Deep slate report 20260919

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_20260919.json	378363
forensic deep audit	loaded	probability, provenance, starter and decision contracts	/app/output/reports/PREDICTION_SYSTEM_DEEP_AUDIT_20260919.json	240964
continuous learning	loaded	pipeline-stage quality, calibration, probability lineage, score distribution, and official starter comparisons	/app/output/learning/continuous_learning_daily_20260919.json	2480431
detailed report	loaded	full game and simulation rendering	/app/output/reports/detailed_20260919.html	355458
ultra detailed player audit	loaded	player-level prediction trace	/app/output/reports/20260919_ultra_detailed_player_audit.html	32516149
structural audit	loaded	all invariant findings	/app/output/reports/audit_report_20260919.html	1837773
integrity revision audit	missing	post-Luna remediation proof for immutable revision artifacts	/app/output/reports/audit_report_20260919_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	probability_lineage	Published probability differs materially without a recorded transformation 3 game(s) have an unexplained ≥2pp calibrated-to-published delta	ATL@HOU, SEA@COL, SF@LAD	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	pipeline_stages	An intermediate stage materially degraded winner direction historical_ml → variance_applied changed winner accuracy by -0.2000 (-3 hits)	ATL@HOU, MIA@SD, WAS@STL	Do thisList the exact games whose winner flipped at this stage, and keep both the before-stage and after-stage picks in the trace. Don't concludeEven if a later stage recovers the pick, this unstable step is still a risk; do not treat it as harmless.
HIGH	calibration	Published confidence buckets were poorly aligned on this slate weighted absolute calibration error 0.3458 across 15 games	ATL@HOU, BOS@TB, CHC@CIN, DET@CHW, KC@PIT, MIA@SD, MIL@BAL, MIN@LAA, NYY@ARI, OAK@CLE, PHI@NYM, SEA@COL, SF@LAD, TOR@TEX, WAS@STL	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 = 0.0082/0.0273/0.0092	ATL@HOU, BOS@TB, CHC@CIN, DET@CHW, KC@PIT, MIA@SD, MIL@BAL, MIN@LAA, NYY@ARI, OAK@CLE, PHI@NYM, SEA@COL, SF@LAD, TOR@TEX, WAS@STL	Do thisTreat exact run scores as low-confidence estimates, and look for stronger features or a better target before tuning the score tails. Don't concludeA small positive R² only means the model beats guessing the average; it does not make the exact scores reliable.
MEDIUM	calibration	Most high-raw picks lost before/after shrinkage 1/3 games with raw pick probability ≥0.65 won	NYY@ARI, SEA@COL, SF@LAD	Do thisTrack how often high-raw-probability picks lose across multiple slates, and keep the raw to calibrated to published lineage. Don't concludeThis one slate is not enough to justify a new per-bucket shrinkage rule.
INFO	coverage	Every settled game received structural QA 15/15 games structurally audited	ATL@HOU, BOS@TB, CHC@CIN, DET@CHW, KC@PIT, MIA@SD, MIL@BAL, MIN@LAA, NYY@ARI, OAK@CLE, PHI@NYM, SEA@COL, SF@LAD, TOR@TEX, WAS@STL	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	ATL@HOU, BOS@TB, CHC@CIN, DET@CHW, KC@PIT, MIA@SD, MIL@BAL, MIN@LAA, NYY@ARI, OAK@CLE, PHI@NYM, SEA@COL, SF@LAD, TOR@TEX, WAS@STL	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	1	2/11	yes	param
pipeline_winner_instability	2	4/11	yes	code/diagnostic
score_margin_compression	0	1/11	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	11	11/11	yes	code/diagnostic

Metric	Direction	Series (old → new)
winner_accuracy	improving	[0.6667, 0.6, 0.6667, 0.5333, 0.6, 0.7, 0.4, 0.6667, 0.4444, 0.6667, 0.7333]
weighted_calibration_error	regressing	[0.1069, 0.1171, 0.2091, 0.1711, 0.0827, 0.2812, 0.2795, 0.1661, 0.1016, 0.1247, 0.3458]
margin_bias	improving	[-2.133, -1.0, -2.933, -2.667, -1.8, -0.4, -1.667, -1.133, -2.111, -2.0, -0.667]
team_run_error_on_losses	regressing	[5.6, 6.5, 5.0, 6.143, 5.0, 6.333, 5.556, 6.4, 3.6, 5.2, 6.5]

Stage	Winner hits	Accuracy	Team-run MAE	Total-run MAE	Margin MAE
historical_ml	10/15	66.7%	2.302	3.627	2.702
variance_applied	7/15	46.7%	2.333	3.733	2.667
enhanced_prediction	11/15	73.3%	2.367	4.067	2.467
embedded_savant_run_environment	11/15	73.3%	2.4	4.133	2.4
fundamental_edge_target_review	11/15	73.3%	2.4	4.133	2.4
capped_target	11/15	73.3%	2.4	4.133	2.4
simulation_result	11/15	73.3%	2.4	4.133	2.4
published_final	11/15	73.3%	2.4	4.133	2.4

Transition	Δ accuracy	Δ hits	Δ team-run MAE	Δ margin MAE	Games whose winner/tie changed
historical_ml → variance_applied	-0.2	-3	0.0312	-0.0357	ATL@HOU: ATL → TIE; MIA@SD: SD → TIE; TOR@TEX: TOR → TIE; WAS@STL: WAS → TIE
variance_applied → enhanced_prediction	0.2666	4	0.0334	-0.2	ATL@HOU: TIE → ATL; DET@CHW: CHW → DET; KC@PIT: KC → PIT; MIA@SD: TIE → SD; TOR@TEX: TIE → TEX; WAS@STL: TIE → WAS
enhanced_prediction → embedded_savant_run_environment	0	0	0.0333	-0.0667
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
ATL@HOU	hit	0.5609	0.5984	0.5984	0.5871	0.562	no recorded transformation explains the publication delta
BOS@TB	hit	0.5686	0.5404	0.5404	0.5404	0.5292	score coherence followed by matchup-recency guardrail
CHC@CIN	hit	0.6443	0.6837	0.6837	0.6837	0.56	score coherence followed by matchup-recency guardrail
DET@CHW	miss	0.5703	0.6277	0.6277	0.6277	0.56	score coherence followed by matchup-recency guardrail
KC@PIT	hit	0.5	0.5482	0.5482	0.5482	0.541	no recorded transformation explains the publication delta
MIA@SD	hit	0.6431	0.6938	0.6938	0.6938	0.58	score coherence followed by matchup-recency guardrail
MIL@BAL	miss	0.5795	0.5462	0.5462	0.5462	0.53	score coherence followed by matchup-recency guardrail
MIN@LAA	hit	0.6036	0.5752	0.5752	0.5752	0.5544	score coherence followed by matchup-recency guardrail
NYY@ARI	miss	0.7133	0.7112	0.7112	0.7112	0.7112	published probability matches the calibrated/governed value
OAK@CLE	hit	0.6147	0.6689	0.6689	0.6589	0.58	score coherence followed by matchup-recency guardrail
PHI@NYM	hit	0.5684	0.5403	0.5403	0.5403	0.5291	score coherence followed by matchup-recency guardrail
SEA@COL	miss	0.7647	0.7642	0.7642	0.7642	0.7213	no recorded transformation explains the publication delta
SF@LAD	hit	0.8181	0.6271	0.6271	0.6271	0.5651	no recorded transformation explains the publication delta
TOR@TEX	hit	0.542	0.611	0.611	0.611	0.611	published probability matches the calibrated/governed value
WAS@STL	hit	0.5218	0.5597	0.5597	0.5597	0.5431	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	1	14	1.36	-0.1

Game	Trained p(home)	Blended p(home)	Blend favoured	Published p(home)	Published side	Artifact says winner changed	Outcome
KC@PIT	0.432	0.493	away	0.541	home	yes	hit

Slate	Games	Winner acc	Mean starter ERA error	Mean lineup runs vs proj	Sides broke on pitching	Sides broke on offense
20260918	15	0.667	3.95	0.30	6	5
20260917	9	0.444	2.16	-0.22	2	3
20260916	15	0.667	2.09	-0.23	4	5
20260915	15	0.400	-1.58	-0.70	1	12
20260914	10	0.700	-0.48	-0.20	2	6
20260913	15	0.600	-0.07	0.03	2	8
20260912	15	0.533	1.35	0.57	6	7
20260911	15	0.667	0.99	0.00	4	9
20260910	5	0.600	-3.19	-1.70		8
20260909	15	0.667	3.36	1.07	8	7
20260908	15	0.733	-0.15	-0.23	7	15
20260907	11	0.545	-0.86	-0.36	5	4

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

ATL@HOUHIT

Predicted: ATL 4-3 (conf -0.12, moderate)Actual: ATL 6-3Brier 0.192total bias -2

Root cause: Strong correct call

Evidence-backed diagnosis

published system edge 0.12 (moderate); this is not the half-scale distance from 50%
team-run absolute error 2; total bias -2
pregame home-minus-away evidence: season -0.896, recent -0.638, pitcher -0.012, composite -0.858
confidence pipeline reached 0.598 before later publication/recency controls; published home probability was 0.438
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

No large single-factor errors.

Team	Starter	Projected	Actual	ER err
ATL	Grant Holmes	2 ER / 15 outs	2 ER / 18 outs	0
HOU	Hayden Wesneski	2 ER / 16 outs	1 ER / 16 outs	-1

Decision internals: run factors home×0.962 / away×0.974; 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:

Mike Yastrzemski (ATL LF): 2-for-4, 1 HR, 1 RBI (projected 0.0 H, 0.0 HR) [run impact 1.87]
Ronald Acuña Jr. (ATL RF): 2-for-3, 1 HR, 4 RBI (projected 1.0 H, 0.0 HR) [run impact 1.73]
Isaac Paredes (HOU 3B): 2-for-3, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.4]
Yordan Alvarez (HOU DH): 1-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.26]
Jose Altuve (HOU 2B): 3-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact 0.94]

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

Grant Holmes (ATL) threw 18 outs, 2 ER = 3 ERA against a projected 3.6 ERA (-0.6 ERA error)

Bullpen behind him: 1 ER over 6 outs.

The HOME lineup he faced produced 3 runs on 10 hits and 1 HR, against a projected 4 runs / 7 hits (-1 runs vs projection).

Who did the damage to him:

Isaac Paredes: 2-for-3, 1 HR, 2 RBI [+1.4 runs vs projection] season xwOBA 0.319, average - a good day, not a model error
Jose Altuve: 3-for-4, 0 HR, 0 RBI [+0.94 runs vs projection] season xwOBA 0.271, weak - a good day, not a model error
Christian Walker: 2-for-4, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.298, below average - a good day, not a model error

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

HOME pitching

Hayden Wesneski (HOU) threw 16 outs, 1 ER = 1.69 ERA against a projected 3.38 ERA (-1.69 ERA error) only 9 prior appearances in the covered season - the projection had little of his own record to stand on

Bullpen behind him: 4 ER over 3 outs.

The AWAY lineup he faced produced 6 runs on 11 hits and 2 HR, against a projected 5 runs / 8 hits (1 runs vs projection).

Who did the damage to him:

Mike Yastrzemski: 2-for-4, 1 HR, 1 RBI [+1.87 runs vs projection] season xwOBA 0.307, average - a good day, not a model error
Ronald Acuña Jr.: 2-for-3, 1 HR, 4 RBI [+1.73 runs vs projection] season xwOBA 0.364, elite - real projection error
Drake Baldwin: 3-for-5, 0 HR, 0 RBI [+0.92 runs vs projection] season xwOBA 0.366, elite - real projection 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	Grant Holmes (Starter)	18 / 15	3 / 3.6	2 / 2	8 / 4	1 / 0	0 / 4	1.58	4 / 1	2.99	venue away: 2.58 vs 4.79 ERA (ok) · opp run factor 0.96 (average) · 30 prior appearances this season
ATL	Dylan Lee (Setup Man)	3 / 3	9 / 0 (+9.00)	1 / 0	2 / 0	0 / 0	1 / 1	0.21	1 / 0	1.03
ATL	Raisel Iglesias (Closer)	3 / 3	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0.25	1 / 0	1.08
HOU	Hayden Wesneski (Starter)	16 / 16	1.69 / 3.38	1 / 2	4 / 2	0 / 0	3 / 1	1.78	3 / 5	3.79	venue home: 2.86 vs 4.39 ERA (ok) · opp run factor 0.97 (strong) · 9 prior appearances this season
HOU	AJ Blubaugh (Middle Reliever (High Leverage))	1 / 4	81 / 6.75 (+74.25)	3 / 1	2 / 2	0 / 0	1 / 0	0.53	0 / 2	0.86
HOU	Bryan Abreu (Setup Man)	2 / 4	13.5 / 0 (+13.50)	1 / 0	2 / 1	1 / 0	0 / 0	0.55	1 / 1	1.33

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
Mike Yastrzemski (LF)	ATL	4 / 3	2 / 0	0 / 0	0 / 0	1 / 0	1 / 0	2 / 0	0 / 0	1 / 1	1.87	0.307 average	no signal
Ronald Acuña Jr. (RF)	ATL	3 / 4	2 / 1	0 / 0	0 / 0	1 / 0	4 / 1	1 / 1	2 / 1	0 / 1	1.73	0.364 elite	projection too low
Isaac Paredes (3B)	HOU	3 / 4	2 / 1	0 / 0	0 / 0	1 / 0	2 / 1	2 / 0	1 / 1	0 / 0	1.4	0.319 average	no signal
Yordan Alvarez (DH)	HOU	4 / 4	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	0 / 1	-1.26	0.443 elite	bad day
Jose Altuve (2B)	HOU	4 / 4	3 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	0.94	0.271 weak	good day
Drake Baldwin (DH)	ATL	5 / 4	3 / 1	1 / 0	0 / 0	0 / 0	0 / 1	1 / 1	0 / 1	1 / 1	0.92	0.366 elite	projection too low
Matt Olson (1B)	ATL	4 / 4	2 / 1	0 / 0	0 / 0	0 / 1	1 / 1	0 / 1	0 / 1	0 / 1	-0.79	0.354 elite	bad day
Michael Harris II (CF)	ATL	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	0 / 1	-0.78	0.34 elite	bad day
Ozzie Albies (2B)	ATL	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	-0.47	0.281 weak	—
Daulton Varsho (CF)	HOU	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	-0.47	0.284 weak	—
Christian Walker (1B)	HOU	4 / 3	2 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0 / 1	0.47	0.298 below average	—
LaMonte Wade Jr. (LF)	HOU	4 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	-0.33	0.274 weak	—
Jeremy Peña (SS)	HOU	4 / 4	1 / 1	0 / 1	1 / 0	0 / 0	0 / 1	1 / 1	0 / 0	2 / 1	0.31	0.325 above average	—
Yainer Diaz (C)	HOU	4 / 4	1 / 1	1 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0 / 0	0.31	0.301 below average	—
Sean Murphy (C)	ATL	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	2 / 0	0 / 1	0.19	0.252 weak	—
Mauricio Dubón (SS)	ATL	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0	0.308 average	—
Austin Riley (3B)	ATL	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 1	0	0.296 below average	—
Cam Smith (RF)	HOU	4 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	2 / 1	0	0.318 average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-0.896	positive favors home; negative favors away
Recent form	-0.638	positive favors home; negative favors away
Starting-pitcher matchup	-0.012	positive favors home; negative favors away
Composite	-0.858	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.35-4.28	n/a	Base ML historical model prediction
variance_applied	4-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-3	0.562	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	4.35-4.28	4-4	-0.078	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)
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
ATL	Mike Yastrzemski	batter / LF	worse	predicted_AB 2.336
ATL	Sean Murphy	batter / C	worse	predicted_SO 1.815
HOU	Cam Smith	batter / RF	worse	predicted_AB 1.564
HOU	LaMonte Wade Jr.	batter / LF	better	predicted_AB 1.479
HOU	Yainer Diaz	batter / C	worse	predicted_AB 1.381
ATL	Ozzie Albies	batter / 2B	worse	predicted_SO 1.379
HOU	Enyel De Los Santos	pitcher / Middle Reliever (High Leverage)	better	predicted_IP 1.333
ATL	Mauricio Dubón	batter / SS	boosted	predicted_AB 1.29

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	HOU pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	HOU
info	CURRENT_FINALIZED_MATCH_BOX	ATL model-prediction batting totals differ from final simulation box score (AB: current=33, sim=41; 2B: current=1, sim=2; R: current=5, sim=4; SO: current=8, sim=10).	ATL
info	CURRENT_PLAYER_MATCH_BOX	ATL current finalized hitter lines are model predictions and differ from the simulation box score: Drake Baldwin: RBI current=1 box=0, R current=1 box=0; Mauricio Dubón: AB current=4 box=5, 2B current=0 box=1, SO current=0 box=1; Matt Olson: H current=1 box=2, 2B current=0 box=1, HR current=1 box=0, RBI current=1 box=0; Ozzie Albies: AB current=4 box=5, RBI current=0 box=1, SO current=1 box=2; Michael Harris II: AB current=4 box=5, H current=1 box=2, 2B current=1 box=0, HR current=0 box=1; Ronald Acuña Jr.: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, SO current=1 box=0; Mike Yastrzemski: AB current=3 box=5, H current=0 box=1, RBI current=0 box=2, R current=0 box=1; Sean Murphy: AB current=3 box=4, H current=1 box=0, SO current=1 box=3	ATL
info	CURRENT_FINALIZED_MATCH_BOX	HOU model-prediction batting totals differ from final simulation box score (AB: current=33, sim=38; H: current=7, sim=6; 2B: current=1, sim=3; HR: current=1, sim=0; RBI: current=4, sim=3; R: current=4, sim=3; BB: current=3, sim=6; SO: current=7, 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: 2B current=0 box=1, R current=0 box=1; Jose Altuve: H current=1 box=0, R current=1 box=0, BB current=0 box=1; Yordan Alvarez: AB current=4 box=3, HR current=1 box=0, SO current=1 box=0, SF current=0 box=1; Jeremy Peña: AB current=4 box=5, H current=1 box=2, RBI current=1 box=0, R current=1 box=0; Yainer Diaz: AB current=4 box=5, H current=1 box=0; Daulton Varsho: 2B current=0 box=1, BB current=0 box=1; Christian Walker: AB current=3 box=4, H current=1 box=0, RBI current=1 box=0, BB current=0 box=1; Cam Smith: AB current=3 box=5, R current=1 box=0, SO current=1 box=0	HOU
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

BOS@TBHIT

Predicted: TB 4-5 (conf +0.06, lean)Actual: TB 1-2Brier 0.222total bias +6

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 6; total bias +6
pregame home-minus-away evidence: season +0.299, recent +0.410, pitcher +0.052, composite +0.562
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

total runs over-projected by 6 (9 vs 3)
BOS offense over-projected by 3
TB offense over-projected by 3

Team	Starter	Projected	Actual	ER err
BOS	Brayan Bello	2 ER / 13 outs	1 ER / 24 outs	-1
TB	Freddy Peralta	3 ER / 17 outs	1 ER / 17 outs	-2

Decision internals: run factors home×1.073 / away×1.021; 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:

Caleb Durbin (BOS 3B): 2-for-4, 0 HR, 0 RBI (projected 0.0 H, 0.0 HR) [run impact 1.56]
Junior Caminero (TB 3B): 1-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.26]
Jarren Duran (BOS CF): 2-for-3, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact 1.09]
Wilyer Abreu (BOS RF): 0-for-2, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.07]

Starter miss: Freddy Peralta (TB) posted 1.59 ERA vs projected 4.76 (1 ER in 17 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

Brayan Bello (BOS) threw 24 outs, 1 ER = 1.12 ERA against a projected 4.15 ERA (-3.03 ERA error)

Bullpen behind him: 0 ER over 2 outs.

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

Who did the damage to him:

Liam Hicks: 2-for-3, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.318, average - a good day, not a model error
Chandler Simpson: 2-for-4, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.264, weak - a good day, not a model error

Verdict: offense: the lineup scored well below its projection

HOME pitching

Freddy Peralta (TB) threw 17 outs, 1 ER = 1.59 ERA against a projected 4.76 ERA (-3.17 ERA error)

Bullpen behind him: 0 ER over 7 outs.

The AWAY lineup he faced produced 1 runs on 8 hits and 0 HR, against a projected 4 runs / 8 hits (-3 runs vs projection).

Who did the damage to him:

Caleb Durbin: 2-for-4, 0 HR, 0 RBI [+1.56 runs vs projection] season xwOBA 0.294, below average - a good day, not a model error
Jarren Duran: 2-for-3, 0 HR, 0 RBI [+1.09 runs vs projection] season xwOBA 0.296, 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)
BOS	Brayan Bello (Starter)	24 / 13	1.12 / 4.15 (-3.03)	1 / 2	5 / 3	1 / 0	0 / 0	2.43	11 / 6	2.94	venue away: 10.03 vs 7.79 ERA (worse here) · opp run factor 1.07 (average) · 24 prior appearances this season
BOS	Garrett Whitlock (Middle Reliever (High Leverage))	2 / 3	0 / 27 (-27.00)	0 / 3	3 / 3	0 / 0	0 / 1	0.18	1 / 0	1.11
TB	Freddy Peralta (Starter)	17 / 17	1.59 / 4.76 (-3.17)	1 / 3	4 / 5	0 / 0	2 / 1	1.39	6 / 4	4.09	venue home: 3.66 vs 5.81 ERA (ok) · opp run factor 1.02 (average) · 30 prior appearances this season
TB	Bryan Baker (Closer)	3 / 3	0 / 9 (-9.00)	0 / 1	2 / 2	0 / 0	0 / 0	0.34	0 / 1	0.78
TB	Cam Booser (Middle Reliever (High Leverage))	1 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0.13	0 / 0	0.91
TB	Kevin Kelly (Setup Man)	3 / 4	0 / 0	0 / 0	1 / 1	0 / 0	0 / 1	0.12	1 / 0	0.93

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
Caleb Durbin (3B)	BOS	4 / 3	2 / 0	2 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0 / 0	0 / 0	1.56	0.294 below average	good day
Junior Caminero (3B)	TB	4 / 4	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	0 / 1	1 / 1	-1.26	0.364 elite	bad day
Jarren Duran (CF)	BOS	3 / 3	2 / 1	0 / 0	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1.09	0.296 below average	good day
Wilyer Abreu (RF)	BOS	2 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	2 / 1	2 / 1	-1.07	0.35 elite	bad day
Mickey Gasper (DH)	BOS	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	2 / 1	-0.47	0.366 elite	—
Isiah Kiner-Falefa (2B)	BOS	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	-0.47	0.306 average	—
Jake Fraley (DH)	TB	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	-0.47	0.278 weak	—
Liam Hicks (C)	TB	3 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0 / 0	0.47	0.318 average	—
Chandler Simpson (LF)	TB	4 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0.47	0.264 weak	—
Jorge Mateo (SS)	TB	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	-0.47	0.283 weak	—
Nick Sogard (1B)	BOS	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	-0.33	0.311 average	—
Roman Anthony (LF)	BOS	4 / 4	1 / 1	1 / 1	0 / 0	0 / 0	1 / 0	0 / 0	0 / 1	1 / 1	-0.33	0.351 elite	—
Richie Palacios (2B)	TB	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 1	1 / 1	-0.33	0.31 average	—
Jonathan Aranda (1B)	TB	4 / 4	1 / 1	0 / 1	0 / 0	1 / 0	1 / 1	1 / 1	0 / 1	2 / 1	0.29	0.346 elite	—
Adley Rutschman (C)	BOS	4 / 4	1 / 1	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	1 / 1	0	0.33 above average	—
Trevor Story (SS)	BOS	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	1 / 1	0	0.254 weak	—
Cedric Mullins (CF)	TB	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0	0.264 weak	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	0.299	positive favors home; negative favors away
Recent form	0.41	positive favors home; negative favors away
Starting-pitcher matchup	0.052	positive favors home; negative favors away
Composite	0.562	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	3.87-4.85	n/a	Base ML historical model prediction
variance_applied	4-5	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-5	0.529	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	3.87-4.85	4-5	0.016	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
TB	Liam Hicks	batter / C	worse	predicted_RBI 2.71
BOS	Roman Anthony	batter / LF	worse	predicted_RBI 2.22
BOS	Adley Rutschman	batter / C	worse	predicted_AB -1.95
TB	Cedric Mullins	batter / CF	worse	predicted_H 1.84
BOS	Wilyer Abreu	batter / RF	worse	predicted_AB -1.47
TB	Chandler Simpson	batter / LF	worse	predicted_H -1.42
BOS	Trevor Story	batter / SS	worse	predicted_SO -1.328
TB	Jonathan Aranda	batter / 1B	worse	predicted_H -1.19

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	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	BOS model-prediction batting totals differ from final simulation box score (AB: current=32, sim=35; 2B: current=2, sim=3; HR: current=1, sim=0; BB: current=3, sim=2; SO: current=7, sim=5).	BOS
info	CURRENT_PLAYER_MATCH_BOX	BOS current finalized hitter lines are model predictions and differ from the simulation box score: Wilyer Abreu: AB current=4 box=3, H current=1 box=2, HR current=1 box=0, RBI 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; Adley Rutschman: AB current=4 box=3, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Mickey Gasper: AB current=3 box=4, RBI current=1 box=0, R current=0 box=1; Jarren Duran: AB current=3 box=4, H current=1 box=0, SO current=1 box=0; Roman Anthony: AB current=4 box=5, H current=1 box=2, 2B current=1 box=2, RBI current=0 box=3; Trevor Story: RBI current=1 box=0, SO current=1 box=0; Caleb Durbin: AB current=3 box=4, H current=0 box=1, 2B current=0 box=1, RBI current=0 box=1	BOS
info	CURRENT_FINALIZED_MATCH_BOX	TB model-prediction batting totals differ from final simulation box score (AB: current=32, sim=31; H: current=9, sim=7; 3B: current=0, sim=1; HR: current=1, sim=0; RBI: current=4, sim=5; R: current=4, sim=5; BB: current=3, sim=1; SO: current=6, sim=8).	TB
info	CURRENT_PLAYER_MATCH_BOX	TB current finalized hitter lines are model predictions and differ from the simulation box score: Jake Fraley: RBI current=0 box=1, R current=0 box=1, BB current=0 box=1; Liam Hicks: H current=1 box=2, 2B current=0 box=1, RBI current=1 box=3, R current=0 box=1; Junior Caminero: H current=1 box=0, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0; Chandler Simpson: H current=1 box=0, R current=1 box=0; Jonathan Aranda: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, R current=1 box=0; Victor Mesa Jr.: AB current=4 box=3, H current=1 box=0, RBI current=1 box=0, R current=1 box=0; Richie Palacios: 3B current=0 box=1, R current=0 box=1, BB current=1 box=0; Jorge Mateo: R current=0 box=1	TB
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

CHC@CINHIT

Predicted: CHC 7-4 (conf -0.12, moderate)Actual: CHC 5-2Brier 0.194total bias +4

Root cause: Solid correct call

Evidence-backed diagnosis

published system edge 0.12 (moderate); this is not the half-scale distance from 50%
team-run absolute error 4; total bias +4
pregame home-minus-away evidence: season -1.607, recent -0.754, pitcher +0.025, composite -1.558
confidence pipeline reached 0.684 before later publication/recency controls; published home probability was 0.440
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 4 (11 vs 7)
Matthew Boyd (CHC) allowed 3 fewer ER than projected (1 vs 4)

Team	Starter	Projected	Actual	ER err
CHC	Matthew Boyd	4 ER / 18 outs	1 ER / 15 outs	-3
CIN	Nick Lodolo	4 ER / 14 outs	3 ER / 21 outs	-1

Decision internals: run factors home×1.011 / away×0.986; 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:

Dansby Swanson (CHC SS): 2-for-4, 1 HR, 1 RBI (projected 0.0 H, 0.0 HR) [run impact 1.87]
Pete Crow-Armstrong (CHC CF): 0-for-5, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.73]
Sal Stewart (CIN 1B): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.4]
Michael Busch (CHC 1B): 2-for-3, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 1.07]
Nico Hoerner (CHC 2B): 2-for-4, 0 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 0.78]

Starter miss: Matthew Boyd (CHC) posted 1.80 ERA vs projected 6.00 (1 ER in 15 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

Matthew Boyd (CHC) threw 15 outs, 1 ER = 1.8 ERA against a projected 6 ERA (-4.2 ERA error)

Bullpen behind him: 0 ER over 9 outs.

The HOME lineup he faced produced 2 runs on 4 hits and 0 HR, against a projected 4 runs / 7 hits (-2 runs vs projection).

Who did the damage to him:

Ke'Bryan Hayes: 1-for-3, 0 HR, 1 RBI [+0.47 runs vs projection] season xwOBA 0.292, below average - a good day, not a model error

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

HOME pitching

Nick Lodolo (CIN) threw 21 outs, 3 ER = 3.86 ERA against a projected 7.71 ERA (-3.85 ERA error)

Bullpen behind him: 0 ER over 0 outs.

The AWAY lineup he faced produced 4 runs on 8 hits and 2 HR, against a projected 5 runs / 8 hits (-1 runs vs projection).

Who did the damage to him:

Dansby Swanson: 2-for-4, 1 HR, 1 RBI [+1.87 runs vs projection] season xwOBA 0.306, average - a good day, not a model error
Michael Busch: 2-for-3, 1 HR, 1 RBI [+1.07 runs vs projection] season xwOBA 0.33, above average - real projection error
Nico Hoerner: 2-for-4, 0 HR, 1 RBI [+0.78 runs vs projection] season xwOBA 0.321, above average - real projection 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)
CHC	Matthew Boyd (Starter)	15 / 18	1.8 / 6 (-4.20)	1 / 4	2 / 2	0 / 1	2 / 3	2.04	4 / 3	3.91	venue away: 2.95 vs 5.53 ERA (ok) · opp run factor 1.01 (weak) · 20 prior appearances this season
CHC	Daniel Palencia (Middle Reliever (High Leverage))	6 / 3	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0.47	3 / 0	0.81
CHC	Ryan Zeferjahn (Setup Man)	3 / 3	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0.67	2 / 1	1.17
CIN	Nick Lodolo (Starter)	21 / 14	3.86 / 7.71 (-3.85)	3 / 4	6 / 8	1 / 3	2 / 3	2.98	1 / 7	4.62	venue home: 4.58 vs 5.97 ERA (ok) · opp run factor 0.99 (strong) · 19 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
Dansby Swanson (SS)	CHC	4 / 3	2 / 0	0 / 0	0 / 0	1 / 0	1 / 0	2 / 1	0 / 0	0 / 1	1.87	0.306 average	no signal
Pete Crow-Armstrong (CF)	CHC	5 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	1 / 1	-1.73	0.376 elite	bad day
Sal Stewart (1B)	CIN	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 0	2 / 1	-1.4	0.348 elite	bad day
Michael Busch (1B)	CHC	3 / 4	2 / 1	0 / 0	0 / 0	1 / 0	1 / 0	1 / 1	0 / 1	1 / 1	1.07	0.33 above average	projection too low
Nico Hoerner (2B)	CHC	4 / 4	2 / 1	1 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 0	0 / 0	0.78	0.321 above average	projection too low
Dane Myers (CF)	CIN	4 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	-0.64	0.312 average	no signal
Tyrone Taylor (LF)	CHC	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0 / 1	-0.47	0.31 average	—
Alex Bregman (3B)	CHC	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	1 / 1	-0.47	0.322 above average	—
Matt Shaw (RF)	CHC	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0 / 1	-0.47	0.305 average	—
Carson Kelly (C)	CHC	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	0 / 0	-0.47	0.317 average	—
Juan Brito (RF)	CIN	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	1 / 1	-0.47	0.285 below average	—
Ke'Bryan Hayes (3B)	CIN	3 / 3	1 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0 / 1	0.47	0.292 below average	—
Eugenio Suárez (DH)	CIN	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	1 / 1	1 / 1	-0.47	0.298 below average	—
Tyler Stephenson (C)	CIN	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	-0.33	0.331 above average	—
Seiya Suzuki (DH)	CHC	3 / 4	2 / 1	0 / 1	0 / 0	0 / 0	1 / 1	0 / 1	1 / 1	1 / 1	0.16	0.337 above average	—
Elly De La Cruz (SS)	CIN	2 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	2 / 1	1 / 1	-0.14	0.376 elite	—
Matt McLain (2B)	CIN	2 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	2 / 1	0	0.298 below average	—
JJ Bleday (LF)	CIN	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 1	0	0.34 elite	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-1.607	positive favors home; negative favors away
Recent form	-0.754	positive favors home; negative favors away
Starting-pitcher matchup	0.025	positive favors home; negative favors away
Composite	-1.558	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.63-3.83	n/a	Base ML historical model prediction
variance_applied	5-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	7-4	0.56	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by 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	4.63-3.83	5-4	0.197	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-4	7-4	2	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by matchup_recency_guardrail)
embedded_savant_run_environment	7-4	7-4	-0	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	7-4	7-4	-0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	7-4	7-4	-0	Capped at max 12 realistic runs
simulation_result	7-4	7-4	-0	Final inning-by-inning simulation result

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	Nico Hoerner	batter / 2B	better	predicted_RBI 2.52
CIN	Eugenio Suárez	batter / DH	better	predicted_RBI 2.43
CHC	Pete Crow-Armstrong	batter / CF	worse	predicted_AB -2.06
CHC	Matt Shaw	batter / RF	boosted	predicted_RBI 1.75
CHC	Tyrone Taylor	batter / LF	worse	predicted_SO 1.525
CHC	Seiya Suzuki	batter / DH	worse	predicted_BB 1.433
CIN	Matt McLain	batter / 2B	worse	predicted_AB -1.405
CIN	Elly De La Cruz	batter / SS	worse	predicted_H -1.29

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	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_FINALIZED_MATCH_BOX	CHC model-prediction batting totals differ from final simulation box score (AB: current=33, sim=37; H: current=8, sim=10; HR: current=1, sim=3; RBI: current=5, sim=7; R: current=5, sim=7; BB: current=4, sim=5; SO: current=7, sim=11).	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=3, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Tyrone Taylor: AB current=3 box=4, H current=1 box=0, RBI current=1 box=0, BB current=0 box=1; Nico Hoerner: AB current=4 box=5, H current=1 box=2, RBI current=0 box=3, R current=0 box=1; Alex Bregman: H current=1 box=2, HR current=0 box=1, R current=0 box=1; Pete Crow-Armstrong: HR current=1 box=0, RBI current=1 box=0, SO current=1 box=2; Matt Shaw: AB current=3 box=4, HR current=0 box=1, RBI current=0 box=2, R current=0 box=1; Michael Busch: AB current=4 box=5, H current=1 box=2, 2B current=0 box=1, BB current=1 box=0; Carson Kelly: RBI current=1 box=0, SO current=0 box=1	CHC
info	CURRENT_FINALIZED_MATCH_BOX	CIN model pitcher projections differ from final simulation pitching box score (ER: current=6, sim=7).	CIN
info	CURRENT_FINALIZED_MATCH_BOX	CIN model-prediction batting totals differ from final simulation box score (AB: current=32, sim=30; H: current=7, sim=3; BB: current=4, sim=3; SO: current=9, sim=5; HBP: 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: Tyler Stephenson: AB current=4 box=3, H current=1 box=2, 2B current=0 box=1, RBI 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; Juan Brito: AB current=4 box=3, H current=1 box=0, RBI current=1 box=0, BB current=0 box=1; Dane Myers: AB current=4 box=3, H current=1 box=0, 2B current=1 box=0, R current=0 box=1; Elly De La Cruz: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, BB current=1 box=0; Matt McLain: AB current=3 box=2, R current=1 box=0, SO current=1 box=0, HBP current=0 box=1; Eugenio Suárez: AB current=3 box=4, HR current=0 box=1, RBI current=1 box=3, R current=0 box=1; JJ Bleday: AB current=3 box=4, H current=1 box=0	CIN
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

DET@CHWMISS

Predicted: DET 6-4 (conf -0.12, moderate)Actual: CHW 1-3Brier 0.314total bias +6

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

Evidence-backed diagnosis

published system edge 0.12 (moderate); this is not the half-scale distance from 50%
team-run absolute error 6; total bias +6
pregame home-minus-away evidence: season -0.479, recent -0.718, pitcher +0.056, composite -0.681
confidence pipeline reached 0.628 before later publication/recency controls; published home probability was 0.440
score-volatility actions: supported_blowout_margin_expansion
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 (10 vs 4)
DET offense over-projected by 5
committed to DET at 0.12 confidence and it lost

Team	Starter	Projected	Actual	ER err
DET	Jackson Jobe	1 ER / 15 outs	3 ER / 18 outs	2
CHW	Sean Burke	1 ER / 15 outs	1 ER / 18 outs	0

Decision internals: run factors home×1.066 / away×1.01; 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:

Munetaka Murakami (CHW 1B): 1-for-4, 0 HR, 1 RBI (projected 1.0 H, 1.0 HR) [run impact -1.26]
Miguel Vargas (CHW 3B): 1-for-2, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 0.95]
Dillon Dingler (DET C): 1-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -0.93]
Kyle Teel (CHW C): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.8]

Pitcher versus batter: what actually broke this game

AWAY pitching

Jackson Jobe (DET) threw 18 outs, 3 ER = 4.5 ERA against a projected 1.8 ERA (2.7 ERA error) only 7 prior appearances in the covered season - the projection had little of his own record to stand on

Bullpen behind him: 0 ER over 0 outs.

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

Who did the damage to him:

Miguel Vargas: 1-for-2, 1 HR, 1 RBI [+0.95 runs vs projection] season xwOBA 0.378, elite - real projection error
Braden Montgomery: 1-for-3, 0 HR, 0 RBI [+0.45 runs vs projection] season xwOBA 0.324, above 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 Sean Burke (CHW) at a projected 1.8 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 1 runs on 8 hits and 0 HR, against a projected 4 runs / 7 hits (-3 runs vs projection).

Who did the damage to him:

Brett Callahan: 1-for-3, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.297, below average - a good day, not a model error
Hao-Yu Lee: 2-for-5, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.303, below average - a good day, not a model error
Ben Malgeri: 0-for-2, 0 HR, 0 RBI [+0.33 runs vs projection] season xwOBA 0.314, 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)
DET	Jackson Jobe (Starter)	18 / 15	4.5 / 1.8	3 / 1	4 / 3	1 / 0	1 / 2	2.08	9 / 6	5.08	venue away: 3.1 vs 5.02 ERA (ok) · opp run factor 1.07 (average) · 7 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
Munetaka Murakami (1B)	CHW	4 / 4	1 / 1	0 / 0	0 / 0	0 / 1	1 / 0	0 / 1	0 / 1	2 / 1	-1.26	0.341 elite	bad day
Miguel Vargas (3B)	CHW	2 / 4	1 / 1	0 / 1	0 / 0	1 / 0	1 / 1	2 / 1	2 / 1	0 / 1	0.95	0.378 elite	projection too low
Dillon Dingler (C)	DET	4 / 4	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 0	1 / 1	-0.93	0.339 above average	bad day
Kyle Teel (C)	CHW	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	2 / 1	-0.8	0.32 average	no signal
Colt Keith (1B)	DET	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	-0.47	0.335 above average	—
Brett Callahan (LF)	DET	3 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0.47	0.297 below average	—
Hao-Yu Lee (2B)	DET	5 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	3 / 1	0.47	0.303 below average	—
Sam Antonacci (LF)	CHW	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	0 / 0	0 / 1	-0.47	0.342 elite	—
Andrew Benintendi (DH)	CHW	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	3 / 1	-0.47	0.309 average	—
Colson Montgomery (SS)	CHW	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	3 / 1	-0.47	0.302 below average	—
Chase Meidroth (2B)	CHW	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	-0.47	0.295 below average	—
Braden Montgomery (RF)	CHW	3 / 3	1 / 0	1 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	0 / 1	0.45	0.324 above average	—
Kevin McGonigle (3B)	DET	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	0 / 1	0 / 1	-0.33	0.361 elite	—
Ben Malgeri (RF)	DET	2 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	1 / 1	0.33	0.314 average	—
Riley Greene (DH)	DET	3 / 4	2 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	-0.17	0.374 elite	—
John Peck (SS)	DET	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	-0.14	0.35 elite	—
Max Clark (CF)	DET	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	0 / 1	0	0.315 average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-0.479	positive favors home; negative favors away
Recent form	-0.718	positive favors home; negative favors away
Starting-pitcher matchup	0.056	positive favors home; negative favors away
Composite	-0.681	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.32-4.64	n/a	Base ML historical model prediction
variance_applied	4-5	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	6-4	0.56	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by matchup_recency_guardrail)
embedded_savant_run_environment	6-4	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	6-4	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	6-4	n/a	Capped at max 12 realistic runs
simulation_result	6-4	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.32-4.64	4-5	-0.675	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-5	6-4	3	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by 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
DET	Ben Malgeri	batter / RF	better	predicted_RBI 2.668
DET	Colt Keith	batter / 1B	better	predicted_H 2.276
DET	Hao-Yu Lee	batter / 2B	worse	predicted_BB 1.747
CHW	Kyle Teel	batter / C	worse	predicted_BB 1.581
DET	Brett Callahan	batter / LF	worse	predicted_AB 1.5
CHW	Sam Antonacci	batter / LF	worse	predicted_SO 1.386
CHW	Brenton Doyle	batter / CF	deboosted	predicted_SO -1.015
CHW	Andrew Benintendi	batter / DH	worse	predicted_AB 0.985

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	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_FINALIZED_MATCH_BOX	CHW model-prediction batting totals differ from final simulation box score (AB: current=31, sim=33; 2B: current=1, sim=2; HR: current=1, sim=0; SO: current=9, sim=8; SF: current=0, sim=1; HBP: current=0, sim=1).	CHW
info	CURRENT_PLAYER_MATCH_BOX	CHW current finalized hitter lines are model predictions and differ from the simulation box score: Sam Antonacci: H current=1 box=0, R current=1 box=0, BB current=0 box=1, SO current=1 box=2; Munetaka Murakami: AB current=4 box=3, HR current=1 box=0, R current=1 box=0, SO current=1 box=2; Kyle Teel: AB current=4 box=3, RBI current=1 box=0, R current=0 box=1, BB current=1 box=2; Andrew Benintendi: AB current=3 box=4, 2B current=0 box=1, RBI current=1 box=0, SO current=1 box=0; Braden Montgomery: AB current=3 box=4, H current=0 box=1, BB current=1 box=0, SO current=1 box=0; Colson Montgomery: AB current=3 box=4, H current=1 box=0, RBI current=1 box=0, SO current=1 box=2; Miguel Vargas: AB current=4 box=5, RBI current=1 box=2, BB current=1 box=0, SO current=1 box=0; Chase Meidroth: RBI current=0 box=1, SF current=0 box=1	CHW
info	CURRENT_FINALIZED_MATCH_BOX	DET model-prediction batting totals differ from final simulation box score (AB: current=31, sim=37; H: current=7, sim=10; 2B: current=1, sim=4; HR: current=1, sim=2; RBI: current=4, sim=6; R: current=4, sim=6; BB: current=3, sim=6; SO: current=9, sim=5).	DET
info	CURRENT_PLAYER_MATCH_BOX	DET current finalized hitter lines are model predictions and differ from the simulation box score: Kevin McGonigle: RBI current=1 box=0, SO current=1 box=0; Colt Keith: AB current=3 box=5, H current=1 box=3, 2B current=0 box=2, HR current=0 box=1; Riley Greene: H current=1 box=2, 2B current=1 box=0; Ben Malgeri: AB current=3 box=5, H current=0 box=3, 2B current=0 box=2, HR current=0 box=1; Brett Callahan: AB current=3 box=4, BB current=0 box=1, SO current=1 box=0; Hao-Yu Lee: AB current=4 box=3, H current=1 box=0, RBI current=1 box=0, R current=0 box=1; Max Clark: AB current=3 box=5, BB current=1 box=0; Dillon Dingler: H current=1 box=0, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0	DET
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

KC@PITHIT

Predicted: PIT 4-5 (conf +0.08, lean)Actual: PIT 5-6Brier 0.211total bias -2

Root cause: Solid correct call

Evidence-backed diagnosis

published system edge 0.08 (lean); this is not the half-scale distance from 50%
team-run absolute error 2; total bias -2
pregame home-minus-away evidence: season +0.777, recent +0.858, pitcher +0.083, composite +1.005
confidence pipeline reached 0.548 before later publication/recency controls; published home probability was 0.541
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

Noah Cameron (KC) allowed 3 more ER than projected (4 vs 1)

Team	Starter	Projected	Actual	ER err
KC	Noah Cameron	1 ER / 17 outs	4 ER / 18 outs	3
PIT	Bubba Chandler	4 ER / 16 outs	3 ER / 12 outs	-1

Decision internals: run factors home×1.02 / away×0.937; 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:

Nick Loftin (KC 3B): 4-for-5, 0 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 1.41]
Brandon Lowe (PIT 2B): 2-for-4, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.4]
Jac Caglianone (KC DH): 2-for-3, 0 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 1.13]
Bryan Reynolds (PIT LF): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -1.11]
Ronny Simon (PIT RF): 1-for-4, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 0.93]

Starter miss: Noah Cameron (KC) posted 6.00 ERA vs projected 1.59 (4 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

Noah Cameron (KC) threw 18 outs, 4 ER = 6 ERA against a projected 1.59 ERA (4.41 ERA error)

Bullpen behind him: 2 ER over 3 outs.

The HOME lineup he faced produced 5 runs on 9 hits and 3 HR, against a projected 6 runs / 9 hits (-1 runs vs projection).

Who did the damage to him:

Brandon Lowe: 2-for-4, 1 HR, 2 RBI [+1.4 runs vs projection] season xwOBA 0.326, above average - real projection error
Ronny Simon: 1-for-4, 1 HR, 2 RBI [+0.93 runs vs projection] season xwOBA 0.315, average - a good day, not a model error
Jake Mangum: 1-for-3, 1 HR, 2 RBI [+0.93 runs vs projection] season xwOBA 0.288, below average - a good day, not a model error

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

HOME pitching

Bubba Chandler (PIT) threw 12 outs, 3 ER = 6.75 ERA against a projected 6.75 ERA (0 ERA error)

Bullpen behind him: 0 ER over 6 outs.

The AWAY lineup he faced produced 5 runs on 13 hits and 1 HR, against a projected 4 runs / 8 hits (1 runs vs projection).

Who did the damage to him:

Nick Loftin: 4-for-5, 0 HR, 1 RBI [+1.41 runs vs projection] season xwOBA 0.29, below average - a good day, not a model error
Jac Caglianone: 2-for-3, 0 HR, 1 RBI [+1.13 runs vs projection] season xwOBA 0.353, elite - real projection 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)
KC	Noah Cameron (Starter)	18 / 17	6 / 1.59 (+4.41)	4 / 1	7 / 2	2 / 1	2 / 2	1.93	6 / 7	4.5	venue away: 3.46 vs 4.58 ERA (ok) · opp run factor 1.02 (average) · 29 prior appearances this season
KC	Steven Cruz (Closer)	3 / 3	18 / 0 (+18.00)	2 / 0	2 / 0	1 / 0	0 / 0	0.34	1 / 1	1.02
PIT	Bubba Chandler (Starter)	12 / 16	6.75 / 6.75	3 / 4	8 / 7	1 / 0	1 / 1	2.23	6 / 6	4.88	venue home: 4.31 vs 4.34 ERA (ok) · opp run factor 0.94 (weak) · 29 prior appearances this season
PIT	Luke Weaver (Setup Man)	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0.2	1 / 1	1.08
PIT	Mason Montgomery (Closer)	3 / 4	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0.4	1 / 0	1.53

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
Nick Loftin (3B)	KC	5 / 4	4 / 1	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0 / 1	1.41	0.29 below average	good day
Brandon Lowe (2B)	PIT	4 / 4	2 / 1	0 / 0	0 / 0	1 / 0	2 / 1	2 / 1	0 / 0	1 / 1	1.4	0.326 above average	projection too low
Jac Caglianone (DH)	KC	3 / 4	2 / 1	0 / 0	0 / 0	0 / 0	1 / 1	2 / 0	2 / 0	0 / 1	1.13	0.353 elite	projection too low
Bryan Reynolds (LF)	PIT	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	-1.11	0.354 elite	bad day
Ronny Simon (RF)	PIT	4 / 3	1 / 1	0 / 0	0 / 0	1 / 0	2 / 0	1 / 1	0 / 0	1 / 1	0.93	0.315 average	no signal
Jake Mangum (CF)	PIT	3 / 3	1 / 1	0 / 0	0 / 0	1 / 0	2 / 0	1 / 0	0 / 0	0 / 1	0.93	0.288 below average	good day
Michael Massey (2B)	KC	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	1 / 1	0 / 0	0 / 0	1 / 1	-0.78	0.309 average	no signal
Kyle Isbel (CF)	KC	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	-0.47	0.257 weak	—
Ryan O'Hearn (1B)	PIT	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	2 / 1	-0.47	0.318 average	—
Konnor Griffin (SS)	PIT	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 0	0 / 1	-0.45	0.306 average	—
Carter Jensen (C)	KC	5 / 4	1 / 1	0 / 0	0 / 0	1 / 1	1 / 1	1 / 1	0 / 1	3 / 1	-0.33	0.312 average	—
Isaac Collins (LF)	KC	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	3 / 1	-0.33	0.295 below average	—
Nick Gonzales (3B)	PIT	4 / 4	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 1	0.31	0.326 above average	—
Bobby Witt Jr. (SS)	KC	5 / 4	3 / 1	0 / 1	0 / 0	0 / 0	0 / 0	2 / 1	0 / 1	0 / 1	0.3	0.378 elite	—
Rafael Flores Jr. (C)	PIT	3 / 4	2 / 1	1 / 0	0 / 0	0 / 1	0 / 1	0 / 1	1 / 1	0 / 1	-0.15	0.341 elite	—
Vinnie Pasquantino (1B)	KC	4 / 4	2 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	0 / 1	1 / 0	0.14	0.332 above average	—
Oneil Cruz (DH)	PIT	3 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	1 / 1	0.14	0.351 elite	—
John Rave (RF)	KC	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	2 / 1	0	0.318 average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	0.777	positive favors home; negative favors away
Recent form	0.858	positive favors home; negative favors away
Starting-pitcher matchup	0.083	positive favors home; negative favors away
Composite	1.005	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	5.02-4.43	n/a	Base ML historical model prediction
variance_applied	5-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-5	0.541	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run)
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	5.02-4.43	5-4	-0.411	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-4	4-5	2	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run)
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
PIT	Konnor Griffin	batter / SS	worse	predicted_AB -1.88
PIT	Oneil Cruz	batter / DH	worse	predicted_SO 1.5
PIT	Jake Mangum	batter / CF	worse	predicted_AB -1.44
KC	Vinnie Pasquantino	batter / 1B	worse	predicted_RBI 1.39
KC	Kyle Isbel	batter / CF	worse	predicted_H 1.28
PIT	Ryan O'Hearn	batter / 1B	better	predicted_RBI 1.258
PIT	Ronny Simon	batter / RF	worse	predicted_AB -1.23
KC	Carter Jensen	batter / C	worse	predicted_H -1.22

Structural audit findings

Severity	Rule	Finding	Subject
info	CURRENT_PITCHER_DUPLICATION	KC pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	KC
info	CURRENT_PITCHER_DUPLICATION	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	KC model-prediction batting totals differ from final simulation box score (AB: current=33, sim=34; H: current=8, sim=7; HR: current=1, sim=0; BB: current=4, sim=1; SO: current=8, sim=9).	KC
info	CURRENT_PLAYER_MATCH_BOX	KC current finalized hitter lines are model predictions and differ from the simulation box score: Jac Caglianone: H current=1 box=2, RBI current=1 box=0, R current=0 box=1; Vinnie Pasquantino: 2B current=0 box=1, RBI current=1 box=2, BB current=1 box=0, SO current=0 box=1; Michael Massey: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Bobby Witt Jr.: 2B current=1 box=0, RBI current=0 box=1, BB current=1 box=0; Carter Jensen: H current=1 box=0, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0; John Rave: RBI current=0 box=1, BB current=0 box=1; Nick Loftin: H current=1 box=0, SO current=1 box=0; Kyle Isbel: AB current=3 box=4, H current=1 box=2, 2B current=0 box=1, SO current=1 box=2	KC
info	CURRENT_FINALIZED_MATCH_BOX	PIT model-prediction batting totals differ from final simulation box score (AB: current=34, sim=31; H: current=9, sim=5; HR: current=1, sim=3; R: current=6, sim=5).	PIT
info	CURRENT_PLAYER_MATCH_BOX	PIT current finalized hitter lines are model predictions and differ from the simulation box score: Oneil Cruz: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, BB current=1 box=0; Konnor Griffin: AB current=4 box=3, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Rafael Flores Jr.: BB current=1 box=0; Ryan O'Hearn: AB current=4 box=3, HR current=0 box=1, RBI current=1 box=2, R current=0 box=1; Brandon Lowe: HR current=0 box=1; Bryan Reynolds: H current=1 box=0, 2B current=1 box=0, R current=1 box=0, BB current=1 box=0; Ronny Simon: AB current=3 box=2, H current=1 box=0, R current=1 box=0, BB current=0 box=1; Jake Mangum: 2B current=0 box=1, RBI current=0 box=1, R current=0 box=1	PIT
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

MIA@SDHIT

Predicted: SD 5-8 (conf +0.16, moderate)Actual: SD 6-7Brier 0.176total bias +0

Root cause: Strong correct call

Evidence-backed diagnosis

published system edge 0.16 (moderate); this is not the half-scale distance from 50%
team-run absolute error 2; total bias +0
pregame home-minus-away evidence: season +0.570, recent +1.406, pitcher -0.070, composite +1.050
confidence pipeline reached 0.694 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

No large single-factor errors.

Team	Starter	Projected	Actual	ER err
MIA	Eury Pérez	5 ER / 15 outs	5 ER / 14 outs	0
SD	Casey Mize	3 ER / 15 outs	5 ER / 9 outs	2

Decision internals: run factors home×1.006 / away×1.076; 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:

Otto Lopez (MIA SS): 1-for-4, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.28]
Samad Taylor (SD LF): 1-for-2, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.26]
Ty France (SD 1B): 3-for-5, 0 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 1.25]
Jakob Marsee (MIA CF): 3-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact 0.94]
Esteury Ruiz (MIA RF): 1-for-3, 0 HR, 1 RBI (projected 0.0 H, 0.0 HR) [run impact 0.8]

Starter miss: Casey Mize (SD) posted 15.00 ERA vs projected 5.40 (5 ER in 9 outs); he faced a strong offense (opponent run factor 1.08, bucket average).

What to do: Consider raising context.opponent_strength_weight; validate across the rolling window before promotion.

Pitcher versus batter: what actually broke this game

AWAY pitching

Eury Pérez (MIA) threw 14 outs, 5 ER = 9.64 ERA against a projected 9 ERA (0.64 ERA error)

Bullpen behind him: 1 ER over 13 outs.

The HOME lineup he faced produced 7 runs on 10 hits and 2 HR, against a projected 5 runs / 7 hits (2 runs vs projection).

Who did the damage to him:

Samad Taylor: 1-for-2, 1 HR, 2 RBI [+1.26 runs vs projection] season xwOBA 0.307, average - a good day, not a model error
Ty France: 3-for-5, 0 HR, 1 RBI [+1.25 runs vs projection] season xwOBA 0.335, above average - real projection error
Jake Cronenworth: 1-for-4, 0 HR, 1 RBI [+0.47 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

HOME pitching

Casey Mize (SD) threw 9 outs, 5 ER = 15 ERA against a projected 5.4 ERA (9.6 ERA error)

Bullpen behind him: 1 ER over 12 outs.

The AWAY lineup he faced produced 5 runs on 9 hits and 1 HR, against a projected 5 runs / 8 hits (0 runs vs projection).

Who did the damage to him:

Otto Lopez: 1-for-4, 1 HR, 2 RBI [+1.28 runs vs projection] season xwOBA 0.327, above average - real projection error
Jakob Marsee: 3-for-4, 0 HR, 0 RBI [+0.94 runs vs projection] season xwOBA 0.312, average - a good day, not a model error
Esteury Ruiz: 1-for-3, 0 HR, 1 RBI [+0.8 runs vs projection] season xwOBA 0.283, weak - a good day, not a model error

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

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
MIA	Eury Pérez (Starter)	14 / 15	9.64 / 9	5 / 5	8 / 7	2 / 0	3 / 3	2.84	4 / 5	5.05	venue away: 4.54 vs 3.45 ERA (worse here) · opp run factor 1.01 (average) · 27 prior appearances this season
MIA	Calvin Faucher (Closer)	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0.55	1 / 1	0.98
MIA	Michael Petersen (Middle Reliever (High Leverage))	4 / 3	6.75 / 0 (+6.75)	1 / 0	2 / 0	0 / 0	0 / 0	0.38	4 / 2	1.25
MIA	Victor Vodnik (Middle Reliever (High Leverage))	6 / 3	0 / 18 (-18.00)	0 / 2	0 / 3	0 / 0	1 / 0	0.58	1 / 0	0.67
SD	Casey Mize (Starter)	9 / 15	15 / 5.4 (+9.60)	5 / 3	5 / 5	1 / 1	5 / 2	1.26	2 / 2	3.34	venue home: 3.15 vs 4.2 ERA (ok) · opp run factor 1.08 (average) · 24 prior appearances this season
SD	Adrian Morejon (Setup Man)	2 / 3	0 / 0	0 / 0	0 / 0	0 / 0	2 / 0	0.23	2 / 2	1.24
SD	Bradgley Rodriguez (Middle Reliever (High Leverage))	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0.19	2 / 2	1.11
SD	Mason Miller (Closer)	5 / 3	5.4 / 0 (+5.40)	1 / 0	1 / 0	1 / 0	0 / 0	0.4	4 / 1	1.89
SD	Yuki Matsui (Middle Reliever (High Leverage))	2 / 3	0 / 18 (-18.00)	0 / 2	1 / 2	0 / 1	1 / 1	0.53	0 / 1	1.02

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
Otto Lopez (SS)	MIA	4 / 4	1 / 1	0 / 1	0 / 0	1 / 0	2 / 0	1 / 1	2 / 0	1 / 1	1.28	0.327 above average	projection too low
Samad Taylor (LF)	SD	2 / 3	1 / 1	0 / 0	0 / 0	1 / 0	2 / 0	1 / 1	1 / 0	1 / 1	1.26	0.307 average	no signal
Ty France (1B)	SD	5 / 4	3 / 1	1 / 0	0 / 0	0 / 0	1 / 1	1 / 0	0 / 0	0 / 1	1.25	0.335 above average	projection too low
Jakob Marsee (CF)	MIA	4 / 4	3 / 1	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	1 / 1	0 / 1	0.94	0.312 average	no signal
Esteury Ruiz (RF)	MIA	3 / 3	1 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 0	1 / 1	0.8	0.283 weak	good day
Manny Machado (3B)	SD	5 / 4	2 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	0 / 1	1 / 1	-0.79	0.333 above average	bad day
Agustín Ramírez (DH)	MIA	5 / 3	2 / 1	1 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 0	3 / 1	0.78	0.286 below average	good day
Xavier Edwards (2B)	MIA	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	2 / 1	-0.47	0.312 average	—
Joe Mack (C)	MIA	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	1 / 1	1 / 0	0 / 0	0 / 1	-0.47	0.285 below average	—
Jake Cronenworth (SS)	SD	4 / 4	1 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	1 / 1	1 / 1	0.47	0.273 weak	—
Jackson Merrill (CF)	SD	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	1 / 0	3 / 1	0.33	0.351 elite	—
Dustin Harris (RF)	SD	5 / 4	1 / 1	1 / 0	0 / 0	0 / 0	0 / 1	1 / 1	0 / 0	1 / 0	0.31	0.326 above average	—
Fernando Tatis Jr. (2B)	SD	4 / 4	1 / 1	0 / 1	0 / 0	1 / 0	1 / 0	1 / 1	0 / 1	1 / 1	0.29	0.377 elite	—
Heriberto Hernández (LF)	MIA	4 / 4	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	2 / 0	2 / 1	-0.27	0.349 elite	—
Javier Sanoja (3B)	MIA	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	1 / 0	1 / 0	-0.14	0.292 below average	—
Luis Campusano (DH)	SD	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	2 / 1	1 / 1	-0.14	0.328 above average	—
Kyle Stowers (1B)	MIA	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	2 / 1	0	0.325 above average	—
Freddy Fermin (C)	SD	4 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0	0.273 weak	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	0.57	positive favors home; negative favors away
Recent form	1.406	positive favors home; negative favors away
Starting-pitcher matchup	-0.07	positive favors home; negative favors away
Composite	1.05	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.16-4.17	n/a	Base ML historical model prediction
variance_applied	4-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-8	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	4.16-4.17	4-4	-0.005	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-4	5-8	3	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by matchup_recency_guardrail)
embedded_savant_run_environment	5-8	5-8	0	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
SD	Dustin Harris	batter / RF	better	predicted_RBI 5.438
SD	Samad Taylor	batter / LF	better	predicted_H 2.95
MIA	Kyle Stowers	batter / 1B	worse	predicted_RBI 2.49
SD	Manny Machado	batter / 3B	boosted	predicted_SO 2.1
MIA	Heriberto Hernández	batter / LF	worse	predicted_AB -1.83
MIA	Agustín Ramírez	batter / DH	better	predicted_RBI 1.59
SD	Ty France	batter / 1B	worse	predicted_H -1.3
SD	Luis Campusano	batter / DH	worse	predicted_SO 1.227

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	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	MIA model pitcher projections differ from final simulation pitching box score (H: current=10, sim=13; ER: current=7, sim=8).	MIA
info	CURRENT_FINALIZED_MATCH_BOX	MIA 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=5).	MIA
info	CURRENT_PLAYER_MATCH_BOX	MIA current finalized hitter lines are model predictions and differ from the simulation box score: Heriberto Hernández: AB current=4 box=3, HR current=1 box=0, RBI current=1 box=0, BB current=0 box=1; Kyle Stowers: HR current=0 box=1, RBI current=1 box=3, BB current=1 box=0, SO current=1 box=2; Javier Sanoja: 2B current=0 box=1, RBI current=1 box=0; Otto Lopez: 2B current=1 box=0, BB current=0 box=1; Xavier Edwards: BB current=1 box=0, SO current=1 box=0; Agustín Ramírez: AB current=3 box=4, HR current=0 box=1, RBI current=0 box=2, R current=0 box=1; Joe Mack: AB current=3 box=4, H current=1 box=0, RBI current=1 box=0, SO current=1 box=2; Jakob Marsee: 2B current=0 box=1, R current=1 box=0, BB current=1 box=0	MIA
info	CURRENT_FINALIZED_MATCH_BOX	SD model-prediction batting totals differ from final simulation box score (AB: current=34, sim=37; H: current=7, sim=13; 2B: current=1, sim=2; HR: current=1, sim=0; RBI: current=4, sim=8; R: current=5, sim=8; BB: current=4, sim=3).	SD
info	CURRENT_PLAYER_MATCH_BOX	SD current finalized hitter lines are model predictions and differ from the simulation box score: Fernando Tatis Jr.: H current=1 box=0, 2B current=1 box=0, SO current=1 box=0; Dustin Harris: AB current=4 box=5, H current=1 box=5, 2B current=0 box=2, RBI current=1 box=6; Luis Campusano: H current=1 box=0, BB current=1 box=0, SO current=1 box=2; Jake Cronenworth: AB current=4 box=3, R current=0 box=1; Jackson Merrill: RBI current=1 box=0; Ty France: H current=1 box=0, RBI current=1 box=0, R current=0 box=1, BB current=0 box=1; Samad Taylor: AB current=3 box=4, H current=1 box=4, SO current=1 box=0; Manny Machado: AB current=4 box=5, H current=1 box=2, HR current=1 box=0, RBI current=1 box=2	SD
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

MIL@BALMISS

Predicted: BAL 4-5 (conf +0.06, lean)Actual: MIL 1-0Brier 0.281total bias +8

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

Evidence-backed diagnosis

published system edge 0.06 (lean); this is not the half-scale distance from 50%
team-run absolute error 8; total bias +8
pregame home-minus-away evidence: season -1.298, recent -0.966, pitcher +0.104, composite -0.547
confidence pipeline reached 0.546 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 8 (9 vs 1)
MIL offense over-projected by 3
BAL offense over-projected by 5
Robert Gasser (MIL) allowed 3 fewer ER than projected (0 vs 3)

Team	Starter	Projected	Actual	ER err
MIL	Robert Gasser	3 ER / 15 outs	0 ER / 21 outs	-3
BAL	Trevor Rogers	3 ER / 18 outs	1 ER / 24 outs	-2

Decision internals: run factors home×1.052 / away×0.948; 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 Chourio (MIL LF): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.73]
Pete Alonso (BAL 1B): 0-for-3, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.73]
Christian Encarnacion-Strand (BAL 3B): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.78]
William Contreras (MIL DH): 1-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.64]
Rece Hinds (BAL RF): 1-for-3, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact 0.64]

Starter miss: Robert Gasser (MIL) posted 0.00 ERA vs projected 5.40 (0 ER in 21 outs); he is materially worse pitching away (5.44 ERA there vs 3.62 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

Robert Gasser (MIL) threw 21 outs, 0 ER = 0 ERA against a projected 5.4 ERA (-5.4 ERA error)

Bullpen behind him: 0 ER over 6 outs.

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

Who did the damage to him:

Rece Hinds: 1-for-3, 0 HR, 0 RBI [+0.64 runs vs projection] season xwOBA 0.17, weak - a good day, not a model error
Leody Taveras: 1-for-2, 0 HR, 0 RBI [+0.33 runs vs projection] season xwOBA 0.268, weak - a good day, not a model error

Verdict: offense: the lineup scored well below its projection

HOME pitching

Trevor Rogers (BAL) threw 24 outs, 1 ER = 1.12 ERA against a projected 4.5 ERA (-3.38 ERA error)

Bullpen behind him: 0 ER over 0 outs.

The AWAY lineup he faced produced 1 runs on 3 hits and 0 HR, against a projected 5 runs / 9 hits (-4 runs vs projection).

Who did the damage to him:

Joey Ortiz: 1-for-3, 0 HR, 0 RBI [+0.62 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)
MIL	Robert Gasser (Starter)	21 / 15	0 / 5.4 (-5.40)	0 / 3	3 / 6	0 / 1	3 / 2	2.15	2 / 7	4.79	venue away: 5.44 vs 3.62 ERA (worse here) · opp run factor 1.05 (average) · 19 prior appearances this season
MIL	Aaron Ashby (Middle Reliever (High Leverage))	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0.56	1 / 0	1.05
MIL	Abner Uribe (Setup Man)	3 / 3	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0.34	1 / 1	1.09
BAL	Trevor Rogers (Starter)	24 / 18	1.12 / 4.5 (-3.38)	1 / 3	3 / 5	0 / 2	1 / 2	0.9	10 / 9	7.1	venue home: 3.79 vs 4.48 ERA (ok) · opp run factor 0.95 (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 Chourio (LF)	MIL	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	2 / 1	-1.73	0.343 elite	bad day
Pete Alonso (1B)	BAL	3 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	1 / 1	-1.73	0.395 elite	bad day
Christian Encarnacion-Strand (3B)	BAL	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	1 / 1	-0.78	0.338 above average	bad day
William Contreras (DH)	MIL	4 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	-0.64	0.336 above average	bad day
Rece Hinds (RF)	BAL	3 / 4	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 2	0.64	0.17 weak	good day
Joey Ortiz (3B)	MIL	3 / 4	1 / 1	0 / 0	1 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 1	0.62	0.3 below average	good day
Luis Lara (RF)	MIL	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	2 / 1	-0.47	0.297 below average	—
David Hamilton (2B)	MIL	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	-0.47	0.271 weak	—
Andrew Vaughn (1B)	MIL	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	-0.47	0.33 above average	—
Cooper Pratt (SS)	MIL	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	2 / 1	-0.47	0.317 average	—
Coby Mayo (DH)	BAL	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	1 / 1	-0.47	0.333 above average	—
Garrett Mitchell (CF)	MIL	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 0	2 / 1	-0.45	0.355 elite	—
Gary Sánchez (C)	MIL	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 0	0 / 1	0 / 1	-0.33	0.341 elite	—
Leody Taveras (CF)	BAL	2 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	0.33	0.268 weak	—
Gunnar Henderson (SS)	BAL	4 / 4	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	-0.02	0.32 average	—
Jeremiah Jackson (LF)	BAL	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0 / 1	0	0.288 below average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-1.298	positive favors home; negative favors away
Recent form	-0.966	positive favors home; negative favors away
Starting-pitcher matchup	0.104	positive favors home; negative favors away
Composite	-0.547	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	3.98-4.76	n/a	Base ML historical model prediction
variance_applied	4-5	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-5	0.53	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run, matchup_recency_guardrail)
embedded_savant_run_environment	4-5	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	3.98-4.76	4-5	0.219	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
BAL	Rece Hinds	batter / RF	worse	predicted_AB 2.5
MIL	Joey Ortiz	batter / 3B	better	predicted_SO 2.342
MIL	Jackson Chourio	batter / LF	worse	predicted_AB -2
BAL	Carlos Narváez	batter / C	better	predicted_BB 1.786
MIL	Cooper Pratt	batter / SS	better	predicted_RBI 1.6
MIL	Gary Sánchez	batter / C	worse	predicted_AB 1.316
MIL	Garrett Mitchell	batter / CF	worse	predicted_H -1.06
BAL	Leody Taveras	batter / CF	worse	predicted_SO 1.024

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	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	BAL model-prediction batting totals differ from final simulation box score (AB: current=32, sim=34; H: current=8, sim=9; 2B: current=1, sim=4; HR: current=1, sim=2; RBI: current=4, sim=5; R: current=4, sim=5; SO: current=10, sim=9).	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, BB current=1 box=0; Rece Hinds: AB current=4 box=5, H current=1 box=2, 2B current=0 box=2; Jeremiah Jackson: AB current=3 box=4, R current=0 box=1; Pete Alonso: HR current=1 box=0, BB current=1 box=0; Gunnar Henderson: H current=1 box=2, 2B current=0 box=1, R current=1 box=0, BB current=1 box=0; José Barrero: 2B current=0 box=1, RBI current=0 box=1, R current=0 box=1, BB current=0 box=1; Christian Encarnacion-Strand: 2B current=1 box=0, HR current=0 box=1; Carlos Narváez: AB current=3 box=2, H current=0 box=1, HR current=0 box=1, RBI current=0 box=1	BAL
info	CURRENT_FINALIZED_MATCH_BOX	MIL model pitcher projections differ from final simulation pitching box score (H: current=7, sim=9; BB: current=2, sim=3).	MIL
info	CURRENT_FINALIZED_MATCH_BOX	MIL model-prediction batting totals differ from final simulation box score (AB: current=32, sim=33; H: current=9, sim=6; 2B: current=2, sim=1; HR: current=1, sim=3; R: current=5, sim=4; BB: current=3, sim=2; SO: current=8, sim=10).	MIL
info	CURRENT_PLAYER_MATCH_BOX	MIL current finalized hitter lines are model predictions and differ from the simulation box score: William Contreras: H current=1 box=2, HR current=0 box=1, BB current=1 box=0, SO current=1 box=0; Joey Ortiz: HR current=0 box=1, RBI current=0 box=1, R current=0 box=1, SO current=1 box=3; Jackson Chourio: HR current=1 box=0, RBI current=1 box=0, BB current=1 box=0; Gary Sánchez: AB current=3 box=4, RBI current=1 box=0, BB current=1 box=0; Luis Lara: H current=1 box=0, R current=1 box=0; David Hamilton: AB current=3 box=2, H current=1 box=0, BB current=0 box=1; Garrett Mitchell: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, R current=1 box=0; Andrew Vaughn: H current=1 box=0, BB current=0 box=1	MIL
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

MIN@LAAHIT

Predicted: LAA 3-4 (conf +0.11, moderate)Actual: LAA 5-6Brier 0.199total bias -4

Root cause: Solid correct call

Evidence-backed diagnosis

published system edge 0.11 (moderate); this is not the half-scale distance from 50%
team-run absolute error 4; total bias -4
pregame home-minus-away evidence: season +0.022, recent +0.338, pitcher +0.062, composite +0.414
confidence pipeline reached 0.575 before later publication/recency controls; published home probability was 0.554
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 4 (7 vs 11)

Team	Starter	Projected	Actual	ER err
MIN	Joe Ryan	0 ER / 15 outs	2 ER / 15 outs	2
LAA	Reid Detmers	2 ER / 18 outs	4 ER / 15 outs	2

Decision internals: run factors home×1.021 / away×0.959; 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:

Ryan Kreidler (MIN SS): 1-for-4, 1 HR, 1 RBI (projected 0.0 H, 0.0 HR) [run impact 1.73]
Jose Siri (LAA CF): 2-for-4, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.73]
Vaughn Grissom (LAA 2B): 1-for-4, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 1.26]
Ryan Jeffers (MIN C): 0-for-5, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -1.11]
Zach Neto (LAA SS): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.07]

Starter miss: Reid Detmers (LAA) posted 7.20 ERA vs projected 3.00 (4 ER in 15 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

Joe Ryan (MIN) threw 15 outs, 2 ER = 3.6 ERA against a projected 0 ERA (3.6 ERA error)

Bullpen behind him: 3 ER over 10 outs.

The HOME lineup he faced produced 6 runs on 9 hits and 2 HR, against a projected 4 runs / 7 hits (2 runs vs projection).

Who did the damage to him:

Jose Siri: 2-for-4, 1 HR, 2 RBI [+1.73 runs vs projection] season xwOBA 0.299, below average - a good day, not a model error
Vaughn Grissom: 1-for-4, 1 HR, 1 RBI [+1.26 runs vs projection] season xwOBA 0.305, average - a good day, not a model error
Josh Lowe: 1-for-5, 0 HR, 0 RBI [+0.78 runs vs projection] season xwOBA 0.287, below average - a good day, not a model error

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

HOME pitching

Reid Detmers (LAA) threw 15 outs, 4 ER = 7.2 ERA against a projected 3 ERA (4.2 ERA error)

Bullpen behind him: 1 ER over 9 outs.

The AWAY lineup he faced produced 4 runs on 11 hits and 1 HR, against a projected 4 runs / 7 hits (0 runs vs projection).

Who did the damage to him:

Ryan Kreidler: 1-for-4, 1 HR, 1 RBI [+1.73 runs vs projection] season xwOBA 0.288, below average - a good day, not a model error
Victor Caratini: 2-for-5, 0 HR, 2 RBI [+0.94 runs vs projection] season xwOBA 0.313, average - a good day, not a model error
Walker Jenkins: 2-for-4, 0 HR, 0 RBI [+0.8 runs vs projection] season xwOBA 0.301, below 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)
MIN	Joe Ryan (Starter)	15 / 15	3.6 / 0 (+3.60)	2 / 0	7 / 4	1 / 0	1 / 0	1	5 / 4	4.11	venue away: 3.92 vs 3.76 ERA (ok) · opp run factor 1.02 (weak) · 25 prior appearances this season
MIN	A.J. Minter (Middle Reliever (High Leverage))	3 / 3	9 / 0 (+9.00)	1 / 0	1 / 0	1 / 0	0 / 1	0.24	0 / 0	0.97
MIN	Jeff Hoffman (Setup Man)	3 / 3	0 / 27 (-27.00)	0 / 3	0 / 2	0 / 1	1 / 0	0.46	1 / 1	1.48
MIN	Yoendrys Gómez (Closer)	1 / 3	54 / 9 (+45.00)	2 / 1	2 / 2	0 / 0	2 / 0	0.48	1 / 2	0.64
MIN	Andrew Morris (Middle Reliever (High Leverage))	3 / 0	0 / n/a	0 / 0	0 / 0	0 / 0	0 / 0	0.36	1 / 0	0.86
LAA	Reid Detmers (Starter)	15 / 18	7.2 / 3 (+4.20)	4 / 2	6 / 4	1 / 2	0 / 0	0.91	5 / 9	7.17	venue home: 3.29 vs 3.47 ERA (ok) · opp run factor 0.96 (average) · 30 prior appearances this season
LAA	Ben Joyce (Closer)	3 / 5	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0.63	0 / 2	1.33
LAA	Ryan Watson (Setup Man)	3 / 4	9 / 6.75	1 / 1	3 / 2	0 / 0	1 / 1	0.2	3 / 1	0.99
LAA	Tayler Saucedo (Middle Reliever (High Leverage))	3 / 0	0 / n/a	0 / 0	0 / 0	0 / 0	0 / 0	0.19	2 / 0	0.84

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
Ryan Kreidler (SS)	MIN	4 / 3	1 / 0	0 / 0	0 / 0	1 / 0	1 / 0	1 / 0	1 / 0	2 / 1	1.73	0.288 below average	good day
Jose Siri (CF)	LAA	4 / 3	2 / 1	0 / 0	0 / 0	1 / 0	2 / 0	1 / 0	1 / 0	2 / 2	1.73	0.299 below average	good day
Vaughn Grissom (2B)	LAA	4 / 4	1 / 1	0 / 0	0 / 0	1 / 0	1 / 1	1 / 0	1 / 0	1 / 1	1.26	0.305 average	no signal
Ryan Jeffers (C)	MIN	5 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	-1.11	0.351 elite	bad day
Zach Neto (SS)	LAA	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	1 / 0	3 / 2	-1.07	0.312 average	no signal
Victor Caratini (1B)	MIN	5 / 3	2 / 0	0 / 0	0 / 0	0 / 0	2 / 0	0 / 0	0 / 0	3 / 1	0.94	0.313 average	no signal
Austin Martin (LF)	MIN	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 1	2 / 1	-0.8	0.307 average	no signal
Walker Jenkins (CF)	MIN	4 / 3	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 1	0.8	0.301 below average	good day
Royce Lewis (2B)	MIN	5 / 4	2 / 1	1 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	1 / 1	0.78	0.307 average	no signal
Josh Lowe (RF)	LAA	5 / 3	1 / 0	1 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	3 / 1	0.78	0.287 below average	good day
Brooks Lee (3B)	MIN	5 / 4	1 / 1	1 / 0	0 / 0	0 / 1	0 / 1	1 / 1	0 / 0	1 / 1	-0.62	0.276 weak	projection too high
Mike Trout (DH)	LAA	4 / 4	2 / 1	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	2 / 1	0 / 1	0.49	0.377 elite	—
Josh Bell (DH)	MIN	4 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	1 / 1	0.47	0.333 above average	—
Moisés Ballesteros (1B)	LAA	6 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	-0.47	0.301 below average	—
Tyler Heineman (C)	LAA	3 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0.47	0.236 weak	—
Wade Meckler (LF)	LAA	5 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	-0.33	0.31 average	—
Denzer Guzman (3B)	LAA	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	2 / 1	1 / 0	0 / 1	0.33	0.267 weak	—
Luke Keaschall (RF)	MIN	5 / 4	1 / 1	1 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0 / 1	1 / 1	-0.02	0.318 average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	0.022	positive favors home; negative favors away
Recent form	0.338	positive favors home; negative favors away
Starting-pitcher matchup	0.062	positive favors home; negative favors away
Composite	0.414	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	3.79-4.81	n/a	Base ML historical model prediction
variance_applied	4-5	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	3-4	0.554	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	3.79-4.81	4-5	-0.024	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
MIN	Ryan Jeffers	batter / C	worse	predicted_2B 1.783
LAA	Wade Meckler	batter / LF	better	predicted_RBI 1.679
LAA	Zach Neto	batter / SS	better	predicted_RBI 1.52
MIN	Luke Keaschall	batter / RF	worse	predicted_AB 1.514
LAA	Jose Siri	batter / CF	better	predicted_AB 1.476
LAA	Josh Lowe	batter / RF	better	predicted_H 1.39
LAA	Mike Trout	batter / DH	worse	predicted_AB 1.378
LAA	Tyler Heineman	batter / C	worse	predicted_AB 1.302

Structural audit findings

Severity	Rule	Finding	Subject
info	CURRENT_PITCHER_DUPLICATION	LAA pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	LAA
info	CURRENT_PITCHER_DUPLICATION	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	LAA model-prediction batting totals differ from final simulation box score (AB: current=33, sim=39; H: current=7, sim=9; 2B: current=1, sim=2; RBI: current=3, sim=4; BB: current=2, sim=1; SO: current=11, sim=7).	LAA
info	CURRENT_PLAYER_MATCH_BOX	LAA current finalized hitter lines are model predictions and differ from the simulation box score: Mike Trout: AB current=4 box=5, 2B current=1 box=0, R current=1 box=0, BB current=1 box=0; Wade Meckler: H current=1 box=2, 2B current=0 box=1, RBI current=0 box=2, SO current=1 box=0; Moisés Ballesteros: SO current=1 box=0; Josh Lowe: AB current=3 box=4, H current=0 box=2, R current=0 box=1; Jose Siri: AB current=3 box=4, H current=1 box=2, 2B current=0 box=1, R current=0 box=1; Denzer Guzman: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, SO current=1 box=2; Zach Neto: AB current=4 box=5, RBI current=1 box=2; Tyler Heineman: AB current=3 box=4, SO current=1 box=0	LAA
info	CURRENT_FINALIZED_MATCH_BOX	MIN model pitcher projections differ from final simulation pitching box score (H: current=8, sim=9).	MIN
info	CURRENT_FINALIZED_MATCH_BOX	MIN model-prediction batting totals differ from final simulation box score (AB: current=33, sim=36; H: current=7, sim=6; 2B: current=1, sim=3; HR: current=1, sim=2; RBI: current=4, sim=3; R: current=4, sim=3; BB: current=3, sim=1; SO: current=9, sim=12).	MIN
info	CURRENT_PLAYER_MATCH_BOX	MIN current finalized hitter lines are model predictions and differ from the simulation box score: Luke Keaschall: AB current=4 box=5, R current=1 box=0, BB current=1 box=0; Austin Martin: AB current=4 box=3, H current=1 box=0; Royce Lewis: H current=1 box=0, RBI current=1 box=0, SO current=1 box=2; Victor Caratini: AB current=3 box=4, SO current=1 box=2; Brooks Lee: 2B current=0 box=1, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0; Walker Jenkins: AB current=3 box=4, HR current=0 box=1, RBI current=0 box=1, R current=0 box=1; Josh Bell: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, SO current=1 box=2; Ryan Jeffers: H current=1 box=2, 2B current=1 box=2, BB current=1 box=0	MIN
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

NYY@ARIMISS

Predicted: NYY 5-2 (conf -0.42, strong)Actual: ARI 3-5Brier 0.506total bias -1

Root cause: Backed the wrong side with real conviction

Evidence-backed diagnosis

published system edge 0.42 (strong); this is not the half-scale distance from 50%
team-run absolute error 5; total bias -1
pregame home-minus-away evidence: season -0.995, recent -1.198, pitcher +0.034, composite -1.312
confidence pipeline reached 0.711 before later publication/recency controls; published home probability was 0.289
score-volatility actions: supported_blowout_margin_expansion
forecast-file integrity: clean (0 defects, 7 informational note(s)); the saved artifact is not why the pick missed

Quantified contributing factors

ARI offense under-projected by 3
committed to NYY at 0.42 confidence and it lost

Team	Starter	Projected	Actual	ER err
NYY	Cam Schlittler	1 ER / 19 outs	1 ER / 18 outs	0
ARI	Brandon Pfaadt	2 ER / 19 outs	1 ER / 13 outs	-1

Decision internals: run factors home×0.983 / away×0.949; 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:

Austin Wells (NYY C): 2-for-3, 1 HR, 1 RBI (projected 0.0 H, 0.0 HR) [run impact 1.87]
Luis García Jr. (NYY 1B): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.4]
Cody Bellinger (NYY LF): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -1.11]
Jordan Lawlar (ARI CF): 1-for-4, 0 HR, 2 RBI (projected 0.0 H, 0.0 HR) [run impact 1.09]
Heliot Ramos (NYY RF): 1-for-4, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 0.93]

Pitcher versus batter: what actually broke this game

AWAY pitching

Cam Schlittler (NYY) threw 18 outs, 1 ER = 1.5 ERA against a projected 1.42 ERA (0.08 ERA error)

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 4 runs / 7 hits (-1 runs vs projection).

Who did the damage to him:

Jordan Lawlar: 1-for-4, 0 HR, 2 RBI [+1.09 runs vs projection] season xwOBA 0.32, 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 Brandon Pfaadt (ARI) at a projected 2.84 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 3 runs on 6 hits and 2 HR, against a projected 6 runs / 8 hits (-3 runs vs projection).

Who did the damage to him:

Austin Wells: 2-for-3, 1 HR, 1 RBI [+1.87 runs vs projection] season xwOBA 0.289, below average - a good day, not a model error
Heliot Ramos: 1-for-4, 1 HR, 1 RBI [+0.93 runs vs projection] season xwOBA 0.328, above average - real projection error
José Caballero: 2-for-2, 0 HR, 0 RBI [+0.8 runs vs projection] season xwOBA 0.28, 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)
NYY	Cam Schlittler (Starter)	18 / 19	1.5 / 1.42	1 / 1	3 / 1	0 / 0	3 / 1	1.71	6 / 8	7.99	venue away: 1.14 vs 2.98 ERA (ok) · opp run factor 0.98 (average) · 31 prior appearances this season
NYY	Brent Headrick (Setup Man)	2 / 3	0 / 0	0 / 0	1 / 1	0 / 0	0 / 1	0.38	1 / 0	1.28
NYY	Paul Blackburn (Middle Reliever (High Leverage))	4 / 2	0 / 0	0 / 0	2 / 0	0 / 0	0 / 0	0.48	2 / 0	0.79

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
Austin Wells (C)	NYY	3 / 3	2 / 0	0 / 0	0 / 0	1 / 0	1 / 0	1 / 0	1 / 1	0 / 1	1.87	0.289 below average	good day
Luis García Jr. (1B)	NYY	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 0	2 / 0	-1.4	0.348 elite	bad day
Cody Bellinger (LF)	NYY	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 0	-1.11	0.341 elite	bad day
Jordan Lawlar (CF)	ARI	4 / 3	1 / 0	0 / 0	1 / 0	0 / 0	2 / 0	0 / 0	0 / 0	1 / 1	1.09	0.32 average	no signal
Heliot Ramos (RF)	NYY	4 / 3	1 / 1	0 / 0	0 / 0	1 / 0	1 / 0	1 / 1	0 / 0	1 / 1	0.93	0.328 above average	projection too low
Ben Rice (DH)	NYY	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	0 / 1	1 / 1	-0.8	0.378 elite	bad day
José Caballero (CF)	NYY	2 / 3	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	0 / 1	0.8	0.28 weak	good day
Lars Nootbaar (LF)	ARI	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	-0.8	0.343 elite	bad day
George Lombard Jr. (SS)	NYY	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	1 / 1	-0.47	0.289 below average	—
Corbin Carroll (RF)	ARI	2 / 4	0 / 1	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	1 / 1	2 / 1	-0.47	0.344 elite	—
Tim Tawa (1B)	ARI	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	0 / 0	2 / 1	-0.47	0.312 average	—
Nolan Arenado (3B)	ARI	4 / 4	2 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	0 / 0	1 / 1	-0.46	0.293 below average	—
Ketel Marte (2B)	ARI	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	1 / 1	1 / 0	1 / 1	-0.45	0.345 elite	—
Geraldo Perdomo (SS)	ARI	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	-0.33	0.329 above average	—
Ryan McMahon (3B)	NYY	1 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	-0.14	0.294 below average	—
Gabriel Moreno (C)	ARI	3 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 0	0 / 1	0.02	0.362 elite	—
Anthony Volpe (2B)	NYY	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 0	0	0.296 below average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-0.995	positive favors home; negative favors away
Recent form	-1.198	positive favors home; negative favors away
Starting-pitcher matchup	0.034	positive favors home; negative favors away
Composite	-1.312	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.9-3.18	n/a	Base ML historical model prediction
variance_applied	5-3	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-2	0.711	Enhanced ML with recent form, H2H, low-score analysis
embedded_savant_run_environment	5-2	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	5-2	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	5-2	n/a	Capped at max 12 realistic runs
simulation_result	5-2	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.9-3.18	5-3	0.28	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-3	5-2	1	Enhanced ML with recent form, H2H, low-score analysis
embedded_savant_run_environment	5-2	5-2	-0	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	5-2	5-2	-0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	5-2	5-2	-0	Capped at max 12 realistic runs
simulation_result	5-2	5-2	-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
NYY	Cam Schlittler	pitcher / Starter	better	predicted_SO 3
NYY	Cody Bellinger	batter / LF	better	predicted_RBI 2.28
NYY	George Lombard Jr.	batter / SS	worse	predicted_AB -2.21
ARI	Corbin Carroll	batter / RF	worse	predicted_SO 2.027
NYY	Paul Blackburn	pitcher / Middle Reliever (High Leverage)	worse	predicted_SO -2
ARI	Jesús Sánchez	batter / DH	worse	predicted_RBI 1.639
ARI	Lars Nootbaar	batter / LF	worse	predicted_BB 1.5
ARI	Geraldo Perdomo	batter / SS	worse	predicted_SO 1.459

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	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	ARI model-prediction batting totals differ from final simulation box score (AB: current=33, sim=30; H: current=7, sim=3; 2B: current=2, sim=0; HR: current=1, sim=0; RBI: current=4, sim=2; R: current=4, sim=2; BB: current=3, sim=2; SO: current=8, sim=9).	ARI
info	CURRENT_PLAYER_MATCH_BOX	ARI current finalized hitter lines are model predictions and differ from the simulation box score: Lars Nootbaar: AB current=4 box=2, H current=1 box=0, BB current=1 box=2; Gabriel Moreno: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, R current=1 box=0; Geraldo Perdomo: H current=1 box=0, BB current=1 box=0, SO current=0 box=2; Corbin Carroll: H current=1 box=0, R current=1 box=0, BB current=1 box=0, SO current=1 box=3; Jesús Sánchez: H current=0 box=2, RBI current=0 box=2, R current=0 box=1, SO current=1 box=0; Ketel Marte: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, R current=1 box=0; Nolan Arenado: AB current=4 box=3, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0; Tim Tawa: H current=1 box=0, RBI current=1 box=0	ARI
info	CURRENT_FINALIZED_MATCH_BOX	NYY model-prediction batting totals differ from final simulation box score (AB: current=31, sim=33; H: current=8, sim=6; RBI: current=4, sim=5; R: current=6, sim=5; BB: current=3, sim=2; SO: current=6, sim=7; HBP: current=0, sim=1).	NYY
info	CURRENT_PLAYER_MATCH_BOX	NYY current finalized hitter lines are model predictions and differ from the simulation box score: Heliot Ramos: AB current=3 box=4, H current=1 box=0, R current=1 box=0, SO current=1 box=2; Ben Rice: H current=1 box=0, BB current=1 box=0; George Lombard Jr.: AB current=4 box=2, H current=1 box=0, RBI current=1 box=0, R current=1 box=0; Anthony Volpe: AB current=3 box=4, R current=0 box=1; Cody Bellinger: H current=1 box=2, 2B current=1 box=0, HR current=0 box=1, RBI current=1 box=3; José Caballero: AB current=3 box=4, RBI current=0 box=1, SO current=1 box=0; Luis García Jr.: H current=1 box=0, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0; Austin Wells: AB current=3 box=4, H current=0 box=1, R current=0 box=1, BB current=1 box=0	NYY
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

OAK@CLEHIT

Predicted: CLE 4-6 (conf +0.16, moderate)Actual: CLE 6-12Brier 0.176total bias -8

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 8; total bias -8
pregame home-minus-away evidence: season +1.262, recent +0.911, pitcher +0.067, composite +1.429
confidence pipeline reached 0.669 before later publication/recency controls; published home probability was 0.580
score-volatility actions: supported_blowout_margin_expansion
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 8 (10 vs 18)
CLE offense under-projected by 6
Jacob Lopez (OAK) allowed 6 more ER than projected (10 vs 4)
Tanner Bibee (CLE) allowed 5 more ER than projected (6 vs 1)

Team	Starter	Projected	Actual	ER err
OAK	Jacob Lopez	4 ER / 14 outs	10 ER / 8 outs	6
CLE	Tanner Bibee	1 ER / 17 outs	6 ER / 14 outs	5

Decision internals: run factors home×1.081 / away×1.014; 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): 4-for-5, 2 HR, 4 RBI (projected 1.0 H, 1.0 HR) [run impact 2.65]
Alika Williams (OAK SS): 2-for-4, 0 HR, 0 RBI (projected 0.0 H, 0.0 HR) [run impact 1.56]
Donovan Walton (OAK 2B): 3-for-3, 1 HR, 4 RBI (projected 1.0 H, 0.0 HR) [run impact 1.54]
Jeff McNeil (OAK 1B): 0-for-5, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.78]
Chase DeLauter (CLE DH): 2-for-5, 1 HR, 3 RBI (projected 1.0 H, 0.0 HR) [run impact 0.76]

Starter miss: Jacob Lopez (OAK) posted 33.75 ERA vs projected 7.71 (10 ER in 8 outs); he faced a strong offense (opponent run factor 1.08, bucket weak).

What to do: Consider raising context.opponent_strength_weight; validate across the rolling window before promotion.

Pitcher versus batter: what actually broke this game

AWAY pitching

Jacob Lopez (OAK) threw 8 outs, 10 ER = 33.75 ERA against a projected 7.71 ERA (26.04 ERA error)

Bullpen behind him: 1 ER over 7 outs.

The HOME lineup he faced produced 12 runs on 15 hits and 3 HR, against a projected 5 runs / 8 hits (7 runs vs projection).

Who did the damage to him:

Angel Martínez: 4-for-5, 2 HR, 4 RBI [+2.65 runs vs projection] season xwOBA 0.304, average - a good day, not a model error
Chase DeLauter: 2-for-5, 1 HR, 3 RBI [+0.76 runs vs projection] season xwOBA 0.344, elite - real projection error
Jo Adell: 2-for-5, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.327, above average - a good day, not a model error

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

HOME pitching

Tanner Bibee (CLE) threw 14 outs, 6 ER = 11.57 ERA against a projected 1.59 ERA (9.98 ERA error)

Bullpen behind him: 0 ER over 0 outs.

The AWAY lineup he faced produced 6 runs on 12 hits and 1 HR, against a projected 4 runs / 8 hits (2 runs vs projection).

Who did the damage to him:

Alika Williams: 2-for-4, 0 HR, 0 RBI [+1.56 runs vs projection] season xwOBA 0.241, weak - a good day, not a model error
Donovan Walton: 3-for-3, 1 HR, 4 RBI [+1.54 runs vs projection] season xwOBA 0.329, above average - real projection error
Zack Gelof: 2-for-5, 0 HR, 1 RBI [+0.47 runs vs projection] season xwOBA 0.314, 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)
OAK	Jacob Lopez (Starter)	8 / 14	33.75 / 7.71 (+26.04)	10 / 4	10 / 8	2 / 2	0 / 1	1.96	1 / 4	4.69	venue away: 4.02 vs 5.64 ERA (ok) · opp run factor 1.08 (weak) · 25 prior appearances this season
OAK	Chris Roycroft (Reliever)	7 / 3	3.86 / 9 (-5.14)	1 / 1	2 / 1	1 / 0	1 / 0	n/a	1 / 2	n/a
CLE	Tanner Bibee (Starter)	14 / 17	11.57 / 1.59 (+9.98)	6 / 1	12 / 2	1 / 1	0 / 0	2.06	2 / 5	3.96	venue home: 4.11 vs 4.22 ERA (ok) · opp run factor 1.01 (weak) · 31 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
Angel Martínez (LF)	CLE	5 / 4	4 / 1	1 / 0	0 / 0	2 / 1	4 / 1	3 / 1	0 / 0	0 / 1	2.65	0.304 average	no signal
Alika Williams (SS)	OAK	4 / 3	2 / 0	2 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 1	1.56	0.241 weak	good day
Donovan Walton (2B)	OAK	3 / 3	3 / 1	0 / 0	0 / 0	1 / 0	4 / 0	1 / 0	0 / 1	0 / 1	1.54	0.329 above average	projection too low
Jeff McNeil (1B)	OAK	5 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	1 / 0	-0.78	0.301 below average	projection too high
Chase DeLauter (DH)	CLE	5 / 4	2 / 1	0 / 1	0 / 0	1 / 0	3 / 1	2 / 0	0 / 1	1 / 1	0.76	0.344 elite	projection too low
Brian Serven (C)	OAK	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	1 / 1	-0.47	0.29 below average	—
Zack Gelof (3B)	OAK	5 / 4	2 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0 / 0	1 / 1	0.47	0.314 average	—
Jo Adell (RF)	CLE	5 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	0 / 0	1 / 1	0.47	0.327 above average	—
Travis Bazzana (2B)	CLE	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	-0.47	0.305 average	—
David Fry (1B)	CLE	4 / 3	1 / 0	1 / 0	0 / 0	0 / 0	2 / 0	1 / 0	0 / 1	1 / 1	0.45	0.265 weak	—
Henry Bolte (CF)	OAK	5 / 4	1 / 1	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	0 / 1	0 / 1	-0.33	0.318 average	—
Carlos Cortes (LF)	OAK	3 / 3	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0.31	0.302 below average	—
Austin Hedges (C)	CLE	4 / 3	1 / 1	1 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 0	1 / 1	0.31	0.276 weak	—
Brayan Rocchio (SS)	CLE	3 / 3	1 / 1	1 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 0	0 / 0	0.31	0.295 below average	—
José Ramírez (3B)	CLE	4 / 4	2 / 1	0 / 1	0 / 0	0 / 0	0 / 1	2 / 1	1 / 1	0 / 1	0.16	0.328 above average	—
Steven Kwan (CF)	CLE	5 / 4	2 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0 / 1	0 / 0	0.14	0.324 above average	—
Shea Langeliers (DH)	OAK	4 / 4	2 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	1 / 0	1 / 1	-0.13	0.361 elite	—
Lawrence Butler (RF)	OAK	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	2 / 1	1 / 1	0 / 1	0	0.324 above average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	1.262	positive favors home; negative favors away
Recent form	0.911	positive favors home; negative favors away
Starting-pitcher matchup	0.067	positive favors home; negative favors away
Composite	1.429	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.3-4.55	n/a	Base ML historical model prediction
variance_applied	4-5	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-6	0.58	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	4.3-4.55	4-5	0.759	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-5	4-6	1	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by matchup_recency_guardrail)
embedded_savant_run_environment	4-6	4-6	0	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
OAK	Jacob Lopez	pitcher / Starter	worse	predicted_IP -3.8
OAK	Seth Johnson	pitcher / Reliever	boosted	predicted_H 2
OAK	Chris Roycroft	pitcher / Reliever	boosted	predicted_SO 2
CLE	Steven Kwan	batter / CF	better	predicted_H 2
CLE	Austin Hedges	batter / C	better	predicted_RBI 1.76
CLE	Angel Martínez	batter / LF	better	predicted_RBI 1.48
CLE	Brayan Rocchio	batter / SS	worse	predicted_SO 1.434
CLE	José Ramírez	batter / 3B	worse	predicted_SO 1.421

Structural audit findings

Severity	Rule	Finding	Subject
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_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_FINALIZED_MATCH_BOX	CLE model-prediction batting totals differ from final simulation box score (AB: current=32, sim=35; H: current=8, sim=11; 2B: current=2, sim=1; HR: current=1, sim=2; RBI: current=4, sim=6; R: current=5, sim=6; BB: current=4, 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=3, BB current=1 box=0; Chase DeLauter: H current=1 box=2, 2B current=1 box=0, RBI current=1 box=0, R current=0 box=1; David Fry: H current=0 box=1; Jo Adell: 2B current=0 box=1, SO current=1 box=0; José Ramírez: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, R current=1 box=0; Travis Bazzana: AB current=3 box=4, H current=1 box=2, RBI current=0 box=1, SO current=1 box=0; Angel Martínez: RBI current=1 box=2; Austin Hedges: AB current=3 box=4, HR current=0 box=1, RBI current=0 box=2, R current=0 box=1	CLE
info	CURRENT_FINALIZED_MATCH_BOX	OAK model-prediction batting totals differ from final simulation box score (H: current=8, sim=5; 2B: current=1, sim=2; BB: current=3, sim=1; SO: current=8, sim=9).	OAK
info	CURRENT_PLAYER_MATCH_BOX	OAK current finalized hitter lines are model predictions and differ from the simulation box score: Henry Bolte: H current=1 box=0, R current=1 box=0, BB current=1 box=0, SO current=1 box=2; Jeff McNeil: AB current=4 box=3, RBI current=1 box=0, R current=0 box=1, BB current=0 box=1; Shea Langeliers: HR current=1 box=0, SO current=1 box=2; Carlos Cortes: H current=1 box=0; Donovan Walton: AB current=3 box=4, H current=1 box=0, BB current=1 box=0; Alika Williams: H current=0 box=1; Brian Serven: HR current=0 box=1, R current=0 box=1; Lawrence Butler: 2B current=0 box=1, RBI current=1 box=2, BB current=1 box=0, SO current=1 box=0	OAK
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

PHI@NYMHIT

Predicted: NYM 5-6 (conf +0.06, lean)Actual: NYM 3-10Brier 0.222total bias -2

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 6; total bias -2
pregame home-minus-away evidence: season -0.382, recent +0.142, pitcher +0.044, composite +0.102
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

NYM offense under-projected by 4

Team	Starter	Projected	Actual	ER err
PHI	Andrew Painter	2 ER / 16 outs	4 ER / 16 outs	2
NYM	Christian Scott	3 ER / 15 outs	3 ER / 14 outs	0

Decision internals: run factors home×1.008 / away×0.964; 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:

Carson Benge (NYM RF): 3-for-4, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 2.2]
Francisco Alvarez (NYM C): 2-for-4, 1 HR, 3 RBI (projected 1.0 H, 0.0 HR) [run impact 1.4]
Brett Baty (NYM 1B): 2-for-3, 0 HR, 1 RBI (projected 0.0 H, 0.0 HR) [run impact 1.27]
Juan Soto (NYM LF): 1-for-5, 0 HR, 1 RBI (projected 1.0 H, 1.0 HR) [run impact -1.26]
Alec Bohm (PHI 1B): 2-for-3, 0 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 0.8]

Starter miss: Andrew Painter (PHI) posted 6.75 ERA vs projected 3.38 (4 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

Andrew Painter (PHI) threw 16 outs, 4 ER = 6.75 ERA against a projected 3.38 ERA (3.37 ERA error)

Bullpen behind him: 0 ER over 2 outs.

The HOME lineup he faced produced 9 runs on 11 hits and 2 HR, against a projected 4 runs / 8 hits (5 runs vs projection).

Who did the damage to him:

Carson Benge: 3-for-4, 1 HR, 2 RBI [+2.2 runs vs projection] season xwOBA 0.338, above average - real projection error
Francisco Alvarez: 2-for-4, 1 HR, 3 RBI [+1.4 runs vs projection] season xwOBA 0.321, above average - real projection error
Brett Baty: 2-for-3, 0 HR, 1 RBI [+1.27 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

HOME pitching

Christian Scott (NYM) threw 14 outs, 3 ER = 5.79 ERA against a projected 5.4 ERA (0.39 ERA error)

Bullpen behind him: 0 ER over 4 outs.

The AWAY lineup he faced produced 3 runs on 9 hits and 0 HR, against a projected 4 runs / 7 hits (-1 runs vs projection).

Who did the damage to him:

Alec Bohm: 2-for-3, 0 HR, 2 RBI [+0.8 runs vs projection] season xwOBA 0.328, above average - real projection error
Trea Turner: 2-for-3, 0 HR, 0 RBI [+0.8 runs vs projection] season xwOBA 0.292, 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)
PHI	Andrew Painter (Starter)	16 / 16	6.75 / 3.38 (+3.37)	4 / 2	4 / 3	2 / 1	1 / 1	1.87	7 / 5	4.63	venue away: 4.71 vs 6.56 ERA (ok) · opp run factor 1.01 (average) · 23 prior appearances this season
PHI	Chase Shugart (Setup Man)	2 / 0	0 / n/a	0 / 0	0 / 0	0 / 0	1 / 0	0.28	0 / 1	0.81
NYM	Christian Scott (Starter)	14 / 15	5.79 / 5.4	3 / 3	6 / 5	0 / 1	1 / 0	2.4	7 / 5	5.68	venue home: 3.92 vs 3.88 ERA (ok) · opp run factor 0.96 (average) · 22 prior appearances this season
NYM	Jonathan Pintaro (Middle Reliever (High Leverage))	1 / 4	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0.68	0 / 0	1.32
NYM	Dedniel Núñez (Middle Reliever (High Leverage))	3 / 1	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0	1 / 1	0.84

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
Carson Benge (RF)	NYM	4 / 4	3 / 1	0 / 0	0 / 0	1 / 0	2 / 1	2 / 1	1 / 0	0 / 1	2.2	0.338 above average	projection too low
Francisco Alvarez (C)	NYM	4 / 3	2 / 1	0 / 0	0 / 0	1 / 0	3 / 0	1 / 0	0 / 0	1 / 1	1.4	0.321 above average	projection too low
Brett Baty (1B)	NYM	3 / 3	2 / 0	0 / 0	0 / 0	0 / 0	1 / 0	2 / 0	1 / 0	1 / 1	1.27	0.312 average	no signal
Juan Soto (LF)	NYM	5 / 4	1 / 1	0 / 0	0 / 0	0 / 1	1 / 1	0 / 1	0 / 1	3 / 0	-1.26	0.416 elite	bad day
Alec Bohm (1B)	PHI	3 / 4	2 / 1	0 / 0	0 / 0	0 / 0	2 / 1	0 / 0	1 / 0	1 / 1	0.8	0.328 above average	projection too low
Trea Turner (SS)	PHI	3 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	2 / 1	0 / 1	0.8	0.292 below average	good day
Francisco Lindor (SS)	NYM	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	1 / 1	1 / 1	1 / 1	1 / 1	-0.78	0.331 above average	bad day
Marcus Semien (2B)	NYM	3 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	2 / 0	2 / 0	0 / 1	0.66	0.303 below average	good day
Kyle Schwarber (DH)	PHI	4 / 4	1 / 1	1 / 0	0 / 0	0 / 1	0 / 1	1 / 1	1 / 1	2 / 1	-0.62	0.364 elite	bad day
Brandon Marsh (LF)	PHI	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	-0.47	0.301 below average	—
Bryson Stott (3B)	PHI	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	2 / 1	-0.33	0.328 above average	—
A.J. Ewing (CF)	NYM	5 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	1 / 1	-0.33	0.32 average	—
Bo Bichette (3B)	NYM	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 0	1 / 0	2 / 1	0.33	0.325 above average	—
Luis Arraez (2B)	PHI	5 / 4	2 / 1	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0.16	0.307 average	—
Ronny Mauricio (DH)	NYM	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 0	1 / 1	-0.14	0.205 weak	—
Bryce Harper (RF)	PHI	4 / 4	1 / 1	1 / 0	0 / 0	0 / 0	1 / 1	1 / 1	0 / 1	1 / 1	-0.02	0.383 elite	—
J.T. Realmuto (C)	PHI	4 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	0	0.321 above average	—
Justin Crawford (CF)	PHI	4 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	0	0.288 below average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-0.382	positive favors home; negative favors away
Recent form	0.142	positive favors home; negative favors away
Starting-pitcher matchup	0.044	positive favors home; negative favors away
Composite	0.102	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	3.93-5.16	n/a	Base ML historical model prediction
variance_applied	4-5	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-6	0.529	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run, matchup_recency_guardrail)
embedded_savant_run_environment	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	3.93-5.16	4-5	-0.23	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
NYM	Francisco Lindor	batter / SS	worse	predicted_RBI 2.52
PHI	Kyle Schwarber	batter / DH	better	predicted_RBI 2.27
PHI	Tim Mayza	pitcher / Reliever	worse	predicted_ER 2
NYM	Francisco Alvarez	batter / C	better	predicted_AB -1.94
PHI	Bryce Harper	batter / RF	worse	predicted_AB 1.441
NYM	Juan Soto	batter / LF	worse	predicted_H -1.29
PHI	Trea Turner	batter / SS	worse	predicted_H -1.21
NYM	Carson Benge	batter / RF	worse	predicted_AB 1.16

Structural audit findings

Severity	Rule	Finding	Subject
info	CURRENT_PITCHER_DUPLICATION	NYM pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	NYM
info	CURRENT_PITCHER_DUPLICATION	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	NYM model-prediction batting totals differ from final simulation box score (AB: current=33, sim=37; H: current=8, sim=7; 2B: current=1, sim=0; HR: current=1, sim=3; RBI: current=4, sim=6; R: current=4, sim=6; BB: current=3, sim=2; SO: current=8, sim=10).	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=5, HR current=0 box=1; A.J. Ewing: AB current=4 box=3, HBP current=0 box=1; Juan Soto: AB current=4 box=5, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0; Brett Baty: AB current=3 box=4, H current=0 box=1, R current=0 box=1; Francisco Lindor: AB current=4 box=5, 2B current=1 box=0, HR current=0 box=1, RBI current=1 box=3; Ronny Mauricio: AB current=3 box=4, H current=1 box=0, SO current=1 box=2; Bo Bichette: AB current=4 box=5, H current=1 box=2, RBI current=1 box=0, R current=0 box=1; Marcus Semien: H current=1 box=0	NYM
info	CURRENT_FINALIZED_MATCH_BOX	PHI model-prediction batting totals differ from final simulation box score (AB: current=33, sim=40; H: current=7, sim=10; 2B: current=1, sim=3; RBI: current=4, sim=5; R: current=4, sim=5; BB: current=4, sim=0; SO: current=8, sim=6).	PHI
info	CURRENT_PLAYER_MATCH_BOX	PHI current finalized hitter lines are model predictions and differ from the simulation box score: Bryce Harper: AB current=4 box=5, H current=1 box=2, RBI current=1 box=0, BB current=1 box=0; Alec Bohm: 2B current=0 box=1, R current=0 box=1, SO current=1 box=0; Bryson Stott: RBI current=1 box=0, R current=1 box=0, BB current=1 box=0; Kyle Schwarber: AB current=4 box=5, H current=1 box=2, RBI current=1 box=3, BB current=1 box=0; Brandon Marsh: AB current=3 box=4; Trea Turner: AB current=4 box=5, H current=1 box=0, R current=1 box=0, BB current=1 box=0; Luis Arraez: AB current=4 box=5, H current=1 box=0, 2B current=1 box=0, SO current=0 box=1; J.T. Realmuto: AB current=3 box=4, H current=0 box=2, 2B current=0 box=1, R current=0 box=1	PHI
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

SEA@COLMISS

Predicted: SEA 10-5 (conf -0.44, strong)Actual: COL 3-5Brier 0.52total bias +7

Root cause: Starting-pitcher projection was wrong

Evidence-backed diagnosis

published system edge 0.44 (strong); this is not the half-scale distance from 50%
team-run absolute error 7; total bias +7
pregame home-minus-away evidence: season -1.008, recent -1.458, pitcher -0.079, composite -1.993
confidence pipeline reached 0.764 before later publication/recency controls; published home probability was 0.279
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 7 (15 vs 8)
SEA offense over-projected by 7
Bryce Miller (SEA) allowed 5 fewer ER than projected (0 vs 5)
committed to SEA at 0.44 confidence and it lost

Team	Starter	Projected	Actual	ER err
SEA	Bryce Miller	5 ER / 15 outs	0 ER / 16 outs	-5
COL	Jose Quintana	5 ER / 13 outs	3 ER / 12 outs	-2

Decision internals: run factors home×0.934 / 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:

Cal Raleigh (SEA C): 2-for-4, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.38]
J.P. Crawford (SEA SS): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.8]
Hunter Goodman (COL C): 1-for-3, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -0.6]

Starter miss: Bryce Miller (SEA) posted 0.00 ERA vs projected 9.00 (0 ER in 16 outs); he is materially worse pitching away (4.69 ERA there vs 3.66 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

Bryce Miller (SEA) threw 16 outs, 0 ER = 0 ERA against a projected 9 ERA (-9 ERA error)

Bullpen behind him: 2 ER over 4 outs.

The HOME lineup he faced produced 4 runs on 8 hits and 0 HR, against a projected 4 runs / 9 hits (0 runs vs projection).

Who did the damage to him:

Zac Veen: 1-for-3, 0 HR, 0 RBI [+0.33 runs vs projection] season xwOBA 0.282, weak - a good day, not a model error

Verdict: offense: the lineup scored well below its projection

HOME pitching

Jose Quintana (COL) threw 12 outs, 3 ER = 6.75 ERA against a projected 10.38 ERA (-3.63 ERA error) only 9 prior appearances in the covered season - the projection had little of his own record to stand on

Bullpen behind him: 0 ER over 9 outs.

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

Who did the damage to him:

Cal Raleigh: 2-for-4, 1 HR, 2 RBI [+1.38 runs vs projection] season xwOBA 0.31, average - a good day, not a model error
Michael Arroyo: 2-for-4, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.38, elite - a good day, not a model error
Julio Rodríguez: 2-for-4, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.333, above 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)
SEA	Bryce Miller (Starter)	16 / 15	0 / 9 (-9.00)	0 / 5	4 / 5	0 / 2	2 / 6	3.17	6 / 6	5.23	venue away: 4.69 vs 3.66 ERA (worse here) · opp run factor 0.93 (weak) · 20 prior appearances this season
SEA	Gabe Speier (Middle Reliever (High Leverage))	3 / 2	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0.24	2 / 1	0.71
SEA	José A. Ferrer (Setup Man)	1 / 2	54 / 0 (+54.00)	2 / 0	2 / 0	0 / 0	0 / 0	0.24	0 / 0	0.65
COL	Jose Quintana (Starter)	12 / 13	6.75 / 10.38 (-3.63)	3 / 5	5 / 6	2 / 1	1 / 2	1.51	2 / 3	2.44	venue home: 4.03 vs 6.75 ERA (ok) · opp run factor 1.01 (weak) · 9 prior appearances this season
COL	Brennan Bernardino (Setup Man)	3 / 3	0 / 0	0 / 0	1 / 1	0 / 0	0 / 0	0.15	2 / 1	0.87
COL	Juan Mejia (Middle Reliever)	6 / 2	0 / 54 (-54.00)	0 / 4	1 / 4	0 / 1	0 / 1	0.81	2 / 0	1.52

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
Cal Raleigh (C)	SEA	4 / 4	2 / 1	1 / 0	0 / 0	1 / 0	2 / 1	1 / 1	0 / 1	1 / 1	1.38	0.31 average	no signal
J.P. Crawford (SS)	SEA	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	3 / 1	-0.8	0.326 above average	bad day
Hunter Goodman (C)	COL	3 / 4	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	1 / 0	0 / 1	-0.6	0.32 average	no signal
Michael Arroyo (2B)	SEA	4 / 3	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0.47	0.38 elite	—
Lazaro Montes (DH)	SEA	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	2 / 1	-0.47	0.37 elite	—
Julio Rodríguez (CF)	SEA	4 / 4	2 / 1	1 / 1	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0 / 1	0.47	0.333 above average	—
Weston Wilson (3B)	SEA	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	-0.47	0.264 weak	—
Josh Naylor (1B)	SEA	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	-0.47	0.313 average	—
Mickey Moniak (DH)	COL	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	1 / 1	-0.47	0.28 weak	—
Ezequiel Tovar (SS)	COL	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	-0.47	0.28 weak	—
Zac Veen (RF)	COL	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	2 / 1	0.33	0.282 weak	—
Cole Carrigg (CF)	COL	4 / 4	1 / 1	0 / 1	0 / 0	0 / 0	2 / 1	0 / 1	0 / 0	1 / 1	-0.31	0.311 average	—
Jake McCarthy (LF)	COL	5 / 4	1 / 1	0 / 1	0 / 0	0 / 0	1 / 1	1 / 1	0 / 0	0 / 1	-0.31	0.327 above average	—
Randy Arozarena (LF)	SEA	4 / 4	1 / 1	0 / 1	0 / 0	1 / 0	1 / 1	1 / 1	0 / 1	1 / 1	0.29	0.347 elite	—
Dominic Canzone (RF)	SEA	2 / 4	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	2 / 0	0 / 1	-0.27	0.355 elite	—
TJ Rumfield (1B)	COL	4 / 4	2 / 1	0 / 0	0 / 0	0 / 0	1 / 1	1 / 0	0 / 1	1 / 1	0.14	0.336 above average	—
Connor Norby (2B)	COL	4 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	1 / 1	0.14	0.308 average	—
Kyle Karros (3B)	COL	2 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	0 / 1	-0.14	0.34 elite	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-1.008	positive favors home; negative favors away
Recent form	-1.458	positive favors home; negative favors away
Starting-pitcher matchup	-0.079	positive favors home; negative favors away
Composite	-1.993	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	6.01-4.13	n/a	Base ML historical model prediction
variance_applied	6-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	9-5	0.721	Enhanced ML with recent form, H2H, low-score analysis
embedded_savant_run_environment	10-5	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	10-5	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	10-5	n/a	Capped at max 12 realistic runs
simulation_result	10-5	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	6.01-4.13	6-4	0.113	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	6-4	9-5	2	Enhanced ML with recent form, H2H, low-score analysis
embedded_savant_run_environment	9-5	10-5	1	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	10-5	10-5	-0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	10-5	10-5	-0	Capped at max 12 realistic runs
simulation_result	10-5	10-5	-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
SEA	Julio Rodríguez	batter / CF	better	predicted_RBI 5.47
SEA	Michael Arroyo	batter / 2B	worse	predicted_AB 2.5
COL	Zac Veen	batter / RF	better	predicted_RBI 2.464
SEA	Weston Wilson	batter / 3B	worse	predicted_H 1.8
COL	Kyle Karros	batter / 3B	worse	predicted_AB -1.72
COL	Jake McCarthy	batter / LF	worse	predicted_H -1.64
SEA	J.P. Crawford	batter / SS	better	predicted_RBI 1.55
SEA	Dominic Canzone	batter / RF	better	predicted_RBI 1.485

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	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_FINALIZED_MATCH_BOX	COL model pitcher projections differ from final simulation pitching box score (ER: current=9, sim=10).	COL
info	CURRENT_FINALIZED_MATCH_BOX	COL model-prediction batting totals differ from final simulation box score (AB: current=34, sim=33; H: current=9, sim=6; 2B: current=2, sim=1; HR: current=1, sim=2; R: current=4, sim=5; BB: current=3, sim=7; SO: current=9, sim=8).	COL
info	CURRENT_PLAYER_MATCH_BOX	COL current finalized hitter lines are model predictions and differ from the simulation box score: TJ Rumfield: H current=1 box=0, RBI current=1 box=0, R current=0 box=1; Cole Carrigg: AB current=4 box=3, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Jake McCarthy: AB current=4 box=5, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Mickey Moniak: H current=1 box=2, 2B current=0 box=1, R current=0 box=1; Zac Veen: HR current=0 box=1, RBI current=0 box=3, R current=0 box=1, BB current=0 box=1; Connor Norby: H current=1 box=2, HR current=0 box=1, RBI current=0 box=1, R current=0 box=1; Hunter Goodman: AB current=4 box=5, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0; Kyle Karros: AB current=4 box=2, R current=0 box=1, BB current=1 box=2	COL
info	CURRENT_FINALIZED_MATCH_BOX	SEA model-prediction batting totals differ from final simulation box score (AB: current=33, sim=39; H: current=9, sim=12; 2B: current=2, sim=3; HR: current=1, sim=2; RBI: current=5, sim=10; R: current=5, sim=10; BB: current=3, sim=4; SO: current=7, sim=5).	SEA
info	CURRENT_PLAYER_MATCH_BOX	SEA current finalized hitter lines are model predictions and differ from the simulation box score: Dominic Canzone: 2B current=0 box=1, HR current=1 box=0, RBI current=1 box=2, BB current=0 box=1; J.P. Crawford: 2B current=0 box=1, RBI current=0 box=2, SO current=1 box=2; Cal Raleigh: RBI current=1 box=0, SO current=1 box=2; Michael Arroyo: AB current=3 box=5, H current=1 box=0; Lazaro Montes: AB current=3 box=4, RBI current=1 box=0, R current=0 box=1, SO current=1 box=0; Julio Rodríguez: AB current=4 box=5, H current=1 box=2, 2B current=1 box=0, HR current=0 box=2; Randy Arozarena: AB current=4 box=5, H current=1 box=3, 2B current=1 box=0, RBI current=1 box=0; Weston Wilson: AB current=3 box=4, H current=1 box=2, 2B current=0 box=1, R current=0 box=1	SEA
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

SF@LADHIT

Predicted: LAD 1-7 (conf +0.13, moderate)Actual: LAD 4-10Brier 0.189total bias -6

Root cause: Right winner, wrong score

Evidence-backed diagnosis

published system edge 0.13 (moderate); this is not the half-scale distance from 50%
team-run absolute error 6; total bias -6
pregame home-minus-away evidence: season +1.503, recent +0.674, pitcher +0.069, composite +2.227
confidence pipeline reached 0.627 before later publication/recency controls; published home probability was 0.565
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 6 (8 vs 14)
SF offense under-projected by 3
LAD offense under-projected by 3
Tarik Skubal (LAD) allowed 4 more ER than projected (4 vs 0)

Team	Starter	Projected	Actual	ER err
SF	Yunior Marte	5 ER / 15 outs	6 ER / 14 outs	1
LAD	Tarik Skubal	0 ER / 19 outs	4 ER / 15 outs	4

Decision internals: run factors home×1 / away×0.931; 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:

Mookie Betts (LAD SS): 4-for-4, 0 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 2.67]
Will Smith (LAD C): 3-for-5, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.87]
Victor Bericoto (SF DH): 2-for-4, 1 HR, 2 RBI (projected 1.0 H, 1.0 HR) [run impact 1.11]
Freddie Freeman (LAD 1B): 0-for-5, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -1.11]
Jonah Cox (SF CF): 0-for-3, 0 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact -0.8]

Starter miss: Tarik Skubal (LAD) posted 7.20 ERA vs projected 0.00 (4 ER in 15 outs); he is materially worse pitching home (3.40 ERA there vs 1.87 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

Yunior Marte (SF) threw 14 outs, 6 ER = 11.57 ERA against a projected 9 ERA (2.57 ERA error) only 3 prior appearances in the covered season - the projection had little of his own record to stand on

Bullpen behind him: 0 ER over 0 outs.

The HOME lineup he faced produced 10 runs on 14 hits and 1 HR, against a projected 5 runs / 7 hits (5 runs vs projection).

Who did the damage to him:

Mookie Betts: 4-for-4, 0 HR, 2 RBI [+2.67 runs vs projection] season xwOBA 0.339, above average - real projection error
Will Smith: 3-for-5, 1 HR, 2 RBI [+1.87 runs vs projection] season xwOBA 0.38, elite - real projection error
Josue De Paula: 1-for-3, 0 HR, 2 RBI [+0.8 runs vs projection] season xwOBA 0.435, elite - real projection error

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

HOME pitching

Tarik Skubal (LAD) threw 15 outs, 4 ER = 7.2 ERA against a projected 0 ERA (7.2 ERA error)

Bullpen behind him: 0 ER over 3 outs.

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

Who did the damage to him:

Victor Bericoto: 2-for-4, 1 HR, 2 RBI [+1.11 runs vs projection] season xwOBA 0.282, weak - a good day, not a model error
Brett Harris: 1-for-3, 0 HR, 0 RBI [+0.78 runs vs projection] season xwOBA 0.257, weak - a good day, not a model error
Jung Hoo Lee: 2-for-4, 0 HR, 1 RBI [+0.47 runs vs projection] season xwOBA 0.314, 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)
SF	Yunior Marte (Starter)	14 / 15	11.57 / 9	6 / 5	8 / 6	1 / 1	3 / 2	1.81	2 / 3	3.15	opp run factor 1 (strong) · thin sample: 3 prior appearances this season (high uncertainty)
LAD	Tarik Skubal (Starter)	15 / 19	7.2 / 0 (+7.20)	4 / 0	8 / 0	1 / 0	2 / 2	1.08	3 / 7	6.73	venue home: 3.4 vs 1.87 ERA (worse here) · opp run factor 0.93 (weak) · 24 prior appearances this season
LAD	Alex Vesia (Middle Reliever (High Leverage))	3 / 2	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	0.34	2 / 1	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
Mookie Betts (SS)	LAD	4 / 4	4 / 1	3 / 0	0 / 0	0 / 0	2 / 1	2 / 1	1 / 0	0 / 0	2.67	0.339 above average	projection too low
Will Smith (C)	LAD	5 / 4	3 / 1	0 / 0	0 / 0	1 / 0	2 / 1	2 / 0	0 / 0	0 / 1	1.87	0.38 elite	projection too low
Victor Bericoto (DH)	SF	4 / 4	2 / 1	1 / 0	0 / 0	1 / 1	2 / 1	2 / 1	1 / 0	2 / 1	1.11	0.282 weak	good day
Freddie Freeman (1B)	LAD	5 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	-1.11	0.366 elite	bad day
Jonah Cox (CF)	SF	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	0 / 1	3 / 2	-0.8	0.271 weak	projection too high
Josue De Paula (DH)	LAD	3 / 3	1 / 0	0 / 0	0 / 0	0 / 0	2 / 0	2 / 1	2 / 1	1 / 0	0.8	0.435 elite	projection too low
Alex Freeland (2B)	LAD	1 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 1	0.8	0.296 below average	good day
Brett Harris (1B)	SF	3 / 3	1 / 0	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0.78	0.257 weak	good day
Jung Hoo Lee (RF)	SF	4 / 4	2 / 1	1 / 1	0 / 0	0 / 0	1 / 1	0 / 1	0 / 0	0 / 1	0.47	0.314 average	—
Shay Whitcomb (3B)	SF	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	-0.47	0.26 weak	—
Andrew Knizner (C)	SF	4 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 1	0.47	0.3 below average	—
Christian Koss (SS)	SF	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	-0.47	0.254 weak	—
Turner Hill (LF)	SF	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	0 / 1	1 / 1	-0.33	0.326 above average	—
Osleivis Basabe (2B)	SF	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	2 / 1	0.33	0.279 weak	—
Max Muncy (3B)	LAD	5 / 4	2 / 1	2 / 0	0 / 0	0 / 1	1 / 1	1 / 1	0 / 1	0 / 1	-0.17	0.35 elite	—
Tommy Edman (CF)	LAD	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	2 / 0	0 / 0	0 / 0	0 / 1	0	0.302 below average	—
Teoscar Hernández (LF)	LAD	5 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	0 / 0	2 / 1	0	0.334 above average	—
Kyle Tucker (RF)	LAD	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	1 / 1	0 / 1	0	0.334 above average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	1.503	positive favors home; negative favors away
Recent form	0.674	positive favors home; negative favors away
Starting-pitcher matchup	0.069	positive favors home; negative favors away
Composite	2.227	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	2.94-5.74	n/a	Base ML historical model prediction
variance_applied	3-6	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	2-6	0.565	Enhanced ML with recent form, H2H, low-score analysis
embedded_savant_run_environment	1-7	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	1-7	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	1-7	n/a	Capped at max 12 realistic runs
simulation_result	1-7	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	2.94-5.74	3-6	0.196	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	3-6	2-6	1	Enhanced ML with recent form, H2H, low-score analysis
embedded_savant_run_environment	2-6	1-7	2	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	1-7	1-7	0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	1-7	1-7	0	Capped at max 12 realistic runs
simulation_result	1-7	1-7	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
LAD	Mookie Betts	batter / SS	deboosted	predicted_RBI 3.14
LAD	Teoscar Hernández	batter / LF	worse	predicted_AB -2.19
LAD	Will Smith	batter / C	worse	predicted_SO 1.359
SF	Shay Whitcomb	batter / 3B	worse	predicted_AB 1.294
LAD	Josue De Paula	batter / DH	better	predicted_H 1.286
SF	Christian Koss	batter / SS	worse	predicted_SO 1.246
LAD	Kyle Tucker	batter / RF	worse	predicted_H -1.22
SF	Victor Bericoto	batter / DH	worse	predicted_AB 1.194

Structural audit findings

Severity	Rule	Finding	Subject
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_PITCHER_DUPLICATION	SF pitcher list appears in both predicted_performance.pitchers and player_predictions_by_position.pitchers; IDs match and were de-duplicated in this audit.	SF
info	CURRENT_FINALIZED_MATCH_BOX	LAD model-prediction batting totals differ from final simulation box score (AB: current=33, sim=31; H: current=7, sim=8; 3B: current=0, sim=1; RBI: current=4, sim=7; R: current=5, sim=7; BB: current=4, sim=3; SO: current=7, sim=4; SF: current=0, sim=1).	LAD
info	CURRENT_PLAYER_MATCH_BOX	LAD current finalized hitter lines are model predictions and differ from the simulation box score: Will Smith: H current=1 box=2, RBI current=1 box=0, R current=0 box=1, SO current=1 box=2; Max Muncy: H current=1 box=0, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0; Mookie Betts: RBI current=1 box=4; Josue De Paula: AB current=3 box=4, H current=0 box=2, HR current=0 box=1, RBI current=0 box=2; Tommy Edman: AB current=4 box=3, H current=1 box=2, 3B current=0 box=1, R current=0 box=1; Freddie Freeman: H current=1 box=0, 2B current=1 box=0, R current=0 box=1, BB current=1 box=0; Teoscar Hernández: AB current=4 box=2, H current=1 box=0, RBI current=1 box=0, BB current=0 box=2; Alex Freeland: AB current=3 box=2, H current=0 box=1, 2B current=0 box=1, RBI current=0 box=1	LAD
info	CURRENT_FINALIZED_MATCH_BOX	SF model pitcher projections differ from final simulation pitching box score (H: current=7, sim=8; ER: current=6, sim=7).	SF
info	CURRENT_FINALIZED_MATCH_BOX	SF model-prediction batting totals differ from final simulation box score (AB: current=32, sim=30; H: current=7, sim=3; HR: current=1, sim=0; RBI: current=4, sim=1; R: current=4, sim=1; BB: current=2, sim=3; SO: current=10, sim=9).	SF
info	CURRENT_PLAYER_MATCH_BOX	SF current finalized hitter lines are model predictions and differ from the simulation box score: Turner Hill: RBI current=1 box=0, BB current=1 box=0; Victor Bericoto: H current=1 box=0, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0; Brett Harris: BB current=0 box=1, SO current=1 box=2; Jung Hoo Lee: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, R current=1 box=0; Osleivis Basabe: H current=1 box=0, SO current=1 box=0; Jonah Cox: AB current=4 box=3, H current=1 box=0, RBI current=1 box=0, R current=1 box=0; Andrew Knizner: AB current=3 box=2, H current=0 box=1, 2B current=0 box=1, RBI current=0 box=1; Christian Koss: H current=1 box=0, R current=1 box=0, SO current=1 box=2	SF
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

TOR@TEXHIT

Predicted: TEX 2-4 (conf +0.22, strong)Actual: TEX 2-6Brier 0.151total bias -2

Root cause: Strong correct call

Evidence-backed diagnosis

published system edge 0.22 (strong); this is not the half-scale distance from 50%
team-run absolute error 2; total bias -2
pregame home-minus-away evidence: season +0.200, recent +0.782, pitcher +0.052, composite +0.779
confidence pipeline reached 0.611 before later publication/recency controls; published home probability was 0.611
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
TOR	José Soriano	4 ER / 18 outs	5 ER / 15 outs	1
TEX	Cal Quantrill	1 ER / 18 outs	0 ER / 4 outs	-1

Decision internals: run factors home×0.997 / away×0.945; 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:

Corey Seager (TEX DH): 1-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.26]
Danny Jansen (TEX C): 1-for-2, 0 HR, 1 RBI (projected 0.0 H, 0.0 HR) [run impact 1.11]
Kazuma Okamoto (TOR 3B): 1-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -0.93]
Josh Jung (TEX 3B): 1-for-3, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 0.93]
George Springer (TOR DH): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.8]

Starter miss: José Soriano (TOR) posted 9.00 ERA vs projected 6.00 (5 ER in 15 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

José Soriano (TOR) threw 15 outs, 5 ER = 9 ERA against a projected 6 ERA (3 ERA error)

Bullpen behind him: 0 ER over 3 outs.

The HOME lineup he faced produced 6 runs on 5 hits and 1 HR, against a projected 4 runs / 7 hits (2 runs vs projection).

Who did the damage to him:

Danny Jansen: 1-for-2, 0 HR, 1 RBI [+1.11 runs vs projection] season xwOBA 0.261, weak - a good day, not a model error
Josh Jung: 1-for-3, 1 HR, 1 RBI [+0.93 runs vs projection] season xwOBA 0.348, elite - real projection error
Wyatt Langford: 1-for-2, 0 HR, 1 RBI [+0.64 runs vs projection] season xwOBA 0.323, above average - real projection error

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

HOME pitching

Cal Quantrill (TEX) threw 4 outs, 0 ER = 0 ERA against a projected 1.5 ERA (-1.5 ERA error)

Bullpen behind him: 0 ER over 6 outs.

The AWAY lineup he faced produced 1 runs on 5 hits and 0 HR, against a projected 4 runs / 8 hits (-3 runs vs projection).

Who did the damage to him:

Alejandro Kirk: 2-for-4, 0 HR, 1 RBI [+0.47 runs vs projection] season xwOBA 0.31, 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)
TOR	José Soriano (Starter)	15 / 18	9 / 6 (+3.00)	5 / 4	4 / 7	1 / 1	1 / 2	2.03	8 / 5	4.2	venue away: 3.39 vs 3.79 ERA (ok) · opp run factor 1 (average) · 30 prior appearances this season
TOR	Brendan Cellucci (Setup Man)	3 / 4	0 / 0	0 / 0	0 / 0	0 / 0	2 / 0	0	1 / 1	1.5
TEX	Cal Quantrill (Starter)	4 / 18	0 / 1.5	0 / 1	0 / 1	0 / 0	0 / 0	0.7	0 / 9	5.25	venue home: 0.97 vs 3.99 ERA (ok) · opp run factor 0.94 (weak) · 31 prior appearances this season
TEX	Jacob Latz (Closer)	3 / 4	0 / 0	0 / 0	0 / 0	0 / 0	0 / 2	0.41	1 / 0	1.65
TEX	Chase Silseth (Middle Reliever (High Leverage))	3 / 2	0 / 13.5 (-13.50)	0 / 1	2 / 1	0 / 1	0 / 0	0.55	1 / 2	1.32

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
Corey Seager (DH)	TEX	4 / 4	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	1 / 1	-1.26	0.354 elite	bad day
Danny Jansen (C)	TEX	2 / 3	1 / 0	1 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 0	0 / 1	1.11	0.261 weak	good day
Kazuma Okamoto (3B)	TOR	4 / 3	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 0	0 / 1	-0.93	0.33 above average	bad day
Josh Jung (3B)	TEX	3 / 4	1 / 1	0 / 0	0 / 0	1 / 0	1 / 0	2 / 1	0 / 0	1 / 1	0.93	0.348 elite	projection too low
George Springer (DH)	TOR	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	1 / 1	-0.8	0.306 average	no signal
Wyatt Langford (LF)	TEX	2 / 3	1 / 1	1 / 0	0 / 0	0 / 0	1 / 1	2 / 1	2 / 1	0 / 1	0.64	0.323 above average	projection too low
Alejandro Kirk (C)	TOR	4 / 4	2 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 0	0 / 0	0 / 1	0.47	0.31 average	—
Nathan Lukes (RF)	TOR	2 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	-0.47	0.285 below average	—
Myles Straw (LF)	TOR	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	-0.47	0.298 below average	—
Andrés Giménez (SS)	TOR	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	0 / 0	0 / 1	-0.47	0.272 weak	—
Ezequiel Duran (SS)	TEX	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	2 / 1	-0.47	0.305 average	—
Jake Burger (1B)	TEX	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	3 / 1	-0.47	0.319 average	—
Brett Bateman (CF)	TOR	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	-0.33	0.301 below average	—
Evan Carter (CF)	TEX	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 1	2 / 1	-0.33	0.276 weak	—
Brandon Nimmo (RF)	TEX	4 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	2 / 1	-0.31	0.343 elite	—
Nicky Lopez (2B)	TEX	1 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 0	-0.14	0.263 weak	—
Vladimir Guerrero Jr. (1B)	TOR	3 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	0 / 0	0.02	0.336 above average	—
Josh Smith (2B)	TOR	1 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0	0.284 weak	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	0.2	positive favors home; negative favors away
Recent form	0.782	positive favors home; negative favors away
Starting-pitcher matchup	0.052	positive favors home; negative favors away
Composite	0.779	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.24-3.96	n/a	Base ML historical model prediction
variance_applied	4-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	2-4	0.611	Enhanced ML with recent form, H2H, low-score analysis
embedded_savant_run_environment	2-4	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	2-4	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	2-4	n/a	Capped at max 12 realistic runs
simulation_result	2-4	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.24-3.96	4-4	0.281	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-4	2-4	2	Enhanced ML with recent form, H2H, low-score analysis
embedded_savant_run_environment	2-4	2-4	0	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	2-4	2-4	0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	2-4	2-4	0	Capped at max 12 realistic runs
simulation_result	2-4	2-4	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
TOR	Brett Bateman	batter / CF	worse	predicted_AB -2.6
TEX	Josh Jung	batter / 3B	worse	predicted_AB -1.73
TOR	Nathan Lukes	batter / RF	boosted	predicted_RBI 1.7
TOR	George Springer	batter / DH	worse	predicted_AB -1.46
TEX	Ezequiel Duran	batter / SS	worse	predicted_RBI 1.4
TEX	Jake Burger	batter / 1B	worse	predicted_H 1.26
TOR	Alejandro Kirk	batter / C	worse	predicted_H -1.24
TEX	Evan Carter	batter / CF	worse	predicted_SO 1.16

Structural audit findings

Severity	Rule	Finding	Subject
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_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	TEX model pitcher projections differ from final simulation pitching box score (H: current=3, sim=2).	TEX
info	CURRENT_FINALIZED_MATCH_BOX	TEX model-prediction batting totals differ from final simulation box score (2B: current=1, sim=2; BB: current=3, sim=2; SO: current=8, sim=6).	TEX
info	CURRENT_PLAYER_MATCH_BOX	TEX current finalized hitter lines are model predictions and differ from the simulation box score: Corey Seager: AB current=4 box=3, 2B current=0 box=1, HR current=1 box=0, RBI current=1 box=0; Brandon Nimmo: RBI current=0 box=1, R current=0 box=1; Wyatt Langford: AB current=3 box=4, H current=1 box=2, BB current=1 box=0, SO current=1 box=0; Josh Jung: AB current=4 box=3, H current=1 box=0, R current=1 box=0, BB current=0 box=1; Ezequiel Duran: HR current=0 box=1, RBI current=1 box=2, R current=0 box=1; Jake Burger: AB current=3 box=4, H current=1 box=2, RBI current=1 box=0, SO current=1 box=0; Nicky Lopez: H current=1 box=0; Danny Jansen: SO current=1 box=0	TEX
info	CURRENT_FINALIZED_MATCH_BOX	TOR model-prediction batting totals differ from final simulation box score (AB: current=32, sim=29; H: current=8, sim=2; 2B: current=1, sim=0; RBI: current=4, sim=2; R: current=4, sim=2; BB: current=2, sim=4; SO: current=6, sim=12).	TOR
info	CURRENT_PLAYER_MATCH_BOX	TOR current finalized hitter lines are model predictions and differ from the simulation box score: Alejandro Kirk: H current=1 box=0, RBI current=1 box=0; Nathan Lukes: HR current=0 box=1, RBI current=0 box=2, R current=0 box=1, SO current=0 box=2; Kazuma Okamoto: AB current=3 box=4, HR current=1 box=0, RBI current=1 box=0, SO current=1 box=2; George Springer: AB current=4 box=3, H current=1 box=0, RBI current=1 box=0, R current=1 box=0; Josh Smith: AB current=3 box=2, BB current=0 box=1; Brett Bateman: AB current=4 box=3, H current=1 box=0, R current=1 box=0, SO current=1 box=2; Myles Straw: AB current=3 box=2, H current=1 box=0, BB current=0 box=1, SO current=0 box=1; Vladimir Guerrero Jr.: H current=1 box=0, 2B current=1 box=0, R current=1 box=0	TOR
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

WAS@STLHIT

Predicted: WAS 5-4 (conf -0.09, lean)Actual: WAS 8-5Brier 0.209total bias -4

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 4; total bias -4
pregame home-minus-away evidence: season -0.255, recent -0.384, pitcher -0.071, composite -0.572
confidence pipeline reached 0.560 before later publication/recency controls; published home probability was 0.457
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)
WAS offense under-projected by 3

Team	Starter	Projected	Actual	ER err
WAS	Andrew Alvarez	1 ER / 17 outs	0 ER / 18 outs	-1
STL	Michael McGreevy	2 ER / 17 outs	2 ER / 12 outs	0

Decision internals: run factors home×0.975 / away×1.046; 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:

Jordan Walker (STL RF): 0-for-6, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.4]
Jorbit Vivas (WAS 3B): 4-for-6, 0 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 1.08]
Daylen Lile (WAS LF): 1-for-6, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 0.93]
Iván Herrera (STL DH): 3-for-5, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact 0.92]
Nasim Nuñez (WAS SS): 1-for-4, 0 HR, 1 RBI (projected 0.0 H, 0.0 HR) [run impact 0.8]

Pitcher versus batter: what actually broke this game

AWAY pitching

Andrew Alvarez (WAS) threw 18 outs, 0 ER = 0 ERA against a projected 1.59 ERA (-1.59 ERA error)

Bullpen behind him: 1 ER over 1 outs.

The HOME lineup he faced produced 5 runs on 12 hits and 0 HR, against a projected 4 runs / 8 hits (1 runs vs projection).

Who did the damage to him:

Iván Herrera: 3-for-5, 0 HR, 0 RBI [+0.92 runs vs projection] season xwOBA 0.331, above average - real projection error
Nathan Church: 2-for-5, 0 HR, 0 RBI [+0.78 runs vs projection] season xwOBA 0.29, below average - a good day, not a model error
José Fermín: 3-for-5, 0 HR, 1 RBI [+0.63 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

Michael McGreevy (STL) threw 12 outs, 2 ER = 4.5 ERA against a projected 3.18 ERA (1.32 ERA error)

Bullpen behind him: 0 ER over 8 outs.

The AWAY lineup he faced produced 4 runs on 11 hits and 1 HR, against a projected 4 runs / 8 hits (0 runs vs projection).

Who did the damage to him:

Jorbit Vivas: 4-for-6, 0 HR, 1 RBI [+1.08 runs vs projection] season xwOBA 0.293, below average - a good day, not a model error
Daylen Lile: 1-for-6, 1 HR, 2 RBI [+0.93 runs vs projection] season xwOBA 0.326, above average - real projection error
Nasim Nuñez: 1-for-4, 0 HR, 1 RBI [+0.8 runs vs projection] season xwOBA 0.268, 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)
WAS	Andrew Alvarez (Starter)	18 / 17	0 / 1.59	0 / 1	5 / 3	0 / 0	1 / 1	2.83	8 / 5	5.43	venue away: 3.79 vs 2.65 ERA (worse here) · opp run factor 0.98 (average) · 23 prior appearances this season
WAS	Erik Tolman (Closer)	1 / 3	27 / 0 (+27.00)	1 / 0	3 / 0	0 / 0	0 / 1	0.51	1 / 1	0.88
STL	Michael McGreevy (Starter)	12 / 17	4.5 / 3.18	2 / 2	5 / 3	1 / 0	1 / 3	1.96	1 / 7	3.85	venue home: 3.86 vs 3.79 ERA (ok) · opp run factor 1.05 (strong) · 29 prior appearances this season
STL	Riley O'Brien (Closer)	3 / 5	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0.58	2 / 0	0.93
STL	Ryne Stanek (Middle Reliever (High Leverage))	2 / 3	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0.47	0 / 0	1.03
STL	George Soriano (Setup Man)	3 / 2	0 / 40.5 (-40.50)	0 / 3	0 / 4	0 / 0	0 / 0	0.43	1 / 0	0.58

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
Jordan Walker (RF)	STL	6 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 0	4 / 1	-1.4	0.344 elite	bad day
Jorbit Vivas (3B)	WAS	6 / 3	4 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 1	1 / 0	1.08	0.293 below average	good day
Daylen Lile (LF)	WAS	6 / 4	1 / 1	0 / 0	0 / 0	1 / 0	2 / 1	1 / 1	0 / 0	3 / 1	0.93	0.326 above average	projection too low
Iván Herrera (DH)	STL	5 / 4	3 / 1	1 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	0.92	0.331 above average	projection too low
Nasim Nuñez (SS)	WAS	4 / 3	1 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	1 / 0	1 / 0	0.8	0.268 weak	good day
James Wood (RF)	WAS	5 / 4	2 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	0 / 1	1 / 1	-0.79	0.412 elite	bad day
CJ Abrams (2B)	WAS	5 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	0 / 1	-0.78	0.329 above average	bad day
Nathan Church (CF)	STL	5 / 3	2 / 1	1 / 0	0 / 0	0 / 0	0 / 1	2 / 0	0 / 0	0 / 0	0.78	0.29 below average	good day
José Fermín (3B)	STL	5 / 4	3 / 1	0 / 1	0 / 0	0 / 0	1 / 1	0 / 0	1 / 1	0 / 0	0.63	0.291 below average	good day
Abimelec Ortiz (DH)	WAS	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	-0.47	0.282 weak	—
Pedro Pagés (C)	STL	4 / 3	2 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 0	1 / 1	0.47	0.267 weak	—
Andrés Chaparro (1B)	WAS	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 0	0 / 1	0.33	0.347 elite	—
Masyn Winn (SS)	STL	5 / 4	1 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 1	1 / 1	-0.33	0.298 below average	—
Keibert Ruiz (C)	WAS	5 / 3	1 / 1	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	-0.31	0.289 below average	—
Leonardo Bernal (1B)	STL	6 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 1	1 / 1	0 / 0	2 / 1	-0.31	0.385 elite	—
Thomas Saggese (2B)	STL	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	2 / 1	-0.14	0.317 average	—
Dylan Crews (CF)	WAS	5 / 4	1 / 1	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	0 / 0	1 / 1	0	0.313 average	—
Victor Scott II (LF)	STL	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0	0.274 weak	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-0.255	positive favors home; negative favors away
Recent form	-0.384	positive favors home; negative favors away
Starting-pitcher matchup	-0.071	positive favors home; negative favors away
Composite	-0.572	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.27-3.99	n/a	Base ML historical model prediction
variance_applied	4-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-4	0.543	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	4.27-3.99	4-4	-0.277	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-4	5-4	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	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
WAS	Dylan Crews	batter / CF	better	predicted_RBI 2.576
STL	Pedro Pagés	batter / C	worse	predicted_RBI 1.86
WAS	James Wood	batter / RF	worse	predicted_SO 1.693
STL	Victor Scott II	batter / LF	better	predicted_H 1.541
WAS	Andrés Chaparro	batter / 1B	worse	predicted_AB 1.457
STL	Nathan Church	batter / CF	worse	predicted_SO 1.38
WAS	CJ Abrams	batter / 2B	better	predicted_RBI 1.326
WAS	Jorbit Vivas	batter / 3B	worse	predicted_AB 1.321

Structural audit findings

Severity	Rule	Finding	Subject
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_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	STL model-prediction batting totals differ from final simulation box score (AB: current=33, sim=35; HR: current=1, sim=0; BB: current=3, sim=2; SO: current=7, sim=10).	STL
info	CURRENT_PLAYER_MATCH_BOX	STL current finalized hitter lines are model predictions and differ from the simulation box score: José Fermín: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, SO current=0 box=1; Leonardo Bernal: 2B current=1 box=0, RBI current=1 box=0, R current=1 box=0; Thomas Saggese: AB current=4 box=3, R current=1 box=0, BB current=0 box=1, SO current=1 box=2; Iván Herrera: H current=1 box=0, R current=1 box=0, BB current=1 box=0, SO current=1 box=0; Masyn Winn: H current=1 box=2, 2B current=0 box=1, R current=0 box=1, BB current=1 box=0; Victor Scott II: AB current=3 box=4, H current=0 box=2, RBI current=0 box=1, R current=0 box=1; Nathan Church: AB current=3 box=4, H current=1 box=0, RBI current=1 box=0, SO current=0 box=2; Jordan Walker: 2B current=0 box=1, HR current=1 box=0	STL
info	CURRENT_FINALIZED_MATCH_BOX	WAS model-prediction batting totals differ from final simulation box score (AB: current=32, sim=35; HR: current=1, sim=0; RBI: current=4, sim=5; R: current=4, sim=5; SO: current=6, sim=7).	WAS
info	CURRENT_PLAYER_MATCH_BOX	WAS current finalized hitter lines are model predictions and differ from the simulation box score: Abimelec Ortiz: H current=1 box=0, SO current=1 box=0; Dylan Crews: AB current=4 box=3, H current=1 box=2, 2B current=0 box=1, RBI current=0 box=3; CJ Abrams: AB current=4 box=3, H current=1 box=2, RBI current=1 box=2, SO current=1 box=0; James Wood: AB current=4 box=5, HR current=1 box=0, RBI current=1 box=0, BB current=1 box=0; Daylen Lile: H current=1 box=2, RBI current=1 box=0, BB current=0 box=1, SO current=1 box=0; Nasim Nuñez: AB current=3 box=4, SO current=0 box=2; Andrés Chaparro: AB current=3 box=4, H current=1 box=0, RBI current=1 box=0; Jorbit Vivas: AB current=3 box=4, H current=1 box=0, BB current=1 box=0	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 20260918. 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	0.2413	0.2434	+0.002099	regressed
accuracy	0.5833	0.6083	+0.025	improved
team_run_mae	2.4333	2.3917	-0.0416	improved
total_run_mae	3.2333	3.2833	+0.05	regressed
calibration_error	0.0572	0.0875	+0.0303	regressed
starter_er_mae	1.686	1.6653	-0.0207	improved
starter_outs_mae	3.3721	3.2782	-0.0939	improved

Deep slate report — 20260919

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)

ATL@HOUHIT

Player-level breakdown (projected vs actual)

Fundamental evidence entering the pick

Score and confidence evolution

Official postgame stage attribution

Largest player-level adjustments from the ultra-detailed trace

Structural audit findings

BOS@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

CHC@CINHIT

Player-level breakdown (projected vs actual)

Fundamental evidence entering the pick

Score and confidence evolution

Official postgame stage attribution

Largest player-level adjustments from the ultra-detailed trace

Structural audit findings

DET@CHWMISS

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

KC@PITHIT

Player-level breakdown (projected vs actual)

Fundamental evidence entering the pick

Score and confidence evolution

Official postgame stage attribution

Largest player-level adjustments from the ultra-detailed trace

Structural audit findings

MIA@SDHIT

Player-level breakdown (projected vs actual)

Fundamental evidence entering the pick

Score and confidence evolution

Official postgame stage attribution

Largest player-level adjustments from the ultra-detailed trace

Structural audit findings

MIL@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

MIN@LAAHIT

Player-level breakdown (projected vs actual)

Fundamental evidence entering the pick

Score and confidence evolution

Official postgame stage attribution

Largest player-level adjustments from the ultra-detailed trace

Structural audit findings

NYY@ARIMISS

Player-level breakdown (projected vs actual)

Fundamental evidence entering the pick

Score and confidence evolution

Official postgame stage attribution

Largest player-level adjustments from the ultra-detailed trace

Structural audit findings

OAK@CLEHIT

Player-level breakdown (projected vs actual)