE or Brier worsens over the registered rolling evaluation.

4. Bounded calibration-confidence shadow — current calibration and high-raw cluster

Patterns: single_slate_calibration_instability has a 2-slate streak and is present; high_raw_probability_cluster is present with a current 0–4 result but only a one-slate consecutive streak.
Exact lever: confidence.base_cap — reduce modestly, for example 0.01 in shadow. Do not simultaneously change all three confidence levers.
Expected metric: Lower weighted calibration error and improve Brier by reducing overconfidence in the 0.60+ region.
Helps: Current high-confidence misses BAL@COL, MIL@CHC, SD@CIN, and STL@LAD.
Hurts: Correct high-conviction or moderate calls could become underconfident, especially MIA@KC’s opposing-side calibration context, and low-confidence successes could lose useful separation.
Gate: Monitor bucket calibration over the rolling window; rollback if low-confidence underconfidence worsens or Brier deteriorates.

5. Pitcher-form/venue shadow — material but less persistent than run-model failures

Pattern: Starter projection was the primary driver in four losses; the lever-tension count is 10 helped versus 0 hurt.
Exact lever: player.pitcher.v4 form weights — increase the recent-form component slightly in shadow, without changing player.pitcher_prior_games simultaneously.
Expected metric: Improve starter_er_mae and starter_outs_mae.
Helps: The tension record names BAL@COL, CHW@HOU, DET@MIN, MIA@KC, MIL@CHC, NYM@TB, SD@CIN, SEA@BOS, SF@PIT, and TOR@CLE.
Hurts: No direct hurt games were listed, but recent-form overweighting could damage stable starter baselines and exacerbate SEA@BOS-type reversals.
Important companion diagnostic: Separately shadow context.venue_adjustment_weight for credible venue-role cases, especially MIA@KC and MIL@CHC; do not bundle this with the form-weight experiment.
Gate: Require improvement in starter MAE without worsening winner Brier or introducing excess score volatility.

6. Single strongest current-slate driver: run-environment underprojection

The strongest single-slate aggregate driver is run-environment compression/underdispersion, because it affected the largest errors and is supported by a two-slate streak and five-slate regressing margin bias.

The bounded score.variance_multiplier recommendation above is therefore the safest direct single-slate parameter experiment. Do not separately tune a one-game DET/MIN-specific factor.

---

9. Bottom line

Known

The slate was 7–8, with poor winner accuracy and especially poor calibration at higher confidence.
Score margins were materially compressed: 13 predicted margins ≤2 versus 9 actual margins ≥4.
The enhancement stage was the worst measured stage for margin and team-run MAE.
Starter projection errors materially contributed to four losses, while several extreme deviations are best classified as variance.
Four high-raw-probability selections lost.
Simulation produced no winner-direction intervention and has done so for five consecutive slates.
Weak run-model signal has persisted for five consecutive slates.
All 15 games were reviewed and structurally audited.
There were no critical or error-level structural findings.
There was one structural warning, one missing manifest artifact, and one unexplained material probability-publication delta.

Uncertain

Whether widening score variance will improve future slates rather than merely fit this run of high-margin outcomes.
Whether pitcher-form or venue adjustments are causal improvements rather than retrospective explanations.
Whether the high-confidence cluster reflects calibration failure, winner variance, or a short-lived sample anomaly.
Whether the simulation is genuinely inert or whether its score contribution is simply not represented in winner-flip metrics.
Whether the CHW@HOU score-path discrepancy is a true artifact defect or an omitted intermediate rendering detail.
The individual content of the 108 informational audit findings is not included.

Safest next action

The single safest next action is:

> Run a non-promoting diagnostic/shadow audit of the simulation path and probability/score lineage, while testing a small increase to score.variance_multiplier only in shadow.

This addresses the five-slate simulation/run-model persistence, directly tests the current score-compression failure, preserves rollback ability, and avoids making an irreversible one-game calibration or winner-side correction.

Severity	Rule	Finding	Subject
info	CURRENT_PITCHER_DUPLICATION	KC pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	KC
info	CURRENT_PITCHER_DUPLICATION	MIA pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	MIA
info	CURRENT_FINALIZED_MATCH_BOX	KC model-prediction batting totals differ from final simulation box score (AB: current=36, sim=29; H: current=8, sim=7; RBI: current=4, sim=5; R: current=4, sim=5; BB: current=3, sim=4; SO: current=7, sim=6; SF: current=0, sim=2).	KC
info	CURRENT_PLAYER_MATCH_BOX	KC current finalized hitter lines are model predictions and differ from the simulation box score: Carter Jensen: AB current=5 box=3, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0; Bobby Witt Jr.: AB current=5 box=3, RBI current=0 box=2, SO current=1 box=0; Nick Loftin: H current=1 box=0, BB current=0 box=1; Jac Caglianone: AB current=5 box=4, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Salvador Perez: AB current=5 box=3, HR current=0 box=1, RBI current=1 box=2, R current=0 box=1; Josh Rojas: AB current=3 box=4, SO current=0 box=1; Vinnie Pasquantino: 2B current=0 box=1, RBI current=1 box=0, BB current=1 box=0, SO current=1 box=2; Isaac Collins: H current=0 box=1, R current=0 box=1, BB current=0 box=1, SO current=1 box=0	KC
info	CURRENT_FINALIZED_MATCH_BOX	MIA model-prediction batting totals differ from final simulation box score (AB: current=31, sim=32; H: current=8, sim=5; 2B: current=2, sim=0; RBI: current=4, sim=3; R: current=4, sim=3; BB: current=3, sim=4).	MIA
info	CURRENT_PLAYER_MATCH_BOX	MIA current finalized hitter lines are model predictions and differ from the simulation box score: Kyle Stowers: AB current=5 box=3, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Griffin Conine: H current=1 box=0, RBI current=1 box=0, BB current=0 box=1, SO current=1 box=0; Otto Lopez: AB current=5 box=4, H current=1 box=2, RBI current=1 box=0, SO current=1 box=0; Heriberto Hernández: AB current=3 box=4, HR current=1 box=0, RBI current=1 box=2, BB current=1 box=0; Jakob Marsee: H current=1 box=0, R current=1 box=0; Javier Sanoja: AB current=3 box=4, 2B current=1 box=0, SO current=0 box=1; Owen Caissie: AB current=3 box=4, HR current=0 box=1, RBI current=0 box=1, R current=0 box=1; Graham Pauley: AB current=3 box=4, SO current=1 box=3	MIA
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

Team	Starter	Projected	Actual	ER err
MIL	Robert Gasser	5 ER / 15 outs	2 ER / 13 outs	-3
CHC	Matthew Boyd	4 ER / 18 outs	3 ER / 16 outs	-1

Team	Pitcher	Outs	ERA	ER	H	HR	BB	SO	Context (as-of)
MIL	Robert Gasser (Starter)	13 / 15	4.15 / 9 (-4.85)	2 / 5	3 / 8	0 / 1	2 / 1	4 / 4	venue away: 5.59 vs 4.05 ERA (worse here) · opp run factor 1.09 (strong) · 16 prior appearances this season
MIL	Aaron Ashby (Setup Man)	3 / 2	0 / 40.5 (-40.50)	0 / 3	0 / 4	0 / 0	0 / 1	1 / 1
CHC	Matthew Boyd (Starter)	16 / 18	5.06 / 6	3 / 4	7 / 3	0 / 2	1 / 3	3 / 8	venue home: 5.44 vs 2.83 ERA (worse here) · opp run factor 1.11 (strong) · 17 prior appearances this season
CHC	Ryan Zeferjahn (Middle Reliever (High Leverage))	2 / 3	27 / 9 (+18.00)	2 / 1	2 / 1	0 / 0	0 / 0	1 / 0

Team	Batter	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
CHC	Pete Crow-Armstrong (CF)	4 / 6	0 / 2	0 / 1	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	3 / 1	-2.51	0.369 elite	bad day
MIL	Joey Ortiz (3B)	4 / 3	2 / 1	0 / 0	0 / 0	1 / 0	2 / 0	1 / 0	0 / 0	0 / 0	1.4	0.295 below average	good day
CHC	Alex Bregman (3B)	4 / 5	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	0 / 1	1 / 1	-1.26	0.318 average	no signal
MIL	William Contreras (C)	5 / 4	2 / 1	2 / 0	0 / 0	0 / 0	1 / 1	2 / 1	0 / 0	1 / 0	1.09	0.333 above average	projection too low
CHC	Matt Shaw (2B)	3 / 3	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	1 / 1	0.64	0.307 average	no signal
MIL	Christian Yelich (DH)	4 / 4	1 / 1	0 / 1	0 / 0	1 / 0	2 / 1	2 / 1	1 / 1	0 / 1	0.62	0.314 average	no signal
MIL	Brice Turang (2B)	4 / 5	2 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	1 / 1	0 / 1	0.47	0.335 above average	—
CHC	Carson Kelly (C)	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	0 / 0	0 / 1	-0.47	0.311 average	—
MIL	Jackson Chourio (LF)	5 / 5	2 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	0 / 0	1 / 1	-0.46	0.334 above average	—
CHC	Seiya Suzuki (RF)	2 / 5	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	1 / 1	2 / 1	1 / 1	-0.45	0.339 above average	—
CHC	Tyrone Taylor (LF)	2 / 3	1 / 1	1 / 0	0 / 0	0 / 0	2 / 1	1 / 0	0 / 0	0 / 0	0.31	0.319 average	—
MIL	Jake Bauers (RF)	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	3 / 1	1 / 1	0.19	0.373 elite	—
MIL	Andrew Vaughn (1B)	3 / 3	1 / 1	1 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 1	0 / 1	-0.02	0.336 above average	—
CHC	Michael Busch (1B)	4 / 5	1 / 1	1 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	1 / 1	-0.02	0.328 above average	—
MIL	Luis Lara (CF)	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 1	0	0.313 average	—
MIL	Cooper Pratt (SS)	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	1 / 1	0	0.313 average	—
CHC	Nico Hoerner (SS)	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0	0.324 above average	—
CHC	BJ Murray Jr. (DH)	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0	0.08 weak	—

Signal	Home-away edge	Meaning
Season quality	-0.009	positive favors home; negative favors away
Recent form	0.482	positive favors home; negative favors away
Starting-pitcher matchup	-0.023	positive favors home; negative favors away
Composite	0.425	positive favors home; negative favors away

Stage	Away-home runs	Confidence	What changed
historical_ml	4.47-4.63	n/a	Base ML historical model prediction
variance_applied	4-5	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-8	0.699	Enhanced ML with recent form, H2H, low-score analysis
embedded_savant_run_environment	6-8	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	6-8	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	6-8	n/a	Capped at max 12 realistic runs
simulation_result	6-8	n/a	Final inning-by-inning simulation result

Stage	Before	After	Toward final pick	What changed
variance_applied	4.47-4.63	4-5	0.846	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-5	5-8	2	Enhanced ML with recent form, H2H, low-score analysis
embedded_savant_run_environment	5-8	6-8	-1	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	6-8	6-8	0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	6-8	6-8	0	Capped at max 12 realistic runs
simulation_result	6-8	6-8	0	Final inning-by-inning simulation result

Team	Player	Role	Net effect	Largest change
CHC	Tyrone Taylor	batter / LF	better	predicted_RBI 3.37
MIL	Jackson Chourio	batter / LF	better	predicted_RBI 2.28
CHC	Seiya Suzuki	batter / RF	worse	predicted_AB -2.24
MIL	Christian Yelich	batter / DH	worse	predicted_SO 1.86
CHC	Nico Hoerner	batter / SS	better	predicted_RBI 1.534
CHC	Pete Crow-Armstrong	batter / CF	worse	predicted_AB -1.53
CHC	Alex Bregman	batter / 3B	worse	predicted_RBI -1.5
CHC	Matt Shaw	batter / 2B	worse	predicted_SO 1.471

Severity	Rule	Finding	Subject
info	CURRENT_PITCHER_DUPLICATION	CHC pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	CHC
info	CURRENT_PITCHER_DUPLICATION	MIL pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	MIL
info	CURRENT_FINALIZED_MATCH_BOX	CHC model pitcher projections differ from final simulation pitching box score (H: current=6, sim=8; HR: current=2, sim=4).	CHC
info	CURRENT_FINALIZED_MATCH_BOX	CHC model-prediction batting totals differ from final simulation box score (AB: current=38, sim=37; H: current=10, sim=13; 2B: current=2, sim=3; HR: current=2, sim=1; RBI: current=6, sim=8; R: current=5, sim=8; BB: current=4, sim=3; SO: current=7, sim=8).	CHC
info	CURRENT_PLAYER_MATCH_BOX	CHC current finalized hitter lines are model predictions and differ from the simulation box score: Pete Crow-Armstrong: AB current=6 box=4, 2B current=1 box=0, HR current=1 box=0; Tyrone Taylor: AB current=3 box=5, H current=1 box=2, 2B current=0 box=1, HR current=0 box=1; Nico Hoerner: H current=1 box=2, 2B current=0 box=1, RBI current=0 box=2, R current=0 box=1; Alex Bregman: H current=1 box=2, HR current=1 box=0, RBI current=1 box=0, BB current=1 box=0; BJ Murray Jr.: AB current=3 box=4, R current=0 box=1, SO current=1 box=2; Matt Shaw: AB current=3 box=4, H current=1 box=0, R current=1 box=0, SO current=1 box=2; Seiya Suzuki: AB current=5 box=3, H current=1 box=2, 2B current=1 box=0, RBI current=1 box=0; Michael Busch: AB current=5 box=4, R current=0 box=1, BB current=1 box=0, SO current=1 box=0	CHC
info	CURRENT_FINALIZED_MATCH_BOX	MIL model-prediction batting totals differ from final simulation box score (AB: current=33, sim=34; H: current=9, sim=8; 2B: current=1, sim=0; 3B: current=0, sim=1; HR: current=1, sim=4; RBI: current=5, sim=6; R: current=5, sim=6; BB: current=4, sim=3).	MIL
info	CURRENT_PLAYER_MATCH_BOX	MIL current finalized hitter lines are model predictions and differ from the simulation box score: Jake Bauers: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, SO current=1 box=0; Andrew Vaughn: 3B current=0 box=1, R current=0 box=1; Jackson Chourio: H current=1 box=2, HR current=1 box=2, RBI current=1 box=3, SO current=1 box=0; Brice Turang: RBI current=1 box=0, BB current=1 box=0, SO current=1 box=2; Joey Ortiz: RBI current=0 box=1, R current=0 box=1, SF current=0 box=1; William Contreras: HR current=0 box=1; Christian Yelich: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, R current=1 box=0; Luis Lara: HR current=0 box=1, RBI current=0 box=1, R current=0 box=1, BB current=0 box=1	MIL
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

Team	Starter	Projected	Actual	ER err
NYM	Sean Manaea	2 ER / 18 outs	5 ER / 17 outs	3
TB	Freddy Peralta	4 ER / 16 outs	1 ER / 18 outs	-3

Team	Pitcher	Outs	ERA	ER	H	HR	BB	SO	Context (as-of)
NYM	Sean Manaea (Starter)	17 / 18	7.94 / 3 (+4.94)	5 / 2	11 / 5	1 / 0	1 / 3	2 / 6	venue away: 4.08 vs 4.13 ERA (ok) · opp run factor 1.07 (average) · 28 prior appearances this season
TB	Freddy Peralta (Starter)	18 / 16	1.5 / 6.75 (-5.25)	1 / 4	3 / 5	1 / 2	0 / 2	7 / 5	venue home: 4.09 vs 6.19 ERA (ok) · opp run factor 0.94 (weak) · 27 prior appearances this season
TB	Tyler Wells (Middle Reliever (High Leverage))	5 / 4	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	1 / 1
TB	Kevin Kelly (Setup Man)	3 / 3	0 / 9 (-9.00)	0 / 1	0 / 2	0 / 1	0 / 0	2 / 0
TB	Garrett Cleavinger (Middle Reliever (High Leverage))	1 / 1	27 / 0 (+27.00)	1 / 0	0 / 1	0 / 0	3 / 0	0 / 1

Team	Batter	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
NYM	Juan Soto (DH)	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	0 / 1	-1.73	0.416 elite	bad day
TB	Yandy Díaz (DH)	4 / 5	2 / 1	1 / 0	0 / 0	1 / 0	1 / 1	2 / 1	1 / 1	1 / 1	1.71	0.355 elite	projection too low
TB	Ryan Vilade (1B)	5 / 3	2 / 1	0 / 0	1 / 0	0 / 0	2 / 1	0 / 0	0 / 0	0 / 1	1.09	0.334 above average	projection too low
TB	Victor Mesa Jr. (RF)	4 / 3	3 / 1	0 / 0	0 / 0	0 / 0	0 / 0	2 / 0	0 / 0	0 / 1	0.94	0.311 average	no signal
NYM	Francisco Lindor (SS)	3 / 5	1 / 1	0 / 0	0 / 0	1 / 0	2 / 0	1 / 1	1 / 1	1 / 1	0.93	0.338 above average	projection too low
TB	Junior Caminero (3B)	5 / 4	2 / 1	0 / 0	0 / 0	0 / 1	1 / 1	0 / 1	0 / 1	0 / 1	-0.79	0.363 elite	bad day
TB	Jonny DeLuca (CF)	3 / 3	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	0 / 1	-0.78	0.265 weak	projection too high
TB	Nick Fortes (C)	4 / 3	2 / 1	1 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 0	1 / 0	0.78	0.277 weak	good day
NYM	A.J. Ewing (CF)	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	2 / 1	-0.47	0.317 average	—
NYM	Francisco Alvarez (C)	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	2 / 1	-0.47	0.326 above average	—
TB	Jorge Mateo (2B)	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	-0.47	0.303 below average	—
NYM	Jared Young (1B)	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	1 / 0	1 / 0	2 / 1	-0.45	0.309 average	—
TB	Taylor Walls (SS)	2 / 3	1 / 1	1 / 0	0 / 0	0 / 0	1 / 0	0 / 1	0 / 1	0 / 1	-0.02	0.249 weak	—
NYM	Carson Benge (RF)	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	0	0.336 above average	—
NYM	Brett Baty (2B)	2 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0	0.306 average	—
NYM	Nick Morabito (LF)	1 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 2	0	0.298 below average	—
NYM	Bo Bichette (3B)	4 / 5	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0	0.325 above average	—
TB	Chandler Simpson (LF)	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 0	0	0.261 weak	—

Signal	Home-away edge	Meaning
Season quality	0.794	positive favors home; negative favors away
Recent form	0.086	positive favors home; negative favors away
Starting-pitcher matchup	0.134	positive favors home; negative favors away
Composite	0.932	positive favors home; negative favors away

Stage	Away-home runs	Confidence	What changed
historical_ml	5.31-4.12	n/a	Base ML historical model prediction
variance_applied	5-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-6	0.527	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by winner_starter_is_below_average, projected_margin_is_one_run, matchup_recency_guardrail)
embedded_savant_run_environment	5-6	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	5-6	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	5-6	n/a	Capped at max 12 realistic runs
simulation_result	5-6	n/a	Final inning-by-inning simulation result

Stage	Before	After	Toward final pick	What changed
variance_applied	5.31-4.12	5-4	0.194	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-4	5-6	2	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by winner_starter_is_below_average, projected_margin_is_one_run, matchup_recency_guardrail)
embedded_savant_run_environment	5-6	5-6	0	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	5-6	5-6	0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	5-6	5-6	0	Capped at max 12 realistic runs
simulation_result	5-6	5-6	0	Final inning-by-inning simulation result

Team	Player	Role	Net effect	Largest change
NYM	Bo Bichette	batter / 3B	better	predicted_RBI 2.55
TB	Ryan Vilade	batter / 1B	better	predicted_AB 2.301
NYM	Brett Baty	batter / 2B	worse	predicted_BB 1.719
TB	Yandy Díaz	batter / DH	better	predicted_H 1.65
TB	Jorge Mateo	batter / 2B	worse	predicted_AB 1.417
NYM	Nick Morabito	batter / LF	better	predicted_AB 1.25
TB	Junior Caminero	batter / 3B	better	predicted_AB -1.25
TB	Victor Mesa Jr.	batter / RF	worse	predicted_SO 1.184

Severity	Rule	Finding	Subject
info	CURRENT_PITCHER_DUPLICATION	NYM pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	NYM
info	CURRENT_PITCHER_DUPLICATION	TB pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	TB
info	CURRENT_FINALIZED_MATCH_BOX	NYM model-prediction batting totals differ from final simulation box score (AB: current=36, sim=35; H: current=7, sim=8; HR: current=1, sim=3; RBI: current=3, sim=5; R: current=3, sim=5; SO: current=10, sim=8).	NYM
info	CURRENT_PLAYER_MATCH_BOX	NYM current finalized hitter lines are model predictions and differ from the simulation box score: Juan Soto: HR current=1 box=0; Carson Benge: SO current=1 box=0; Brett Baty: AB current=3 box=2, BB current=0 box=2, SO current=1 box=0; Francisco Lindor: H current=1 box=0, R current=1 box=0, BB current=1 box=0, SO current=1 box=2; A.J. Ewing: BB current=1 box=0, SO current=1 box=2; Nick Morabito: AB current=3 box=4, H current=0 box=1, HR current=0 box=1, RBI current=0 box=1; Jared Young: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Francisco Alvarez: RBI current=1 box=0, R current=0 box=1	NYM
info	CURRENT_FINALIZED_MATCH_BOX	TB model-prediction batting totals differ from final simulation box score (AB: current=31, sim=34; H: current=9, sim=10; 3B: current=0, sim=1; HR: current=1, sim=2; RBI: current=4, sim=6; R: current=5, sim=6; BB: current=3, sim=5; HBP: current=0, sim=1).	TB
info	CURRENT_PLAYER_MATCH_BOX	TB current finalized hitter lines are model predictions and differ from the simulation box score: Ryan Vilade: AB current=3 box=5, HR current=0 box=1, RBI current=1 box=2, R current=0 box=1; Jorge Mateo: AB current=3 box=4; Jonny DeLuca: AB current=3 box=4, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Junior Caminero: AB current=4 box=3, SO current=1 box=0, HBP current=0 box=1; Yandy Díaz: AB current=5 box=4, H current=1 box=3, RBI current=1 box=2, SO current=1 box=0; Nick Fortes: 2B current=0 box=1, RBI current=0 box=1, R current=0 box=1, BB current=0 box=1; Chandler Simpson: AB current=4 box=3, H current=1 box=2, R current=0 box=1, BB current=0 box=1; Victor Mesa Jr.: AB current=3 box=4, H current=1 box=0, SO current=1 box=2	TB
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

Team	Starter	Projected	Actual	ER err
NYY	Gerrit Cole	1 ER / 18 outs	3 ER / 16 outs	2
LAA	Grayson Rodriguez	3 ER / 15 outs	4 ER / 18 outs	1

Team	Pitcher	Outs	ERA	ER	H	HR	BB	SO	Context (as-of)
NYY	Gerrit Cole (Starter)	16 / 18	5.06 / 1.5 (+3.56)	3 / 1	8 / 9	0 / 1	1 / 0	5 / 6	venue away: 3.04 vs 3.35 ERA (ok) · opp run factor 0.98 (weak) · 17 prior appearances this season
NYY	Brent Headrick (Setup Man)	4 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0
NYY	Paul Blackburn (Middle Reliever (High Leverage))	4 / 3	0 / 18 (-18.00)	0 / 2	1 / 1	0 / 0	0 / 4	2 / 0
LAA	Grayson Rodriguez (Starter)	18 / 15	6 / 5.4	4 / 3	7 / 4	2 / 0	2 / 3	7 / 8	venue home: 7.78 vs 3.9 ERA (worse here) · opp run factor 1.01 (average) · 15 prior appearances this season

Team	Batter	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
NYY	Spencer Jones (RF)	3 / 3	2 / 1	0 / 0	0 / 0	1 / 0	3 / 1	2 / 0	2 / 0	1 / 1	2.06	0.329 above average	projection too low
NYY	Jazz Chisholm Jr. (2B)	5 / 3	3 / 1	0 / 0	0 / 0	1 / 0	1 / 0	2 / 1	0 / 0	2 / 1	1.87	0.294 below average	good day
NYY	Ben Rice (DH)	5 / 5	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	0 / 1	3 / 1	-1.26	0.378 elite	bad day
LAA	Mike Trout (CF)	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	2 / 1	-0.8	0.376 elite	bad day
NYY	Luis García Jr. (1B)	5 / 4	2 / 1	1 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	0 / 1	0.78	0.348 elite	projection too low
LAA	Vaughn Grissom (1B)	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	2 / 1	-0.78	0.316 average	no signal
NYY	Ryan McMahon (3B)	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 0	0 / 1	0.66	0.308 average	no signal
LAA	Christian Moore (2B)	4 / 4	1 / 1	0 / 0	1 / 0	0 / 0	2 / 1	0 / 0	0 / 0	1 / 1	0.62	0.238 weak	good day
LAA	Wade Meckler (LF)	4 / 3	3 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0 / 1	0 / 1	0.61	0.312 average	no signal
LAA	Zach Neto (SS)	4 / 4	2 / 1	1 / 0	0 / 0	0 / 1	0 / 1	1 / 1	0 / 1	1 / 1	-0.48	0.305 average	—
NYY	José Caballero (SS)	5 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	3 / 1	-0.47	0.28 weak	—
LAA	Tyler Heineman (C)	3 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0.47	0.243 weak	—
LAA	Josh Lowe (RF)	1 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	0.33	0.29 below average	—
NYY	Cody Bellinger (LF)	5 / 4	3 / 1	0 / 1	0 / 0	0 / 0	0 / 1	2 / 1	0 / 1	1 / 0	0.3	0.342 elite	—
NYY	Trent Grisham (CF)	5 / 4	1 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	0 / 1	2 / 1	0.14	0.337 above average	—
NYY	Austin Wells (C)	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0 / 1	0	0.291 below average	—
LAA	Moisés Ballesteros (DH)	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 1	0	0.291 below average	—
LAA	Oswald Peraza (3B)	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0	0.271 weak	—

Deep slate report — 20260901

Slate summary

What drove the losses

Source coverage receipt

High-level causal issue register

Cross-slate trends (persistent-pattern memory)

Escalated persistent patterns

Finding streaks across the window

Metric direction across the window

Pipeline stage regression

Probability lineage and calibration

Score distribution and margin compression

Per-game evidence coverage

Every game, in depth (15)

ATL@WASMISS

Player-level breakdown (projected vs actual)

Fundamental evidence entering the pick

Score and confidence evolution

Official postgame stage attribution

Largest player-level adjustments from the ultra-detailed trace

Structural audit findings

BAL@COLMISS

Player-level breakdown (projected vs actual)

Fundamental evidence entering the pick

Score and confidence evolution

Official postgame stage attribution

Largest player-level adjustments from the ultra-detailed trace

Structural audit findings

CHW@HOUHIT

Player-level breakdown (projected vs actual)

Fundamental evidence entering the pick

Score and confidence evolution

Official postgame stage attribution

Largest player-level adjustments from the ultra-detailed trace

Structural audit findings

DET@MINMISS

Player-level breakdown (projected vs actual)

Fundamental evidence entering the pick

Score and confidence evolution

Official postgame stage attribution

Largest player-level adjustments from the ultra-detailed trace

Structural audit findings

MIA@KCMISS

Player-level breakdown (projected vs actual)

Fundamental evidence entering the pick

Score and confidence evolution

Official postgame stage attribution

Largest player-level adjustments from the ultra-detailed trace

Structural audit findings

MIL@CHCMISS

Player-level breakdown (projected vs actual)

Fundamental evidence entering the pick

Score and confidence evolution

Official postgame stage attribution

Largest player-level adjustments from the ultra-detailed trace

Structural audit findings

NYM@TBHIT

Player-level breakdown (projected vs actual)

Fundamental evidence entering the pick

Score and confidence evolution

Official postgame stage attribution

Largest player-level adjustments from the ultra-detailed trace

Structural audit findings

NYY@LAAHIT

Player-level breakdown (projected vs actual)

Fundamental evidence entering the pick

Score and confidence evolution

Official postgame stage attribution

Largest player-level adjustments from the ultra-detailed trace

Structural audit findings

OAK@TEXHIT

Player-level breakdown (projected vs actual)

Fundamental evidence entering the pick

Score and confidence evolution

Official postgame stage attribution

Largest player-level adjustments from the ultra-detailed trace

Structural audit findings

PHI@ARIHIT

Player-level breakdown (projected vs actual)

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-1.225	positive favors home; negative favors away
Recent form	-0.49	positive favors home; negative favors away
Starting-pitcher matchup	-0.034	positive favors home; negative favors away
Composite	-1.232	positive favors home; negative favors away

Stage	Away-home runs	Confidence	What changed
historical_ml	4.3-2.98	n/a	Base ML historical model prediction
variance_applied	4-3	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-4	0.54	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by matchup_recency_guardrail)
embedded_savant_run_environment	6-4	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	6-4	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	6-4	n/a	Capped at max 12 realistic runs
simulation_result	6-4	n/a	Final inning-by-inning simulation result

Stage	Before	After	Toward final pick	What changed
variance_applied	4.3-2.98	4-3	-0.324	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-3	5-4	-0	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by matchup_recency_guardrail)
embedded_savant_run_environment	5-4	6-4	1	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	6-4	6-4	-0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	6-4	6-4	-0	Capped at max 12 realistic runs
simulation_result	6-4	6-4	-0	Final inning-by-inning simulation result

Severity	Rule	Finding	Subject
info	CURRENT_PITCHER_DUPLICATION	LAA pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	LAA
info	CURRENT_PITCHER_DUPLICATION	NYY pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	NYY
info	CURRENT_FINALIZED_MATCH_BOX	LAA model-prediction batting totals differ from final simulation box score (AB: current=32, sim=38; H: current=7, sim=11; R: current=3, sim=4; BB: current=3, sim=6; SO: current=9, sim=6).	LAA
info	CURRENT_PLAYER_MATCH_BOX	LAA current finalized hitter lines are model predictions and differ from the simulation box score: Vaughn Grissom: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, BB current=0 box=1; Wade Meckler: AB current=3 box=4, H current=1 box=2, SO current=1 box=0; Moisés Ballesteros: AB current=4 box=5, H current=1 box=3, HR current=0 box=1, RBI current=0 box=2; Josh Lowe: AB current=3 box=5, H current=0 box=2, RBI current=0 box=1, R current=0 box=1; Oswald Peraza: AB current=3 box=5, H current=1 box=0, SO current=1 box=3; Mike Trout: AB current=4 box=3, 2B current=0 box=1, RBI current=1 box=0, R current=1 box=0; Christian Moore: R current=0 box=1, BB current=0 box=1, SO current=1 box=0; Zach Neto: HR current=1 box=0, RBI current=1 box=0, R current=1 box=0	LAA
info	CURRENT_FINALIZED_MATCH_BOX	NYY model-prediction batting totals differ from final simulation box score (AB: current=32, sim=33; H: current=8, sim=6; 2B: current=1, sim=2; RBI: current=5, sim=6; R: current=5, sim=6; BB: current=3, sim=7; SO: current=8, sim=13).	NYY
info	CURRENT_PLAYER_MATCH_BOX	NYY current finalized hitter lines are model predictions and differ from the simulation box score: Spencer Jones: AB current=3 box=5, H current=1 box=2, 2B current=0 box=2, RBI current=1 box=0; Cody Bellinger: AB current=4 box=3, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Jazz Chisholm Jr.: H current=1 box=0, R current=1 box=0, BB current=0 box=1, SO current=1 box=2; Ben Rice: H current=1 box=2, RBI current=1 box=3, BB current=1 box=0, SO current=1 box=3; Luis García Jr.: AB current=4 box=5, RBI current=1 box=2, SO current=1 box=2; Ryan McMahon: RBI current=0 box=1, R current=0 box=1, BB current=0 box=1; Trent Grisham: AB current=4 box=2, BB current=1 box=2; Austin Wells: H current=1 box=0, RBI current=1 box=0, BB current=0 box=1	NYY
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

Team	Starter	Projected	Actual	ER err
OAK	Brady Basso	4 ER / 13 outs	2 ER / 7 outs	-2
TEX	MacKenzie Gore	3 ER / 16 outs	1 ER / 16 outs	-2

Team	Pitcher	Outs	ERA	ER	H	HR	BB	SO	Context (as-of)
OAK	Brady Basso (Starter)	7 / 13	7.71 / 8.31	2 / 4	4 / 5	0 / 0	0 / 0	2 / 4	opp run factor 0.98 (weak) · 20 prior appearances this season
OAK	Luis Medina (Middle Reliever (High Leverage))	3 / 5	9 / 0 (+9.00)	1 / 0	2 / 1	0 / 0	0 / 0	1 / 0
TEX	MacKenzie Gore (Starter)	16 / 16	1.69 / 5.06 (-3.37)	1 / 3	5 / 3	0 / 0	4 / 3	4 / 4	venue home: 2.99 vs 5.62 ERA (ok) · opp run factor 0.98 (weak) · 28 prior appearances this season
TEX	Jacob Latz (Closer)	2 / 4	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	2 / 1
TEX	Chase Silseth (Middle Reliever (High Leverage))	2 / 3	0 / 9 (-9.00)	0 / 1	0 / 2	0 / 0	1 / 0	2 / 1

Team	Batter	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
TEX	Brandon Nimmo (RF)	5 / 4	4 / 1	1 / 0	0 / 0	0 / 0	5 / 1	0 / 1	0 / 0	0 / 1	1.72	0.353 elite	projection too low
TEX	Justin Foscue (DH)	2 / 3	0 / 1	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	0 / 0	0 / 1	-1.4	0.331 above average	bad day
OAK	Jeff McNeil (2B)	4 / 4	2 / 1	0 / 1	0 / 0	1 / 0	2 / 1	1 / 0	0 / 0	0 / 1	1.09	0.307 average	no signal
TEX	Corey Seager (SS)	3 / 4	2 / 1	2 / 0	0 / 0	0 / 0	1 / 0	3 / 1	1 / 1	1 / 1	1.09	0.351 elite	projection too low
TEX	Jake Burger (1B)	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	1 / 1	-0.78	0.312 average	no signal
TEX	Elias Díaz (C)	3 / 3	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	2 / 1	0.64	0.278 weak	good day
OAK	Zack Gelof (LF)	3 / 4	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	2 / 1	2 / 1	1 / 1	-0.6	0.316 average	no signal
OAK	Tommy White (1B)	0 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	-0.47	0.298 below average	—
OAK	Jonah Heim (C)	5 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0.47	0.292 below average	—
TEX	Evan Carter (CF)	3 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 1	0.47	0.28 weak	—
TEX	Ezequiel Duran (2B)	2 / 4	0 / 1	0 / 1	0 / 0	0 / 0	1 / 1	0 / 1	2 / 1	1 / 1	-0.45	0.306 average	—
TEX	Wyatt Langford (LF)	5 / 4	2 / 1	1 / 0	0 / 0	0 / 0	0 / 0	2 / 1	0 / 1	1 / 1	0.45	0.313 average	—
OAK	Henry Bolte (DH)	5 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	3 / 1	0.33	0.311 average	—
OAK	Lawrence Butler (RF)	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	2 / 1	0 / 1	0 / 1	1 / 1	-0.33	0.324 above average	—
OAK	Alika Williams (SS)	2 / 3	1 / 1	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	-0.31	0.245 weak	—
OAK	Max Muncy (3B)	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	-0.14	0.293 below average	—
OAK	Denzel Clarke (CF)	2 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0	0.2 weak	—
TEX	Cody Freeman (3B)	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	1 / 0	0	0.231 weak	—

Stage	Away-home runs	Confidence	What changed
historical_ml	3.79-4.83	n/a	Base ML historical model prediction
variance_applied	4-5	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-5	0.549	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run, matchup_recency_guardrail)
embedded_savant_run_environment	4-5	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	4-5	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	4-5	n/a	Capped at max 12 realistic runs
simulation_result	4-5	n/a	Final inning-by-inning simulation result

Stage	Before	After	Toward final pick	What changed
variance_applied	3.79-4.83	4-5	-0.049	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-5	4-5	0	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run, matchup_recency_guardrail)
embedded_savant_run_environment	4-5	4-5	0	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	4-5	4-5	0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	4-5	4-5	0	Capped at max 12 realistic runs
simulation_result	4-5	4-5	0	Final inning-by-inning simulation result

Team	Player	Role	Net effect	Largest change
OAK	Zack Gelof	batter / LF	worse	predicted_RBI 1.542
TEX	Justin Foscue	batter / DH	better	predicted_AB 1.5
TEX	Jake Burger	batter / 1B	better	predicted_RBI 1.48
OAK	Henry Bolte	batter / DH	worse	predicted_AB 1.451
OAK	Alika Williams	batter / SS	better	predicted_SO 1.298
TEX	Corey Seager	batter / SS	worse	predicted_H 1.09
OAK	Jonah Heim	batter / C	worse	predicted_AB -1.02
OAK	Tommy White	batter / 1B	worse	predicted_H -0.974