Deep slate report 20260909

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_20260909.json	376114
forensic deep audit	loaded	probability, provenance, starter and decision contracts	/app/output/reports/PREDICTION_SYSTEM_DEEP_AUDIT_20260909.json	241363
continuous learning	loaded	pipeline-stage quality, calibration, probability lineage, score distribution, and official starter comparisons	/app/output/learning/continuous_learning_daily_20260909.json	2537069
detailed report	loaded	full game and simulation rendering	/app/output/reports/detailed_20260909.html	353038
ultra detailed player audit	loaded	player-level prediction trace	/app/output/reports/20260909_ultra_detailed_player_audit.html	32410370
structural audit	loaded	all invariant findings	/app/output/reports/audit_report_20260909.html	1865922
integrity revision audit	missing	post-Luna remediation proof for immutable revision artifacts	/app/output/reports/audit_report_20260909_integrity_revision.html	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
CRITICAL	artifact_integrity	Critical canonical artifact integrity failure structural audit totals: {'critical': 1, 'error': 0, 'warning': 0, 'info': 114}	20260909_WAS@SD	Do thisExclude this slate from parameter learning until the artifact-generation step that wrote the broken file is fixed. Don't concludeA malformed artifact is a file bug, not proof the prediction model was wrong; do not blame accuracy changes on it.
CRITICAL	probability_lineage	Published probability differs materially without a recorded transformation 2 game(s) have an unexplained ≥2pp calibrated-to-published delta	PIT@CHW, TB@ATL	Do thisBlock any promotion until the missing probability transformation is recorded and reconciled. Don't concludeDo not assume the probability governor was bypassed until you have traced the pick-side math.
HIGH	model_validation	Validated run models explain very little holdout variance home/away/total R² = 0.0118/0.0210/0.0207	ARI@KC, CHC@MIL, CIN@LAD, CLE@BAL, COL@NYY, HOU@PHI, LAA@BOS, MIN@DET, NYM@MIA, PIT@CHW, STL@SF, TB@ATL, TEX@SEA, TOR@OAK, WAS@SD	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.
INFO	coverage	Every settled game received structural QA 15/15 games structurally audited	ARI@KC, CHC@MIL, CIN@LAD, CLE@BAL, COL@NYY, HOU@PHI, LAA@BOS, MIN@DET, NYM@MIA, PIT@CHW, STL@SF, TB@ATL, TEX@SEA, TOR@OAK, WAS@SD	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	simulation	Simulation produced no winner-direction intervention on this slate flip candidates 0; published flips 0	ARI@KC, CHC@MIL, CIN@LAD, CLE@BAL, COL@NYY, HOU@PHI, LAA@BOS, MIN@DET, NYM@MIA, PIT@CHW, STL@SF, TB@ATL, TEX@SEA, TOR@OAK, WAS@SD	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	11 consecutive · 11 in window	simulation.allow_score	confirm the simulation module is running, not silently no-oping	simulation_inert_on_slate has persisted 11 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	11 consecutive · 11 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 11 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	0	3/11	no	param
pipeline_winner_instability	0	4/11	no	code/diagnostic
score_margin_compression	0	3/11	no	param
simulation_inert_on_slate	11	11/11	yes	code/diagnostic
single_slate_calibration_instability	0	5/11	no	param
weak_run_model_signal	11	11/11	yes	code/diagnostic

Metric	Direction	Series (old → new)
winner_accuracy	improving	[0.5, 0.6667, 0.4667, 0.3333, 0.5556, 0.625, 0.5333, 0.6, 0.5455, 0.7333, 0.6667]
weighted_calibration_error	improving	[0.0693, 0.1661, 0.2476, 0.21, 0.1474, 0.1316, 0.1, 0.0937, 0.2611, 0.1825, 0.1069]
margin_bias	regressing	[-2.786, -3.5, -2.533, -3.067, -1.778, -0.688, -1.467, -1.467, -0.182, -0.867, -2.133]
team_run_error_on_losses	improving	[7.0, 8.25, 7.125, 6.4, 6.75, 6.0, 5.286, 5.333, 3.8, 6.75, 5.6]

Stage	Winner hits	Accuracy	Team-run MAE	Total-run MAE	Margin MAE
historical_ml	10/15	66.7%	3.101	4.83	3.185
variance_applied	9/15	60%	3.133	4.8	3.2
enhanced_prediction	10/15	66.7%	3.267	5.2	3.333
embedded_savant_run_environment	10/15	66.7%	3.267	5.2	3.2
fundamental_edge_target_review	10/15	66.7%	3.267	5.2	3.2
capped_target	10/15	66.7%	3.267	5.2	3.2
simulation_result	10/15	66.7%	3.267	5.2	3.2
published_final	10/15	66.7%	3.267	5.2	3.2

Transition	Δ accuracy	Δ hits	Δ team-run MAE	Δ margin MAE	Games whose winner/tie changed
historical_ml → variance_applied	-0.0667	-1	0.0324	0.0153	CLE@BAL: CLE → TIE; NYM@MIA: MIA → TIE; TEX@SEA: SEA → TIE; TOR@OAK: TOR → TIE
variance_applied → enhanced_prediction	0.0667	1	0.1334	0.1333	CLE@BAL: TIE → CLE; NYM@MIA: TIE → NYM; TEX@SEA: TIE → TEX; TOR@OAK: TIE → TOR
enhanced_prediction → embedded_savant_run_environment	0	0	0	-0.1333
embedded_savant_run_environment → fundamental_edge_target_review	0	0	0	0
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
ARI@KC	hit	0.5823	0.5474	0.5474	0.5474	0.53	score coherence followed by matchup-recency guardrail
CHC@MIL	hit	0.5411	0.5339	0.5339	0.5339	0.5245	score coherence followed by matchup-recency guardrail
CIN@LAD	hit	0.691	0.6629	0.6629	0.6617	0.58	score coherence followed by matchup-recency guardrail
CLE@BAL	miss	0.5282	0.5802	0.5802	0.5802	0.558	score coherence followed by matchup-recency guardrail
COL@NYY	hit	0.7754	0.7839	0.7839	0.7839	0.7839	published probability matches the calibrated/governed value
HOU@PHI	hit	0.5438	0.5273	0.5273	0.5273	0.5197	score coherence followed by matchup-recency guardrail
LAA@BOS	miss	0.5671	0.5397	0.5397	0.5397	0.5287	score coherence followed by matchup-recency guardrail
MIN@DET	hit	0.5546	0.5331	0.5331	0.5331	0.5239	score coherence followed by matchup-recency guardrail
NYM@MIA	hit	0.5464	0.6118	0.6118	0.6118	0.5416	score coherence followed by matchup-recency guardrail
PIT@CHW	hit	0.6887	0.6496	0.6496	0.6158	0.5984	no recorded transformation explains the publication delta
STL@SF	miss	0.5472	0.5291	0.5291	0.5291	0.521	score coherence followed by matchup-recency guardrail
TB@ATL	hit	0.6188	0.5996	0.5996	0.5908	0.5772	no recorded transformation explains the publication delta
TEX@SEA	miss	0.5169	0.6398	0.6398	0.6398	0.5487	score coherence followed by matchup-recency guardrail
TOR@OAK	miss	0.5523	0.55	0.55	0.55	0.5362	score coherence followed by matchup-recency guardrail
WAS@SD	hit	0.5641	0.5546	0.5546	0.5546	0.5245	score coherence followed by matchup-recency guardrail

Sides analyzed	Broke on pitching	Broke on offense	Both	Variance	Mean starter ERA error	Mean lineup runs vs proj
30	8	7	4	11	3.36	1.07

Slate	Games	Winner acc	Mean starter ERA error	Mean lineup runs vs proj	Sides broke on pitching	Sides broke on offense
20260908	15	0.733	-0.15	-0.23	7	15
20260907	11	0.545	-0.86	-0.36	5	4
20260906	15	0.600	0.74	-0.30	6	10
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	1.48	1.07	9	3
20260831	12	0.667	0.83	0.08	2	7
20260830	14	0.500	-0.36	-0.29	3	7
20260829	17	0.941	—	—
20260828	15	0.467	—	—

Game	Pred margin	Actual margin	Pred − actual
ARI@KC	2	3	-1
CHC@MIL	1	2	-1
CIN@LAD	3	13	-10
CLE@BAL	1	4	-3
COL@NYY	5	5	0
HOU@PHI	1	4	-3
LAA@BOS	1	2	-1
MIN@DET	1	5	-4
NYM@MIA	3	1	2
PIT@CHW	1	2	-1
STL@SF	1	1	0
TB@ATL	1	5	-4
TEX@SEA	2	1	1
TOR@OAK	1	2	-1
WAS@SD	1	7	-6

ARI@KCHIT

Predicted: KC 5-7 (conf +0.06, lean)Actual: KC 2-5Brier 0.221total bias +5

Root cause: Right winner, wrong score

Evidence-backed diagnosis

published system edge 0.06 (lean); this is not the half-scale distance from 50%
team-run absolute error 5; total bias +5
pregame home-minus-away evidence: season -0.640, recent -0.730, pitcher -0.014, composite -0.205
confidence pipeline reached 0.547 before later publication/recency controls; published home probability was 0.530
score-volatility actions: no_change
forecast-file integrity: clean (0 defects, 8 informational note(s)); the saved artifact is not why the pick missed

Quantified contributing factors

total runs over-projected by 5 (12 vs 7)
ARI offense over-projected by 3

Team	Starter	Projected	Actual	ER err
ARI	Zac Gallen	3 ER / 15 outs	1 ER / 10 outs	-2
KC	Daniel Lynch IV	1 ER / 15 outs	2 ER / 15 outs	1

Decision internals: run factors home×0.977 / away×0.99; 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:

Bobby Witt Jr. (KC SS): 2-for-4, 1 HR, 3 RBI (projected 1.0 H, 0.0 HR) [run impact 0.76]

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

No settled pitching line for this club. The forecast named Zac Gallen (ARI) at a projected 5.4 ERA, but the per-player feed carries no pitching row for him in this game, so the actual-versus-projected comparison cannot be made on this side. That is a source-coverage gap in the settled box, not a forecasting failure.

Bullpen behind him: 0 ER over 0 outs.

The HOME lineup he faced produced 5 runs on 7 hits and 2 HR, against a projected 4 runs / 9 hits (1 runs vs projection).

Who did the damage to him:

Bobby Witt Jr.: 2-for-4, 1 HR, 3 RBI [+0.76 runs vs projection] season xwOBA 0.385, elite - real projection error

Verdict: neither side broke materially - the result is closer to variance

HOME pitching

Daniel Lynch IV (KC) threw 15 outs, 2 ER = 3.6 ERA against a projected 1.8 ERA (1.8 ERA error)

Bullpen behind him: 0 ER over 6 outs.

The AWAY lineup he faced produced 0 runs on 0 hits and 0 HR, against a projected 0 runs / 0 hits (0 runs vs projection).

No individual hitter beat his projection by a meaningful margin.

Verdict: neither side broke materially - the result is closer to variance

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. exp BB and exp SO are the continuous rate expectations. The integer beside each is a simulation draw reconciled to the opposing simulated batting box, so it is not the line that ranks pitchers: on the 2026-09-07 slate the integers published 39 walks and 121 strikeouts against 51 and 130 real, while these expectations landed on 50.4 and 132.6.)

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
KC	Daniel Lynch IV (Starter)	15 / 15	3.6 / 1.8	2 / 1	4 / 4	1 / 1	2 / 1	1.44	3 / 7	1.82	venue home: 6.6 vs 7.94 ERA (ok) · opp run factor 0.99 (average) · 52 prior appearances this season
KC	Nate Pearson (Setup Man)	3 / 4	0 / 6.75 (-6.75)	0 / 1	1 / 2	0 / 0	1 / 1	0.51	1 / 0	1.17
KC	Craig Kimbrel (Middle Reliever (High Leverage))	3 / 5	0 / 5.4 (-5.40)	0 / 1	2 / 3	0 / 1	0 / 0	0.46	1 / 0	0.8

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

Batter	Team	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
Bobby Witt Jr. (SS)	KC	4 / 4	2 / 1	0 / 1	0 / 0	1 / 0	3 / 0	1 / 1	0 / 1	0 / 0	0.76	0.385 elite	projection too low
Vinnie Pasquantino (1B)	KC	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	1 / 0	-0.47	0.329 above average	—
Salvador Perez (DH)	KC	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0 / 1	-0.47	0.31 average	—
Nick Loftin (3B)	KC	2 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 0	1 / 0	1 / 0	-0.45	0.294 below average	—
Carter Jensen (C)	KC	3 / 4	1 / 1	0 / 0	0 / 0	1 / 1	1 / 1	2 / 1	1 / 1	0 / 1	0	0.306 average	—
Michael Massey (2B)	KC	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0	0.305 average	—
John Rave (RF)	KC	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0 / 0	1 / 1	0	0.313 average	—
Kyle Isbel (CF)	KC	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0	0.26 weak	—
Isaac Collins (LF)	KC	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 1	0	0.295 below average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-0.64	positive favors home; negative favors away
Recent form	-0.73	positive favors home; negative favors away
Starting-pitcher matchup	-0.014	positive favors home; negative favors away
Composite	-0.205	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	3.99-5.91	n/a	Base ML historical model prediction
variance_applied	4-6	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-6	0.53	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by matchup_recency_guardrail)
embedded_savant_run_environment	5-7	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	5-7	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	5-7	n/a	Capped at max 12 realistic runs
simulation_result	5-7	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	3.99-5.91	4-6	0.085	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 matchup_recency_guardrail)
embedded_savant_run_environment	5-6	5-7	1	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	5-7	5-7	0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	5-7	5-7	0	Capped at max 12 realistic runs
simulation_result	5-7	5-7	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
KC	Carter Jensen	batter / C	worse	predicted_RBI 3.28
ARI	Jordan Lawlar	batter / CF	better	predicted_RBI 1.87
KC	Kyle Isbel	batter / CF	better	predicted_RBI 1.77
ARI	Lourdes Gurriel Jr.	batter / LF	worse	predicted_H 1.53
ARI	Ildemaro Vargas	batter / 2B	worse	predicted_AB 1.463
ARI	Geraldo Perdomo	batter / SS	worse	predicted_BB 1.404
KC	Vinnie Pasquantino	batter / 1B	worse	predicted_SO 1.357
ARI	Nolan Arenado	batter / 3B	better	predicted_SO 1.206

Structural audit findings

Severity	Rule	Finding	Subject
info	CURRENT_PITCHER_DUPLICATION	ARI pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	ARI
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_FINALIZED_MATCH_BOX	ARI model pitcher projections differ from final simulation pitching box score (ER: current=6, sim=7).	ARI
info	CURRENT_FINALIZED_MATCH_BOX	ARI model-prediction batting totals differ from final simulation box score (AB: current=33, sim=38; H: current=8, sim=11; 2B: current=2, sim=0; HR: current=1, sim=3; R: current=4, sim=5; BB: current=3, sim=2; SO: current=6, sim=8).	ARI
info	CURRENT_PLAYER_MATCH_BOX	ARI current finalized hitter lines are model predictions and differ from the simulation box score: Ildemaro Vargas: AB current=4 box=5, 2B current=1 box=0, RBI current=1 box=0, R current=1 box=0; Ketel Marte: AB current=4 box=5, H current=1 box=2, 2B current=1 box=0, RBI current=1 box=0; Lourdes Gurriel Jr.: H current=1 box=2; Geraldo Perdomo: AB current=4 box=3, H current=1 box=2, R current=0 box=1, BB current=1 box=2; Lars Nootbaar: AB current=3 box=4, H current=0 box=1, HR current=0 box=1, RBI current=0 box=1; Nolan Arenado: AB current=4 box=5, BB current=1 box=0, SO current=1 box=2; James McCann: AB current=3 box=4, H current=1 box=0, RBI current=1 box=0, SO current=1 box=2; Jordan Lawlar: AB current=3 box=4, HR current=0 box=1, RBI current=0 box=2, R current=0 box=1	ARI
info	CURRENT_FINALIZED_MATCH_BOX	KC model-prediction batting totals differ from final simulation box score (AB: current=33, sim=37; H: current=9, sim=10; 3B: current=0, sim=1; RBI: current=4, sim=7; R: current=4, sim=7; BB: current=3, sim=1; SO: current=6, sim=11).	KC
info	CURRENT_PLAYER_MATCH_BOX	KC current finalized hitter lines are model predictions and differ from the simulation box score: Vinnie Pasquantino: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, BB current=1 box=0; Salvador Perez: 2B current=0 box=1, RBI current=1 box=0, R current=0 box=1, SO current=1 box=2; Bobby Witt Jr.: AB current=4 box=5, H current=1 box=2, 2B current=1 box=0, BB current=1 box=0; Carter Jensen: AB current=4 box=5, H current=1 box=2, HR current=1 box=0, RBI current=1 box=4; Michael Massey: H current=1 box=2, 2B current=0 box=1, R current=0 box=1; John Rave: AB current=3 box=4, H current=1 box=0, R current=1 box=0, SO current=1 box=0; Kyle Isbel: AB current=3 box=4, H current=1 box=2, 3B current=0 box=1, HR current=0 box=1; Nick Loftin: 2B current=1 box=0, R current=0 box=1, SO current=0 box=1	KC
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

CHC@MILHIT

Predicted: MIL 5-6 (conf +0.05, lean)Actual: MIL 6-8Brier 0.226total bias -3

Root cause: Solid correct call

Evidence-backed diagnosis

published system edge 0.05 (lean); this is not the half-scale distance from 50%
team-run absolute error 3; total bias -3
pregame home-minus-away evidence: season +0.228, recent +0.182, pitcher +0.136, composite +0.649
confidence pipeline reached 0.534 before later publication/recency controls; published home probability was 0.524
score-volatility actions: no_change
forecast-file integrity: clean (0 defects, 8 informational note(s)); the saved artifact is not why the pick missed

Quantified contributing factors

total runs under-projected by 3 (11 vs 14)
Logan Henderson (MIL) allowed 3 fewer ER than projected (0 vs 3)

Team	Starter	Projected	Actual	ER err
CHC	Kevin Gausman	5 ER / 17 outs	6 ER / 12 outs	1
MIL	Logan Henderson	3 ER / 19 outs	0 ER / 15 outs	-3

Decision internals: run factors home×1.095 / away×0.96; 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:

William Contreras (MIL C): 2-for-3, 1 HR, 3 RBI (projected 1.0 H, 0.0 HR) [run impact 1.73]
Jackson Chourio (MIL LF): 3-for-5, 1 HR, 2 RBI (projected 1.0 H, 1.0 HR) [run impact 0.94]
Pete Crow-Armstrong (CHC CF): 1-for-3, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -0.93]
Garrett Mitchell (MIL CF): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.8]
Seiya Suzuki (CHC RF): 0-for-3, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.78]

Starter miss: Kevin Gausman (CHC) posted 13.50 ERA vs projected 7.94 (6 ER in 12 outs); he is materially worse pitching away (4.95 ERA there vs 3.91 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

AWAY pitching

Kevin Gausman (CHC) threw 12 outs, 6 ER = 13.5 ERA against a projected 7.94 ERA (5.56 ERA error)

Bullpen behind him: 0 ER over 3 outs.

The HOME lineup he faced produced 8 runs on 9 hits and 3 HR, against a projected 5 runs / 8 hits (3 runs vs projection).

Who did the damage to him:

William Contreras: 2-for-3, 1 HR, 3 RBI [+1.73 runs vs projection] season xwOBA 0.331, above average - real projection error
Jackson Chourio: 3-for-5, 1 HR, 2 RBI [+0.94 runs vs projection] season xwOBA 0.336, above average - real projection error
Christian Yelich: 1-for-4, 1 HR, 3 RBI [+0.6 runs vs projection] season xwOBA 0.316, average - a good day, not a model error

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

HOME pitching

Logan Henderson (MIL) threw 15 outs, 0 ER = 0 ERA against a projected 4.26 ERA (-4.26 ERA error)

Bullpen behind him: 1 ER over 2 outs.

The AWAY lineup he faced produced 6 runs on 5 hits and 0 HR, against a projected 5 runs / 7 hits (1 runs vs projection).

Who did the damage to him:

Michael Conforto: 1-for-3, 0 HR, 2 RBI [+0.78 runs vs projection] season xwOBA 0.338, above average - real projection error
Alex Bregman: 1-for-3, 0 HR, 0 RBI [+0.64 runs vs projection] season xwOBA 0.318, average - a good day, not a model error
Ian Happ: 0-for-3, 0 HR, 1 RBI [+0.33 runs vs projection] season xwOBA 0.313, average - a good day, not a model error

Verdict: neither side broke materially - the result is closer to variance

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
CHC	Kevin Gausman (Starter)	12 / 17	13.5 / 7.94 (+5.56)	6 / 5	8 / 7	3 / 0	2 / 4	1.75	6 / 5	6	venue away: 4.95 vs 3.91 ERA (worse here) · opp run factor 1.1 (strong) · 29 prior appearances this season
CHC	Trent Thornton (Setup Man)	3 / 3	0 / 9 (-9.00)	0 / 1	0 / 2	0 / 0	0 / 1	0.48	2 / 2	0.66
MIL	Logan Henderson (Starter)	15 / 19	0 / 4.26 (-4.26)	0 / 3	2 / 7	0 / 1	1 / 4	0.94	6 / 4	5.88	venue home: 2.08 vs 2.84 ERA (ok) · opp run factor 0.96 (strong) · 15 prior appearances this season
MIL	Abner Uribe (Setup Man)	2 / 3	13.5 / 0 (+13.50)	1 / 0	2 / 0	0 / 0	0 / 0	0.4	1 / 0	0.92

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

Batter	Team	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
William Contreras (C)	MIL	3 / 4	2 / 1	0 / 0	0 / 0	1 / 0	3 / 1	2 / 1	1 / 0	1 / 1	1.73	0.331 above average	projection too low
Jackson Chourio (LF)	MIL	5 / 4	3 / 1	0 / 0	0 / 0	1 / 1	2 / 1	1 / 1	0 / 0	1 / 1	0.94	0.336 above average	projection too low
Pete Crow-Armstrong (CF)	CHC	3 / 4	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	2 / 1	1 / 1	1 / 1	-0.93	0.372 elite	bad day
Garrett Mitchell (CF)	MIL	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	2 / 1	-0.8	0.351 elite	bad day
Seiya Suzuki (RF)	CHC	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	1 / 1	1 / 1	1 / 1	-0.78	0.34 elite	bad day
Michael Busch (1B)	CHC	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	1 / 0	1 / 0	1 / 1	1 / 1	-0.78	0.329 above average	bad day
Michael Conforto (LF)	CHC	3 / 3	1 / 0	1 / 0	0 / 0	0 / 0	2 / 0	0 / 0	0 / 0	0 / 1	0.78	0.338 above average	projection too low
Brice Turang (2B)	MIL	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	3 / 1	1 / 1	0 / 1	-0.78	0.338 above average	bad day
Alex Bregman (3B)	CHC	3 / 4	1 / 1	1 / 0	0 / 0	0 / 0	0 / 1	0 / 1	1 / 0	1 / 1	0.64	0.318 average	no signal
Christian Yelich (DH)	MIL	4 / 4	1 / 1	0 / 0	0 / 0	1 / 0	3 / 0	1 / 1	0 / 1	3 / 1	0.6	0.316 average	no signal
Pedro Ramirez (2B)	CHC	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	1 / 1	-0.47	0.293 below average	—
Ian Happ (DH)	CHC	3 / 4	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	1 / 0	2 / 2	0.33	0.313 average	—
Nico Hoerner (SS)	CHC	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 0	0 / 1	0	0.323 above average	—
Carson Kelly (C)	CHC	3 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	1 / 1	0	0.31 average	—
Andrew Vaughn (1B)	MIL	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	1 / 1	0	0.335 above average	—
Joey Ortiz (3B)	MIL	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0	0.295 below average	—
Sal Frelick (RF)	MIL	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0	0.283 weak	—
Cooper Pratt (SS)	MIL	4 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0	0.309 average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	0.228	positive favors home; negative favors away
Recent form	0.182	positive favors home; negative favors away
Starting-pitcher matchup	0.136	positive favors home; negative favors away
Composite	0.649	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.19-5.01	n/a	Base ML historical model prediction
variance_applied	4-5	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-6	0.524	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by 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

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.19-5.01	4-5	0.183	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-5	5-6	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	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
CHC	Pete Crow-Armstrong	batter / CF	better	predicted_AB -2.88
MIL	Sal Frelick	batter / RF	deboosted	predicted_BB 2.737
MIL	Andrew Vaughn	batter / 1B	worse	predicted_H 2.29
MIL	Jackson Chourio	batter / LF	worse	predicted_H -1.72
CHC	Pedro Ramirez	batter / 2B	deboosted	predicted_H 1.6
CHC	Michael Conforto	batter / LF	worse	predicted_AB 1.5
CHC	Ian Happ	batter / DH	worse	predicted_AB 1.375
MIL	Christian Yelich	batter / DH	worse	predicted_AB 1.248

Structural audit findings

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=10, sim=12).	CHC
info	CURRENT_FINALIZED_MATCH_BOX	CHC model-prediction batting totals differ from final simulation box score (AB: current=35, sim=36; H: current=7, sim=9; 2B: current=2, sim=0; RBI: current=4, sim=5; BB: current=3, sim=5; SO: current=10, sim=4).	CHC
info	CURRENT_PLAYER_MATCH_BOX	CHC current finalized hitter lines are model predictions and differ from the simulation box score: Seiya Suzuki: AB current=4 box=5, H current=1 box=2, 2B current=1 box=0, RBI current=1 box=0; Ian Happ: AB current=4 box=5, R current=1 box=0, SO current=2 box=0; Nico Hoerner: AB current=4 box=3, R current=0 box=1, BB current=0 box=1, SO current=1 box=0; Michael Busch: AB current=4 box=5, 2B current=1 box=0, BB current=1 box=0; Michael Conforto: AB current=3 box=4, SO current=1 box=2; Alex Bregman: BB current=0 box=1, SO current=1 box=0; Pete Crow-Armstrong: AB current=4 box=3, H current=1 box=2, RBI current=1 box=3, BB current=1 box=2; Pedro Ramirez: AB current=4 box=3, H current=1 box=2, RBI current=0 box=1, R current=0 box=1	CHC
info	CURRENT_FINALIZED_MATCH_BOX	MIL model-prediction batting totals differ from final simulation box score (AB: current=31, sim=37; H: current=8, sim=12; 2B: current=1, sim=3; 3B: current=0, sim=1; HR: current=1, sim=0; RBI: current=4, sim=6; R: current=5, sim=6; BB: current=3, sim=5).	MIL
info	CURRENT_PLAYER_MATCH_BOX	MIL current finalized hitter lines are model predictions and differ from the simulation box score: Andrew Vaughn: AB current=3 box=5, H current=1 box=3, 3B current=0 box=1, RBI current=1 box=0; Jackson Chourio: H current=1 box=0, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0; Brice Turang: H current=1 box=2, RBI current=1 box=2, SO current=1 box=0; Joey Ortiz: AB current=3 box=4, H current=1 box=2, 2B current=0 box=1, RBI current=0 box=1; William Contreras: AB current=4 box=5, SO current=1 box=0; Garrett Mitchell: AB current=3 box=4, 2B current=0 box=1, R current=1 box=0, BB current=1 box=0; Christian Yelich: AB current=4 box=5, R current=1 box=0, BB current=1 box=0, SO current=1 box=2; Sal Frelick: AB current=3 box=2, RBI current=0 box=1, R current=0 box=1, BB current=0 box=3	MIL
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

CIN@LADHIT

Predicted: LAD 5-8 (conf +0.16, moderate)Actual: LAD 1-14Brier 0.176total bias -2

Root cause: Right winner, wrong score

Evidence-backed diagnosis

published system edge 0.16 (moderate); this is not the half-scale distance from 50%
team-run absolute error 10; total bias -2
pregame home-minus-away evidence: season +1.460, recent +0.762, pitcher +0.102, composite +1.757
confidence pipeline reached 0.663 before later publication/recency controls; published home probability was 0.580
score-volatility actions: supported_blowout_margin_expansion
forecast-file integrity: clean (0 defects, 8 informational note(s)); the saved artifact is not why the pick missed

Quantified contributing factors

CIN offense over-projected by 4
LAD offense under-projected by 6
Rhett Lowder (CIN) allowed 3 more ER than projected (7 vs 4)

Team	Starter	Projected	Actual	ER err
CIN	Rhett Lowder	4 ER / 14 outs	7 ER / 10 outs	3
LAD	Yoshinobu Yamamoto	1 ER / 19 outs	1 ER / 21 outs	0

Decision internals: run factors home×1.097 / away×0.995; 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:

Enrique Hernández (LAD 1B): 2-for-4, 1 HR, 4 RBI (projected 0.0 H, 0.0 HR) [run impact 2.18]
Mookie Betts (LAD SS): 3-for-5, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.87]
Kyle Tucker (LAD RF): 3-for-5, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 1.85]
Alex Freeland (LAD 2B): 1-for-3, 0 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 0.97]
Sal Stewart (CIN 1B): 2-for-3, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -0.79]

Starter miss: Rhett Lowder (CIN) posted 18.90 ERA vs projected 7.71 (7 ER in 10 outs); he is materially worse pitching away (6.79 ERA there vs 4.29 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

AWAY pitching

Rhett Lowder (CIN) threw 10 outs, 7 ER = 18.9 ERA against a projected 7.71 ERA (11.19 ERA error)

Bullpen behind him: 2 ER over 2 outs.

The HOME lineup he faced produced 14 runs on 13 hits and 4 HR, against a projected 4 runs / 8 hits (10 runs vs projection).

Who did the damage to him:

Enrique Hernández: 2-for-4, 1 HR, 4 RBI [+2.18 runs vs projection] season xwOBA 0.26, weak - a good day, not a model error
Mookie Betts: 3-for-5, 1 HR, 2 RBI [+1.87 runs vs projection] season xwOBA 0.331, above average - real projection error
Kyle Tucker: 3-for-5, 1 HR, 1 RBI [+1.85 runs vs projection] season xwOBA 0.324, above average - real projection error

Verdict: both: the starter was hit hard AND the lineup underproduced

HOME pitching

Yoshinobu Yamamoto (LAD) threw 21 outs, 1 ER = 1.29 ERA against a projected 1.42 ERA (-0.13 ERA error)

Bullpen behind him: 0 ER over 0 outs.

The AWAY lineup he faced produced 1 runs on 5 hits and 1 HR, against a projected 4 runs / 6 hits (-3 runs vs projection).

Who did the damage to him:

Will Banfield: 1-for-3, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.07, weak - a good day, not a model error
Ke'Bryan Hayes: 1-for-3, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.305, average - a good day, not a model error

Verdict: neither side broke materially - the result is closer to variance

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
CIN	Rhett Lowder (Starter)	10 / 14	18.9 / 7.71 (+11.19)	7 / 4	7 / 6	1 / 2	2 / 2	1.94	4 / 3	3.8	venue away: 6.79 vs 4.29 ERA (worse here) · opp run factor 1.1 (strong) · 25 prior appearances this season
CIN	Luis Mey (Setup Man)	2 / 6	27 / 13.5 (+13.50)	2 / 3	2 / 2	0 / 1	1 / 2	0.64	2 / 2	1.42
LAD	Yoshinobu Yamamoto (Starter)	21 / 19	1.29 / 1.42	1 / 1	3 / 3	1 / 0	1 / 0	1.68	10 / 7	7.39	venue home: 3.28 vs 1.94 ERA (worse here) · opp run factor 1 (weak) · 25 prior appearances this season

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

Batter	Team	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
Enrique Hernández (1B)	LAD	4 / 3	2 / 0	1 / 0	0 / 0	1 / 0	4 / 0	2 / 0	0 / 0	1 / 1	2.18	0.26 weak	good day
Mookie Betts (SS)	LAD	5 / 4	3 / 1	0 / 0	0 / 0	1 / 0	2 / 1	3 / 1	0 / 0	1 / 0	1.87	0.331 above average	projection too low
Kyle Tucker (RF)	LAD	5 / 4	3 / 1	1 / 0	0 / 0	1 / 0	1 / 0	2 / 1	0 / 1	1 / 0	1.85	0.324 above average	projection too low
Alex Freeland (2B)	LAD	3 / 3	1 / 1	1 / 0	0 / 0	0 / 0	1 / 0	2 / 0	2 / 0	1 / 1	0.97	0.283 weak	good day
Sal Stewart (1B)	CIN	3 / 4	2 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	0 / 1	-0.79	0.35 elite	bad day
JJ Bleday (DH)	CIN	2 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	2 / 1	-0.47	0.343 elite	—
Héctor Rodríguez (RF)	CIN	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	1 / 1	-0.47	0.277 weak	—
Juan Brito (2B)	CIN	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	-0.47	0.277 weak	—
Will Banfield (C)	CIN	3 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0.47	0.07 weak	—
Carlos Jorge (CF)	CIN	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	-0.47	0.26 weak	—
Ke'Bryan Hayes (3B)	CIN	3 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	0.47	0.305 average	—
Teoscar Hernández (LF)	LAD	3 / 4	2 / 1	0 / 0	0 / 0	0 / 0	2 / 1	1 / 0	0 / 0	1 / 1	0.47	0.332 above average	—
Tommy Edman (CF)	LAD	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 0	1 / 1	-0.45	0.314 average	—
Max Muncy (3B)	LAD	5 / 4	1 / 1	0 / 0	0 / 0	1 / 1	2 / 1	1 / 1	0 / 1	3 / 1	-0.33	0.35 elite	—
Will Smith (DH)	LAD	3 / 4	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	2 / 0	1 / 1	0 / 1	0.31	0.379 elite	—
Elly De La Cruz (SS)	CIN	3 / 4	1 / 1	0 / 1	0 / 0	1 / 0	1 / 1	1 / 1	0 / 1	1 / 1	0.29	0.371 elite	—
Hunter Feduccia (C)	LAD	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 1	-0.14	0.316 average	—
TJ Friedl (LF)	CIN	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	0	0.248 weak	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	1.46	positive favors home; negative favors away
Recent form	0.762	positive favors home; negative favors away
Starting-pitcher matchup	0.102	positive favors home; negative favors away
Composite	1.757	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	3.76-6.09	n/a	Base ML historical model prediction
variance_applied	4-6	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-7	0.58	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by matchup_recency_guardrail)
embedded_savant_run_environment	5-8	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	5-8	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	5-8	n/a	Capped at max 12 realistic runs
simulation_result	5-8	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	3.76-6.09	4-6	-0.329	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-6	5-7	0	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by matchup_recency_guardrail)
embedded_savant_run_environment	5-7	5-8	1	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	5-8	5-8	0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	5-8	5-8	0	Capped at max 12 realistic runs
simulation_result	5-8	5-8	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
CIN	JJ Bleday	batter / DH	better	predicted_H 2.212
LAD	Teoscar Hernández	batter / LF	better	predicted_RBI 2.16
CIN	Sal Stewart	batter / 1B	worse	predicted_SO 2.085
CIN	Will Banfield	batter / C	better	predicted_RBI 1.91
LAD	Will Smith	batter / DH	better	predicted_RBI 1.526
LAD	Tommy Edman	batter / CF	better	predicted_RBI 1.5
LAD	Enrique Hernández	batter / 1B	boosted	predicted_AB 1.5
LAD	Hunter Feduccia	batter / C	worse	predicted_AB 1.481

Structural audit findings

Severity	Rule	Finding	Subject
info	CURRENT_PITCHER_DUPLICATION	CIN pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	CIN
info	CURRENT_PITCHER_DUPLICATION	LAD pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	LAD
info	CURRENT_FINALIZED_MATCH_BOX	CIN model pitcher projections differ from final simulation pitching box score (ER: current=7, sim=8).	CIN
info	CURRENT_FINALIZED_MATCH_BOX	CIN model-prediction batting totals differ from final simulation box score (AB: current=32, sim=34; H: current=6, sim=8; 2B: current=1, sim=4; RBI: current=4, sim=5; R: current=4, sim=5; BB: current=3, sim=0; SO: current=9, sim=11; SF: current=0, sim=1).	CIN
info	CURRENT_PLAYER_MATCH_BOX	CIN current finalized hitter lines are model predictions and differ from the simulation box score: JJ Bleday: H current=1 box=3, 2B current=0 box=2, HR current=0 box=1, RBI current=1 box=2; Héctor Rodríguez: RBI current=1 box=0, SO current=1 box=2; Juan Brito: 2B current=0 box=1, R current=1 box=0; Will Banfield: H current=0 box=1, 2B current=0 box=1, RBI current=0 box=2, R current=0 box=1; TJ Friedl: AB current=3 box=4, H current=0 box=1, RBI current=0 box=1, R current=0 box=1; Sal Stewart: H current=1 box=0, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0; Carlos Jorge: AB current=3 box=4, H current=1 box=0; Elly De La Cruz: 2B current=1 box=0, RBI current=1 box=0, BB current=1 box=0	CIN
info	CURRENT_FINALIZED_MATCH_BOX	LAD model-prediction batting totals differ from final simulation box score (AB: current=33, sim=32; 2B: current=1, sim=0; 3B: current=0, sim=1; HR: current=1, sim=3; RBI: current=4, sim=8; R: current=4, sim=8; BB: current=3, sim=4).	LAD
info	CURRENT_PLAYER_MATCH_BOX	LAD current finalized hitter lines are model predictions and differ from the simulation box score: Will Smith: HR current=0 box=1, RBI current=0 box=2, R current=0 box=1, BB current=1 box=0; Tommy Edman: 2B current=1 box=0, 3B current=0 box=1, RBI current=1 box=2, SO current=1 box=0; Max Muncy: AB current=4 box=3, SO current=1 box=0; Teoscar Hernández: AB current=4 box=3, HR current=0 box=1, RBI current=1 box=3, R current=0 box=1; Mookie Betts: H current=1 box=2, RBI current=1 box=0; Enrique Hernández: AB current=3 box=4, H current=0 box=1, R current=0 box=1; Kyle Tucker: AB current=4 box=3, H current=1 box=0, SO current=0 box=2; Hunter Feduccia: AB current=3 box=4, H current=1 box=0	LAD
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

CLE@BALMISS

Predicted: CLE 4-3 (conf -0.12, moderate)Actual: BAL 5-9Brier 0.311total bias -7

Root cause: Starting-pitcher projection was wrong

Evidence-backed diagnosis

published system edge 0.12 (moderate); this is not the half-scale distance from 50%
team-run absolute error 7; total bias -7
pregame home-minus-away evidence: season -0.277, recent -0.542, pitcher -0.080, composite -0.643
confidence pipeline reached 0.580 before later publication/recency controls; published home probability was 0.442
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 7 (7 vs 14)
BAL offense under-projected by 6
Foster Griffin (CLE) allowed 5 more ER than projected (8 vs 3)
Shane Baz (BAL) allowed 4 more ER than projected (5 vs 1)
committed to CLE at 0.12 confidence and it lost

Team	Starter	Projected	Actual	ER err
CLE	Foster Griffin	3 ER / 17 outs	8 ER / 7 outs	5
BAL	Shane Baz	1 ER / 17 outs	5 ER / 17 outs	4

Decision internals: run factors home×0.982 / away×1.062; 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:

Coby Mayo (BAL DH): 1-for-3, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 1.26]
Carlos Narváez (BAL C): 2-for-4, 0 HR, 0 RBI (projected 0.0 H, 0.0 HR) [run impact 0.94]
Patrick Bailey (CLE C): 1-for-3, 0 HR, 1 RBI (projected 0.0 H, 0.0 HR) [run impact 0.78]
José Ramírez (CLE 3B): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.78]
Christian Encarnacion-Strand (BAL 3B): 0-for-3, 0 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact -0.78]

Starter miss: Foster Griffin (CLE) posted 30.86 ERA vs projected 4.76 (8 ER in 7 outs); no credible venue, opponent-strength or experience signal explains it — an outlier start (variance).

What to do: Variance — no model change is warranted for this start.

Pitcher versus batter: what actually broke this game

AWAY pitching

Foster Griffin (CLE) threw 7 outs, 8 ER = 30.86 ERA against a projected 4.76 ERA (26.1 ERA error)

Bullpen behind him: 0 ER over 0 outs.

The HOME lineup he faced produced 8 runs on 8 hits and 1 HR, against a projected 2 runs / 7 hits (6 runs vs projection).

Who did the damage to him:

Coby Mayo: 1-for-3, 1 HR, 1 RBI [+1.26 runs vs projection] season xwOBA 0.322, above average - real projection error
Carlos Narváez: 2-for-4, 0 HR, 0 RBI [+0.94 runs vs projection] season xwOBA 0.265, weak - a good day, not a model error

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

HOME pitching

Shane Baz (BAL) threw 17 outs, 5 ER = 7.94 ERA against a projected 1.59 ERA (6.35 ERA error)

Bullpen behind him: 0 ER over 10 outs.

The AWAY lineup he faced produced 5 runs on 9 hits and 2 HR, against a projected 5 runs / 8 hits (0 runs vs projection).

Who did the damage to him:

Patrick Bailey: 1-for-3, 0 HR, 1 RBI [+0.78 runs vs projection] season xwOBA 0.309, average - a good day, not a model error
Chase DeLauter: 2-for-5, 1 HR, 2 RBI [+0.76 runs vs projection] season xwOBA 0.343, elite - real projection error
Angel Genao: 2-for-4, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.281, weak - a good day, not a model error

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

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
CLE	Foster Griffin (Starter)	7 / 17	30.86 / 4.76 (+26.10)	8 / 3	7 / 6	0 / 1	2 / 2	1.77	4 / 5	3.69	venue away: 3.14 vs 3.28 ERA (ok) · opp run factor 0.98 (average) · 28 prior appearances this season
BAL	Shane Baz (Starter)	17 / 17	7.94 / 1.59 (+6.35)	5 / 1	6 / 4	2 / 1	1 / 4	2.08	5 / 2	6.18	venue home: 3.63 vs 4.18 ERA (ok) · opp run factor 1.06 (weak) · 28 prior appearances this season
BAL	Josh Walker (Middle Reliever (High Leverage))	4 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0.37	0 / 1	1.34
BAL	Rico Garcia (Setup Man)	5 / 2	0 / 0	0 / 0	2 / 0	0 / 0	0 / 0	0.2	3 / 1	0.81
BAL	Yennier Cano (Middle Reliever (High Leverage))	1 / 2	0 / 40.5 (-40.50)	0 / 3	1 / 1	0 / 1	1 / 1	0.11	0 / 0	0.76

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

Batter	Team	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
Coby Mayo (DH)	BAL	3 / 4	1 / 1	0 / 0	0 / 0	1 / 0	1 / 1	2 / 1	1 / 0	1 / 1	1.26	0.322 above average	projection too low
Carlos Narváez (C)	BAL	4 / 3	2 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 1	0.94	0.265 weak	good day
Patrick Bailey (C)	CLE	3 / 3	1 / 0	1 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	1 / 1	0.78	0.309 average	no signal
José Ramírez (3B)	CLE	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	1 / 1	-0.78	0.332 above average	bad day
Christian Encarnacion-Strand (3B)	BAL	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	1 / 1	1 / 0	0 / 0	2 / 1	-0.78	0.349 elite	bad day
Chase DeLauter (RF)	CLE	5 / 4	2 / 1	0 / 1	0 / 0	1 / 0	2 / 1	1 / 1	0 / 1	0 / 1	0.76	0.343 elite	projection too low
Pete Alonso (1B)	BAL	3 / 4	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	2 / 1	2 / 1	-0.6	0.394 elite	bad day
Angel Genao (2B)	CLE	4 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0 / 0	0 / 1	0.47	0.281 weak	—
Brayan Rocchio (SS)	CLE	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	-0.47	0.294 below average	—
Rece Hinds (LF)	BAL	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	1 / 1	-0.47	0.184 weak	—
Jackson Holliday (2B)	BAL	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	1 / 1	1 / 0	0 / 1	0 / 2	-0.33	0.302 below average	—
Steven Kwan (CF)	CLE	5 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	1 / 0	0.14	0.321 above average	—
Travis Bazzana (DH)	CLE	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 1	-0.14	0.301 below average	—
Leody Taveras (CF)	BAL	4 / 3	2 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 1	1 / 1	0.14	0.271 weak	—
Jo Adell (LF)	CLE	5 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	3 / 1	0	0.326 above average	—
Nathaniel Lowe (1B)	CLE	4 / 4	1 / 1	0 / 0	0 / 0	1 / 1	2 / 1	1 / 1	0 / 0	1 / 1	0	0.336 above average	—
Jeremiah Jackson (RF)	BAL	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 0	2 / 1	0	0.286 below average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-0.277	positive favors home; negative favors away
Recent form	-0.542	positive favors home; negative favors away
Starting-pitcher matchup	-0.08	positive favors home; negative favors away
Composite	-0.643	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.41-4.3	n/a	Base ML historical model prediction
variance_applied	4-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-3	0.558	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

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.41-4.3	4-4	-0.106	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-4	4-3	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	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

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
CLE	José Ramírez	batter / 3B	better	predicted_AB -2.63
CLE	Nathaniel Lowe	batter / 1B	worse	predicted_RBI 2.44
BAL	Gunnar Henderson	batter / SS	worse	predicted_SO 2.035
BAL	Carlos Narváez	batter / C	better	predicted_H 1.84
BAL	Jeremiah Jackson	batter / RF	worse	predicted_RBI 1.573
BAL	Rece Hinds	batter / LF	better	predicted_AB 1.333
CLE	Chase DeLauter	batter / RF	worse	predicted_AB 1.25
CLE	Patrick Bailey	batter / C	worse	predicted_SO 1.169

Structural audit findings

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	CLE pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	CLE
info	CURRENT_FINALIZED_MATCH_BOX	BAL model-prediction batting totals differ from final simulation box score (AB: current=32, sim=34; H: current=8, sim=7; RBI: current=4, sim=3; SO: current=10, sim=7).	BAL
info	CURRENT_PLAYER_MATCH_BOX	BAL current finalized hitter lines are model predictions and differ from the simulation box score: Coby Mayo: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, SO current=1 box=0; Jackson Holliday: R current=0 box=1, SO current=2 box=0; Rece Hinds: AB current=3 box=4; Christian Encarnacion-Strand: RBI current=1 box=0, SO current=1 box=0; Gunnar Henderson: H current=1 box=0, R current=1 box=0, SO current=1 box=3; Jeremiah Jackson: AB current=3 box=4, HR current=0 box=1, RBI current=0 box=2, R current=0 box=1; Pete Alonso: AB current=4 box=3, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0; Carlos Narváez: H current=0 box=2, R current=0 box=1, BB current=0 box=1, SO current=1 box=0	BAL
info	CURRENT_FINALIZED_MATCH_BOX	CLE model-prediction batting totals differ from final simulation box score (AB: current=33, sim=32; H: current=8, sim=5; 2B: current=2, sim=0; HR: current=1, sim=2; R: current=5, sim=4; BB: current=3, sim=5; SO: current=7, sim=5; HBP: current=0, sim=1).	CLE
info	CURRENT_PLAYER_MATCH_BOX	CLE current finalized hitter lines are model predictions and differ from the simulation box score: Steven Kwan: H current=1 box=0, R current=1 box=0, SO current=0 box=1; Jo Adell: H current=1 box=2, RBI current=1 box=0, R current=0 box=1, SO current=1 box=0; Travis Bazzana: H current=1 box=0, SO current=1 box=0, HBP current=0 box=1; Angel Genao: AB current=4 box=3, H current=1 box=0, R current=1 box=0, BB current=0 box=1; Patrick Bailey: AB current=3 box=4, H current=0 box=1, R current=0 box=1, SO current=1 box=2; Nathaniel Lowe: RBI current=1 box=3; Chase DeLauter: AB current=4 box=5, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; José Ramírez: AB current=4 box=2, 2B current=1 box=0, HR current=0 box=1, BB current=1 box=2	CLE
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

COL@NYYHIT

Predicted: NYY 4-9 (conf +0.57, strong)Actual: NYY 1-6Brier 0.047total bias +6

Root cause: Right winner, wrong score

Evidence-backed diagnosis

published system edge 0.57 (strong); this is not the half-scale distance from 50%
team-run absolute error 6; total bias +6
pregame home-minus-away evidence: season +1.571, recent +0.950, pitcher -0.048, composite +1.897
confidence pipeline reached 0.784 before later publication/recency controls; published home probability was 0.784
score-volatility actions: no_change
forecast-file integrity: clean (0 defects, 8 informational note(s)); the saved artifact is not why the pick missed

Quantified contributing factors

total runs over-projected by 6 (13 vs 7)
COL offense over-projected by 3
NYY offense over-projected by 3
Will Warren (NYY) allowed 3 fewer ER than projected (1 vs 4)

Team	Starter	Projected	Actual	ER err
COL	Tomoyuki Sugano	3 ER / 14 outs	5 ER / 18 outs	2
NYY	Will Warren	4 ER / 15 outs	1 ER / 18 outs	-3

Decision internals: run factors home×0.985 / away×1.033; 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:

Ben Rice (NYY DH): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.73]
Connor Norby (COL 2B): 2-for-3, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 1.71]
Hunter Goodman (COL DH): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.4]
Luis García Jr. (NYY 1B): 2-for-4, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 1.4]
Spencer Jones (NYY CF): 1-for-2, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact 0.97]

Starter miss: Will Warren (NYY) posted 1.50 ERA vs projected 7.20 (1 ER in 18 outs); no credible venue, opponent-strength or experience signal explains it — an outlier start (variance).

What to do: Variance — no model change is warranted for this start.

Pitcher versus batter: what actually broke this game

AWAY pitching

Tomoyuki Sugano (COL) threw 18 outs, 5 ER = 7.5 ERA against a projected 5.79 ERA (1.71 ERA error)

Bullpen behind him: 0 ER over 0 outs.

The HOME lineup he faced produced 6 runs on 7 hits and 3 HR, against a projected 5 runs / 8 hits (1 runs vs projection).

Who did the damage to him:

Luis García Jr.: 2-for-4, 1 HR, 1 RBI [+1.4 runs vs projection] season xwOBA 0.348, elite - real projection error
Spencer Jones: 1-for-2, 0 HR, 0 RBI [+0.97 runs vs projection] season xwOBA 0.324, above average - real projection error
George Lombard Jr.: 1-for-4, 1 HR, 4 RBI [+0.93 runs vs projection] season xwOBA 0.283, weak - a good day, not a model error

Verdict: offense: the lineup scored well below its projection

HOME pitching

Will Warren (NYY) threw 18 outs, 1 ER = 1.5 ERA against a projected 7.2 ERA (-5.7 ERA error)

Bullpen behind him: 0 ER over 3 outs.

The AWAY lineup he faced produced 1 runs on 5 hits and 1 HR, against a projected 4 runs / 8 hits (-3 runs vs projection).

Who did the damage to him:

Connor Norby: 2-for-3, 1 HR, 1 RBI [+1.71 runs vs projection] season xwOBA 0.298, below average - a good day, not a model error
Zac Veen: 1-for-2, 0 HR, 0 RBI [+0.8 runs vs projection] season xwOBA 0.291, below average - a good day, not a model error

Verdict: neither side broke materially - the result is closer to variance

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
COL	Tomoyuki Sugano (Starter)	18 / 14	7.5 / 5.79	5 / 3	4 / 3	2 / 1	3 / 2	0.97	5 / 7	3.49	venue away: 5.46 vs 4.96 ERA (worse here) · opp run factor 0.99 (average) · 25 prior appearances this season
NYY	Will Warren (Starter)	18 / 15	1.5 / 7.2 (-5.70)	1 / 4	4 / 3	1 / 0	2 / 2	1.93	8 / 5	5.4	venue home: 4.06 vs 4.45 ERA (ok) · opp run factor 1.03 (average) · 28 prior appearances this season
NYY	Paul Blackburn (Middle Reliever (High Leverage))	3 / 4	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0.43	1 / 0	1.4

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

Batter	Team	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
Ben Rice (DH)	NYY	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	1 / 1	-1.73	0.376 elite	bad day
Connor Norby (2B)	COL	3 / 3	2 / 1	1 / 0	0 / 0	1 / 0	1 / 0	1 / 0	0 / 0	1 / 1	1.71	0.298 below average	good day
Hunter Goodman (DH)	COL	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 0	0 / 1	-1.4	0.323 above average	bad day
Luis García Jr. (1B)	NYY	4 / 4	2 / 1	1 / 1	0 / 0	1 / 0	1 / 1	2 / 0	0 / 0	2 / 1	1.4	0.348 elite	projection too low
Spencer Jones (CF)	NYY	2 / 4	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	1 / 1	2 / 0	0 / 1	0.97	0.324 above average	projection too low
George Lombard Jr. (SS)	NYY	4 / 3	1 / 1	0 / 0	0 / 0	1 / 0	4 / 0	1 / 0	0 / 0	0 / 1	0.93	0.283 weak	good day
Zac Veen (LF)	COL	2 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0.8	0.291 below average	good day
Aaron Judge (RF)	NYY	4 / 4	2 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	0 / 1	1 / 1	-0.79	0.411 elite	bad day
Cole Carrigg (SS)	COL	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	2 / 1	-0.78	0.317 average	no signal
Cody Bellinger (LF)	NYY	4 / 4	1 / 1	0 / 0	0 / 0	1 / 0	1 / 1	1 / 1	0 / 1	1 / 0	0.6	0.339 above average	projection too low
Mickey Moniak (RF)	COL	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	3 / 1	-0.47	0.282 weak	—
Brett Sullivan (C)	COL	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	2 / 1	-0.47	0.288 below average	—
TJ Rumfield (1B)	COL	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	0 / 1	-0.47	0.335 above average	—
José Caballero (2B)	NYY	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	-0.47	0.28 weak	—
Ryan McMahon (3B)	NYY	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	-0.47	0.304 average	—
Austin Wells (C)	NYY	2 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	0.33	0.29 below average	—
Jake McCarthy (CF)	COL	4 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	1 / 1	-0.31	0.322 above average	—
Kyle Karros (3B)	COL	4 / 3	1 / 1	0 / 0	1 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	0.29	0.337 above average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	1.571	positive favors home; negative favors away
Recent form	0.95	positive favors home; negative favors away
Starting-pitcher matchup	-0.048	positive favors home; negative favors away
Composite	1.897	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.04-5.58	n/a	Base ML historical model prediction
variance_applied	4-6	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-8	0.784	Enhanced ML with recent form, H2H, low-score analysis
embedded_savant_run_environment	4-9	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	4-9	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	4-9	n/a	Capped at max 12 realistic runs
simulation_result	4-9	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.04-5.58	4-6	0.457	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-6	4-8	2	Enhanced ML with recent form, H2H, low-score analysis
embedded_savant_run_environment	4-8	4-9	1	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	4-9	4-9	0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	4-9	4-9	0	Capped at max 12 realistic runs
simulation_result	4-9	4-9	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
NYY	Ben Rice	batter / DH	better	predicted_RBI 2.49
NYY	Aaron Judge	batter / RF	better	predicted_HR 1.79
COL	Brett Sullivan	batter / C	deboosted	predicted_RBI 1.56
NYY	Ryan McMahon	batter / 3B	better	predicted_H 1.35
NYY	George Lombard Jr.	batter / SS	better	predicted_H 1.31
NYY	José Caballero	batter / 2B	better	predicted_H 1.226
NYY	Austin Wells	batter / C	worse	predicted_AB 1.187
COL	Jake McCarthy	batter / CF	worse	predicted_H -1.13

Structural audit findings

Severity	Rule	Finding	Subject
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_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	COL model pitcher projections differ from final simulation pitching box score (ER: current=8, sim=9).	COL
info	CURRENT_FINALIZED_MATCH_BOX	COL model-prediction batting totals differ from final simulation box score (AB: current=32, sim=30; H: current=8, sim=3; 2B: current=2, sim=0; 3B: current=0, sim=2; HR: current=1, sim=0; RBI: current=5, sim=4; BB: current=3, sim=4; SO: current=9, sim=5).	COL
info	CURRENT_PLAYER_MATCH_BOX	COL current finalized hitter lines are model predictions and differ from the simulation box score: Jake McCarthy: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, R current=1 box=0; Cole Carrigg: AB current=4 box=3, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Hunter Goodman: AB current=4 box=3, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0; Mickey Moniak: R current=0 box=1, SO current=1 box=0; Connor Norby: H current=1 box=0, SO current=1 box=0; Brett Sullivan: AB current=3 box=4, 3B current=0 box=1, RBI current=1 box=2, R current=0 box=1; Kyle Karros: H current=1 box=0, SO current=1 box=0; TJ Rumfield: AB current=4 box=3, 3B current=0 box=1, RBI current=0 box=1, SO current=1 box=0	COL
info	CURRENT_FINALIZED_MATCH_BOX	NYY model-prediction batting totals differ from final simulation box score (AB: current=32, sim=36; H: current=8, sim=12; HR: current=2, sim=5; RBI: current=4, sim=9; R: current=5, sim=9; SO: current=8, sim=9).	NYY
info	CURRENT_PLAYER_MATCH_BOX	NYY current finalized hitter lines are model predictions and differ from the simulation box score: Cody Bellinger: AB current=4 box=5, BB current=1 box=0, SO current=0 box=1; Aaron Judge: AB current=4 box=3, H current=1 box=2, HR current=1 box=2, RBI current=1 box=2; Luis García Jr.: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, R current=0 box=1; Ben Rice: AB current=4 box=5, H current=1 box=3, 2B current=0 box=1, HR current=1 box=0; Spencer Jones: H current=1 box=0, SO current=1 box=2; José Caballero: AB current=3 box=4, H current=1 box=2, HR current=0 box=1, RBI current=0 box=1; George Lombard Jr.: H current=1 box=2, HR current=0 box=1, RBI current=0 box=1, R current=0 box=1; Ryan McMahon: AB current=3 box=4, H current=1 box=2, HR current=0 box=1, RBI current=0 box=1	NYY
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

HOU@PHIHIT

Predicted: PHI 3-4 (conf +0.04, toss_up)Actual: PHI 7-11Brier 0.231total bias -11

Root cause: Toss-up that went our way

Evidence-backed diagnosis

published system edge 0.04 (toss_up); this is not the half-scale distance from 50%
team-run absolute error 11; total bias -11
pregame home-minus-away evidence: season +0.378, recent -0.314, pitcher +0.077, composite +0.479
confidence pipeline reached 0.527 before later publication/recency controls; published home probability was 0.520
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 11 (7 vs 18)
HOU offense under-projected by 4
PHI offense under-projected by 7
Cristopher Sánchez (PHI) allowed 4 more ER than projected (6 vs 2)

Team	Starter	Projected	Actual	ER err
HOU	Hunter Brown	4 ER / 17 outs	4 ER / 15 outs	0
PHI	Cristopher Sánchez	2 ER / 19 outs	6 ER / 18 outs	4

Decision internals: run factors home×0.998 / away×0.921; 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:

Cam Smith (HOU RF): 2-for-4, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.4]
Isaac Paredes (HOU 3B): 2-for-4, 1 HR, 3 RBI (projected 1.0 H, 0.0 HR) [run impact 1.07]
Trea Turner (PHI SS): 2-for-5, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.07]
Luis Arraez (PHI 2B): 2-for-4, 0 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 0.8]
Bryan De La Cruz (PHI RF): 2-for-3, 0 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 0.8]

Starter miss: Cristopher Sánchez (PHI) posted 9.00 ERA vs projected 2.84 (6 ER in 18 outs); no credible venue, opponent-strength or experience signal explains it — an outlier start (variance).

What to do: Variance — no model change is warranted for this start.

Pitcher versus batter: what actually broke this game

AWAY pitching

Hunter Brown (HOU) threw 15 outs, 4 ER = 7.2 ERA against a projected 6.35 ERA (0.85 ERA error)

Bullpen behind him: 0 ER over 0 outs.

The HOME lineup he faced produced 10 runs on 11 hits and 2 HR, against a projected 4 runs / 8 hits (6 runs vs projection).

Who did the damage to him:

Trea Turner: 2-for-5, 1 HR, 2 RBI [+1.07 runs vs projection] season xwOBA 0.294, below average - a good day, not a model error
Luis Arraez: 2-for-4, 0 HR, 1 RBI [+0.8 runs vs projection] season xwOBA 0.305, average - a good day, not a model error
Bryan De La Cruz: 2-for-3, 0 HR, 1 RBI [+0.8 runs vs projection] season xwOBA 0.281, weak - a good day, not a model error

Verdict: neither side broke materially - the result is closer to variance

HOME pitching

Cristopher Sánchez (PHI) threw 18 outs, 6 ER = 9 ERA against a projected 2.84 ERA (6.16 ERA error)

Bullpen behind him: 1 ER over 6 outs.

The AWAY lineup he faced produced 7 runs on 12 hits and 2 HR, against a projected 4 runs / 8 hits (3 runs vs projection).

Who did the damage to him:

Cam Smith: 2-for-4, 1 HR, 2 RBI [+1.4 runs vs projection] season xwOBA 0.319, average - a good day, not a model error
Isaac Paredes: 2-for-4, 1 HR, 3 RBI [+1.07 runs vs projection] season xwOBA 0.32, average - a good day, not a model error
Jeremy Peña: 2-for-5, 0 HR, 0 RBI [+0.78 runs vs projection] season xwOBA 0.325, above average - real projection error

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

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
HOU	Hunter Brown (Starter)	15 / 17	7.2 / 6.35	4 / 4	7 / 2	1 / 2	0 / 3	2.86	7 / 7	5.59	venue away: 2.67 vs 3.71 ERA (ok) · opp run factor 1 (average) · 16 prior appearances this season
PHI	Cristopher Sánchez (Starter)	18 / 19	9 / 2.84 (+6.16)	6 / 2	10 / 4	2 / 0	0 / 2	1.44	3 / 7	6.22	venue home: 1.83 vs 3.68 ERA (ok) · opp run factor 0.92 (average) · 29 prior appearances this season
PHI	Jhoan Duran (Closer)	3 / 3	0 / 9 (-9.00)	0 / 1	0 / 3	0 / 0	0 / 0	0.17	2 / 1	1.2
PHI	Jonathan Bowlan (Setup Man)	3 / 3	9 / 0 (+9.00)	1 / 0	2 / 1	0 / 0	0 / 0	0.21	1 / 1	1.24

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

Batter	Team	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
Cam Smith (RF)	HOU	4 / 3	2 / 1	0 / 0	0 / 0	1 / 0	2 / 0	2 / 0	0 / 0	0 / 1	1.4	0.319 average	no signal
Isaac Paredes (3B)	HOU	4 / 4	2 / 1	0 / 0	0 / 0	1 / 0	3 / 1	1 / 0	0 / 1	1 / 1	1.07	0.32 average	no signal
Trea Turner (SS)	PHI	5 / 4	2 / 1	0 / 0	0 / 0	1 / 0	2 / 0	2 / 1	0 / 1	1 / 1	1.07	0.294 below average	good day
Luis Arraez (2B)	PHI	4 / 4	2 / 1	1 / 1	0 / 0	0 / 0	1 / 0	1 / 0	1 / 0	0 / 0	0.8	0.305 average	no signal
Bryan De La Cruz (RF)	PHI	3 / 3	2 / 1	0 / 0	0 / 0	0 / 0	1 / 1	2 / 0	1 / 0	0 / 1	0.8	0.281 weak	good day
Jeremy Peña (SS)	HOU	5 / 4	2 / 1	1 / 0	0 / 0	0 / 0	0 / 1	1 / 1	0 / 0	2 / 1	0.78	0.325 above average	projection too low
Yainer Diaz (C)	HOU	4 / 4	2 / 1	1 / 0	0 / 0	0 / 0	1 / 0	0 / 1	0 / 0	0 / 1	0.78	0.3 below average	good day
Christian Walker (1B)	HOU	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0 / 1	-0.78	0.303 below average	projection too high
Yordan Alvarez (DH)	HOU	5 / 4	2 / 1	1 / 0	0 / 0	0 / 1	1 / 1	1 / 1	0 / 1	0 / 1	-0.48	0.447 elite	—
Bryson Stott (3B)	PHI	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	1 / 1	1 / 1	-0.47	0.33 above average	—
J.T. Realmuto (C)	PHI	4 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	2 / 1	0.47	0.321 above average	—
Kyle Schwarber (1B)	PHI	5 / 4	1 / 1	0 / 0	0 / 0	1 / 1	4 / 1	1 / 1	0 / 1	3 / 1	-0.33	0.369 elite	—
Jose Altuve (2B)	HOU	4 / 4	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0 / 0	1 / 1	0.31	0.274 weak	—
Jorge Barrosa (LF)	HOU	2 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0	0.235 weak	—
Daulton Varsho (CF)	HOU	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 1	0	0.292 below average	—
Bryce Harper (DH)	PHI	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	2 / 1	1 / 1	1 / 1	3 / 1	0	0.391 elite	—
Justin Crawford (CF)	PHI	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 1	0	0.281 weak	—
Brandon Marsh (LF)	PHI	2 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0	0.306 average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	0.378	positive favors home; negative favors away
Recent form	-0.314	positive favors home; negative favors away
Starting-pitcher matchup	0.077	positive favors home; negative favors away
Composite	0.479	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.33-5.11	n/a	Base ML historical model prediction
variance_applied	4-5	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	3-4	0.52	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run, matchup_recency_guardrail)
embedded_savant_run_environment	3-4	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	3-4	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	3-4	n/a	Capped at max 12 realistic runs
simulation_result	3-4	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.33-5.11	4-5	0.213	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-5	3-4	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	3-4	3-4	0	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	3-4	3-4	0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	3-4	3-4	0	Capped at max 12 realistic runs
simulation_result	3-4	3-4	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
PHI	Trea Turner	batter / SS	worse	predicted_RBI 2.45
PHI	Luis Arraez	batter / 2B	worse	predicted_AB -2.42
HOU	Yainer Diaz	batter / C	worse	predicted_SO 2.387
PHI	Brandon Marsh	batter / LF	worse	predicted_AB -1.544
HOU	Daulton Varsho	batter / CF	worse	predicted_AB -1.54
HOU	Jorge Barrosa	batter / LF	worse	predicted_AB 1.5
PHI	Bryan De La Cruz	batter / RF	worse	predicted_SO 1.381
PHI	Bryson Stott	batter / 3B	worse	predicted_SO 1.343

Structural audit findings

Severity	Rule	Finding	Subject
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_PITCHER_DUPLICATION	PHI pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	PHI
info	CURRENT_FINALIZED_MATCH_BOX	HOU model-prediction batting totals differ from final simulation box score (AB: current=34, sim=35; 2B: current=1, sim=3; HR: current=1, sim=0; RBI: current=4, sim=3; R: current=4, sim=3; BB: current=2, sim=4).	HOU
info	CURRENT_PLAYER_MATCH_BOX	HOU current finalized hitter lines are model predictions and differ from the simulation box score: Isaac Paredes: AB current=4 box=5, RBI current=1 box=0, BB current=1 box=0; Jose Altuve: H current=1 box=2, 2B current=0 box=1, RBI current=0 box=1; Yordan Alvarez: H current=1 box=2, 2B current=0 box=1, HR current=1 box=0, SO current=1 box=2; Jeremy Peña: BB current=0 box=1, SO current=1 box=2; Yainer Diaz: H current=1 box=0, R current=1 box=0, SO current=1 box=3; Jorge Barrosa: AB current=3 box=4, H current=0 box=1, SO current=1 box=0; Christian Walker: AB current=4 box=3, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Daulton Varsho: AB current=4 box=3, 2B current=0 box=1, BB current=0 box=1, SO current=1 box=0	HOU
info	CURRENT_FINALIZED_MATCH_BOX	PHI model-prediction batting totals differ from final simulation box score (AB: current=32, sim=26; H: current=8, sim=2; 2B: current=1, sim=0; HR: current=1, sim=2).	PHI
info	CURRENT_PLAYER_MATCH_BOX	PHI current finalized hitter lines are model predictions and differ from the simulation box score: Bryce Harper: HR current=0 box=1, BB current=1 box=0, SO current=1 box=0; Kyle Schwarber: AB current=4 box=3, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0; Bryson Stott: AB current=4 box=3, H current=1 box=0, RBI current=1 box=0, R current=1 box=0; Luis Arraez: AB current=4 box=2, H current=1 box=0, 2B current=1 box=0, R current=0 box=1; Trea Turner: AB current=4 box=3, HR current=0 box=1, RBI current=0 box=3; Justin Crawford: H current=1 box=0, SO current=1 box=0; Brandon Marsh: AB current=3 box=2, H current=1 box=0, BB current=0 box=1, SO current=1 box=2; Bryan De La Cruz: H current=1 box=0, RBI current=1 box=0, SO current=1 box=2	PHI
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

LAA@BOSMISS

Predicted: BOS 4-5 (conf +0.06, lean)Actual: LAA 6-4Brier 0.28total bias -1

Root cause: Several errors of similar size

Evidence-backed diagnosis

published system edge 0.06 (lean); this is not the half-scale distance from 50%
team-run absolute error 3; total bias -1
pregame home-minus-away evidence: season +0.917, recent +0.330, pitcher -0.086, composite +0.578
confidence pipeline reached 0.540 before later publication/recency controls; published home probability was 0.529
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

Jake Bennett (BOS) allowed 3 more ER than projected (6 vs 3)

Team	Starter	Projected	Actual	ER err
LAA	Ryan Johnson	4 ER / 16 outs	3 ER / 18 outs	-1
BOS	Jake Bennett	3 ER / 17 outs	6 ER / 16 outs	3

Decision internals: run factors home×0.931 / away×1.016; 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:

Denzer Guzman (LAA 3B): 3-for-4, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 1.87]
Jose Siri (LAA RF): 2-for-4, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.4]
Mickey Gasper (BOS C): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.8]
Jarren Duran (BOS LF): 2-for-3, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact 0.8]
Vaughn Grissom (LAA 1B): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.78]

Starter miss: Jake Bennett (BOS) posted 10.12 ERA vs projected 4.76 (6 ER in 16 outs); no credible venue, opponent-strength or experience signal explains it — an outlier start (variance).

What to do: Variance — no model change is warranted for this start.

Pitcher versus batter: what actually broke this game

AWAY pitching

Ryan Johnson (LAA) threw 18 outs, 3 ER = 4.5 ERA against a projected 6.75 ERA (-2.25 ERA error)

Bullpen behind him: 1 ER over 4 outs.

The HOME lineup he faced produced 4 runs on 8 hits and 2 HR, against a projected 4 runs / 8 hits (0 runs vs projection).

Who did the damage to him:

Jarren Duran: 2-for-3, 0 HR, 0 RBI [+0.8 runs vs projection] season xwOBA 0.3, below average - a good day, not a model error
Nate Eaton: 1-for-4, 0 HR, 1 RBI [+0.47 runs vs projection] season xwOBA 0.291, below average - a good day, not a model error

Verdict: neither side broke materially - the result is closer to variance

HOME pitching

Jake Bennett (BOS) threw 16 outs, 6 ER = 10.12 ERA against a projected 4.76 ERA (5.36 ERA error)

Bullpen behind him: 0 ER over 0 outs.

The AWAY lineup he faced produced 6 runs on 12 hits and 3 HR, against a projected 3 runs / 8 hits (3 runs vs projection).

Who did the damage to him:

Denzer Guzman: 3-for-4, 1 HR, 1 RBI [+1.87 runs vs projection] season xwOBA 0.263, weak - a good day, not a model error
Jose Siri: 2-for-4, 1 HR, 2 RBI [+1.4 runs vs projection] season xwOBA 0.302, below average - a good day, not a model error
Oswald Peraza: 2-for-4, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.271, weak - a good day, not a model error

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

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
LAA	Ryan Johnson (Starter)	18 / 16	4.5 / 6.75	3 / 4	5 / 4	2 / 0	1 / 3	2.47	5 / 7	4.46	venue away: 5.72 vs 3.18 ERA (worse here) · opp run factor 0.93 (average) · 18 prior appearances this season
LAA	Ben Joyce (Closer)	3 / 3	9 / 9	1 / 1	1 / 1	0 / 0	1 / 1	0.56	1 / 0	1.41
LAA	Tayler Saucedo (Middle Reliever (High Leverage))	1 / 5	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0.16	1 / 1	0.83
BOS	Jake Bennett (Starter)	16 / 17	10.12 / 4.76 (+5.36)	6 / 3	9 / 7	3 / 1	1 / 1	1.04	2 / 6	5.58	venue home: 3.52 vs 3.19 ERA (ok) · opp run factor 1.02 (weak) · 17 prior appearances this season

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

Batter	Team	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
Denzer Guzman (3B)	LAA	4 / 3	3 / 1	0 / 0	0 / 0	1 / 0	1 / 0	2 / 1	0 / 0	0 / 1	1.87	0.263 weak	good day
Jose Siri (RF)	LAA	4 / 4	2 / 1	0 / 0	0 / 0	1 / 0	2 / 1	1 / 0	0 / 0	0 / 1	1.4	0.302 below average	good day
Mickey Gasper (C)	BOS	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	-0.8	0.374 elite	bad day
Jarren Duran (LF)	BOS	3 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 1	2 / 0	1 / 0	0 / 1	0.8	0.3 below average	good day
Vaughn Grissom (1B)	LAA	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0 / 1	-0.78	0.315 average	no signal
Oswald Peraza (2B)	LAA	4 / 3	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 1	0.47	0.271 weak	—
Tyler Heineman (C)	LAA	3 / 3	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0.47	0.239 weak	—
Nate Eaton (CF)	BOS	4 / 3	1 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0 / 1	0.47	0.291 below average	—
Isiah Kiner-Falefa (2B)	BOS	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	-0.47	0.315 average	—
Kyren Paris (LF)	LAA	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0.33	0.247 weak	—
Mike Trout (CF)	LAA	4 / 4	1 / 1	1 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	0 / 1	-0.33	0.375 elite	—
Nick Sogard (1B)	BOS	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	2 / 1	-0.33	0.321 above average	—
Trevor Story (SS)	BOS	4 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	1 / 1	-0.31	0.246 weak	—
Roman Anthony (DH)	BOS	5 / 4	1 / 1	0 / 1	0 / 0	1 / 0	2 / 0	1 / 0	0 / 1	3 / 1	0.29	0.352 elite	—
Christian Moore (DH)	LAA	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 0	3 / 1	-0.14	0.249 weak	—
Zach Neto (SS)	LAA	3 / 4	2 / 1	0 / 0	0 / 0	1 / 1	3 / 1	1 / 1	0 / 1	0 / 1	0.14	0.307 average	—
Caleb Durbin (3B)	BOS	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	1 / 0	0	0.291 below average	—
Wilyer Abreu (RF)	BOS	3 / 4	1 / 1	0 / 0	0 / 0	1 / 1	1 / 1	1 / 1	1 / 1	0 / 1	0	0.347 elite	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	0.917	positive favors home; negative favors away
Recent form	0.33	positive favors home; negative favors away
Starting-pitcher matchup	-0.086	positive favors home; negative favors away
Composite	0.578	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.09-5.51	n/a	Base ML historical model prediction
variance_applied	4-6	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-5	0.529	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

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.09-5.51	4-6	0.583	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-6	4-5	-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	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

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
LAA	Kyren Paris	batter / LF	better	predicted_SO -2
BOS	Roman Anthony	batter / DH	better	predicted_RBI 1.825
LAA	Oswald Peraza	batter / 2B	better	predicted_RBI 1.82
LAA	Vaughn Grissom	batter / 1B	worse	predicted_AB -1.636
LAA	Zach Neto	batter / SS	worse	predicted_H -1.59
LAA	Jose Siri	batter / RF	worse	predicted_AB 1.493
LAA	Tyler Heineman	batter / C	worse	predicted_AB 1.305
LAA	Christian Moore	batter / DH	better	predicted_SO -1.182

Structural audit findings

Severity	Rule	Finding	Subject
info	CURRENT_PITCHER_DUPLICATION	BOS pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	BOS
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_FINALIZED_MATCH_BOX	BOS model-prediction batting totals differ from final simulation box score (AB: current=32, sim=29; H: current=8, sim=5; 2B: current=2, sim=3; 3B: current=0, sim=1; HR: current=1, sim=0; RBI: current=4, sim=5; R: current=4, sim=5; SO: current=7, sim=8).	BOS
info	CURRENT_PLAYER_MATCH_BOX	BOS current finalized hitter lines are model predictions and differ from the simulation box score: Mickey Gasper: H current=1 box=0, SO current=1 box=0; Roman Anthony: RBI current=0 box=2, R current=0 box=1, BB current=1 box=0; Nick Sogard: H current=1 box=0, R current=1 box=0, BB current=1 box=0, SO current=1 box=0; Caleb Durbin: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, BB current=0 box=1; Trevor Story: H current=1 box=2, R current=0 box=1, SO current=1 box=2; Nate Eaton: AB current=3 box=2, BB current=0 box=1, SO current=1 box=0; Wilyer Abreu: 3B current=0 box=1, HR current=1 box=0, BB current=1 box=0, SO current=1 box=2; Jarren Duran: AB current=4 box=3, 2B current=0 box=1, R current=0 box=1	BOS
info	CURRENT_FINALIZED_MATCH_BOX	LAA model-prediction batting totals differ from final simulation box score (AB: current=32, sim=34; R: current=3, sim=4; BB: current=2, sim=1; SO: current=9, sim=10; SF: current=0, sim=1).	LAA
info	CURRENT_PLAYER_MATCH_BOX	LAA current finalized hitter lines are model predictions and differ from the simulation box score: Vaughn Grissom: AB current=4 box=2, H current=1 box=0, 2B current=1 box=0, R current=0 box=1; Jose Siri: H current=1 box=2, RBI current=1 box=0, R current=0 box=1; Christian Moore: 2B current=0 box=1, RBI current=1 box=0, SO current=1 box=0; Kyren Paris: AB current=3 box=4, SO current=1 box=0; Zach Neto: H current=1 box=0, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0; Mike Trout: H current=1 box=2, 2B current=1 box=0, RBI current=0 box=1, BB current=1 box=0; Denzer Guzman: AB current=3 box=4, H current=1 box=0, R current=1 box=0, SO current=1 box=2; Oswald Peraza: AB current=3 box=4, H current=1 box=2, 2B current=0 box=1, HR current=0 box=1	LAA
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

MIN@DETHIT

Predicted: DET 4-5 (conf +0.05, lean)Actual: DET 2-7Brier 0.227total bias +0

Root cause: Solid correct call

Evidence-backed diagnosis

published system edge 0.05 (lean); this is not the half-scale distance from 50%
team-run absolute error 4; total bias +0
pregame home-minus-away evidence: season +0.198, recent -0.702, pitcher +0.010, composite +0.136
confidence pipeline reached 0.533 before later publication/recency controls; published home probability was 0.524
score-volatility actions: no_change
forecast-file integrity: clean (0 defects, 8 informational note(s)); the saved artifact is not why the pick missed

Quantified contributing factors

No large single-factor errors.

Team	Starter	Projected	Actual	ER err
MIN	Zebby Matthews	4 ER / 18 outs	5 ER / 7 outs	1
DET	Keider Montero	3 ER / 17 outs	2 ER / 12 outs	-1

Decision internals: run factors home×0.965 / away×0.955; 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:

Josh Bell (MIN DH): 2-for-4, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 1.4]
Luke Keaschall (MIN RF): 0-for-3, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -1.11]
Colt Keith (DET 1B): 1-for-3, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 0.93]
Dillon Dingler (DET C): 1-for-3, 0 HR, 2 RBI (projected 1.0 H, 1.0 HR) [run impact -0.93]

Starter miss: Zebby Matthews (MIN) posted 19.29 ERA vs projected 6.00 (5 ER in 7 outs); he is materially worse pitching away (6.45 ERA there vs 2.84 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

AWAY pitching

Zebby Matthews (MIN) threw 7 outs, 5 ER = 19.29 ERA against a projected 6 ERA (13.29 ERA error)

Bullpen behind him: 0 ER over 0 outs.

The HOME lineup he faced produced 4 runs on 6 hits and 1 HR, against a projected 4 runs / 7 hits (0 runs vs projection).

Who did the damage to him:

Colt Keith: 1-for-3, 1 HR, 2 RBI [+0.93 runs vs projection] season xwOBA 0.334, above average - real projection error
Max Clark: 1-for-3, 0 HR, 0 RBI [+0.33 runs vs projection] season xwOBA 0.307, average - a good day, not a model error
John Peck: 1-for-3, 0 HR, 1 RBI [+0.33 runs vs projection] season xwOBA 0.329, above average - a good day, not a model error

Verdict: both: the starter was hit hard AND the lineup underproduced

HOME pitching

Keider Montero (DET) threw 12 outs, 2 ER = 4.5 ERA against a projected 4.76 ERA (-0.26 ERA error)

Bullpen behind him: 0 ER over 9 outs.

The AWAY lineup he faced produced 2 runs on 5 hits and 2 HR, against a projected 4 runs / 7 hits (-2 runs vs projection).

Who did the damage to him:

Josh Bell: 2-for-4, 1 HR, 1 RBI [+1.4 runs vs projection] season xwOBA 0.334, above average - real projection error
Victor Caratini: 1-for-3, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.319, average - a good day, not a model error
Walker Jenkins: 0-for-2, 0 HR, 0 RBI [+0.33 runs vs projection] season xwOBA 0.273, weak - a good day, not a model error

Verdict: neither side broke materially - the result is closer to variance

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
MIN	Zebby Matthews (Starter)	7 / 18	19.29 / 6 (+13.29)	5 / 4	2 / 5	1 / 0	4 / 1	1.19	2 / 6	4.28	venue away: 6.45 vs 2.84 ERA (worse here) · opp run factor 0.97 (average) · 20 prior appearances this season
DET	Keider Montero (Starter)	12 / 17	4.5 / 4.76	2 / 3	2 / 3	2 / 1	2 / 3	1.37	2 / 7	3.14	venue home: 3.24 vs 4.31 ERA (ok) · opp run factor 0.96 (average) · 29 prior appearances this season
DET	Kyle Finnegan (Middle Reliever (High Leverage))	3 / 3	0 / 9 (-9.00)	0 / 1	1 / 1	0 / 0	0 / 0	0.54	0 / 0	0.84
DET	Tyler Holton (Setup Man)	6 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0.34	2 / 2	0.78

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

Batter	Team	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
Josh Bell (DH)	MIN	4 / 4	2 / 1	0 / 0	0 / 0	1 / 0	1 / 1	1 / 1	0 / 0	0 / 1	1.4	0.334 above average	projection too low
Luke Keaschall (RF)	MIN	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	-1.11	0.316 average	no signal
Colt Keith (1B)	DET	3 / 3	1 / 1	0 / 0	0 / 0	1 / 0	2 / 0	1 / 0	0 / 0	1 / 1	0.93	0.334 above average	projection too low
Dillon Dingler (C)	DET	3 / 4	1 / 1	0 / 0	0 / 0	0 / 1	2 / 1	0 / 1	0 / 0	1 / 1	-0.93	0.341 elite	bad day
Brooks Lee (3B)	MIN	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	1 / 0	-0.47	0.276 weak	—
Kaelen Culpepper (SS)	MIN	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	0 / 1	-0.47	0.309 average	—
Royce Lewis (1B)	MIN	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	-0.47	0.313 average	—
Victor Caratini (C)	MIN	3 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0.47	0.319 average	—
Riley Greene (DH)	DET	2 / 4	0 / 1	0 / 1	0 / 0	0 / 0	1 / 1	0 / 1	2 / 1	0 / 1	-0.45	0.37 elite	—
Walker Jenkins (CF)	MIN	2 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	2 / 1	0.33	0.273 weak	—
Max Clark (CF)	DET	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	1 / 0	1 / 1	0.33	0.307 average	—
John Peck (SS)	DET	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	1 / 0	1 / 1	0.33	0.329 above average	—
Kevin McGonigle (3B)	DET	4 / 4	1 / 1	1 / 0	0 / 0	0 / 0	0 / 1	1 / 1	1 / 1	2 / 0	0.31	0.351 elite	—
Brett Callahan (LF)	DET	4 / 3	1 / 1	1 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	2 / 1	0.31	0.314 average	—
Kody Clemens (2B)	MIN	4 / 4	1 / 1	0 / 0	0 / 0	1 / 1	1 / 1	1 / 1	0 / 0	1 / 1	0	0.34 elite	—
Trevor Larnach (LF)	MIN	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	1 / 1	0	0.324 above average	—
Ben Malgeri (RF)	DET	4 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0	0.326 above average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	0.198	positive favors home; negative favors away
Recent form	-0.702	positive favors home; negative favors away
Starting-pitcher matchup	0.01	positive favors home; negative favors away
Composite	0.136	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.16-4.79	n/a	Base ML historical model prediction
variance_applied	4-5	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-5	0.524	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

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.16-4.79	4-5	0.369	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

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
MIN	Trevor Larnach	batter / LF	better	predicted_RBI 1.62
DET	Brett Callahan	batter / LF	deboosted	predicted_AB 1.5
DET	Riley Greene	batter / DH	better	predicted_RBI 1.475
MIN	Luke Keaschall	batter / RF	worse	predicted_SO 1.4
MIN	Kaelen Culpepper	batter / SS	worse	predicted_AB -1.29
MIN	Royce Lewis	batter / 1B	worse	predicted_AB -1.2
MIN	Josh Bell	batter / DH	worse	predicted_SO 1.187
MIN	Victor Caratini	batter / C	worse	predicted_AB -1.151

Structural audit findings

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=31, sim=32; H: current=8, sim=7; RBI: current=4, sim=5; BB: current=3, sim=1; SO: current=8, sim=7).	DET
info	CURRENT_PLAYER_MATCH_BOX	DET current finalized hitter lines are model predictions and differ from the simulation box score: Kevin McGonigle: BB current=1 box=0, SO current=0 box=1; Colt Keith: R current=0 box=1, BB current=0 box=1; Riley Greene: RBI current=1 box=2, BB current=1 box=0, SO current=1 box=0; Max Clark: H current=1 box=0, R current=1 box=0, SO current=1 box=2; Ben Malgeri: H current=0 box=1, SO current=1 box=0; Dillon Dingler: HR current=1 box=0; Brett Callahan: AB current=3 box=4, HR current=0 box=1, RBI current=0 box=1, R current=0 box=1; Gleyber Torres: RBI current=1 box=0, R current=1 box=0, BB current=1 box=0	DET
info	CURRENT_FINALIZED_MATCH_BOX	MIN model pitcher projections differ from final simulation pitching box score (H: current=6, sim=7).	MIN
info	CURRENT_FINALIZED_MATCH_BOX	MIN model-prediction batting totals differ from final simulation box score (AB: current=33, sim=31; H: current=7, sim=4; 2B: current=1, sim=0; BB: current=2, sim=3; SO: current=6, sim=10).	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; Josh Bell: RBI current=1 box=0, SO current=1 box=2; Kody Clemens: H current=1 box=0, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0; Trevor Larnach: AB current=3 box=4, HR current=0 box=1, RBI current=0 box=2, R current=0 box=1; Brooks Lee: R current=0 box=1, SO current=0 box=1; Kaelen Culpepper: AB current=4 box=3, H current=1 box=0, RBI current=1 box=0, R current=1 box=0; Royce Lewis: AB current=4 box=3, H current=1 box=0, BB current=0 box=1; Victor Caratini: AB current=3 box=2, BB current=0 box=1, SO current=0 box=1	MIN
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

NYM@MIAHIT

Predicted: NYM 7-4 (conf -0.08, lean)Actual: NYM 15-14Brier 0.21total bias -18

Root cause: Right winner, wrong score

Evidence-backed diagnosis

published system edge 0.08 (lean); this is not the half-scale distance from 50%
team-run absolute error 18; total bias -18
pregame home-minus-away evidence: season +0.068, recent -0.450, pitcher -0.083, composite -0.798
confidence pipeline reached 0.612 before later publication/recency controls; published home probability was 0.458
score-volatility actions: supported_blowout_margin_expansion
forecast-file integrity: clean (0 defects, 9 informational note(s)); the saved artifact is not why the pick missed

Quantified contributing factors

total runs under-projected by 18 (11 vs 29)
NYM offense under-projected by 8
MIA offense under-projected by 10
Janson Junk (MIA) allowed 5 more ER than projected (8 vs 3)

Team	Starter	Projected	Actual	ER err
NYM	Robert Stock	3 ER / 13 outs	3 ER / 12 outs	0
MIA	Janson Junk	3 ER / 15 outs	8 ER / 9 outs	5

Decision internals: run factors home×0.938 / away×1.022; 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:

Joe Mack (MIA C): 3-for-5, 1 HR, 3 RBI (projected 1.0 H, 0.0 HR) [run impact 1.87]
Owen Caissie (MIA RF): 2-for-5, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.73]
Otto Lopez (MIA SS): 2-for-3, 0 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.42]
Graham Pauley (MIA 3B): 3-for-6, 0 HR, 2 RBI (projected 0.0 H, 0.0 HR) [run impact 1.41]
Juan Soto (NYM LF): 0-for-5, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.4]

Starter miss: Janson Junk (MIA) posted 24.00 ERA vs projected 5.40 (8 ER in 9 outs); no credible venue, opponent-strength or experience signal explains it — an outlier start (variance).

What to do: Variance — no model change is warranted for this start.

Pitcher versus batter: what actually broke this game

AWAY pitching

Robert Stock (NYM) threw 12 outs, 3 ER = 6.75 ERA against a projected 6.23 ERA (0.52 ERA error) only 7 prior appearances in the covered season - the projection had little of his own record to stand on

Bullpen behind him: 6 ER over 9 outs.

The HOME lineup he faced produced 14 runs on 20 hits and 2 HR, against a projected 5 runs / 8 hits (9 runs vs projection).

Who did the damage to him:

Joe Mack: 3-for-5, 1 HR, 3 RBI [+1.87 runs vs projection] season xwOBA 0.29, below average - a good day, not a model error
Owen Caissie: 2-for-5, 1 HR, 2 RBI [+1.73 runs vs projection] season xwOBA 0.293, below average - a good day, not a model error
Otto Lopez: 2-for-3, 0 HR, 2 RBI [+1.42 runs vs projection] season xwOBA 0.324, above average - real projection error

Verdict: neither side broke materially - the result is closer to variance

HOME pitching

Janson Junk (MIA) threw 9 outs, 8 ER = 24 ERA against a projected 5.4 ERA (18.6 ERA error)

Bullpen behind him: 4 ER over 12 outs.

The AWAY lineup he faced produced 12 runs on 12 hits and 1 HR, against a projected 5 runs / 8 hits (7 runs vs projection).

Who did the damage to him:

Luis Torrens: 3-for-5, 0 HR, 4 RBI [+1.25 runs vs projection] season xwOBA 0.295, below average - a good day, not a model error
Bo Bichette: 1-for-4, 0 HR, 1 RBI [+0.66 runs vs projection] season xwOBA 0.328, above average - real projection error
A.J. Ewing: 3-for-4, 0 HR, 2 RBI [+0.61 runs vs projection] season xwOBA 0.321, above average - real projection error

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

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
NYM	Robert Stock (Starter)	12 / 13	6.75 / 6.23	3 / 3	5 / 5	0 / 1	2 / 2	2.46	4 / 4	3.94	venue away: 8.64 vs 4.58 ERA (ok) · opp run factor 0.94 (average) · 7 prior appearances this season
NYM	Jonathan Pintaro (Middle Reliever (High Leverage))	3 / 4	18 / 0 (+18.00)	2 / 0	3 / 0	0 / 0	0 / 0	0.7	0 / 1	1.17
NYM	Daniel Duarte (Setup Man)	5 / 4	5.4 / 0 (+5.40)	1 / 0	2 / 0	1 / 0	1 / 0	0.22	2 / 1	0.94
NYM	Nathan Lavender (Middle Reliever (High Leverage))	1 / 3	81 / 0 (+81.00)	3 / 0	4 / 0	0 / 0	0 / 0	0.22	0 / 0	1.15
MIA	Janson Junk (Starter)	9 / 15	24 / 5.4 (+18.60)	8 / 3	8 / 4	1 / 1	1 / 2	1.2	0 / 3	3	venue home: 3.53 vs 4.78 ERA (ok) · opp run factor 1.02 (average) · 22 prior appearances this season
MIA	Calvin Faucher (Closer)	3 / 3	0 / 0	0 / 0	1 / 0	0 / 0	1 / 0	0.45	2 / 0	1.03
MIA	Michael Petersen (Middle Reliever (High Leverage))	3 / 3	0 / 27 (-27.00)	0 / 3	0 / 2	0 / 1	0 / 2	0.53	1 / 0	1.21
MIA	Tyler Zuber (Setup Man)	3 / 3	27 / 0 (+27.00)	3 / 0	3 / 1	1 / 0	1 / 0	0.27	1 / 2	1.23
MIA	Victor Vodnik (Middle Reliever (High Leverage))	3 / 3	9 / 0 (+9.00)	1 / 0	3 / 0	0 / 0	0 / 0	0.48	1 / 0	0.76

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

Batter	Team	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
Joe Mack (C)	MIA	5 / 3	3 / 1	0 / 0	0 / 0	1 / 0	3 / 0	2 / 0	0 / 0	1 / 1	1.87	0.29 below average	good day
Owen Caissie (RF)	MIA	5 / 3	2 / 1	0 / 0	0 / 0	1 / 0	2 / 1	2 / 0	1 / 0	1 / 1	1.73	0.293 below average	good day
Otto Lopez (SS)	MIA	3 / 4	2 / 1	0 / 0	1 / 0	0 / 0	2 / 0	3 / 1	1 / 0	0 / 1	1.42	0.324 above average	projection too low
Graham Pauley (3B)	MIA	6 / 3	3 / 0	0 / 0	0 / 0	0 / 0	2 / 0	1 / 0	0 / 0	1 / 1	1.41	0.238 weak	good day
Juan Soto (LF)	NYM	5 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	1 / 1	1 / 0	-1.4	0.419 elite	bad day
Luis Torrens (C)	NYM	5 / 3	3 / 1	1 / 0	0 / 0	0 / 0	4 / 0	2 / 0	0 / 0	1 / 0	1.25	0.295 below average	good day
Jakob Marsee (CF)	MIA	6 / 4	4 / 1	0 / 0	0 / 0	0 / 0	3 / 0	1 / 1	0 / 1	0 / 1	1.08	0.309 average	no signal
Griffin Conine (1B)	MIA	6 / 4	3 / 1	1 / 0	0 / 0	0 / 0	2 / 0	2 / 1	0 / 1	1 / 1	0.92	0.367 elite	projection too low
Heriberto Hernández (LF)	MIA	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	2 / 0	1 / 1	-0.74	0.355 elite	bad day
Bo Bichette (3B)	NYM	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	2 / 0	1 / 0	0.66	0.328 above average	projection too low
A.J. Ewing (CF)	NYM	4 / 4	3 / 1	0 / 0	0 / 0	0 / 0	2 / 0	2 / 0	0 / 1	1 / 1	0.61	0.321 above average	projection too low
Francisco Lindor (SS)	NYM	5 / 4	1 / 1	0 / 0	0 / 0	1 / 0	3 / 1	2 / 1	0 / 1	1 / 1	0.6	0.336 above average	projection too low
Jared Young (DH)	NYM	5 / 4	2 / 1	0 / 1	0 / 0	0 / 0	0 / 1	3 / 1	1 / 0	0 / 1	0.49	0.31 average	—
Brett Baty (1B)	NYM	1 / 3	1 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0 / 1	0.47	0.313 average	—
Marcus Semien (2B)	NYM	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	2 / 0	1 / 0	0 / 1	0.33	0.304 average	—
Javier Sanoja (2B)	MIA	6 / 4	2 / 1	0 / 1	0 / 0	0 / 0	0 / 1	1 / 0	0 / 0	0 / 0	0.16	0.289 below average	—
Carson Benge (RF)	NYM	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	1 / 0	1 / 1	-0.14	0.335 above average	—
Kyle Stowers (DH)	MIA	6 / 4	1 / 1	1 / 0	0 / 0	0 / 0	0 / 1	1 / 1	0 / 1	2 / 1	-0.02	0.333 above average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	0.068	positive favors home; negative favors away
Recent form	-0.45	positive favors home; negative favors away
Starting-pitcher matchup	-0.083	positive favors home; negative favors away
Composite	-0.798	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.23-4.31	n/a	Base ML historical model prediction
variance_applied	4-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	6-4	0.542	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by winner_starter_is_below_average, starter_quality_favours_the_other_side, matchup_recency_guardrail)
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

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.23-4.31	4-4	0.075	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-4	6-4	2	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by winner_starter_is_below_average, starter_quality_favours_the_other_side, matchup_recency_guardrail)
embedded_savant_run_environment	6-4	7-4	1	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

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
MIA	Otto Lopez	batter / SS	worse	predicted_AB -2.74
NYM	Marcus Semien	batter / 2B	better	predicted_RBI 2.65
NYM	Luis Torrens	batter / C	better	predicted_RBI 2.62
NYM	Francisco Lindor	batter / SS	worse	predicted_SO 2.116
NYM	Juan Soto	batter / LF	worse	predicted_AB -1.52
MIA	Heriberto Hernández	batter / LF	worse	predicted_AB -1.33
MIA	Graham Pauley	batter / 3B	better	predicted_AB 1.326
MIA	Owen Caissie	batter / RF	worse	predicted_AB 1.196

Structural audit findings

Severity	Rule	Finding	Subject
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_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_FINALIZED_MATCH_BOX	MIA model pitcher projections differ from final simulation pitching box score (ER: current=6, sim=7).	MIA
info	CURRENT_FINALIZED_MATCH_BOX	MIA model-prediction batting totals differ from final simulation box score (AB: current=33, sim=32; H: current=8, sim=5; R: current=5, sim=4; BB: current=3, sim=2; SO: current=8, sim=6; HBP: current=0, sim=3).	MIA
info	CURRENT_PLAYER_MATCH_BOX	MIA current finalized hitter lines are model predictions and differ from the simulation box score: Kyle Stowers: 2B current=0 box=1, BB current=1 box=0, HBP current=0 box=1; Griffin Conine: AB current=4 box=3, H current=1 box=0, RBI current=0 box=1; Javier Sanoja: 2B current=1 box=0, RBI current=1 box=0; Heriberto Hernández: AB current=4 box=3, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0; Otto Lopez: AB current=4 box=2, H current=1 box=0, R current=1 box=0, SO current=1 box=0; Owen Caissie: AB current=3 box=4, RBI current=1 box=0, SO current=1 box=0; Graham Pauley: AB current=3 box=4, H current=0 box=1, RBI current=0 box=1, R current=0 box=1; Jakob Marsee: H current=1 box=0, R current=1 box=0, BB current=1 box=0	MIA
info	CURRENT_FINALIZED_MATCH_BOX	NYM model pitcher projections differ from final simulation pitching box score (H: current=6, sim=5).	NYM
info	CURRENT_FINALIZED_MATCH_BOX	NYM model-prediction batting totals differ from final simulation box score (AB: current=33, sim=34; H: current=8, sim=7; 2B: current=1, sim=2; HR: current=1, sim=2; RBI: current=4, sim=7; R: current=5, sim=7; BB: current=3, sim=4; SO: current=6, sim=5).	NYM
info	CURRENT_PLAYER_MATCH_BOX	NYM current finalized hitter lines are model predictions and differ from the simulation box score: Carson Benge: AB current=4 box=3, H current=1 box=0, BB current=0 box=1, SO current=1 box=0; Jared Young: SO current=1 box=0; Marcus Semien: AB current=3 box=4, HR current=0 box=1, RBI current=0 box=3, R current=0 box=1; A.J. Ewing: AB current=4 box=3, H current=1 box=0; Brett Baty: AB current=3 box=4, H current=0 box=1, R current=0 box=1, SO current=1 box=0; Luis Torrens: AB current=3 box=4, H current=1 box=2, HR current=0 box=1, RBI current=0 box=3; Bo Bichette: RBI current=1 box=0; Francisco Lindor: AB current=4 box=5, H current=1 box=0, RBI current=1 box=0, R current=1 box=0	NYM
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

PIT@CHWHIT

Predicted: PIT 5-4 (conf -0.20, moderate)Actual: PIT 4-2Brier 0.161total bias +3

Root cause: Strong correct call

Evidence-backed diagnosis

published system edge 0.20 (moderate); this is not the half-scale distance from 50%
team-run absolute error 3; total bias +3
pregame home-minus-away evidence: season -0.097, recent -0.838, pitcher +0.056, composite -0.320
confidence pipeline reached 0.650 before later publication/recency controls; published home probability was 0.402
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 over-projected by 3 (9 vs 6)
Davis Martin (CHW) allowed 3 fewer ER than projected (1 vs 4)

Team	Starter	Projected	Actual	ER err
PIT	Lake Bachar	2 ER / 15 outs	0 ER / 5 outs	-2
CHW	Davis Martin	4 ER / 15 outs	1 ER / 15 outs	-3

Decision internals: run factors home×0.989 / away×0.933; 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:

Colson Montgomery (CHW 3B): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.73]
Bryan Reynolds (PIT LF): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -1.11]
Munetaka Murakami (CHW 1B): 1-for-3, 1 HR, 1 RBI (projected 0.0 H, 0.0 HR) [run impact 1.07]
Esmerlyn Valdez (PIT RF): 1-for-2, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 0.93]
Braden Montgomery (CHW RF): 1-for-2, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact 0.64]

Starter miss: Lake Bachar (PIT) posted 0.00 ERA vs projected 3.60 (0 ER in 5 outs); he is materially worse pitching away (3.00 ERA there vs 0.93 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

AWAY pitching

Lake Bachar (PIT) threw 5 outs, 0 ER = 0 ERA against a projected 3.6 ERA (-3.6 ERA error)

Bullpen behind him: 1 ER over 9 outs.

The HOME lineup he faced produced 2 runs on 6 hits and 1 HR, against a projected 5 runs / 6 hits (-3 runs vs projection).

Who did the damage to him:

Munetaka Murakami: 1-for-3, 1 HR, 1 RBI [+1.07 runs vs projection] season xwOBA 0.344, elite - real projection error
Braden Montgomery: 1-for-2, 0 HR, 0 RBI [+0.64 runs vs projection] season xwOBA 0.328, above average - real projection error
Kyle Teel: 0-for-2, 0 HR, 0 RBI [+0.33 runs vs projection] season xwOBA 0.311, average - a good day, not a model error

Verdict: offense: the lineup scored well below its projection

HOME pitching

No settled pitching line for this club. The forecast named Davis Martin (CHW) at a projected 7.2 ERA, but the per-player feed carries no pitching row for him in this game, so the actual-versus-projected comparison cannot be made on this side. That is a source-coverage gap in the settled box, not a forecasting failure.

Bullpen behind him: 0 ER over 0 outs.

The AWAY lineup he faced produced 2 runs on 9 hits and 1 HR, against a projected 5 runs / 8 hits (-3 runs vs projection).

Who did the damage to him:

Esmerlyn Valdez: 1-for-2, 1 HR, 1 RBI [+0.93 runs vs projection] season xwOBA 0.33, above average - real projection error
Jacob Gonzalez: 1-for-2, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.305, average - a good day, not a model error

Verdict: offense: the lineup scored well below its projection

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
PIT	Lake Bachar (Starter)	5 / 15	0 / 3.6 (-3.60)	0 / 2	2 / 5	0 / 1	1 / 2	2.7	3 / 6	4.81	venue away: 3 vs 0.93 ERA (worse here) · opp run factor 0.99 (average) · 47 prior appearances this season
PIT	Gregory Soto (Middle Reliever (High Leverage))	3 / 4	0 / 6.75 (-6.75)	0 / 1	1 / 1	0 / 1	0 / 0	0.34	2 / 2	1.08
PIT	Luke Weaver (Setup Man)	3 / 5	9 / 5.4 (+3.60)	1 / 1	3 / 2	1 / 0	0 / 0	0.4	1 / 0	1.14
PIT	Mason Montgomery (Closer)	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 2	0.37	2 / 1	1.49

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

Batter	Team	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
Colson Montgomery (3B)	CHW	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	3 / 1	-1.73	0.305 average	no signal
Bryan Reynolds (LF)	PIT	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	-1.11	0.361 elite	bad day
Munetaka Murakami (1B)	CHW	3 / 4	1 / 0	0 / 0	0 / 0	1 / 0	1 / 1	1 / 1	0 / 1	2 / 1	1.07	0.344 elite	projection too low
Esmerlyn Valdez (RF)	PIT	2 / 4	1 / 1	0 / 0	0 / 0	1 / 0	1 / 1	1 / 1	1 / 1	0 / 1	0.93	0.33 above average	projection too low
Braden Montgomery (RF)	CHW	2 / 4	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	1 / 1	2 / 1	1 / 1	0.64	0.328 above average	projection too low
Jacob Gonzalez (3B)	PIT	2 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0.47	0.305 average	—
Spencer Horwitz (1B)	PIT	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	1 / 1	-0.33	0.311 average	—
Sam Antonacci (2B)	CHW	5 / 4	1 / 1	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	0 / 1	1 / 1	-0.33	0.34 elite	—
Kyle Teel (C)	CHW	2 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 0	1 / 1	0.33	0.311 average	—
Andrew Benintendi (DH)	CHW	2 / 4	1 / 1	1 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	1 / 1	0.31	0.315 average	—
Tristan Peters (CF)	CHW	3 / 3	1 / 1	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	-0.31	0.319 average	—
Brandon Lowe (2B)	PIT	4 / 4	2 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	1 / 0	2 / 1	-0.13	0.324 above average	—
Konnor Griffin (SS)	PIT	4 / 4	1 / 1	0 / 1	0 / 0	0 / 0	2 / 0	0 / 1	1 / 0	1 / 1	0.02	0.304 average	—
Nick Gonzales (DH)	PIT	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0	0.327 above average	—
Oneil Cruz (CF)	PIT	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	1 / 1	0	0.352 elite	—
Rafael Flores Jr. (C)	PIT	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 0	0 / 0	0 / 1	0	0.35 elite	—
Chase Meidroth (SS)	CHW	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	0	0.292 below average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-0.097	positive favors home; negative favors away
Recent form	-0.838	positive favors home; negative favors away
Starting-pitcher matchup	0.056	positive favors home; negative favors away
Composite	-0.32	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	5.63-3.3	n/a	Base ML historical model prediction
variance_applied	6-3	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-4	0.598	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run)
embedded_savant_run_environment	5-4	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	5-4	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	5-4	n/a	Capped at max 12 realistic runs
simulation_result	5-4	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	5.63-3.3	6-3	0.671	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	6-3	5-4	-2	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run)
embedded_savant_run_environment	5-4	5-4	-0	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	5-4	5-4	-0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	5-4	5-4	-0	Capped at max 12 realistic runs
simulation_result	5-4	5-4	-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
PIT	Jacob Gonzalez	batter / 3B	better	predicted_RBI 2.623
PIT	Spencer Horwitz	batter / 1B	worse	predicted_AB -2.09
CHW	Andrew Benintendi	batter / DH	worse	predicted_AB 1.953
CHW	Randal Grichuk	batter / LF	better	predicted_AB 1.5
PIT	Esmerlyn Valdez	batter / RF	worse	predicted_SO -1.493
CHW	Munetaka Murakami	batter / 1B	worse	predicted_SO 1.468
CHW	Sam Antonacci	batter / 2B	better	predicted_SO 1.366
PIT	Oneil Cruz	batter / CF	worse	predicted_H -1.26

Structural audit findings

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	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_FINALIZED_MATCH_BOX	CHW model-prediction batting totals differ from final simulation box score (AB: current=32, sim=35; H: current=7, sim=8; 2B: current=2, sim=4; HR: current=1, sim=2; R: current=5, sim=4).	CHW
info	CURRENT_PLAYER_MATCH_BOX	CHW current finalized hitter lines are model predictions and differ from the simulation box score: Sam Antonacci: 2B current=0 box=1, R current=1 box=0, SO current=1 box=2; Munetaka Murakami: RBI current=1 box=0, R current=1 box=0, SO current=1 box=3; Randal Grichuk: AB current=3 box=4, SO current=1 box=0; Colson Montgomery: H current=1 box=2, 2B current=0 box=1, HR current=1 box=0, RBI current=1 box=0; Chase Meidroth: H current=1 box=0, R current=1 box=0, BB current=0 box=1, SO current=1 box=0; Braden Montgomery: H current=1 box=2, 2B current=0 box=1, RBI current=0 box=1, BB current=1 box=0; Andrew Benintendi: AB current=4 box=5, HR current=0 box=1, RBI current=1 box=2, R current=0 box=1; Kyle Teel: AB current=3 box=4, H current=0 box=1, HR current=0 box=1, R current=0 box=1	CHW
info	CURRENT_FINALIZED_MATCH_BOX	PIT model-prediction batting totals differ from final simulation box score (AB: current=34, sim=33; H: current=8, sim=6; 2B: current=2, sim=1; HR: current=1, sim=2; RBI: current=4, sim=5; BB: current=3, sim=4; SO: current=9, sim=8; HBP: current=0, 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: RBI current=0 box=1, R current=0 box=1; Konnor Griffin: H current=1 box=0, 2B current=1 box=0, R current=1 box=0, BB current=0 box=1; Esmerlyn Valdez: AB current=4 box=3, H current=1 box=0, RBI current=1 box=0, R current=1 box=0; Brandon Lowe: AB current=4 box=5, HR current=1 box=0, RBI current=1 box=0; Oneil Cruz: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, SO current=1 box=2; Jacob Gonzalez: AB current=3 box=4, H current=0 box=1, HR current=0 box=1, RBI current=0 box=3; Bryan Reynolds: H current=1 box=2, HR current=0 box=1, RBI current=0 box=1, BB current=1 box=0; Spencer Horwitz: AB current=4 box=2, H current=1 box=0, BB current=1 box=2, SO current=1 box=0	PIT
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

STL@SFMISS

Predicted: STL 5-4 (conf -0.04, lean)Actual: SF 6-7Brier 0.271total bias -4

Root cause: Small edge on the losing side

Evidence-backed diagnosis

published system edge 0.04 (lean); this is not the half-scale distance from 50%
team-run absolute error 4; total bias -4
pregame home-minus-away evidence: season -0.252, recent +0.630, pitcher -0.069, composite -0.313
confidence pipeline reached 0.529 before later publication/recency controls; published home probability was 0.479
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 4 (9 vs 13)
SF offense under-projected by 3
Blade Tidwell (SF) allowed 3 more ER than projected (4 vs 1)

Team	Starter	Projected	Actual	ER err
STL	Andre Pallante	2 ER / 17 outs	1 ER / 15 outs	-1
SF	Blade Tidwell	1 ER / 15 outs	4 ER / 12 outs	3

Decision internals: run factors home×0.944 / away×1.013; 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:

Nolan Gorman (STL 3B): 2-for-4, 0 HR, 1 RBI (projected 0.0 H, 0.0 HR) [run impact 0.94]
Bryce Eldridge (SF 1B): 2-for-4, 0 HR, 3 RBI (projected 1.0 H, 1.0 HR) [run impact -0.79]
Joshua Baez (STL RF): 1-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -0.62]
Leonardo Bernal (STL C): 1-for-4, 1 HR, 3 RBI (projected 1.0 H, 0.0 HR) [run impact 0.6]

Starter miss: Blade Tidwell (SF) posted 9.00 ERA vs projected 1.80 (4 ER in 12 outs); he is materially worse pitching home (7.80 ERA there vs 2.30 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

AWAY pitching

Andre Pallante (STL) threw 15 outs, 1 ER = 1.8 ERA against a projected 3.18 ERA (-1.38 ERA error)

Bullpen behind him: 2 ER over 4 outs.

The HOME lineup he faced produced 6 runs on 10 hits and 0 HR, against a projected 4 runs / 8 hits (2 runs vs projection).

Who did the damage to him:

Christian Koss: 2-for-3, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.246, weak - a good day, not a model error
Jung Hoo Lee: 2-for-4, 0 HR, 2 RBI [+0.47 runs vs projection] season xwOBA 0.315, average - a good day, not a model error
Shay Whitcomb: 1-for-4, 0 HR, 1 RBI [+0.31 runs vs projection] season xwOBA 0.249, weak - a good day, not a model error

Verdict: neither side broke materially - the result is closer to variance

HOME pitching

Blade Tidwell (SF) threw 12 outs, 4 ER = 9 ERA against a projected 1.8 ERA (7.2 ERA error)

Bullpen behind him: 0 ER over 3 outs.

The AWAY lineup he faced produced 6 runs on 7 hits and 1 HR, against a projected 4 runs / 6 hits (2 runs vs projection).

Who did the damage to him:

Nolan Gorman: 2-for-4, 0 HR, 1 RBI [+0.94 runs vs projection] season xwOBA 0.277, weak - a good day, not a model error
Leonardo Bernal: 1-for-4, 1 HR, 3 RBI [+0.6 runs vs projection] season xwOBA 0.344, elite - real projection error
Thomas Saggese: 0-for-3, 0 HR, 0 RBI [+0.33 runs vs projection] season xwOBA 0.31, average - a good day, not a model error

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

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
STL	Andre Pallante (Starter)	15 / 17	1.8 / 3.18	1 / 2	5 / 2	0 / 0	0 / 3	2.18	4 / 2	3.37	venue away: 2.83 vs 4.02 ERA (ok) · opp run factor 0.94 (weak) · 25 prior appearances this season
STL	George Soriano (Setup Man)	1 / 2	54 / 0 (+54.00)	2 / 0	3 / 0	0 / 0	0 / 0	0.35	0 / 0	0.72
STL	Justin Bruihl (Middle Reliever (High Leverage))	3 / 2	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0.36	3 / 0	0.74
SF	Blade Tidwell (Starter)	12 / 15	9 / 1.8 (+7.20)	4 / 1	5 / 5	1 / 1	2 / 1	2.5	2 / 3	4.54	venue home: 7.8 vs 2.3 ERA (worse here) · opp run factor 1.01 (average) · 14 prior appearances this season
SF	Trent Harris (Setup Man)	3 / 3	0 / 0	0 / 0	1 / 0	0 / 0	0 / 1	0.43	0 / 1	1.41

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

Batter	Team	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
Nolan Gorman (3B)	STL	4 / 3	2 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	1 / 1	0.94	0.277 weak	good day
Bryce Eldridge (1B)	SF	4 / 4	2 / 1	0 / 0	0 / 0	0 / 1	3 / 1	1 / 1	0 / 1	1 / 1	-0.79	0.358 elite	bad day
Joshua Baez (RF)	STL	4 / 4	1 / 1	1 / 0	0 / 0	0 / 1	0 / 1	1 / 1	0 / 0	1 / 1	-0.62	0.316 average	no signal
Leonardo Bernal (C)	STL	4 / 4	1 / 1	0 / 0	0 / 0	1 / 0	3 / 1	1 / 1	0 / 1	0 / 1	0.6	0.344 elite	projection too low
Turner Hill (DH)	SF	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	1 / 1	2 / 0	-0.47	0.333 above average	—
Christian Koss (SS)	SF	3 / 3	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	0.47	0.246 weak	—
Jung Hoo Lee (RF)	SF	4 / 4	2 / 1	1 / 1	0 / 0	0 / 0	2 / 0	1 / 0	0 / 0	0 / 0	0.47	0.315 average	—
Alec Burleson (1B)	STL	2 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	1 / 0	2 / 1	0 / 1	-0.45	0.376 elite	—
Thomas Saggese (2B)	STL	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	1 / 0	1 / 1	0.33	0.31 average	—
Iván Herrera (DH)	STL	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	0 / 1	1 / 1	-0.33	0.333 above average	—
Drew Cavanaugh (C)	SF	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	1 / 1	-0.33	0.319 average	—
Shay Whitcomb (3B)	SF	4 / 3	1 / 1	1 / 0	0 / 0	0 / 0	1 / 1	0 / 0	0 / 0	0 / 1	0.31	0.249 weak	—
Drew Gilbert (CF)	SF	2 / 4	1 / 1	1 / 0	0 / 0	0 / 0	0 / 1	1 / 0	0 / 0	0 / 1	0.31	0.324 above average	—
Bryan Torres (LF)	STL	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0	0.3 below average	—
Nathan Church (CF)	STL	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0	0.282 weak	—
Jonah Cox (2B)	SF	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0 / 0	1 / 1	0	0.26 weak	—
Grant McCray (LF)	SF	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	3 / 1	0	0.323 above average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-0.252	positive favors home; negative favors away
Recent form	0.63	positive favors home; negative favors away
Starting-pitcher matchup	-0.069	positive favors home; negative favors away
Composite	-0.313	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.62-3.84	n/a	Base ML historical model prediction
variance_applied	5-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-4	0.521	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run, matchup_recency_guardrail)
embedded_savant_run_environment	5-4	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	5-4	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	5-4	n/a	Capped at max 12 realistic runs
simulation_result	5-4	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.62-3.84	5-4	0.217	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-4	5-4	-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	5-4	5-4	-0	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	5-4	5-4	-0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	5-4	5-4	-0	Capped at max 12 realistic runs
simulation_result	5-4	5-4	-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	Jonah Cox	batter / 2B	worse	predicted_BB 1.861
STL	Iván Herrera	batter / DH	deboosted	predicted_HR 1.84
SF	Jung Hoo Lee	batter / RF	worse	predicted_RBI 1.74
STL	Masyn Winn	batter / SS	better	predicted_RBI 1.6
STL	Bryan Torres	batter / LF	better	predicted_AB 1.5
SF	Shay Whitcomb	batter / 3B	worse	predicted_AB 1.407
SF	Grant McCray	batter / LF	deboosted	predicted_AB 1.348
STL	Alec Burleson	batter / 1B	worse	predicted_SO 1.27

Structural audit findings

Severity	Rule	Finding	Subject
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_PITCHER_DUPLICATION	STL pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	STL
info	CURRENT_FINALIZED_MATCH_BOX	SF model-prediction batting totals differ from final simulation box score (H: current=8, sim=4; 2B: current=1, sim=2; BB: current=3, sim=5; SO: current=7, sim=5).	SF
info	CURRENT_PLAYER_MATCH_BOX	SF current finalized hitter lines are model predictions and differ from the simulation box score: Turner Hill: AB current=4 box=3, 2B current=0 box=1, RBI current=1 box=0; Bryce Eldridge: AB current=4 box=3, 2B current=0 box=1, HR current=1 box=0, SO current=1 box=2; Jonah Cox: AB current=3 box=2, H current=1 box=0, R current=1 box=0, BB current=0 box=2; Shay Whitcomb: AB current=3 box=4, H current=1 box=0, RBI current=1 box=0; Grant McCray: AB current=3 box=4, SO current=1 box=0; Christian Koss: AB current=3 box=4, H current=1 box=0, R current=1 box=0, SO current=1 box=0; Drew Gilbert: R current=0 box=1; Drew Cavanaugh: H current=1 box=0, SO current=1 box=0	SF
info	CURRENT_FINALIZED_MATCH_BOX	STL model-prediction batting totals differ from final simulation box score (AB: current=31, sim=36; H: current=7, sim=9; 2B: current=1, sim=2; HR: current=1, sim=3; RBI: current=4, sim=5; R: current=4, sim=5; BB: current=3, sim=4; SO: current=7, sim=4).	STL
info	CURRENT_PLAYER_MATCH_BOX	STL current finalized hitter lines are model predictions and differ from the simulation box score: Bryan Torres: AB current=3 box=4, R current=0 box=1, BB current=0 box=1; Joshua Baez: HR current=1 box=0, RBI current=1 box=0, R current=1 box=0, SO current=1 box=0; Leonardo Bernal: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, SO current=1 box=0; Thomas Saggese: AB current=3 box=4, H current=0 box=2, 2B current=0 box=1, RBI current=0 box=1; Iván Herrera: AB current=4 box=5, H current=1 box=2, HR current=0 box=2, RBI current=1 box=2; Alec Burleson: AB current=4 box=5, RBI current=1 box=0, BB current=1 box=0, SO current=1 box=2; Nolan Gorman: AB current=3 box=4, SO current=1 box=0; Masyn Winn: HR current=0 box=1, RBI current=0 box=2, R current=0 box=1, BB current=0 box=1	STL
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

TB@ATLHIT

Predicted: TB 4-3 (conf -0.15, moderate)Actual: TB 7-2Brier 0.179total bias -2

Root cause: Solid correct call

Evidence-backed diagnosis

published system edge 0.15 (moderate); this is not the half-scale distance from 50%
team-run absolute error 4; total bias -2
pregame home-minus-away evidence: season +0.188, recent -0.546, pitcher +0.001, composite -0.211
confidence pipeline reached 0.600 before later publication/recency controls; published home probability was 0.423
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

TB offense under-projected by 3
Reynaldo López (ATL) allowed 3 more ER than projected (6 vs 3)

Team	Starter	Projected	Actual	ER err
TB	Griffin Jax	0 ER / 15 outs	1 ER / 15 outs	1
ATL	Reynaldo López	3 ER / 15 outs	6 ER / 14 outs	3

Decision internals: run factors home×0.991 / away×0.99; 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:

Matt Olson (ATL 1B): 0-for-2, 0 HR, 1 RBI (projected 1.0 H, 1.0 HR) [run impact -1.4]
Junior Caminero (TB 3B): 1-for-5, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.26]
Jonathan Aranda (TB 1B): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -1.11]
Drake Baldwin (ATL DH): 1-for-3, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 0.93]
Chandler Simpson (TB LF): 2-for-3, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact 0.8]

Starter miss: Reynaldo López (ATL) posted 11.57 ERA vs projected 5.40 (6 ER in 14 outs); no credible venue, opponent-strength or experience signal explains it — an outlier start (variance).

What to do: Variance — no model change is warranted for this start.

Pitcher versus batter: what actually broke this game

AWAY pitching

Griffin Jax (TB) threw 15 outs, 1 ER = 1.8 ERA against a projected 0 ERA (1.8 ERA error)

Bullpen behind him: 0 ER over 0 outs.

The HOME lineup he faced produced 2 runs on 3 hits and 1 HR, against a projected 4 runs / 8 hits (-2 runs vs projection).

Who did the damage to him:

Drake Baldwin: 1-for-3, 1 HR, 1 RBI [+0.93 runs vs projection] season xwOBA 0.363, elite - real projection error

Verdict: neither side broke materially - the result is closer to variance

HOME pitching

Reynaldo López (ATL) threw 14 outs, 6 ER = 11.57 ERA against a projected 5.4 ERA (6.17 ERA error)

Bullpen behind him: 0 ER over 0 outs.

The AWAY lineup he faced produced 5 runs on 8 hits and 0 HR, against a projected 4 runs / 7 hits (1 runs vs projection).

Who did the damage to him:

Chandler Simpson: 2-for-3, 0 HR, 0 RBI [+0.8 runs vs projection] season xwOBA 0.262, weak - a good day, not a model error
Cedric Mullins: 1-for-4, 0 HR, 1 RBI [+0.47 runs vs projection] season xwOBA 0.267, weak - a good day, not a model error

Verdict: both: the starter was hit hard AND the lineup underproduced

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
TB	Griffin Jax (Starter)	15 / 15	1.8 / 0	1 / 0	2 / 2	1 / 0	1 / 0	1.36	2 / 5	4.39	venue away: 3.32 vs 3.42 ERA (ok) · opp run factor 0.99 (strong) · 30 prior appearances this season
ATL	Reynaldo López (Starter)	14 / 15	11.57 / 5.4 (+6.17)	6 / 3	7 / 3	1 / 1	3 / 5	1.47	5 / 2	4.64	venue home: 3.91 vs 3.7 ERA (ok) · opp run factor 0.99 (average) · 24 prior appearances this season

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

Batter	Team	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
Matt Olson (1B)	ATL	2 / 4	0 / 1	0 / 0	0 / 0	0 / 1	1 / 1	0 / 1	1 / 1	0 / 1	-1.4	0.353 elite	bad day
Junior Caminero (3B)	TB	5 / 4	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	1 / 1	-1.26	0.365 elite	bad day
Jonathan Aranda (1B)	TB	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	1 / 1	0 / 1	2 / 1	-1.11	0.349 elite	bad day
Drake Baldwin (DH)	ATL	3 / 4	1 / 1	0 / 0	0 / 0	1 / 0	1 / 1	2 / 1	1 / 1	0 / 1	0.93	0.363 elite	projection too low
Chandler Simpson (LF)	TB	3 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	2 / 0	1 / 0	0 / 0	0.8	0.262 weak	good day
Cedric Mullins (CF)	TB	4 / 3	1 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 0	1 / 1	0.47	0.267 weak	—
Austin Riley (3B)	ATL	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	-0.47	0.299 below average	—
Ozzie Albies (2B)	ATL	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	-0.47	0.283 weak	—
Sean Murphy (C)	ATL	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	-0.47	0.264 weak	—
Michael Harris II (CF)	ATL	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 0	1 / 0	1 / 1	-0.45	0.34 elite	—
Ronald Acuña Jr. (RF)	ATL	4 / 4	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	-0.02	0.364 elite	—
Liam Hicks (C)	TB	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 0	0 / 0	1 / 0	0	0.318 average	—
Yandy Díaz (DH)	TB	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	0 / 0	0 / 0	0	0.356 elite	—
Richie Palacios (2B)	TB	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	1 / 1	0 / 1	0	0.305 average	—
Jorge Mateo (SS)	TB	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	2 / 0	0 / 1	0 / 0	1 / 1	0	0.297 below average	—
Mauricio Dubón (SS)	ATL	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0 / 0	0	0.304 average	—
Mike Yastrzemski (LF)	ATL	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0	0.308 average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	0.188	positive favors home; negative favors away
Recent form	-0.546	positive favors home; negative favors away
Starting-pitcher matchup	0.001	positive favors home; negative favors away
Composite	-0.211	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.87-3.52	n/a	Base ML historical model prediction
variance_applied	5-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-3	0.577	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run)
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

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.87-3.52	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)
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

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
TB	Yandy Díaz	batter / DH	better	predicted_AB -1.64
TB	Jorge Mateo	batter / SS	worse	predicted_AB 1.338
TB	Liam Hicks	batter / C	worse	predicted_RBI 1.31
TB	Jonathan Aranda	batter / 1B	worse	predicted_H -1.21
ATL	Michael Harris II	batter / CF	worse	predicted_H -1.16
ATL	Drake Baldwin	batter / DH	worse	predicted_H -1.083
TB	Richie Palacios	batter / 2B	worse	predicted_AB 1.04
ATL	Mauricio Dubón	batter / SS	worse	predicted_H -0.993

Structural audit findings

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	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	ATL model-prediction batting totals differ from final simulation box score (AB: current=33, sim=31; H: current=8, sim=4; 2B: current=1, sim=2; HR: current=1, sim=0; RBI: current=4, sim=3; R: current=4, sim=3; BB: current=3, sim=1).	ATL
info	CURRENT_PLAYER_MATCH_BOX	ATL current finalized hitter lines are model predictions and differ from the simulation box score: Drake Baldwin: AB current=4 box=3, H current=1 box=0, RBI current=1 box=0, R current=1 box=0; Mauricio Dubón: H current=1 box=0, RBI current=1 box=0; Mike Yastrzemski: H current=0 box=1, 2B current=0 box=1; Matt Olson: H current=1 box=0, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0; Ronald Acuña Jr.: R current=1 box=0, BB current=1 box=0; Michael Harris II: H current=1 box=0, 2B current=1 box=0, R current=0 box=1; Austin Riley: H current=1 box=0; Ozzie Albies: AB current=4 box=3, 2B current=0 box=1, RBI current=0 box=1	ATL
info	CURRENT_FINALIZED_MATCH_BOX	TB model-prediction batting totals differ from final simulation box score (AB: current=33, sim=34; H: current=8, sim=7; HR: current=1, sim=2; BB: current=3, sim=5; SO: current=6, sim=3).	TB
info	CURRENT_PLAYER_MATCH_BOX	TB current finalized hitter lines are model predictions and differ from the simulation box score: Liam Hicks: HR current=0 box=1, RBI current=1 box=2, R current=0 box=1, SO current=0 box=1; Jonathan Aranda: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, R current=1 box=0; Junior Caminero: HR current=1 box=0, RBI current=1 box=0; Yandy Díaz: H current=1 box=2, HR current=0 box=1, BB current=0 box=1; Victor Mesa Jr.: AB current=4 box=3, BB current=0 box=1, SO current=1 box=0; Chandler Simpson: 2B current=0 box=1, RBI current=0 box=1, R current=0 box=1; Cedric Mullins: BB current=0 box=1, SO current=1 box=0; Richie Palacios: AB current=3 box=4, BB current=1 box=0, SO current=1 box=0	TB
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

TEX@SEAMISS

Predicted: TEX 6-4 (conf -0.10, lean)Actual: SEA 2-3Brier 0.301total bias +5

Root cause: Run environment far lower than projected (over-dispersed)

Evidence-backed diagnosis

published system edge 0.10 (lean); this is not the half-scale distance from 50%
team-run absolute error 5; total bias +5
pregame home-minus-away evidence: season -0.429, recent -0.866, pitcher -0.010, composite -0.835
confidence pipeline reached 0.640 before later publication/recency controls; published home probability was 0.451
score-volatility actions: supported_blowout_margin_expansion
forecast-file integrity: clean (0 defects, 8 informational note(s)); the saved artifact is not why the pick missed

Quantified contributing factors

total runs over-projected by 5 (10 vs 5)
TEX offense over-projected by 4

Team	Starter	Projected	Actual	ER err
TEX	Cody Bradford	3 ER / 15 outs	3 ER / 12 outs	0
SEA	Kade Anderson	3 ER / 15 outs	1 ER / 18 outs	-2

Decision internals: run factors home×0.993 / away×1.003; 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:

Dominic Canzone (SEA DH): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.4]
Ezequiel Duran (TEX CF): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -1.11]
Corey Seager (TEX SS): 2-for-4, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 1.07]
Wyatt Langford (TEX LF): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.8]

Starter miss: Kade Anderson (SEA) posted 1.50 ERA vs projected 5.40 (1 ER in 18 outs). Thin prior record (3 prior appearances this season): high projection uncertainty.

What to do: Reduce game confidence for starters the feed has little on (scale by prior appearances); prefer priors and validate any run-param change in shadow.

Pitcher versus batter: what actually broke this game

AWAY pitching

Cody Bradford (TEX) threw 12 outs, 3 ER = 6.75 ERA against a projected 5.4 ERA (1.35 ERA error) only 6 prior appearances in the covered season - the projection had little of his own record to stand on

Bullpen behind him: 0 ER over 6 outs.

The HOME lineup he faced produced 3 runs on 5 hits and 0 HR, against a projected 5 runs / 8 hits (-2 runs vs projection).

Who did the damage to him:

Cole Young: 1-for-3, 0 HR, 2 RBI [+0.31 runs vs projection] season xwOBA 0.315, average - a good day, not a model error

Verdict: offense: the lineup scored well below its projection

HOME pitching

Kade Anderson (SEA) threw 18 outs, 1 ER = 1.5 ERA against a projected 5.4 ERA (-3.9 ERA error) only 3 prior appearances in the covered season - the projection had little of his own record to stand on

Bullpen behind him: 1 ER over 6 outs.

The AWAY lineup he faced produced 2 runs on 7 hits and 2 HR, against a projected 4 runs / 8 hits (-2 runs vs projection).

Who did the damage to him:

Corey Seager: 2-for-4, 1 HR, 1 RBI [+1.07 runs vs projection] season xwOBA 0.354, elite - real projection error
Brandon Nimmo: 2-for-4, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.35, elite - a good day, not a model error
Cody Freeman: 1-for-2, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.23, weak - a good day, not a model error

Verdict: offense: the lineup scored well below its projection

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
TEX	Cody Bradford (Starter)	12 / 15	6.75 / 5.4	3 / 3	4 / 4	0 / 0	2 / 2	0.97	1 / 4	2.45	venue away: 4.32 vs 4.61 ERA (ok) · opp run factor 0.99 (weak) · 6 prior appearances this season
TEX	Tyler Alexander (Middle Reliever (High Leverage))	6 / 3	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0.25	1 / 0	0.87
SEA	Kade Anderson (Starter)	18 / 15	1.5 / 5.4 (-3.90)	1 / 3	6 / 5	1 / 1	1 / 3	2.64	5 / 7	3.7	venue home: 5.91 vs 4.15 ERA (ok) · opp run factor 1 (average) · thin sample: 3 prior appearances this season (high uncertainty)
SEA	Andrés Muñoz (Closer)	3 / 4	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0.42	2 / 2	1.6
SEA	Gabe Speier (Middle Reliever (High Leverage))	3 / 3	9 / 27 (-18.00)	1 / 3	1 / 4	1 / 0	0 / 1	0.36	1 / 1	0.91

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

Batter	Team	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
Dominic Canzone (DH)	SEA	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 0	0 / 1	-1.4	0.353 elite	bad day
Ezequiel Duran (CF)	TEX	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	-1.11	0.306 average	no signal
Corey Seager (SS)	TEX	4 / 4	2 / 1	0 / 0	0 / 0	1 / 0	1 / 1	1 / 1	0 / 1	1 / 1	1.07	0.354 elite	projection too low
Wyatt Langford (LF)	TEX	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	3 / 1	-0.8	0.316 average	no signal
Logan O'Hoppe (DH)	TEX	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	-0.47	0.268 weak	—
Brandon Nimmo (RF)	TEX	4 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0.47	0.35 elite	—
Elias Díaz (C)	TEX	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	-0.47	0.275 weak	—
Cody Freeman (3B)	TEX	2 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0.47	0.23 weak	—
Cal Raleigh (C)	SEA	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	1 / 1	1 / 1	-0.47	0.315 average	—
J.P. Crawford (SS)	SEA	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	-0.47	0.324 above average	—
Jake Burger (1B)	TEX	3 / 3	1 / 1	0 / 0	0 / 0	1 / 1	1 / 1	1 / 1	1 / 0	1 / 1	0.33	0.316 average	—
Justin Foscue (2B)	TEX	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 0	0 / 1	0.33	0.333 above average	—
Cole Young (2B)	SEA	3 / 3	1 / 1	1 / 0	0 / 0	0 / 0	2 / 0	0 / 1	0 / 0	1 / 1	0.31	0.315 average	—
Randy Arozarena (LF)	SEA	4 / 4	1 / 1	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	1 / 1	0	0.348 elite	—
Taylor Ward (RF)	SEA	2 / 3	1 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	1 / 1	0 / 1	0	0.333 above average	—
Julio Rodríguez (CF)	SEA	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	1 / 1	0	0.334 above average	—
Josh Naylor (1B)	SEA	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0	0.315 average	—
Weston Wilson (3B)	SEA	2 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0	0.256 weak	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-0.429	positive favors home; negative favors away
Recent form	-0.866	positive favors home; negative favors away
Starting-pitcher matchup	-0.01	positive favors home; negative favors away
Composite	-0.835	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	3.9-4.49	n/a	Base ML historical model prediction
variance_applied	4-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	6-4	0.549	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by winner_starter_is_below_average, matchup_recency_guardrail)
embedded_savant_run_environment	6-4	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

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	3.9-4.49	4-4	0.59	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-4	6-4	2	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by winner_starter_is_below_average, matchup_recency_guardrail)
embedded_savant_run_environment	6-4	6-4	-0	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

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
TEX	Ezequiel Duran	batter / CF	worse	predicted_SO 3.067
TEX	Justin Foscue	batter / 2B	worse	predicted_AB 2.5
TEX	Brandon Nimmo	batter / RF	better	predicted_RBI 2.49
TEX	Logan O'Hoppe	batter / DH	worse	predicted_SO 2.165
SEA	Weston Wilson	batter / 3B	better	predicted_H 1.87
SEA	Cal Raleigh	batter / C	worse	predicted_SO 1.697
TEX	Corey Seager	batter / SS	worse	predicted_BB 1.488
SEA	Dominic Canzone	batter / DH	worse	predicted_AB -1.46

Structural audit findings

Severity	Rule	Finding	Subject
info	CURRENT_PITCHER_DUPLICATION	SEA pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	SEA
info	CURRENT_PITCHER_DUPLICATION	TEX pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	TEX
info	CURRENT_FINALIZED_MATCH_BOX	SEA model-prediction batting totals differ from final simulation box score (H: current=8, sim=6; 2B: current=1, sim=0; HR: current=1, sim=2; R: current=5, sim=4; SO: current=8, sim=11; HBP: current=0, sim=1).	SEA
info	CURRENT_PLAYER_MATCH_BOX	SEA current finalized hitter lines are model predictions and differ from the simulation box score: Cal Raleigh: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, BB current=1 box=0; Dominic Canzone: AB current=4 box=3, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0; Randy Arozarena: AB current=4 box=3, 2B current=1 box=0, RBI current=1 box=2, HBP current=0 box=1; Taylor Ward: AB current=3 box=4, R current=0 box=1, BB current=1 box=0, SO current=1 box=0; J.P. Crawford: H current=1 box=0; Julio Rodríguez: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, SO current=1 box=2; Josh Naylor: H current=1 box=2, HR current=0 box=1, RBI current=0 box=1, R current=0 box=1; Cole Young: AB current=3 box=4, H current=1 box=0, R current=1 box=0	SEA
info	CURRENT_FINALIZED_MATCH_BOX	TEX model pitcher projections differ from final simulation pitching box score (HR: current=1, sim=2).	TEX
info	CURRENT_FINALIZED_MATCH_BOX	TEX model-prediction batting totals differ from final simulation box score (AB: current=31, sim=36; H: current=8, sim=9; 3B: current=0, sim=1; RBI: current=4, sim=6; R: current=4, sim=6; BB: current=3, sim=5; SO: current=8, sim=13).	TEX
info	CURRENT_PLAYER_MATCH_BOX	TEX current finalized hitter lines are model predictions and differ from the simulation box score: Ezequiel Duran: AB current=4 box=5, 2B current=1 box=0, RBI current=1 box=0, BB current=1 box=0; Corey Seager: AB current=4 box=3, RBI current=1 box=0, BB current=1 box=2; Jake Burger: AB current=3 box=4, HR current=1 box=0, SO current=1 box=2; Logan O'Hoppe: AB current=3 box=4, 2B current=0 box=1, SO current=1 box=3; Justin Foscue: AB current=3 box=5, R current=0 box=1; Wyatt Langford: AB current=4 box=5, 3B current=0 box=1, R current=1 box=0, BB current=1 box=0; Brandon Nimmo: AB current=4 box=3, H current=1 box=2, HR current=0 box=1, RBI current=0 box=3; Elias Díaz: AB current=3 box=4, H current=1 box=0, SO current=1 box=0	TEX
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

TOR@OAKMISS

Predicted: TOR 6-5 (conf -0.07, lean)Actual: OAK 0-2Brier 0.287total bias +9

Root cause: Run environment far lower than projected (over-dispersed)

Evidence-backed diagnosis

published system edge 0.07 (lean); this is not the half-scale distance from 50%
team-run absolute error 9; total bias +9
pregame home-minus-away evidence: season -0.945, recent -0.690, pitcher +0.012, composite -0.847
confidence pipeline reached 0.550 before later publication/recency controls; published home probability was 0.464
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 over-projected by 9 (11 vs 2)
TOR offense over-projected by 6
OAK offense over-projected by 3
Brady Basso (OAK) allowed 3 fewer ER than projected (0 vs 3)

Team	Starter	Projected	Actual	ER err
TOR	Braydon Fisher	2 ER / 15 outs	0 ER / 3 outs	-2
OAK	Brady Basso	3 ER / 13 outs	0 ER / 12 outs	-3

Decision internals: run factors home×0.99 / away×0.979; 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:

Kazuma Okamoto (TOR 3B): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.4]
Zack Gelof (OAK 3B): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.4]
Carlos Cortes (OAK LF): 0-for-3, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -1.11]
Brandon Valenzuela (TOR C): 0-for-3, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.8]
Max Muncy (OAK DH): 2-for-2, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact 0.8]

Starter miss: Brady Basso (OAK) posted 0.00 ERA vs projected 6.23 (0 ER in 12 outs); no credible venue, opponent-strength or experience signal explains it — an outlier start (variance).

What to do: Variance — no model change is warranted for this start.

Pitcher versus batter: what actually broke this game

AWAY pitching

Braydon Fisher (TOR) threw 3 outs, 0 ER = 0 ERA against a projected 3.6 ERA (-3.6 ERA error)

Bullpen behind him: 0 ER over 6 outs.

The HOME lineup he faced produced 2 runs on 4 hits and 1 HR, against a projected 4 runs / 8 hits (-2 runs vs projection).

Who did the damage to him:

Max Muncy: 2-for-2, 0 HR, 0 RBI [+0.8 runs vs projection] season xwOBA 0.288, below average - a good day, not a model error

Verdict: offense: the lineup scored well below its projection

HOME pitching

Brady Basso (OAK) threw 12 outs, 0 ER = 0 ERA against a projected 6.23 ERA (-6.23 ERA error)

Bullpen behind him: 0 ER over 15 outs.

The AWAY lineup he faced produced 0 runs on 5 hits and 0 HR, against a projected 4 runs / 10 hits (-4 runs vs projection).

No individual hitter beat his projection by a meaningful margin.

Verdict: offense: the lineup scored well below its projection

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
TOR	Braydon Fisher (Starter)	3 / 15	0 / 3.6 (-3.60)	0 / 2	0 / 6	0 / 0	0 / 0	3.23	1 / 5	2.94	venue away: 0 vs 1.69 ERA (ok) · opp run factor 0.99 (weak) · 69 prior appearances this season
TOR	Louis Varland (Closer)	3 / 5	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0.33	2 / 0	1.01
TOR	Tyler Rogers (Setup Man)	3 / 3	0 / 9 (-9.00)	0 / 1	0 / 2	0 / 0	1 / 1	0.28	0 / 1	0.54
OAK	Brady Basso (Starter)	12 / 13	0 / 6.23 (-6.23)	0 / 3	3 / 5	0 / 1	2 / 1	0.49	4 / 2	3.43	opp run factor 0.98 (weak) · 22 prior appearances this season
OAK	Luis Medina (Middle Reliever (High Leverage))	3 / 4	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0.63	2 / 1	1.92
OAK	Scott Blewett (Middle Reliever (High Leverage))	6 / 4	0 / 0	0 / 0	2 / 0	0 / 0	1 / 0	0.49	1 / 1	2.03
OAK	Elvis Alvarado (Closer)	3 / 3	0 / 27 (-27.00)	0 / 3	1 / 2	0 / 1	0 / 1	0.5	2 / 0	1.15
OAK	Hogan Harris (Setup Man)	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0.55	0 / 1	1.39

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

Batter	Team	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
Kazuma Okamoto (3B)	TOR	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 0	3 / 1	-1.4	0.324 above average	bad day
Zack Gelof (3B)	OAK	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 0	2 / 1	-1.4	0.311 average	no signal
Carlos Cortes (LF)	OAK	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	0 / 1	-1.11	0.307 average	no signal
Brandon Valenzuela (C)	TOR	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	2 / 1	-0.8	0.312 average	no signal
Max Muncy (DH)	OAK	2 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 0	0 / 1	0.8	0.288 below average	good day
Lawrence Butler (RF)	OAK	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	0 / 1	-0.8	0.318 average	no signal
Vladimir Guerrero Jr. (1B)	TOR	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	0 / 0	-0.78	0.333 above average	bad day
Myles Straw (CF)	TOR	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	-0.47	0.298 below average	—
Brian Serven (C)	OAK	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	1 / 1	-0.47	0.302 below average	—
Alika Williams (SS)	OAK	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	-0.47	0.239 weak	—
Henry Bolte (CF)	OAK	4 / 4	1 / 1	0 / 1	0 / 0	1 / 0	1 / 0	1 / 1	0 / 1	2 / 1	0.29	0.315 average	—
Brett Bateman (RF)	TOR	4 / 4	1 / 2	1 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	-0.16	0.301 below average	—
Ernie Clement (2B)	TOR	3 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	0 / 0	0.02	0.263 weak	—
Charles McAdoo (LF)	TOR	2 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	1 / 1	0	0.275 weak	—
Alejandro Kirk (DH)	TOR	3 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	1 / 1	0	0.302 below average	—
Andrés Giménez (SS)	TOR	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0 / 1	0	0.274 weak	—
Tommy White (1B)	OAK	3 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 1	0	0.298 below average	—
Donovan Walton (2B)	OAK	2 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	0 / 0	0	0.34 elite	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-0.945	positive favors home; negative favors away
Recent form	-0.69	positive favors home; negative favors away
Starting-pitcher matchup	0.012	positive favors home; negative favors away
Composite	-0.847	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	5.44-4.78	n/a	Base ML historical model prediction
variance_applied	5-5	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	6-5	0.536	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

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	5.44-4.78	5-5	-0.658	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 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

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
TOR	Andrés Giménez	batter / SS	better	predicted_RBI 2.62
TOR	Kazuma Okamoto	batter / 3B	worse	predicted_SO 1.791
OAK	Henry Bolte	batter / CF	better	predicted_RBI 1.677
OAK	Carlos Cortes	batter / LF	boosted	predicted_RBI 1.626
TOR	Charles McAdoo	batter / LF	better	predicted_RBI 1.615
TOR	Vladimir Guerrero Jr.	batter / 1B	worse	predicted_AB -1.56
TOR	Myles Straw	batter / CF	deboosted	predicted_AB 1.5
OAK	Donovan Walton	batter / 2B	worse	predicted_AB 1.254

Structural audit findings

Severity	Rule	Finding	Subject
info	CURRENT_PITCHER_DUPLICATION	OAK pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	OAK
info	CURRENT_PITCHER_DUPLICATION	TOR pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	TOR
info	CURRENT_FINALIZED_MATCH_BOX	OAK model-prediction batting totals differ from final simulation box score (AB: current=33, sim=37; H: current=8, sim=10; 2B: current=2, sim=1; 3B: current=0, sim=2; HR: current=1, sim=0; R: current=4, sim=5; BB: current=4, sim=2; SO: current=8, sim=6).	OAK
info	CURRENT_PLAYER_MATCH_BOX	OAK current finalized hitter lines are model predictions and differ from the simulation box score: Henry Bolte: AB current=4 box=5, H current=1 box=2, RBI current=0 box=2, BB current=1 box=0; Carlos Cortes: H current=1 box=2, 2B current=1 box=0, 3B current=0 box=1, RBI current=1 box=2; Tommy White: RBI current=0 box=1, R current=0 box=1, SO current=1 box=0; Max Muncy: RBI current=1 box=0, SO current=1 box=0; Brian Serven: RBI current=1 box=0, BB current=0 box=1, SO current=1 box=0; Alika Williams: AB current=3 box=4, SO current=1 box=0; Donovan Walton: AB current=3 box=4, R current=1 box=0, BB current=1 box=0, SO current=0 box=1; Zack Gelof: AB current=4 box=5, H current=1 box=2, 3B current=0 box=1, HR current=1 box=0	OAK
info	CURRENT_FINALIZED_MATCH_BOX	TOR model-prediction batting totals differ from final simulation box score (AB: current=33, sim=34; H: current=10, sim=7; HR: current=1, sim=2; RBI: current=5, sim=6; R: current=4, sim=6; BB: current=3, sim=2; SO: current=6, sim=5).	TOR
info	CURRENT_PLAYER_MATCH_BOX	TOR current finalized hitter lines are model predictions and differ from the simulation box score: Charles McAdoo: HR current=0 box=1, RBI current=1 box=2, R current=0 box=1, SO current=1 box=0; Brett Bateman: RBI current=0 box=1, SO current=1 box=0; Alejandro Kirk: RBI current=1 box=0, R current=0 box=1, BB current=1 box=0, SO current=1 box=0; Kazuma Okamoto: AB current=4 box=3, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0; Brandon Valenzuela: AB current=3 box=4, 2B current=0 box=1, R current=0 box=1, BB current=1 box=0; Ernie Clement: H current=1 box=0, 2B current=1 box=0, R current=1 box=0; Andrés Giménez: AB current=3 box=4, HR current=0 box=1, RBI current=1 box=3, R current=0 box=1; Vladimir Guerrero Jr.: AB current=4 box=3, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0	TOR
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

WAS@SDHIT

Predicted: SD 4-5 (conf +0.05, lean)Actual: SD 2-9Brier 0.226total bias -2

Root cause: Right winner, wrong score

Evidence-backed diagnosis

published system edge 0.05 (lean); this is not the half-scale distance from 50%
team-run absolute error 6; total bias -2
pregame home-minus-away evidence: season +0.194, recent +0.374, pitcher -0.045, composite +0.281
confidence pipeline reached 0.555 before later publication/recency controls; published home probability was 0.524
score-volatility actions: no_change
forecast-file integrity: 1 structural defect(s) on the saved artifact; see the per-game file-integrity table (this is a file check, not why the pick missed)

Quantified contributing factors

SD offense under-projected by 4

Team	Starter	Projected	Actual	ER err
WAS	Jackson Kent	5 ER / 14 outs	5 ER / 9 outs	0
SD	Walker Buehler	1 ER / 16 outs	2 ER / 7 outs	1

Decision internals: run factors home×0.996 / away×1.041; 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:

Jackson Merrill (SD CF): 4-for-5, 2 HR, 6 RBI (projected 1.0 H, 0.0 HR) [run impact 3.89]
Manny Machado (SD 3B): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.4]
Ty France (SD 1B): 3-for-5, 0 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 1.25]
James Wood (WAS DH): 1-for-3, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -0.93]
CJ Abrams (WAS 2B): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.8]

Starter miss: Walker Buehler (SD) posted 7.71 ERA vs projected 1.69 (2 ER in 7 outs); no credible venue, opponent-strength or experience signal explains it — an outlier start (variance).

What to do: Variance — no model change is warranted for this start.

Pitcher versus batter: what actually broke this game

AWAY pitching

Jackson Kent (WAS) threw 9 outs, 5 ER = 15 ERA against a projected 9.64 ERA (5.36 ERA error) only 5 prior appearances in the covered season - the projection had little of his own record to stand on

Bullpen behind him: 2 ER over 6 outs.

The HOME lineup he faced produced 9 runs on 16 hits and 2 HR, against a projected 4 runs / 7 hits (5 runs vs projection).

Who did the damage to him:

Jackson Merrill: 4-for-5, 2 HR, 6 RBI [+3.89 runs vs projection] season xwOBA 0.341, elite - real projection error
Ty France: 3-for-5, 0 HR, 1 RBI [+1.25 runs vs projection] season xwOBA 0.334, above average - real projection error
Samad Taylor: 2-for-4, 0 HR, 0 RBI [+0.78 runs vs projection] season xwOBA 0.308, average - a good day, not a model error

Verdict: both: the starter was hit hard AND the lineup underproduced

HOME pitching

Walker Buehler (SD) threw 7 outs, 2 ER = 7.71 ERA against a projected 1.69 ERA (6.02 ERA error)

Bullpen behind him: 0 ER over 0 outs.

The AWAY lineup he faced produced 2 runs on 4 hits and 1 HR, against a projected 4 runs / 8 hits (-2 runs vs projection).

Who did the damage to him:

Keibert Ruiz: 1-for-4, 1 HR, 1 RBI [+0.62 runs vs projection] season xwOBA 0.285, below average - a good day, not a model error
Dylan Crews: 1-for-3, 0 HR, 1 RBI [+0.33 runs vs projection] season xwOBA 0.312, average - a good day, not a model error

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

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
WAS	Jackson Kent (Starter)	9 / 14	15 / 9.64 (+5.36)	5 / 5	8 / 8	0 / 0	1 / 2	3.91	2 / 4	4.34	venue away: 7.42 vs 3.27 ERA (ok) · opp run factor 1 (average) · 5 prior appearances this season
WAS	Brad Lord (Middle Reliever (High Leverage))	6 / 4	9 / 0 (+9.00)	2 / 0	4 / 0	1 / 0	0 / 2	0.77	0 / 3	1.68
SD	Walker Buehler (Starter)	7 / 16	7.71 / 1.69 (+6.02)	2 / 1	3 / 2	1 / 0	2 / 2	1.57	4 / 9	4.52	venue home: 4 vs 5.23 ERA (ok) · opp run factor 1.04 (strong) · 28 prior appearances this season

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

Batter	Team	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
Jackson Merrill (CF)	SD	5 / 4	4 / 1	0 / 0	1 / 0	2 / 0	6 / 1	2 / 1	0 / 0	0 / 1	3.89	0.341 elite	projection too low
Manny Machado (3B)	SD	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	1 / 1	2 / 1	-1.4	0.338 above average	bad day
Ty France (1B)	SD	5 / 4	3 / 1	1 / 0	0 / 0	0 / 0	1 / 1	2 / 0	0 / 0	1 / 1	1.25	0.334 above average	projection too low
James Wood (DH)	WAS	3 / 4	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	1 / 1	1 / 1	-0.93	0.416 elite	bad day
CJ Abrams (2B)	WAS	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	3 / 1	-0.8	0.328 above average	bad day
Samad Taylor (LF)	SD	4 / 3	2 / 1	1 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0 / 0	1 / 1	0.78	0.308 average	no signal
Keibert Ruiz (C)	WAS	4 / 3	1 / 1	0 / 1	0 / 0	1 / 0	1 / 0	1 / 0	0 / 0	0 / 0	0.62	0.285 below average	good day
Andrés Chaparro (3B)	WAS	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	2 / 1	-0.47	0.349 elite	—
Daylen Lile (LF)	WAS	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	0 / 1	-0.47	0.327 above average	—
Yohandy Morales (1B)	WAS	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	-0.47	0.372 elite	—
Austin Hays (DH)	SD	2 / 3	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 1	0.47	0.273 weak	—
Xander Bogaerts (SS)	SD	4 / 4	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0 / 0	0.47	0.321 above average	—
Dylan Crews (RF)	WAS	3 / 4	1 / 1	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	1 / 0	1 / 1	0.33	0.312 average	—
Fernando Tatis Jr. (2B)	SD	5 / 4	3 / 1	0 / 1	0 / 0	0 / 0	1 / 1	1 / 1	0 / 1	0 / 1	0.3	0.375 elite	—
Nasim Nuñez (SS)	WAS	1 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0 / 1	0	0.266 weak	—
Jacob Young (CF)	WAS	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0	0.317 average	—
Jase Bowen (RF)	SD	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	0	0.27 weak	—
Luis Campusano (C)	SD	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	0	0.343 elite	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	0.194	positive favors home; negative favors away
Recent form	0.374	positive favors home; negative favors away
Starting-pitcher matchup	-0.045	positive favors home; negative favors away
Composite	0.281	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	3.92-4.56	n/a	Base ML historical model prediction
variance_applied	4-5	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-5	0.524	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

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	3.92-4.56	4-5	0.366	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

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
SD	Jase Bowen	batter / RF	better	predicted_RBI 3.792
WAS	Jackson Kent	pitcher / Starter	neutral	predicted_SO 2
WAS	Jacob Young	batter / CF	boosted	predicted_RBI 1.86
SD	Samad Taylor	batter / LF	worse	predicted_H 1.78
WAS	Dylan Crews	batter / RF	worse	predicted_AB -1.65
WAS	Keibert Ruiz	batter / C	better	predicted_SO 1.604
WAS	Andrés Chaparro	batter / 3B	worse	predicted_AB 1.5
WAS	Yohandy Morales	batter / 1B	worse	predicted_AB 1.333

Structural audit findings

Severity	Rule	Finding	Subject
critical	RUNS_MATCH_FINAL	simulation:SD runs mismatch: batting R=5, final=5, opponent pitching R=4.	SD
info	CURRENT_PITCHER_DUPLICATION	SD pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	SD
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	SD model-prediction batting totals differ from final simulation box score (H: current=7, sim=8; 2B: current=1, sim=2; 3B: current=0, sim=1; HR: current=1, sim=0; RBI: current=4, sim=5; R: current=4, sim=5; SO: current=8, sim=9).	SD
info	CURRENT_PLAYER_MATCH_BOX	SD current finalized hitter lines are model predictions and differ from the simulation box score: Fernando Tatis Jr.: 2B current=1 box=0, RBI current=1 box=0, BB current=1 box=0, SO current=1 box=2; Jackson Merrill: 2B current=0 box=1, RBI current=1 box=0, R current=1 box=0, SO current=1 box=0; Ty France: RBI current=1 box=0, R current=0 box=1; Austin Hays: AB current=3 box=4, H current=1 box=0; Manny Machado: AB current=4 box=3, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0; Jase Bowen: H current=0 box=2, 2B current=0 box=1, 3B current=0 box=1, RBI current=0 box=4; Xander Bogaerts: BB current=1 box=0, SO current=0 box=1; Luis Campusano: RBI current=0 box=1, R current=0 box=1	SD
info	CURRENT_FINALIZED_MATCH_BOX	WAS model pitcher projections differ from final simulation pitching box score (R: current=5, sim=4; ER: current=5, sim=4).	WAS
info	CURRENT_FINALIZED_MATCH_BOX	WAS model-prediction batting totals differ from final simulation box score (AB: current=33, sim=35; 2B: current=1, sim=2; BB: current=3, sim=2; SO: current=8, sim=13).	WAS
info	CURRENT_PLAYER_MATCH_BOX	WAS current finalized hitter lines are model predictions and differ from the simulation box score: CJ Abrams: AB current=4 box=3, 2B current=0 box=1, RBI current=1 box=0, SO current=1 box=2; James Wood: AB current=4 box=5, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0; Andrés Chaparro: H current=1 box=0, RBI current=1 box=0; Daylen Lile: H current=1 box=0, RBI current=1 box=0, R current=1 box=0; Nasim Nuñez: AB current=3 box=4, H current=0 box=2, RBI current=0 box=1, R current=0 box=1; Dylan Crews: AB current=4 box=3, H current=1 box=0, R current=1 box=0, BB current=0 box=1; Keibert Ruiz: AB current=3 box=4, H current=1 box=2, 2B current=1 box=0, HR current=0 box=1; Jacob Young: AB current=3 box=4, 2B current=0 box=1, RBI current=0 box=2, R current=0 box=1	WAS
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.

Day over day: did it improve or worsen

Compared against packet 20260908. Last recorded change: v4.30.0 (2026-09-08) Hitter shrinkage is derived from the between-hitter spread rather than fitted to slate error, so an established extreme profile is no longer flattened toward the league.

Metric	Prior	Current	Δ	Result
brier_score	0.2542	0.2444	-0.009719	improved
accuracy	0.55	0.5833	+0.0333	improved
team_run_mae	2.5792	2.55	-0.0292	improved
total_run_mae	3.8917	4	+0.1083	regressed
calibration_error	0.0641	0.069	+0.0049	regressed
starter_er_mae	2.0207	1.9879	-0.0328	improved
starter_outs_mae	3.0083	3.1498	+0.1415	regressed

Deep slate report — 20260909

Process, end to end

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)

ARI@KCHIT

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

CHC@MILHIT

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

CIN@LADHIT

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

CLE@BALMISS

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

COL@NYYHIT

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

HOU@PHIHIT

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

LAA@BOSMISS

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

MIN@DETHIT

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@MIAHIT

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

PIT@CHWHIT

Player-level breakdown (projected vs actual)