Deep slate report 20261003

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_20261003.json	102919
forensic deep audit	loaded	probability, provenance, starter and decision contracts	/app/output/reports/PREDICTION_SYSTEM_DEEP_AUDIT_20261003.json	75649
continuous learning	loaded	pipeline-stage quality, calibration, probability lineage, score distribution, and official starter comparisons	/app/output/learning/continuous_learning_daily_20261003.json	692548
detailed report	loaded	full game and simulation rendering	/app/output/reports/detailed_20261003.html	98172
ultra detailed player audit	loaded	player-level prediction trace	/app/output/reports/20261003_ultra_detailed_player_audit.html	8625378
structural audit	loaded	all invariant findings	/app/output/reports/audit_report_20261003.html	519900
integrity revision audit	missing	post-Luna remediation proof for immutable revision artifacts	/app/output/reports/audit_report_20261003_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 (4)

Severity	Domain	Finding and evidence	Affected	Action & guardrail
HIGH	calibration	Published confidence buckets were poorly aligned on this slate weighted absolute calibration error 0.4705 across 4 games	ATL@LAD, CHW@CLE, NYY@TB, SD@MIL	Do thisWatch the same confidence buckets over the full rolling window before changing any calibration weight. Don't concludeA single slate has far too few games to establish a stable calibration curve.
HIGH	model_validation	Validated run models explain very little holdout variance home/away/total R² = 0.0021/0.0309/0.0024	ATL@LAD, CHW@CLE, NYY@TB, SD@MIL	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 4/4 games structurally audited	ATL@LAD, CHW@CLE, NYY@TB, SD@MIL	Do thisKeep the per-game audit receipt in every report. Don't concludeFull audit coverage does not cancel the findings raised inside the audited games.
INFO	probability_lineage	Material calibrated-to-published movements have recorded lineage 1 material movement(s), all explained by named score-coherence and/or recency transformations	SD@MIL	Do thisKeep showing every probability layer and the named transformation between them. Don't concludeA published probability above the calibrated one is not a governor bypass when a recorded transformation explains it.
INFO	simulation	Simulation produced no winner-direction intervention on this slate flip candidates 0; published flips 0	ATL@LAD, CHW@CLE, NYY@TB, SD@MIL	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.

Finding	Consecutive slates	Incidence	Present now	Fix type
pipeline_winner_instability	0	4/11	no	code/diagnostic
score_margin_compression	0	5/11	no	param
simulation_inert_on_slate	11	11/11	yes	code/diagnostic
single_slate_calibration_instability	1	6/11	yes	param
weak_run_model_signal	1	9/11	yes	code/diagnostic

Metric	Direction	Series (old → new)
winner_accuracy	improving	[0.5, 0.6667, 0.6471, 0.5385, 0.6429, None, 1.0, 0.75, 1.0, None, 1.0]
weighted_calibration_error	regressing	[0.1108, 0.1701, 0.1302, 0.1524, 0.1031, None, 0.4792, 0.2317, 0.4832, None, 0.4705]
margin_bias	improving	[-1.188, -1.083, -1.706, -3.0, -1.571, None, -4.25, -2.75, -3.0, None, -0.25]
team_run_error_on_losses	improving	[5.75, 3.5, 5.333, 6.167, 4.8, None, None, 2.0, None, None, None]

Stage	Winner hits	Accuracy	Team-run MAE	Total-run MAE	Margin MAE
historical_ml	4/4	100%	2.628	5.255	1.118
variance_applied	4/4	100%	2.625	5.25	0.75
enhanced_prediction	4/4	100%	2.125	4.25	0.75
embedded_savant_run_environment	4/4	100%	1.875	3.75	0.75
fundamental_edge_target_review	4/4	100%	1.875	3.75	0.75
capped_target	4/4	100%	1.875	3.75	0.75
simulation_result	4/4	100%	1.875	3.75	0.75
published_final	4/4	100%	1.875	3.75	0.75

Transition	Δ team-run MAE	Δ margin MAE
historical_ml → variance_applied	-0.0026	-0.3678
variance_applied → enhanced_prediction	-0.5	0
enhanced_prediction → embedded_savant_run_environment	-0.25	0
embedded_savant_run_environment → fundamental_edge_target_review	0	0
fundamental_edge_target_review → capped_target	0	0
capped_target → simulation_result	0	0
simulation_result → published_final	0	0

Game	Outcome	Raw model	Calibrated	Governor	Score coherence	Published	Explanation
ATL@LAD	hit	0.5848	0.5646	0.5646	0.5646	0.5467	score coherence followed by matchup-recency guardrail
CHW@CLE	hit	0.5439	0.5271	0.5271	0.5271	0.523	published probability matches the calibrated/governed value
NYY@TB	hit	0.5409	0.5254	0.5254	0.5254	0.5184	score coherence followed by matchup-recency guardrail
SD@MIL	hit	0.6451	0.5799	0.5799	0.5761	0.53	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
8	0	4	1	3	-0.95	-2.25

Slate	Games	Winner acc	Mean starter ERA error	Mean lineup runs vs proj	Sides broke on pitching	Sides broke on offense
20261002		—	—	—
20261001	1	1.000	0.00	0.50
20260930	4	0.750	0.78	0.12		1
20260929	4	1.000	-0.76	-0.50		3
20260928		—	—	—
20260927	14	0.643	-2.16	-0.11	2	5
20260926	13	0.538	0.55	0.69	3	4
20260925	17	0.647	1.13	0.50	8	10
20260924	12	0.667	1.68	-0.67	4	11
20260923	16	0.500	-0.49	-0.91	4	11
20260922	15	0.400	-0.32	-0.97	3	11
20260921	3	0.333	-1.08	-0.50

Game	Pred margin	Actual margin	Pred − actual
ATL@LAD	1	2	-1
CHW@CLE	2	3	-1
NYY@TB	1	1	0
SD@MIL	2	1	1

ATL@LADHIT

Predicted: LAD 4-5 (conf +0.09, lean)Actual: LAD 3-5Brier 0.205total bias +1

Root cause: Solid correct call

Evidence-backed diagnosis

published system edge 0.09 (lean); this is not the half-scale distance from 50%
team-run absolute error 1; total bias +1
pregame home-minus-away evidence: season +0.610, recent +0.314, pitcher -0.028, composite +0.532
confidence pipeline reached 0.565 before later publication/recency controls; published home probability was 0.547
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

Dylan Dodd (ATL) allowed 3 fewer ER than projected (0 vs 3)

Team	Starter	Projected	Actual	ER err
ATL	Dylan Dodd	3 ER / 17 outs	0 ER / 9 outs	-3
LAD	Tarik Skubal	1 ER / 19 outs	1 ER / 18 outs	0

Decision internals: run factors home×1 / away×1.028; 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-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.4]
Teoscar Hernández (LAD LF): 2-for-4, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.4]
Max Muncy (LAD 3B): 1-for-4, 1 HR, 1 RBI (projected 0.0 H, 0.0 HR) [run impact 1.4]
Ozzie Albies (ATL 2B): 1-for-3, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.26]
Shohei Ohtani (LAD DH): 1-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.26]

Starter miss: Dylan Dodd (ATL) posted 0.00 ERA vs projected 4.76 (0 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

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

Dylan Dodd (ATL) threw 9 outs, 0 ER = 0 ERA against a projected 4.76 ERA (-4.76 ERA error)

Bullpen behind him: 0 ER over 0 outs.

The HOME lineup he faced produced 5 runs on 8 hits and 3 HR, against a projected 5 runs / 7 hits (0 runs vs projection).

Who did the damage to him:

Teoscar Hernández: 2-for-4, 1 HR, 2 RBI [+1.4 runs vs projection] season xwOBA 0.332, above average - real projection error
Max Muncy: 1-for-4, 1 HR, 1 RBI [+1.4 runs vs projection] season xwOBA 0.349, elite - real projection error
Kyle Tucker: 2-for-4, 1 HR, 2 RBI [+1.07 runs vs projection] season xwOBA 0.331, above average - real projection error

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

HOME pitching

Tarik Skubal (LAD) threw 18 outs, 1 ER = 1.5 ERA against a projected 1.42 ERA (0.08 ERA error)

Bullpen behind him: 2 ER over 3 outs.

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

Who did the damage to him:

Ozzie Albies: 1-for-3, 1 HR, 2 RBI [+1.26 runs vs projection] season xwOBA 0.273, weak - a good day, not a model error
Sean Murphy: 1-for-3, 1 HR, 1 RBI [+0.93 runs vs projection] season xwOBA 0.273, weak - a good day, not a model error
Ha-Seong Kim: 0-for-1, 0 HR, 0 RBI [+0.33 runs vs projection] season xwOBA 0.222, weak - a good day, not a model error

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)
ATL	Dylan Dodd (Starter)	9 / 17	0 / 4.76 (-4.76)	0 / 3	2 / 3	0 / 1	0 / 3	0	4 / 7	3.5	opp run factor 1 (strong) · 49 prior appearances this season
LAD	Tarik Skubal (Starter)	18 / 19	1.5 / 1.42	1 / 1	2 / 5	1 / 0	2 / 1	1.24	7 / 7	6.26	venue home: 3.63 vs 1.7 ERA (worse here) · opp run factor 1.03 (strong) · 26 prior appearances this season
LAD	Tanner Scott (Closer)	2 / 3	27 / 18 (+9.00)	2 / 2	2 / 1	1 / 1	0 / 1	0.29	0 / 0	1.27
LAD	Evan Phillips (Middle Reliever (High Leverage))	1 / 1	0 / 27 (-27.00)	0 / 1	0 / 2	0 / 0	1 / 1	0.2	0 / 1	1.22

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	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 0	1 / 1	-1.4	0.351 elite	bad day
Teoscar Hernández (LF)	LAD	4 / 4	2 / 1	0 / 0	0 / 0	1 / 0	2 / 0	2 / 1	0 / 0	2 / 1	1.4	0.332 above average	projection too low
Max Muncy (3B)	LAD	4 / 3	1 / 0	0 / 0	0 / 0	1 / 0	1 / 1	1 / 0	0 / 0	0 / 1	1.4	0.349 elite	projection too low
Ozzie Albies (2B)	ATL	3 / 4	1 / 1	0 / 0	0 / 0	1 / 0	2 / 1	1 / 0	1 / 0	0 / 1	1.26	0.273 weak	good day
Shohei Ohtani (DH)	LAD	4 / 4	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	0 / 1	-1.26	0.397 elite	bad day
Kyle Tucker (RF)	LAD	4 / 4	2 / 1	0 / 0	0 / 0	1 / 0	2 / 0	1 / 1	0 / 1	2 / 1	1.07	0.331 above average	projection too low
Sean Murphy (C)	ATL	3 / 3	1 / 1	0 / 0	0 / 0	1 / 0	1 / 0	1 / 0	0 / 0	1 / 1	0.93	0.273 weak	good day
Will Smith (C)	LAD	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	-0.8	0.389 elite	bad day
Michael Harris II (CF)	ATL	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	1 / 1	-0.78	0.342 elite	bad day
Austin Riley (3B)	ATL	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 2	-0.47	0.299 below average	—
Andy Pages (CF)	LAD	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	3 / 1	-0.47	0.343 elite	—
Drake Baldwin (DH)	ATL	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	1 / 1	-0.33	0.37 elite	—
Ronald Acuña Jr. (RF)	ATL	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	0 / 1	-0.33	0.366 elite	—
Ha-Seong Kim (SS)	ATL	1 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0.33	0.222 weak	—
Freddie Freeman (1B)	LAD	3 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 1	0.33	0.363 elite	—
Tommy Edman (2B)	LAD	2 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0.33	0.303 average	—
Mauricio Dubón (LF)	ATL	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	-0.14	0.304 average	—
Mookie Betts (SS)	LAD	3 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 0	1 / 0	0.02	0.339 above average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	0.61	positive favors home; negative favors away
Recent form	0.314	positive favors home; negative favors away
Starting-pitcher matchup	-0.028	positive favors home; negative favors away
Composite	0.532	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.47-5.89	n/a	Base ML historical model prediction
variance_applied	4-6	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-5	0.547	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	4.47-5.89	4-6	0.58	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
ATL	Ha-Seong Kim	batter / SS	better	predicted_H 2.76
ATL	Dylan Dodd	pitcher / Starter	better	predicted_ER -2
LAD	Tarik Skubal	pitcher / Starter	neutral	predicted_SO 2
LAD	Tanner Scott	pitcher / Closer	worse	predicted_SO -2
LAD	Teoscar Hernández	batter / LF	worse	predicted_SO 1.919
LAD	Tommy Edman	batter / 2B	worse	predicted_AB -1.8
ATL	Ronald Acuña Jr.	batter / RF	worse	predicted_RBI 1.54
LAD	Mookie Betts	batter / SS	worse	predicted_AB -1.54

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	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	ATL model pitcher projections differ from final simulation pitching box score (H: current=5, sim=6; BB: current=4, sim=5).	ATL
info	CURRENT_FINALIZED_MATCH_BOX	ATL model-prediction batting totals differ from final simulation box score (AB: current=33, sim=36; H: current=8, sim=9; 2B: current=1, sim=0; 3B: current=0, sim=2; BB: current=2, sim=3; SO: current=10, sim=9; HBP: current=0, sim=2).	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; Ozzie Albies: RBI current=1 box=0, BB current=0 box=1, SO current=1 box=0; Michael Harris II: AB current=4 box=5, 2B current=1 box=0, RBI current=1 box=0, R current=1 box=0; Matt Olson: HR current=1 box=0, RBI current=1 box=2, BB current=0 box=1; Ronald Acuña Jr.: AB current=4 box=5, HR current=0 box=1, RBI current=0 box=2, BB current=1 box=0; Mauricio Dubón: H current=1 box=2, 3B current=0 box=1, R current=0 box=1; Sean Murphy: AB current=3 box=4, H current=1 box=0, SO current=1 box=2; Austin Riley: AB current=3 box=4, H current=1 box=0, SO current=2 box=1	ATL
info	CURRENT_FINALIZED_MATCH_BOX	LAD model-prediction batting totals differ from final simulation box score (AB: current=33, sim=30; H: current=7, sim=6; HR: current=1, sim=2; RBI: current=4, sim=5; BB: current=3, sim=5; SO: current=8, sim=9).	LAD
info	CURRENT_PLAYER_MATCH_BOX	LAD current finalized hitter lines are model predictions and differ from the simulation box score: Shohei Ohtani: AB current=4 box=3, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0; Teoscar Hernández: AB current=4 box=3, H current=1 box=0, R current=1 box=0, BB current=0 box=1; Mookie Betts: AB current=4 box=3, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Will Smith: H current=1 box=2, 2B current=0 box=1, HR current=0 box=1, RBI current=0 box=2; Freddie Freeman: AB current=4 box=3, H current=1 box=2, RBI current=0 box=1, R current=0 box=1; Max Muncy: AB current=3 box=4, RBI current=1 box=0, SO current=1 box=2; Kyle Tucker: H current=1 box=0, R current=1 box=0, BB current=1 box=0; Tommy Edman: AB current=3 box=2, BB current=0 box=1, SO current=1 box=2	LAD
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

CHW@CLEHIT

Predicted: CHW 4-2 (conf -0.05, lean)Actual: CHW 3-0Brier 0.228total 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 -0.161, recent -0.030, pitcher +0.006, composite -0.443
confidence pipeline reached 0.527 before later publication/recency controls; published home probability was 0.477
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 (6 vs 3)

Team	Starter	Projected	Actual	ER err
CHW	Hagen Smith	0 ER / 15 outs	0 ER / 8 outs	0
CLE	Parker Messick	3 ER / 18 outs	2 ER / 13 outs	-1

Decision internals: run factors home×0.991 / away×0.985; 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:

Angel Martínez (CLE LF): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.4]
Chase DeLauter (CLE RF): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.8]
Jo Adell (CLE DH): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.78]
José Ramírez (CLE 3B): 0-for-3, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.78]

Pitcher versus batter: what actually broke this game

AWAY pitching

No settled pitching line for this club. The forecast named Hagen Smith (CHW) at a projected 0 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 0 runs on 1 hits and 0 HR, against a projected 5 runs / 7 hits (-5 runs vs projection).

Who did the damage to him:

Austin Hedges: 0-for-1, 0 HR, 0 RBI [+0.33 runs vs projection] season xwOBA 0.28, weak - a good day, not a model error
Travis Bazzana: 1-for-4, 0 HR, 0 RBI [+0.31 runs vs projection] season xwOBA 0.299, below average - a good day, not a model error

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

HOME pitching

Parker Messick (CLE) threw 13 outs, 2 ER = 4.15 ERA against a projected 4.5 ERA (-0.35 ERA error)

Bullpen behind him: 1 ER over 9 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: offense: the lineup scored well below its projection

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
CLE	Parker Messick (Starter)	13 / 18	4.15 / 4.5	2 / 3	4 / 5	1 / 0	2 / 1	1.36	6 / 5	5.1	venue home: 2.97 vs 2.22 ERA (worse here) · opp run factor 0.99 (average) · 32 prior appearances this season
CLE	Cade Smith (Closer)	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0.29	1 / 0	1.25
CLE	Erik Sabrowski (Middle Reliever (High Leverage))	5 / 6	5.4 / 4.5	1 / 1	1 / 1	0 / 0	1 / 1	0.68	2 / 0	1.31
CLE	Hunter Gaddis (Setup Man)	1 / 0	0 / n/a	0 / 0	0 / 1	0 / 0	0 / 1	0.31	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
Angel Martínez (LF)	CLE	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 0	2 / 1	-1.4	0.299 below average	projection too high
Chase DeLauter (RF)	CLE	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	0 / 1	-0.8	0.346 elite	bad day
Jo Adell (DH)	CLE	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	4 / 1	-0.78	0.329 above average	bad day
José Ramírez (3B)	CLE	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	1 / 1	-0.78	0.325 above average	bad day
Steven Kwan (CF)	CLE	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	0 / 0	-0.47	0.329 above average	—
David Fry (1B)	CLE	1 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	-0.33	0.271 weak	—
Austin Hedges (C)	CLE	1 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0.33	0.28 weak	—
Travis Bazzana (2B)	CLE	4 / 3	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0.31	0.299 below average	—
Brayan Rocchio (SS)	CLE	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	1 / 1	-0.14	0.294 below average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-0.161	positive favors home; negative favors away
Recent form	-0.03	positive favors home; negative favors away
Starting-pitcher matchup	0.006	positive favors home; negative favors away
Composite	-0.443	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.88-3.62	n/a	Base ML historical model prediction
variance_applied	5-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-3	0.523	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by matchup_recency_guardrail)
embedded_savant_run_environment	4-2	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	4-2	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	4-2	n/a	Capped at max 12 realistic runs
simulation_result	4-2	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.88-3.62	5-4	-0.259	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 matchup_recency_guardrail)
embedded_savant_run_environment	4-3	4-2	1	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	4-2	4-2	-0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	4-2	4-2	-0	Capped at max 12 realistic runs
simulation_result	4-2	4-2	-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
CHW	Randal Grichuk	batter / LF	better	predicted_AB 2.5
CHW	Brenton Doyle	batter / CF	worse	predicted_AB 2.139
CHW	Braden Montgomery	batter / RF	better	predicted_2B 1.83
CHW	Munetaka Murakami	batter / 1B	better	predicted_SO -1.558
CHW	Colson Montgomery	batter / SS	worse	predicted_SO -1.484
CLE	Steven Kwan	batter / CF	worse	predicted_AB 1.372
CHW	Chase Meidroth	batter / 2B	worse	predicted_H -1.36
CHW	Kyle Teel	batter / C	worse	predicted_SO -1.35

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	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	CHW model-prediction batting totals differ from final simulation box score (AB: current=32, sim=37; 2B: current=2, sim=4; HR: current=1, sim=0; R: current=5, sim=4; SO: current=11, sim=5; HBP: current=0, sim=2).	CHW
info	CURRENT_PLAYER_MATCH_BOX	CHW current finalized hitter lines are model predictions and differ from the simulation box score: Randal Grichuk: AB current=4 box=5, H current=1 box=2, RBI current=1 box=2, R current=0 box=1; Munetaka Murakami: AB current=4 box=5, H current=1 box=2, HR current=1 box=0, RBI current=1 box=0; Chase Meidroth: AB current=4 box=5, H current=1 box=0, 2B current=1 box=0, R current=1 box=0; Miguel Vargas: H current=1 box=0, R current=1 box=0, SO current=1 box=2; Braden Montgomery: AB current=3 box=4, H current=1 box=2, 2B current=0 box=2, RBI current=0 box=1; Brenton Doyle: AB current=4 box=5, H current=1 box=0; Tommy Pham: AB current=3 box=2, H current=0 box=1, 2B current=0 box=1, RBI current=0 box=1; Kyle Teel: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, BB current=0 box=1	CHW
info	CURRENT_FINALIZED_MATCH_BOX	CLE model-prediction batting totals differ from final simulation box score (AB: current=32, sim=35; H: current=7, sim=5; 2B: current=2, sim=0; RBI: current=4, sim=2; R: current=5, sim=2; BB: current=4, sim=2; SO: current=8, sim=9).	CLE
info	CURRENT_PLAYER_MATCH_BOX	CLE current finalized hitter lines are model predictions and differ from the simulation box score: Steven Kwan: AB current=4 box=5, R current=1 box=0, BB current=1 box=0, SO current=0 box=1; Jo Adell: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, R current=1 box=0; David Fry: SO current=1 box=2; José Ramírez: 2B current=1 box=0, RBI current=1 box=0, R current=1 box=0, BB current=1 box=0; Angel Martínez: RBI current=1 box=2, SO current=1 box=2; Chase DeLauter: H current=1 box=0, RBI current=1 box=0, BB current=1 box=0; Travis Bazzana: AB current=3 box=4, H current=1 box=2, R current=0 box=1, SO current=1 box=0; Austin Hedges: BB current=0 box=1	CLE
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

NYY@TBHIT

Predicted: TB 3-4 (conf +0.04, toss_up)Actual: TB 0-1Brier 0.232total bias +6

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 6; total bias +6
pregame home-minus-away evidence: season -0.176, recent -0.182, pitcher +0.027, composite +0.097
confidence pipeline reached 0.525 before later publication/recency controls; published home probability was 0.518
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 6 (7 vs 1)
NYY offense over-projected by 3
TB offense over-projected by 3

Team	Starter	Projected	Actual	ER err
NYY	Gerrit Cole	3 ER / 17 outs	1 ER / 15 outs	-2
TB	Drew Rasmussen	2 ER / 18 outs	0 ER / 24 outs	-2

Decision internals: run factors home×1.015 / away×0.988; 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-3, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.4]
Junior Caminero (TB 3B): 0-for-2, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.07]
Cody Bellinger (NYY LF): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.8]
Luis García Jr. (NYY 1B): 0-for-3, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.78]
Austin Wells (NYY C): 1-for-3, 0 HR, 0 RBI (projected 0.0 H, 0.0 HR) [run impact 0.78]

Starter miss: Drew Rasmussen (TB) posted 0.00 ERA vs projected 3.00 (0 ER in 24 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

Gerrit Cole (NYY) threw 15 outs, 1 ER = 1.8 ERA against a projected 4.76 ERA (-2.96 ERA error)

Bullpen behind him: 0 ER over 1 outs.

The HOME lineup he faced produced 1 runs on 4 hits and 1 HR, against a projected 3 runs / 7 hits (-2 runs vs projection).

Who did the damage to him:

Liam Hicks: 1-for-3, 0 HR, 0 RBI [+0.64 runs vs projection] season xwOBA 0.319, average - a good day, not a model error
Jonathan Aranda: 1-for-4, 1 HR, 1 RBI [+0.6 runs vs projection] season xwOBA 0.343, elite - real projection error
Richie Palacios: 1-for-3, 0 HR, 0 RBI [+0.31 runs vs projection] season xwOBA 0.306, average - a good day, not a model error

Verdict: offense: the lineup scored well below its projection

HOME pitching

Drew Rasmussen (TB) threw 24 outs, 0 ER = 0 ERA against a projected 3 ERA (-3 ERA error)

Bullpen behind him: 0 ER over 3 outs.

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

Who did the damage to him:

Austin Wells: 1-for-3, 0 HR, 0 RBI [+0.78 runs vs projection] season xwOBA 0.3, below 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)
NYY	Gerrit Cole (Starter)	15 / 17	1.8 / 4.76	1 / 3	5 / 7	1 / 0	1 / 3	0.98	5 / 6	5.87	venue away: 3.41 vs 3.86 ERA (ok) · opp run factor 1.01 (average) · 22 prior appearances this season
NYY	Brent Headrick (Setup Man)	1 / 5	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0.31	0 / 1	0.9
TB	Drew Rasmussen (Starter)	24 / 18	0 / 3 (-3.00)	0 / 2	1 / 1	0 / 0	2 / 3	1.25	10 / 5	4.77	venue home: 2.87 vs 2.47 ERA (ok) · opp run factor 0.99 (strong) · 30 prior appearances this season
TB	Bryan Baker (Closer)	3 / 3	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	0.29	2 / 0	0.69

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	3 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	1 / 1	2 / 1	-1.4	0.378 elite	bad day
Junior Caminero (3B)	TB	2 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	2 / 1	1 / 1	-1.07	0.364 elite	bad day
Cody Bellinger (LF)	NYY	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	2 / 1	-0.8	0.335 above average	bad day
Luis García Jr. (1B)	NYY	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0 / 1	-0.78	0.341 elite	bad day
Austin Wells (C)	NYY	3 / 3	1 / 0	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0.78	0.3 below average	good day
Jonny DeLuca (CF)	TB	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	-0.78	0.261 weak	projection too high
Liam Hicks (C)	TB	3 / 4	1 / 1	1 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 0	0 / 0	0.64	0.319 average	no signal
Jonathan Aranda (1B)	TB	4 / 4	1 / 1	0 / 0	0 / 0	1 / 0	1 / 1	1 / 0	0 / 1	0 / 1	0.6	0.343 elite	projection too low
Spencer Jones (RF)	NYY	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	2 / 1	-0.47	0.328 above average	—
George Lombard Jr. (SS)	NYY	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	-0.47	0.298 below average	—
Trent Grisham (CF)	NYY	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	-0.47	0.337 above average	—
Chandler Simpson (LF)	TB	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	-0.47	0.263 weak	—
Richie Palacios (2B)	TB	3 / 3	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0.31	0.306 average	—
Jazz Chisholm Jr. (2B)	NYY	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	1 / 1	-0.14	0.294 below average	—
Ryan McMahon (3B)	NYY	2 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0	0.297 below average	—
Yandy Díaz (DH)	TB	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	0 / 1	0	0.346 elite	—
Taylor Walls (SS)	TB	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	0	0.251 weak	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-0.176	positive favors home; negative favors away
Recent form	-0.182	positive favors home; negative favors away
Starting-pitcher matchup	0.027	positive favors home; negative favors away
Composite	0.097	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.33-4.77	n/a	Base ML historical model prediction
variance_applied	4-5	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	3-4	0.518	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	4.33-4.77	4-5	0.565	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
TB	Jonathan Aranda	batter / 1B	worse	predicted_SO 1.948
NYY	Cody Bellinger	batter / LF	boosted	predicted_AB -1.88
TB	Yandy Díaz	batter / DH	deboosted	predicted_2B 1.83
NYY	Trent Grisham	batter / CF	better	predicted_RBI 1.54
TB	Liam Hicks	batter / C	worse	predicted_AB -1.51
TB	Richie Palacios	batter / 2B	better	predicted_RBI 1.47
NYY	Ben Rice	batter / DH	worse	predicted_H -1.28
TB	Victor Mesa Jr.	batter / RF	boosted	predicted_SO 1.254

Structural audit findings

Severity	Rule	Finding	Subject
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_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	NYY model-prediction batting totals differ from final simulation box score (AB: current=32, sim=31; H: current=7, sim=4; 2B: current=1, sim=2; 3B: current=0, sim=1; RBI: current=4, sim=3; R: current=5, sim=3; BB: current=2, sim=4; SO: current=9, sim=5).	NYY
info	CURRENT_PLAYER_MATCH_BOX	NYY current finalized hitter lines are model predictions and differ from the simulation box score: Spencer Jones: H current=1 box=0, RBI current=1 box=0, R current=1 box=0; Cody Bellinger: AB current=4 box=3, 2B current=0 box=1; Luis García Jr.: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, SO current=1 box=0; Ben Rice: H current=1 box=0, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0; George Lombard Jr.: AB current=4 box=3, 3B current=0 box=1, BB current=0 box=1, SO current=1 box=0; Jazz Chisholm Jr.: H current=1 box=0, R current=1 box=0, BB current=0 box=1; Trent Grisham: H current=1 box=2, 2B current=0 box=1, HR current=0 box=1, RBI current=0 box=2; Austin Wells: AB current=3 box=4	NYY
info	CURRENT_FINALIZED_MATCH_BOX	TB model-prediction batting totals differ from final simulation box score (AB: current=34, sim=32; 2B: current=1, sim=3; 3B: current=0, sim=1; HR: current=1, sim=0; BB: current=2, 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: AB current=4 box=3, H current=1 box=0, RBI current=1 box=0, BB current=0 box=1; Jonathan Aranda: H current=1 box=0, RBI current=1 box=0, BB current=1 box=0, SO current=1 box=3; Yandy Díaz: H current=1 box=3, 2B current=0 box=2, SO current=1 box=0; Jonny DeLuca: H current=1 box=0, 2B current=1 box=0; Junior Caminero: AB current=4 box=3, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0; Chandler Simpson: H current=1 box=2, R current=0 box=1; Richie Palacios: AB current=3 box=4, H current=1 box=2, 2B current=0 box=1, RBI current=0 box=2; Victor Mesa Jr.: 3B current=0 box=1, RBI current=0 box=1, SO current=1 box=2	TB
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

SD@MILHIT

Predicted: MIL 4-6 (conf +0.06, lean)Actual: MIL 2-3Brier 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 +0.723, recent -0.334, pitcher +0.063, composite +0.813
confidence pipeline reached 0.580 before later publication/recency controls; published home probability was 0.530
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 5 (10 vs 5)
MIL offense over-projected by 3

Team	Starter	Projected	Actual	ER err
SD	Robbie Ray	3 ER / 15 outs	2 ER / 8 outs	-1
MIL	Jacob Misiorowski	1 ER / 18 outs	2 ER / 12 outs	1

Decision internals: run factors home×1.025 / away×0.962; 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): 1-for-3, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 0.93]
Dustin Harris (SD LF): 2-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact 0.8]
Cooper Pratt (MIL SS): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.8]
Garrett Mitchell (MIL CF): 1-for-3, 0 HR, 0 RBI (projected 0.0 H, 0.0 HR) [run impact 0.8]
Brice Turang (MIL 2B): 0-for-3, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.78]

Starter miss: Jacob Misiorowski (MIL) posted 4.50 ERA vs projected 1.50 (2 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

Robbie Ray (SD) threw 8 outs, 2 ER = 6.75 ERA against a projected 5.4 ERA (1.35 ERA error)

Bullpen behind him: 1 ER over 10 outs.

The HOME lineup he faced produced 3 runs on 6 hits and 2 HR, against a projected 5 runs / 8 hits (-2 runs vs projection).

Who did the damage to him:

William Contreras: 1-for-3, 1 HR, 1 RBI [+0.93 runs vs projection] season xwOBA 0.334, above average - real projection error
Garrett Mitchell: 1-for-3, 0 HR, 0 RBI [+0.8 runs vs projection] season xwOBA 0.353, elite - real projection error

Verdict: offense: the lineup scored well below its projection

HOME pitching

Jacob Misiorowski (MIL) threw 12 outs, 2 ER = 4.5 ERA against a projected 1.5 ERA (3 ERA error)

Bullpen behind him: 0 ER over 10 outs.

The AWAY lineup he faced produced 2 runs on 9 hits and 0 HR, against a projected 5 runs / 7 hits (-3 runs vs projection).

Who did the damage to him:

Dustin Harris: 2-for-4, 0 HR, 0 RBI [+0.8 runs vs projection] season xwOBA 0.314, average - a good day, not a model error
Ethan Salas: 1-for-3, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.227, 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)
SD	Robbie Ray (Starter)	8 / 15	6.75 / 5.4	2 / 3	4 / 6	1 / 1	2 / 1	2.92	1 / 4	3.71	venue away: 4.29 vs 3.2 ERA (worse here) · opp run factor 1.03 (strong) · 31 prior appearances this season
SD	Bradgley Rodriguez (Middle Reliever (High Leverage))	4 / 5	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0.43	2 / 0	1.37
SD	Adrian Morejon (Setup Man)	6 / 1	4.5 / 54 (-49.50)	1 / 2	3 / 1	1 / 1	0 / 1	0.55	2 / 1	1.42
MIL	Jacob Misiorowski (Starter)	12 / 18	4.5 / 1.5 (+3.00)	2 / 1	4 / 4	0 / 0	1 / 4	1.91	6 / 6	7.58	venue home: 1.52 vs 2.15 ERA (ok) · opp run factor 0.96 (average) · 30 prior appearances this season
MIL	Aaron Ashby (Middle Reliever (High Leverage))	5 / 3	0 / 18 (-18.00)	0 / 2	0 / 3	0 / 0	0 / 0	0.59	1 / 0	1.64
MIL	Abner Uribe (Setup Man)	1 / 3	0 / 0	0 / 0	2 / 2	0 / 0	0 / 0	0.28	0 / 0	1.14
MIL	Trevor Megill (Closer)	4 / 3	0 / 9 (-9.00)	0 / 1	1 / 1	0 / 1	0 / 0	0.21	2 / 1	1.2

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	1 / 1	0 / 0	0 / 0	1 / 0	1 / 1	1 / 0	1 / 1	0 / 1	0.93	0.334 above average	projection too low
Dustin Harris (LF)	SD	4 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	0 / 1	0.8	0.314 average	no signal
Cooper Pratt (SS)	MIL	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	2 / 1	-0.8	0.315 average	no signal
Garrett Mitchell (CF)	MIL	3 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	1 / 0	1 / 1	0.8	0.353 elite	projection too low
Brice Turang (2B)	MIL	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	1 / 1	-0.78	0.344 elite	bad day
Jackson Merrill (CF)	SD	5 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	2 / 2	-0.47	0.347 elite	—
Ethan Salas (C)	SD	3 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 1	0.47	0.227 weak	—
Joey Ortiz (3B)	MIL	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	-0.47	0.297 below average	—
Jackson Chourio (LF)	MIL	4 / 4	1 / 1	0 / 0	0 / 0	1 / 1	2 / 1	1 / 1	0 / 1	0 / 1	-0.33	0.344 elite	—
Fernando Tatis Jr. (RF)	SD	5 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	2 / 1	-0.31	0.379 elite	—
Manny Machado (3B)	SD	4 / 4	2 / 1	1 / 0	0 / 0	0 / 1	0 / 1	0 / 1	1 / 1	1 / 1	-0.15	0.328 above average	—
Jake Bauers (1B)	MIL	3 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	0.14	0.371 elite	—
Luis Lara (RF)	MIL	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	-0.14	0.293 below average	—
Ty France (1B)	SD	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	1 / 1	0	0.327 above average	—
Luis Campusano (DH)	SD	3 / 4	1 / 1	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 1	0 / 1	0	0.327 above average	—
Jake Cronenworth (2B)	SD	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	2 / 1	0	0.275 weak	—
Xander Bogaerts (SS)	SD	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	1 / 1	0 / 1	0	0.325 above average	—
Christian Yelich (DH)	MIL	3 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	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	0.723	positive favors home; negative favors away
Recent form	-0.334	positive favors home; negative favors away
Starting-pitcher matchup	0.063	positive favors home; negative favors away
Composite	0.813	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	3.74-6.33	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	4-6	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	4-6	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	4-6	n/a	Capped at max 12 realistic runs
simulation_result	4-6	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	3.74-6.33	4-6	-0.586	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	4-6	1	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	4-6	4-6	0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	4-6	4-6	0	Capped at max 12 realistic runs
simulation_result	4-6	4-6	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
MIL	Jacob Misiorowski	pitcher / Starter	better	predicted_SO 2
MIL	Abner Uribe	pitcher / Setup Man	worse	predicted_SO -2
SD	Ty France	batter / 1B	better	predicted_H 1.79
MIL	Brice Turang	batter / 2B	better	predicted_AB -1.66
SD	Jackson Merrill	batter / CF	worse	predicted_H -1.5
SD	Ethan Salas	batter / C	worse	predicted_AB 1.5
MIL	Jake Bauers	batter / 1B	better	predicted_RBI 1.49
SD	Dustin Harris	batter / LF	worse	predicted_AB 1.45

Structural audit findings

Severity	Rule	Finding	Subject
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_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_FINALIZED_MATCH_BOX	MIL model-prediction batting totals differ from final simulation box score (AB: current=33, sim=32; 3B: current=0, sim=1; HR: current=1, sim=3; RBI: current=5, sim=6; R: current=5, sim=6; BB: current=5, sim=3; SO: current=9, sim=5).	MIL
info	CURRENT_PLAYER_MATCH_BOX	MIL current finalized hitter lines are model predictions and differ from the simulation box score: Jake Bauers: H current=1 box=2, HR current=0 box=1, RBI current=1 box=2, BB current=1 box=0; Jackson Chourio: HR current=1 box=0, BB current=1 box=0, SO current=1 box=0; William Contreras: 2B current=0 box=1, RBI current=1 box=0, R current=0 box=1, BB current=1 box=0; Luis Lara: H current=1 box=0, SO current=1 box=0; Brice Turang: AB current=4 box=3, 2B current=1 box=0, HR current=0 box=1, RBI current=1 box=2; Joey Ortiz: AB current=3 box=2, 3B current=0 box=1, R current=0 box=1, BB current=0 box=1; Cooper Pratt: H current=1 box=2, HR current=0 box=1, RBI current=0 box=1, R current=0 box=1; Christian Yelich: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, SO current=1 box=0	MIL
info	CURRENT_FINALIZED_MATCH_BOX	SD model-prediction batting totals differ from final simulation box score (AB: current=33, sim=37; H: current=7, sim=10; 2B: current=1, sim=2; 3B: current=0, sim=1; R: current=5, sim=4; BB: current=3, sim=4; SO: current=10, sim=7).	SD
info	CURRENT_PLAYER_MATCH_BOX	SD current finalized hitter lines are model predictions and differ from the simulation box score: Dustin Harris: AB current=4 box=5, H current=1 box=0, R current=1 box=0; Manny Machado: H current=1 box=2, 2B current=0 box=1, HR current=1 box=0, RBI current=1 box=2; Ty France: AB current=4 box=5, H current=1 box=3, 2B current=0 box=1, 3B current=0 box=1; Fernando Tatis Jr.: H current=1 box=2, 2B current=1 box=0, HR current=0 box=1, BB current=0 box=1; Luis Campusano: AB current=4 box=3, R current=0 box=1, SO current=1 box=0; Jackson Merrill: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, BB current=0 box=1; Ethan Salas: AB current=3 box=4, H current=0 box=1; Jake Cronenworth: AB current=3 box=4, H current=0 box=1	SD
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 20261002. 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	0.2435	0.2435	+1.2e-05	regressed
accuracy	0.5877	0.5965	+0.0088	improved
team_run_mae	2.5351	2.5263	-0.0088	improved
total_run_mae	3.6316	3.7018	+0.0702	regressed
calibration_error	0.0896	0.0995	+0.0099	regressed
starter_er_mae	1.7243	1.7072	-0.0171	improved
starter_outs_mae	4.2103	4.2477	+0.0374	regressed

Deep slate report — 20261003

Process, end to end

Slate summary

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 (4)

ATL@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

CHW@CLEHIT

Player-level breakdown (projected vs actual)

Fundamental evidence entering the pick

Score and confidence evolution

Official postgame stage attribution

Largest player-level adjustments from the ultra-detailed trace

Structural audit findings

NYY@TBHIT

Player-level breakdown (projected vs actual)

Fundamental evidence entering the pick

Score and confidence evolution

Official postgame stage attribution

Largest player-level adjustments from the ultra-detailed trace

Structural audit findings

SD@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

Cross-game lever tension

score.variance_multiplier / score.variance_floor (widen run distributions)

player.pitcher.v4 form weights / player.pitcher_prior_games

recent.* / season.* run-mean weighting

Day over day: did it improve or worsen

The adjustment the system made in response