Deep slate report 20260918

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_20260918.json	378918
forensic deep audit	loaded	probability, provenance, starter and decision contracts	/app/output/reports/PREDICTION_SYSTEM_DEEP_AUDIT_20260918.json	241072
continuous learning	loaded	pipeline-stage quality, calibration, probability lineage, score distribution, and official starter comparisons	/app/output/learning/continuous_learning_daily_20260918.json	2578226
detailed report	loaded	full game and simulation rendering	/app/output/reports/detailed_20260918.html	353538
ultra detailed player audit	loaded	player-level prediction trace	/app/output/reports/20260918_ultra_detailed_player_audit.html	33798915
structural audit	loaded	all invariant findings	/app/output/reports/audit_report_20260918.html	1808590
integrity revision audit	missing	post-Luna remediation proof for immutable revision artifacts	/app/output/reports/audit_report_20260918_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 2 game(s) have an unexplained ≥2pp calibrated-to-published delta	KC@PIT, MIL@BAL	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, DET@CHW, 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	model_validation	Validated run models explain very little holdout variance home/away/total R² = 0.0157/0.0327/0.0073	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.
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	0	1/11	no	param
pipeline_winner_instability	1	3/11	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	0	6/11	no	param
weak_run_model_signal	11	11/11	yes	code/diagnostic

Metric	Direction	Series (old → new)
winner_accuracy	improving	[0.7333, 0.6667, 0.6, 0.6667, 0.5333, 0.6, 0.7, 0.4, 0.6667, 0.4444, 0.6667]
weighted_calibration_error	improving	[0.1825, 0.1069, 0.1171, 0.2091, 0.1711, 0.0827, 0.2812, 0.2795, 0.1661, 0.1016, 0.1247]
margin_bias	regressing	[-0.867, -2.133, -1.0, -2.933, -2.667, -1.8, -0.4, -1.667, -1.133, -2.111, -2.0]
team_run_error_on_losses	improving	[6.75, 5.6, 6.5, 5.0, 6.143, 5.0, 6.333, 5.556, 6.4, 3.6, 5.2]

Stage	Winner hits	Accuracy	Team-run MAE	Total-run MAE	Margin MAE
historical_ml	11/15	73.3%	2.217	2.434	3.534
variance_applied	8/15	53.3%	2.233	2.6	3.4
enhanced_prediction	10/15	66.7%	2.4	2.667	3.733
embedded_savant_run_environment	9/15	60%	2.4	2.667	3.733
fundamental_edge_target_review	10/15	66.7%	2.4	2.8	3.733
capped_target	10/15	66.7%	2.4	2.8	3.733
simulation_result	10/15	66.7%	2.4	2.8	3.733
published_final	10/15	66.7%	2.4	2.8	3.733

Transition	Δ accuracy	Δ hits	Δ team-run MAE	Δ margin MAE	Games whose winner/tie changed
historical_ml → variance_applied	-0.2	-3	0.0166	-0.1339	ATL@HOU: ATL → TIE; DET@CHW: DET → TIE; NYY@ARI: ARI → TIE; PHI@NYM: PHI → TIE; WAS@STL: WAS → TIE
variance_applied → enhanced_prediction	0.1334	2	0.1667	0.3333	ATL@HOU: TIE → ATL; CHC@CIN: CIN → CHC; DET@CHW: TIE → DET; MIN@LAA: MIN → LAA; NYY@ARI: TIE → NYY; PHI@NYM: TIE → PHI; TOR@TEX: TOR → TEX; WAS@STL: TIE → STL
enhanced_prediction → embedded_savant_run_environment	-0.0667	-1	0	0	MIN@LAA: LAA → TIE; TOR@TEX: TEX → TIE
embedded_savant_run_environment → fundamental_edge_target_review	0.0667	1	0	0	MIN@LAA: TIE → LAA; TOR@TEX: TIE → TEX
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	0.541	0.5725	0.5725	0.5725	0.5616	no recorded transformation explains the publication delta
BOS@TB	miss	0.618	0.6244	0.6244	0.6001	0.5723	score coherence followed by matchup-recency guardrail
CHC@CIN	miss	0.5404	0.6582	0.6582	0.6536	0.53	score coherence followed by matchup-recency guardrail
DET@CHW	hit	0.5763	0.6129	0.6129	0.6129	0.5435	score coherence followed by matchup-recency guardrail
KC@PIT	hit	0.6541	0.627	0.627	0.6014	0.5862	no recorded transformation explains the publication delta
MIA@SD	hit	0.6803	0.6605	0.6605	0.6547	0.58	score coherence followed by matchup-recency guardrail
MIL@BAL	hit	0.6786	0.7376	0.7376	0.7272	0.6657	no recorded transformation explains the publication delta
MIN@LAA	miss	0.508	0.5594	0.5594	0.5594	0.5343	score coherence followed by matchup-recency guardrail
NYY@ARI	hit	0.5	0.505	0.505	0.505	0.505	published probability matches the calibrated/governed value
OAK@CLE	hit	0.7538	0.698	0.698	0.698	0.58	score coherence followed by matchup-recency guardrail
PHI@NYM	miss	0.5233	0.516	0.516	0.516	0.5115	published probability matches the calibrated/governed value
SEA@COL	hit	0.776	0.7605	0.7605	0.7605	0.7605	published probability matches the calibrated/governed value
SF@LAD	hit	0.626	0.6	0.6	0.5879	0.5383	score coherence followed by matchup-recency guardrail
TOR@TEX	hit	0.5	0.505	0.505	0.505	0.505	published probability matches the calibrated/governed value
WAS@STL	miss	0.5093	0.511	0.511	0.511	0.5094	published probability matches the calibrated/governed value

Sides analyzed	Broke on pitching	Broke on offense	Both	Variance	Mean starter ERA error	Mean lineup runs vs proj
30	6	5	3	16	3.95	0.3

Game	Trained p(home)	Blended p(home)	Blend favoured	Published p(home)	Published side	Artifact says winner changed	Outcome
NYY@ARI	0.63	0.563	home	0.495	away	yes	hit
TOR@TEX	0.315	0.406	away	0.505	home	yes	hit

Slate	Games	Winner acc	Mean starter ERA error	Mean lineup runs vs proj	Sides broke on pitching	Sides broke on offense
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
20260906	15	0.600	0.74	-0.30	6	10

Game	Pred margin	Actual margin	Pred − actual
ATL@HOU	1	4	-3
BOS@TB	1	2	-1
CHC@CIN	2	2	0
DET@CHW	2	3	-1
KC@PIT	1	3	-2
MIA@SD	2	6	-4
MIL@BAL	3	1	2
MIN@LAA	1	3	-2
NYY@ARI	1	7	-6
OAK@CLE	4	2	2
PHI@NYM	1	3	-2
SEA@COL	4	1	3
SF@LAD	2	6	-4
TOR@TEX	1	6	-5
WAS@STL	1	8	-7

ATL@HOUHIT

Predicted: ATL 4-3 (conf -0.12, moderate)Actual: ATL 6-2Brier 0.192total bias -1

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 3; total bias -1
pregame home-minus-away evidence: season -0.784, recent -0.068, pitcher +0.090, composite -0.431
confidence pipeline reached 0.573 before later publication/recency controls; published home probability was 0.438
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
ATL	Tyler Mahle	3 ER / 18 outs	1 ER / 18 outs	-2
HOU	Peter Lambert	2 ER / 16 outs	2 ER / 16 outs	0

Decision internals: run factors home×1.045 / away×0.955; savant lineup adjustment applied

Player-level breakdown (projected vs actual)

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

Top deviations that drove the game:

Michael Harris II (ATL CF): 3-for-5, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.56]
Drake Baldwin (ATL DH): 2-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact 0.78]
Mauricio Dubón (ATL SS): 2-for-5, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact 0.78]
Jeremy Peña (HOU SS): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.78]
Matt Olson (ATL 1B): 3-for-5, 1 HR, 2 RBI (projected 1.0 H, 1.0 HR) [run impact 0.61]

Starter miss: Tyler Mahle (ATL) posted 1.50 ERA vs projected 4.50 (1 ER in 18 outs); he is materially worse pitching away (5.61 ERA there vs 2.29 on the other side), a venue-role weakness the model under-weighted.

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

Pitcher versus batter: what actually broke this game

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

Tyler Mahle (ATL) threw 18 outs, 1 ER = 1.5 ERA against a projected 4.5 ERA (-3 ERA error)

Bullpen behind him: 1 ER over 9 outs.

The HOME lineup he faced produced 2 runs on 5 hits and 0 HR, against a projected 4 runs / 7 hits (-2 runs vs projection).

No individual hitter beat his projection by a meaningful margin.

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

HOME pitching

Peter Lambert (HOU) threw 16 outs, 2 ER = 3.38 ERA against a projected 3.38 ERA (0 ERA error)

Bullpen behind him: 2 ER over 3 outs.

The AWAY lineup he faced produced 6 runs on 11 hits and 3 HR, against a projected 5 runs / 7 hits (1 runs vs projection).

Who did the damage to him:

Michael Harris II: 3-for-5, 1 HR, 2 RBI [+1.56 runs vs projection] season xwOBA 0.339, elite - real projection error
Drake Baldwin: 2-for-4, 0 HR, 0 RBI [+0.78 runs vs projection] season xwOBA 0.364, elite - real projection error
Mauricio Dubón: 2-for-5, 0 HR, 0 RBI [+0.78 runs vs projection] season xwOBA 0.305, average - a good day, not a model error

Verdict: offense: the lineup scored well below its projection

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	Tyler Mahle (Starter)	18 / 18	1.5 / 4.5 (-3.00)	1 / 3	3 / 3	0 / 0	1 / 3	1.06	6 / 4	4.7	venue away: 5.61 vs 2.29 ERA (worse here) · opp run factor 1.04 (average) · 25 prior appearances this season
ATL	Didier Fuentes (Middle Reliever (High Leverage))	6 / 3	4.5 / 0 (+4.50)	1 / 0	2 / 0	0 / 0	1 / 0	0.27	3 / 0	1.58
ATL	Raisel Iglesias (Closer)	3 / 3	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	0.25	3 / 1	1.01
HOU	Peter Lambert (Starter)	16 / 16	3.38 / 3.38	2 / 2	6 / 5	1 / 1	1 / 2	2.1	6 / 2	4.68	venue home: 4.03 vs 3.17 ERA (worse here) · opp run factor 0.96 (strong) · 26 prior appearances this season
HOU	AJ Blubaugh (Middle Reliever (High Leverage))	3 / 4	18 / 13.5 (+4.50)	2 / 2	1 / 1	1 / 1	1 / 1	0.66	1 / 0	1.21

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
Michael Harris II (CF)	ATL	5 / 4	3 / 1	0 / 1	0 / 0	1 / 0	2 / 1	1 / 1	0 / 0	1 / 1	1.56	0.339 elite	projection too low
Drake Baldwin (DH)	ATL	4 / 4	2 / 1	1 / 0	0 / 0	0 / 0	0 / 1	2 / 1	1 / 1	1 / 1	0.78	0.364 elite	projection too low
Mauricio Dubón (SS)	ATL	5 / 4	2 / 1	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0.78	0.305 average	no signal
Jeremy Peña (SS)	HOU	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	2 / 1	-0.78	0.328 above average	bad day
Matt Olson (1B)	ATL	5 / 4	3 / 1	0 / 0	0 / 0	1 / 1	2 / 1	2 / 1	0 / 1	0 / 1	0.61	0.352 elite	projection too low
Ronald Acuña Jr. (RF)	ATL	5 / 4	1 / 1	0 / 0	0 / 0	1 / 0	2 / 1	1 / 1	0 / 1	1 / 1	0.6	0.364 elite	projection too low
Austin Riley (3B)	ATL	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	3 / 1	-0.47	0.297 below average	—
Taylor Trammell (RF)	HOU	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	3 / 1	-0.47	0.294 below average	—
Christian Walker (1B)	HOU	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	2 / 1	-0.47	0.3 below average	—
Daulton Varsho (CF)	HOU	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	-0.47	0.286 below average	—
Yordan Alvarez (DH)	HOU	2 / 4	1 / 1	1 / 0	0 / 0	0 / 1	0 / 1	1 / 1	2 / 1	0 / 1	-0.29	0.444 elite	—
Ozzie Albies (2B)	ATL	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	1 / 1	-0.14	0.282 weak	—
Isaac Paredes (3B)	HOU	4 / 4	2 / 1	0 / 0	0 / 0	0 / 0	2 / 1	0 / 1	0 / 1	1 / 0	0.14	0.319 average	—
LaMonte Wade Jr. (LF)	HOU	3 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	0 / 1	0.14	0.271 weak	—
Mike Yastrzemski (LF)	ATL	4 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	0	0.309 average	—
Sean Murphy (C)	ATL	4 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0	0.256 weak	—
Jose Altuve (2B)	HOU	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	0	0.271 weak	—
Christian Vázquez (C)	HOU	2 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0	0.234 weak	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-0.784	positive favors home; negative favors away
Recent form	-0.068	positive favors home; negative favors away
Starting-pitcher matchup	0.09	positive favors home; negative favors away
Composite	-0.431	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.35-4.07	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	4.35-4.07	4-4	-0.278	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	Michael Harris II	batter / CF	worse	predicted_AB -2.55
HOU	Isaac Paredes	batter / 3B	worse	predicted_AB -1.93
ATL	Mike Yastrzemski	batter / LF	better	predicted_RBI 1.86
HOU	Daulton Varsho	batter / CF	worse	predicted_AB -1.57
HOU	LaMonte Wade Jr.	batter / LF	worse	predicted_AB 1.489
HOU	Jeremy Peña	batter / SS	worse	predicted_RBI 1.29
ATL	Austin Riley	batter / 3B	worse	predicted_SO -1.24
ATL	Sean Murphy	batter / C	worse	predicted_SO -1.231

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 pitcher projections differ from final simulation pitching box score (H: current=4, sim=5).	ATL
info	CURRENT_FINALIZED_MATCH_BOX	ATL model-prediction batting totals differ from final simulation box score (AB: current=33, sim=35; H: current=7, sim=8; 2B: current=1, sim=2; HR: current=1, sim=2; R: current=5, sim=4; SO: current=9, sim=2; HBP: current=0, sim=1).	ATL
info	CURRENT_PLAYER_MATCH_BOX	ATL current finalized hitter lines are model predictions and differ from the simulation box score: Drake Baldwin: AB current=4 box=5, H current=1 box=2, 2B current=0 box=1, HR current=0 box=1; Mauricio Dubón: SO current=1 box=0; Ronald Acuña Jr.: AB current=4 box=5, RBI current=1 box=0, R current=1 box=0, BB current=1 box=0; Ozzie Albies: H current=1 box=2, SO current=1 box=0; Michael Harris II: AB current=4 box=3, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Matt Olson: 2B current=0 box=1, HR current=1 box=0, BB current=1 box=0, SO current=1 box=0; Mike Yastrzemski: H current=0 box=1, HR current=0 box=1, RBI current=0 box=2, R current=0 box=1; Sean Murphy: BB current=0 box=1, SO current=1 box=0	ATL
info	CURRENT_FINALIZED_MATCH_BOX	HOU model-prediction batting totals differ from final simulation box score (AB: current=33, sim=32; H: current=7, sim=5; 2B: current=1, sim=2; HR: current=1, sim=0; RBI: current=4, sim=3; R: current=4, sim=3; BB: current=3, sim=4; SO: current=8, sim=6).	HOU
info	CURRENT_PLAYER_MATCH_BOX	HOU current finalized hitter lines are model predictions and differ from the simulation box score: Jose Altuve: RBI current=0 box=1; Isaac Paredes: AB current=4 box=2, H current=1 box=0, RBI current=1 box=0, R current=1 box=0; Jeremy Peña: AB current=4 box=5, RBI current=1 box=2, SO current=1 box=2; Yordan Alvarez: 2B current=0 box=1, HR current=1 box=0, RBI current=1 box=0, SO current=1 box=0; Taylor Trammell: AB current=3 box=2, H current=1 box=0, BB current=0 box=1, SO current=1 box=0; Christian Walker: H current=1 box=0, RBI current=1 box=0, SO current=1 box=0; Daulton Varsho: AB current=4 box=3, H current=1 box=0, BB current=0 box=1; LaMonte Wade Jr.: AB current=3 box=4, H current=0 box=1, BB current=1 box=0, SO current=1 box=0	HOU
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

BOS@TBMISS

Predicted: TB 4-5 (conf +0.14, moderate)Actual: BOS 4-2Brier 0.328total bias +3

Root cause: Backed the wrong side with real conviction

Evidence-backed diagnosis

published system edge 0.14 (moderate); this is not the half-scale distance from 50%
team-run absolute error 3; total bias +3
pregame home-minus-away evidence: season +0.330, recent +0.570, pitcher +0.093, composite +0.709
confidence pipeline reached 0.624 before later publication/recency controls; published home probability was 0.572
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 3 (9 vs 6)
TB offense over-projected by 3
Ian Seymour (TB) allowed 3 more ER than projected (4 vs 1)
committed to TB at 0.14 confidence and it lost

Team	Starter	Projected	Actual	ER err
BOS	Ranger Suarez	2 ER / 17 outs	1 ER / 17 outs	-1
TB	Ian Seymour	1 ER / 15 outs	4 ER / 15 outs	3

Decision internals: run factors home×1.105 / away×1.012; 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:

Wilyer Abreu (BOS RF): 0-for-5, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.73]
Nick Sogard (BOS 1B): 0-for-5, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -1.11]
Roman Anthony (BOS LF): 2-for-5, 1 HR, 3 RBI (projected 1.0 H, 0.0 HR) [run impact 1.07]
Jahmai Jones (BOS DH): 1-for-3, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 0.93]
Nate Eaton (BOS CF): 1-for-3, 0 HR, 0 RBI (projected 0.0 H, 0.0 HR) [run impact 0.8]

Starter miss: Ian Seymour (TB) posted 7.20 ERA vs projected 1.80 (4 ER in 15 outs); he is materially worse pitching home (5.02 ERA there vs 3.75 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

Ranger Suarez (BOS) threw 17 outs, 1 ER = 1.59 ERA against a projected 3.18 ERA (-1.59 ERA error)

Bullpen behind him: 1 ER over 3 outs.

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

Who did the damage to him:

Nick Fortes: 1-for-1, 0 HR, 0 RBI [+0.64 runs vs projection] season xwOBA 0.276, weak - a good day, not a model error
Ryan Vilade: 1-for-4, 0 HR, 0 RBI [+0.33 runs vs projection] season xwOBA 0.333, above average - a good day, not a model error
Jonathan Aranda: 1-for-5, 0 HR, 1 RBI [+0.31 runs vs projection] season xwOBA 0.347, elite - a good day, not a model error

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

HOME pitching

Ian Seymour (TB) threw 15 outs, 4 ER = 7.2 ERA against a projected 1.8 ERA (5.4 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 3 runs / 8 hits (1 runs vs projection).

Who did the damage to him:

Roman Anthony: 2-for-5, 1 HR, 3 RBI [+1.07 runs vs projection] season xwOBA 0.346, elite - real projection error
Jahmai Jones: 1-for-3, 1 HR, 1 RBI [+0.93 runs vs projection] season xwOBA 0.325, above average - real projection error
Nate Eaton: 1-for-3, 0 HR, 0 RBI [+0.8 runs vs projection] season xwOBA 0.273, weak - a good day, not a model error

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

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
BOS	Ranger Suarez (Starter)	17 / 17	1.59 / 3.18	1 / 2	3 / 6	0 / 1	5 / 2	1.82	5 / 10	4.49	venue away: 2.95 vs 3.81 ERA (ok) · opp run factor 1.1 (strong) · 26 prior appearances this season
BOS	Erik Miller (Setup Man)	3 / 1	9 / 0 (+9.00)	1 / 0	1 / 0	0 / 0	2 / 0	0.82	2 / 1	1.47
TB	Ian Seymour (Starter)	15 / 15	7.2 / 1.8 (+5.40)	4 / 1	7 / 3	2 / 0	2 / 0	0.79	5 / 6	6.12	venue home: 5.02 vs 3.75 ERA (worse here) · opp run factor 1.01 (average) · 44 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
Wilyer Abreu (RF)	BOS	5 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	3 / 1	-1.73	0.352 elite	bad day
Nick Sogard (1B)	BOS	5 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	2 / 1	-1.11	0.317 average	no signal
Roman Anthony (LF)	BOS	5 / 4	2 / 1	0 / 0	0 / 0	1 / 0	3 / 0	1 / 0	0 / 1	1 / 1	1.07	0.346 elite	projection too low
Jahmai Jones (DH)	BOS	3 / 3	1 / 1	0 / 0	0 / 0	1 / 0	1 / 0	1 / 0	0 / 0	1 / 1	0.93	0.325 above average	projection too low
Nate Eaton (CF)	BOS	3 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	0.8	0.273 weak	good day
Jonny DeLuca (CF)	TB	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	-0.78	0.264 weak	projection too high
Nick Fortes (C)	TB	1 / 3	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0.64	0.276 weak	good day
Caleb Durbin (3B)	BOS	3 / 3	2 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	0 / 0	0 / 1	0.47	0.294 below average	—
Isiah Kiner-Falefa (2B)	BOS	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	2 / 1	-0.47	0.311 average	—
Chandler Simpson (LF)	TB	5 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	-0.47	0.265 weak	—
Yandy Díaz (DH)	TB	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	1 / 1	2 / 1	1 / 1	-0.45	0.347 elite	—
Ryan Vilade (2B)	TB	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 0	2 / 1	0.33	0.333 above average	—
Jonathan Aranda (1B)	TB	5 / 4	1 / 1	1 / 0	0 / 0	0 / 0	1 / 1	0 / 1	0 / 0	0 / 1	0.31	0.347 elite	—
Junior Caminero (3B)	TB	3 / 4	1 / 1	1 / 0	0 / 0	0 / 1	1 / 1	0 / 1	2 / 1	1 / 1	-0.29	0.363 elite	—
Taylor Walls (SS)	TB	1 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	-0.14	0.25 weak	—
Trevor Story (SS)	BOS	3 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 1	1 / 0	1 / 0	0 / 1	0.02	0.253 weak	—
Connor Wong (C)	BOS	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	0	0.284 below average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	0.33	positive favors home; negative favors away
Recent form	0.57	positive favors home; negative favors away
Starting-pitcher matchup	0.093	positive favors home; negative favors away
Composite	0.709	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.02-4.73	n/a	Base ML historical model prediction
variance_applied	4-5	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-5	0.572	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run, matchup_recency_guardrail)
embedded_savant_run_environment	4-5	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	4-5	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	4-5	n/a	Capped at max 12 realistic runs
simulation_result	4-5	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.02-4.73	4-5	0.284	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	Ian Seymour	pitcher / Starter	better	predicted_SO 3
BOS	Nick Sogard	batter / 1B	better	predicted_RBI 2.673
TB	Chandler Simpson	batter / LF	worse	predicted_SO 2.671
BOS	Connor Wong	batter / C	deboosted	predicted_H 2.37
TB	Cam Booser	pitcher / Middle Reliever (High Leverage)	worse	predicted_SO -2
BOS	Trevor Story	batter / SS	worse	predicted_2B 1.754
BOS	Nate Eaton	batter / CF	worse	predicted_H 1.688
TB	Ryan Vilade	batter / 2B	boosted	predicted_RBI 1.519

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=31, sim=38; H: current=8, sim=11; 2B: current=2, sim=4; R: current=3, sim=4; BB: current=3, sim=0; SO: current=9, sim=6; HBP: current=0, sim=2).	BOS
info	CURRENT_PLAYER_MATCH_BOX	BOS current finalized hitter lines are model predictions and differ from the simulation box score: Nick Sogard: AB current=4 box=5, H current=1 box=3, RBI current=0 box=3, BB current=1 box=0; Trevor Story: H current=1 box=2, 2B current=1 box=2, RBI current=1 box=0, R current=0 box=1; Caleb Durbin: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, HBP current=0 box=1; Isiah Kiner-Falefa: AB current=3 box=4, HR current=0 box=1, R current=0 box=1; Nate Eaton: AB current=3 box=4, H current=0 box=2, SO current=1 box=0; Jahmai Jones: AB current=3 box=4, H current=1 box=0, SO current=1 box=2, HBP current=0 box=1; Roman Anthony: AB current=4 box=5, H current=1 box=0, BB current=1 box=0, SO current=1 box=0; Wilyer Abreu: AB current=4 box=5, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0	BOS
info	CURRENT_FINALIZED_MATCH_BOX	TB model-prediction batting totals differ from final simulation box score (AB: current=32, sim=34; 2B: current=2, sim=1; RBI: current=4, sim=5; BB: current=2, sim=4; SO: current=7, sim=14).	TB
info	CURRENT_PLAYER_MATCH_BOX	TB current finalized hitter lines are model predictions and differ from the simulation box score: Ryan Vilade: AB current=3 box=4, HR current=0 box=1, RBI current=1 box=2, R current=0 box=1; Yandy Díaz: 2B current=1 box=0; Jonny DeLuca: H current=1 box=2, 2B current=1 box=0, R current=0 box=1, SO current=1 box=2; Chandler Simpson: R current=1 box=0, SO current=0 box=3; Junior Caminero: HR current=1 box=0, RBI current=1 box=0, BB current=1 box=0, SO current=1 box=2; Jonathan Aranda: H current=1 box=2, 2B current=0 box=1, RBI current=1 box=2, BB current=0 box=1; Nick Fortes: H current=1 box=0, BB current=0 box=1, SO current=0 box=1; Victor Mesa Jr.: H current=1 box=0, R current=1 box=0, BB current=0 box=1, SO current=1 box=2	TB
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

CHC@CINMISS

Predicted: CHC 6-4 (conf -0.06, lean)Actual: CIN 4-6Brier 0.281total bias +0

Root cause: Small edge on the losing side

Evidence-backed diagnosis

published system edge 0.06 (lean); this is not the half-scale distance from 50%
team-run absolute error 4; total bias +0
pregame home-minus-away evidence: season -1.677, recent -1.002, pitcher -0.063, composite -1.671
confidence pipeline reached 0.658 before later publication/recency controls; published home probability was 0.470
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

Clay Holmes (CHC) allowed 3 more ER than projected (6 vs 3)

Team	Starter	Projected	Actual	ER err
CHC	Clay Holmes	3 ER / 17 outs	6 ER / 16 outs	3
CIN	Chase Burns	3 ER / 15 outs	2 ER / 9 outs	-1

Decision internals: run factors home×0.947 / 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:

Pete Crow-Armstrong (CHC CF): 0-for-5, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.73]
Michael Conforto (CHC DH): 3-for-4, 0 HR, 0 RBI (projected 0.0 H, 0.0 HR) [run impact 1.41]
Carlos Jorge (CIN CF): 1-for-3, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact 0.95]
Carson Kelly (CHC C): 2-for-2, 0 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 0.8]
Dansby Swanson (CHC SS): 1-for-3, 0 HR, 0 RBI (projected 0.0 H, 0.0 HR) [run impact 0.8]

Starter miss: Clay Holmes (CHC) posted 10.12 ERA vs projected 4.76 (6 ER in 16 outs); he is materially worse pitching away (3.77 ERA there vs 1.94 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

Clay Holmes (CHC) threw 16 outs, 6 ER = 10.12 ERA against a projected 4.76 ERA (5.36 ERA error)

Bullpen behind him: 0 ER over 3 outs.

The HOME lineup he faced produced 6 runs on 8 hits and 0 HR, against a projected 3 runs / 7 hits (3 runs vs projection).

Who did the damage to him:

Carlos Jorge: 1-for-3, 0 HR, 0 RBI [+0.95 runs vs projection] season xwOBA 0.208, weak - a good day, not a model error

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

HOME pitching

Chase Burns (CIN) threw 9 outs, 2 ER = 6 ERA against a projected 5.4 ERA (0.6 ERA error)

Bullpen behind him: 0 ER over 6 outs.

The AWAY lineup he faced produced 4 runs on 9 hits and 1 HR, against a projected 6 runs / 7 hits (-2 runs vs projection).

Who did the damage to him:

Michael Conforto: 3-for-4, 0 HR, 0 RBI [+1.41 runs vs projection] season xwOBA 0.328, above average - real projection error
Carson Kelly: 2-for-2, 0 HR, 1 RBI [+0.8 runs vs projection] season xwOBA 0.313, average - a good day, not a model error
Dansby Swanson: 1-for-3, 0 HR, 0 RBI [+0.8 runs vs projection] season xwOBA 0.306, average - a good day, not a model error

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

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
CHC	Clay Holmes (Starter)	16 / 17	10.12 / 4.76 (+5.36)	6 / 3	3 / 2	0 / 1	5 / 1	2.77	4 / 5	4.28	venue away: 3.77 vs 1.94 ERA (worse here) · opp run factor 0.95 (weak) · 16 prior appearances this season
CHC	Jacob Webb (Closer)	3 / 4	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0.51	2 / 0	0.98
CIN	Chase Burns (Starter)	9 / 15	6 / 5.4	2 / 3	4 / 5	1 / 2	2 / 2	2.06	4 / 6	7.46	venue home: 2.09 vs 3.46 ERA (ok) · opp run factor 1.01 (strong) · 28 prior appearances this season
CIN	Brock Burke (Middle Reliever (High Leverage))	3 / 5	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0.47	1 / 0	1.02
CIN	Emilio Pagán (Closer)	3 / 4	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	0.27	1 / 1	1.25

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
Pete Crow-Armstrong (CF)	CHC	5 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	3 / 1	-1.73	0.378 elite	bad day
Michael Conforto (DH)	CHC	4 / 3	3 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1.41	0.328 above average	projection too low
Carlos Jorge (CF)	CIN	3 / 4	1 / 1	0 / 0	1 / 0	0 / 0	0 / 0	2 / 0	1 / 0	0 / 1	0.95	0.208 weak	good day
Carson Kelly (C)	CHC	2 / 3	2 / 1	0 / 0	0 / 0	0 / 0	1 / 1	1 / 1	1 / 0	0 / 1	0.8	0.313 average	no signal
Dansby Swanson (SS)	CHC	3 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	1 / 0	1 / 1	0.8	0.306 average	no signal
Seiya Suzuki (RF)	CHC	5 / 4	1 / 1	0 / 1	0 / 0	0 / 0	2 / 1	0 / 1	0 / 1	3 / 1	-0.64	0.339 elite	bad day
Ian Happ (LF)	CHC	4 / 3	1 / 1	0 / 0	0 / 0	1 / 0	1 / 0	1 / 0	0 / 1	1 / 1	0.6	0.317 average	no signal
Alex Bregman (3B)	CHC	5 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	0 / 1	-0.47	0.321 above average	—
Héctor Rodríguez (RF)	CIN	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	-0.47	0.256 weak	—
Tyler Stephenson (DH)	CIN	3 / 4	1 / 1	0 / 1	0 / 0	0 / 0	1 / 1	0 / 0	1 / 1	2 / 1	-0.31	0.332 above average	—
Michael Busch (1B)	CHC	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	2 / 1	1 / 1	-0.14	0.33 above average	—
Eugenio Suárez (3B)	CIN	4 / 3	1 / 0	0 / 0	0 / 0	0 / 0	1 / 1	1 / 0	0 / 1	1 / 1	0.14	0.298 below average	—
Elly De La Cruz (SS)	CIN	2 / 4	2 / 1	0 / 0	0 / 0	0 / 1	0 / 1	2 / 1	2 / 1	0 / 1	-0.13	0.375 elite	—
Nico Hoerner (2B)	CHC	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 1	0	0.322 above average	—
Sal Stewart (1B)	CIN	3 / 4	1 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	0 / 0	0 / 1	0	0.349 elite	—
JJ Bleday (LF)	CIN	3 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	1 / 1	0	0.339 elite	—
Jose Trevino (C)	CIN	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	2 / 0	0 / 0	0 / 0	1 / 1	0	0.217 weak	—
Matt McLain (2B)	CIN	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	0	0.3 below average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-1.677	positive favors home; negative favors away
Recent form	-1.002	positive favors home; negative favors away
Starting-pitcher matchup	-0.063	positive favors home; negative favors away
Composite	-1.671	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.05-4.87	n/a	Base ML historical model prediction
variance_applied	4-5	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	6-4	0.53	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	4.05-4.87	4-5	-0.187	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
CHC	Clay Holmes	pitcher / Starter	neutral	predicted_SO 3
CIN	Chase Burns	pitcher / Starter	neutral	predicted_SO 3
CHC	Dansby Swanson	batter / SS	better	predicted_RBI 2.58
CIN	Tyler Stephenson	batter / DH	worse	predicted_RBI 2.558
CIN	Elly De La Cruz	batter / SS	worse	predicted_AB -2.31
CHC	Michael Busch	batter / 1B	better	predicted_H 2.11
CHC	Jacob Webb	pitcher / Closer	neutral	predicted_SO -2
CIN	Brock Burke	pitcher / Middle Reliever (High Leverage)	neutral	predicted_SO -2

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=32, sim=36; H: current=7, sim=9; HR: current=1, sim=3; RBI: current=4, sim=6; BB: current=4, sim=5; SO: current=9, sim=8).	CHC
info	CURRENT_PLAYER_MATCH_BOX	CHC current finalized hitter lines are model predictions and differ from the simulation box score: Pete Crow-Armstrong: AB current=4 box=5, 2B current=0 box=1, HR current=1 box=0, RBI current=1 box=0; Seiya Suzuki: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, R current=1 box=0; Michael Busch: H current=1 box=3, HR current=0 box=1, RBI current=0 box=1; Alex Bregman: RBI current=1 box=0, BB current=0 box=1; Ian Happ: R current=0 box=1, BB current=1 box=2; Nico Hoerner: H current=1 box=0; Michael Conforto: AB current=3 box=4, H current=0 box=2, HR current=0 box=1, RBI current=0 box=2; Carson Kelly: AB current=3 box=4, H current=1 box=0, RBI current=1 box=0, R current=1 box=0	CHC
info	CURRENT_FINALIZED_MATCH_BOX	CIN model-prediction batting totals differ from final simulation box score (AB: current=33, sim=29; H: current=7, sim=3; 2B: current=1, sim=2; R: current=3, sim=4; BB: current=4, sim=3; SO: current=9, sim=5; SF: current=0, sim=1; 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, 2B current=1 box=0, HR current=0 box=1, RBI current=1 box=3; Carlos Jorge: AB current=4 box=3, H current=1 box=0, RBI current=0 box=1, R current=0 box=1; Eugenio Suárez: RBI current=1 box=0, SO current=1 box=0; Héctor Rodríguez: H current=1 box=0; Sal Stewart: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, SO current=1 box=0; JJ Bleday: H current=1 box=0, BB current=1 box=0; Elly De La Cruz: AB current=4 box=2, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0; Jose Trevino: H current=1 box=2, 2B current=0 box=2, R current=0 box=1, SO current=1 box=0	CIN
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

DET@CHWHIT

Predicted: DET 6-4 (conf -0.09, lean)Actual: DET 11-8Brier 0.208total bias -9

Root cause: Right winner, wrong score

Evidence-backed diagnosis

published system edge 0.09 (lean); this is not the half-scale distance from 50%
team-run absolute error 9; total bias -9
pregame home-minus-away evidence: season -0.451, recent -0.726, pitcher -0.048, composite -0.895
confidence pipeline reached 0.613 before later publication/recency controls; published home probability was 0.457
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 9 (10 vs 19)
DET offense under-projected by 5
CHW offense under-projected by 4

Team	Starter	Projected	Actual	ER err
DET	Andrew Sears	2 ER / 14 outs	0 ER / 2 outs	-2
CHW	David Sandlin	3 ER / 16 outs	5 ER / 12 outs	2

Decision internals: run factors home×0.986 / away×1.034; 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:

Hao-Yu Lee (DET 3B): 4-for-5, 1 HR, 3 RBI (projected 1.0 H, 0.0 HR) [run impact 2.65]
Munetaka Murakami (CHW 1B): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.4]
Miguel Vargas (CHW 3B): 2-for-5, 1 HR, 3 RBI (projected 1.0 H, 0.0 HR) [run impact 1.38]
Colson Montgomery (CHW SS): 2-for-4, 0 HR, 3 RBI (projected 0.0 H, 0.0 HR) [run impact 1.25]
Kevin McGonigle (DET SS): 2-for-5, 0 HR, 4 RBI (projected 1.0 H, 0.0 HR) [run impact 1.09]

Starter miss: Andrew Sears (DET) posted 0.00 ERA vs projected 3.86 (0 ER in 2 outs). Thin prior record (4 prior appearances this season): high projection uncertainty.

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

Pitcher versus batter: what actually broke this game

AWAY pitching

Andrew Sears (DET) threw 2 outs, 0 ER = 0 ERA against a projected 3.86 ERA (-3.86 ERA error) only 4 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 HOME lineup he faced produced 5 runs on 4 hits and 1 HR, against a projected 4 runs / 5 hits (1 runs vs projection).

Who did the damage to him:

Miguel Vargas: 2-for-5, 1 HR, 3 RBI [+1.38 runs vs projection] season xwOBA 0.378, elite - real projection error
Colson Montgomery: 2-for-4, 0 HR, 3 RBI [+1.25 runs vs projection] season xwOBA 0.302, below average - a good day, not a model error

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

HOME pitching

No settled pitching line for this club. The forecast named David Sandlin (CHW) at a projected 5.06 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 10 runs on 14 hits and 2 HR, against a projected 5 runs / 7 hits (5 runs vs projection).

Who did the damage to him:

Hao-Yu Lee: 4-for-5, 1 HR, 3 RBI [+2.65 runs vs projection] season xwOBA 0.297, below average - a good day, not a model error
Kevin McGonigle: 2-for-5, 0 HR, 4 RBI [+1.09 runs vs projection] season xwOBA 0.357, elite - real projection error
Spencer Torkelson: 2-for-4, 0 HR, 0 RBI [+0.94 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	Andrew Sears (Starter)	2 / 14	0 / 3.86 (-3.86)	0 / 2	1 / 1	0 / 0	3 / 1	1.64	0 / 6	4.43	venue away: 1.59 vs 8.31 ERA (ok) · opp run factor 0.99 (average) · thin sample: 4 prior appearances this season (high uncertainty)
DET	Drew Sommers (Middle Reliever (High Leverage))	6 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0.32	5 / 0	1.31
DET	Kenley Jansen (Closer)	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0.56	1 / 1	1.08

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
Hao-Yu Lee (3B)	DET	5 / 3	4 / 1	1 / 0	0 / 0	1 / 0	3 / 1	2 / 0	0 / 0	0 / 1	2.65	0.297 below average	good day
Munetaka Murakami (1B)	CHW	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	1 / 1	2 / 1	-1.4	0.342 elite	bad day
Miguel Vargas (3B)	CHW	5 / 4	2 / 1	1 / 0	0 / 0	1 / 0	3 / 1	1 / 1	0 / 1	1 / 1	1.38	0.378 elite	projection too low
Colson Montgomery (SS)	CHW	4 / 3	2 / 0	1 / 0	0 / 0	0 / 0	3 / 0	1 / 0	0 / 0	2 / 1	1.25	0.302 below average	good day
Kevin McGonigle (SS)	DET	5 / 4	2 / 1	0 / 0	1 / 0	0 / 0	4 / 0	1 / 1	1 / 1	0 / 0	1.09	0.357 elite	projection too low
Spencer Torkelson (1B)	DET	4 / 3	2 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	1 / 1	0.94	0.314 average	no signal
Riley Greene (LF)	DET	4 / 4	1 / 1	0 / 1	0 / 0	1 / 0	2 / 1	1 / 1	1 / 1	2 / 1	0.62	0.372 elite	projection too low
Kyle Teel (C)	CHW	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	0 / 0	1 / 1	-0.47	0.32 average	—
Chase Meidroth (2B)	CHW	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	1 / 1	1 / 1	-0.47	0.295 below average	—
Max Clark (CF)	DET	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	1 / 0	0 / 1	0.33	0.31 average	—
Colt Keith (DH)	DET	2 / 3	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 1	0.31	0.337 above average	—
Eduardo Valencia (C)	DET	4 / 4	2 / 1	1 / 0	0 / 0	0 / 1	0 / 1	3 / 1	0 / 0	1 / 1	-0.15	0.338 above average	—
Tristan Peters (RF)	CHW	1 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	1 / 1	-0.14	0.316 average	—
Gleyber Torres (2B)	DET	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	0 / 1	0	0.325 above average	—
Zach McKinstry (RF)	DET	2 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	0	0.284 below average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-0.451	positive favors home; negative favors away
Recent form	-0.726	positive favors home; negative favors away
Starting-pitcher matchup	-0.048	positive favors home; negative favors away
Composite	-0.895	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.48-4.29	n/a	Base ML historical model prediction
variance_applied	4-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	6-4	0.543	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by winner_starter_is_below_average, matchup_recency_guardrail)
embedded_savant_run_environment	6-4	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	6-4	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	6-4	n/a	Capped at max 12 realistic runs
simulation_result	6-4	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.48-4.29	4-4	-0.189	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-4	6-4	2	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by winner_starter_is_below_average, matchup_recency_guardrail)
embedded_savant_run_environment	6-4	6-4	-0	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	6-4	6-4	-0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	6-4	6-4	-0	Capped at max 12 realistic runs
simulation_result	6-4	6-4	-0	Final inning-by-inning simulation result

Largest player-level adjustments from the ultra-detailed trace

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

Team	Player	Role	Net effect	Largest change
CHW	Colson Montgomery	batter / SS	worse	predicted_RBI 2.64
DET	Hao-Yu Lee	batter / 3B	better	predicted_RBI 2.571
DET	Colt Keith	batter / DH	better	predicted_AB 2.173
CHW	Tommy Pham	batter / DH	worse	predicted_AB 1.5
CHW	Randal Grichuk	batter / LF	boosted	predicted_AB 1.5
DET	Max Clark	batter / CF	worse	predicted_RBI 1.47
DET	Zach McKinstry	batter / RF	boosted	predicted_H 1.43
DET	Kevin McGonigle	batter / SS	worse	predicted_SO 1.39

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=32; H: current=8, sim=5; 2B: current=1, sim=2; BB: current=3, sim=4).	CHW
info	CURRENT_PLAYER_MATCH_BOX	CHW current finalized hitter lines are model predictions and differ from the simulation box score: Tommy Pham: AB current=3 box=4, H current=1 box=0, SO current=1 box=2; Randal Grichuk: RBI current=1 box=0, R current=0 box=1; Munetaka Murakami: AB current=4 box=3, HR current=1 box=0; Brenton Doyle: H current=1 box=2, 2B current=0 box=1, R current=0 box=1, BB current=0 box=1; Miguel Vargas: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, BB current=1 box=0; Tristan Peters: H current=1 box=0, R current=1 box=0, BB current=0 box=1, SO current=1 box=0; Kyle Teel: H current=1 box=0, RBI current=1 box=0, BB current=0 box=1; Chase Meidroth: H current=1 box=0, R current=1 box=0, BB current=1 box=0	CHW
info	CURRENT_FINALIZED_MATCH_BOX	DET model-prediction batting totals differ from final simulation box score (AB: current=31, sim=35; H: current=7, sim=8; 2B: current=1, sim=4; RBI: current=4, sim=6; R: current=5, sim=6; SO: current=8, sim=12).	DET
info	CURRENT_PLAYER_MATCH_BOX	DET current finalized hitter lines are model predictions and differ from the simulation box score: Colt Keith: AB current=3 box=5, R current=0 box=1, SO current=1 box=0; Riley Greene: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, R current=1 box=0; Gleyber Torres: 2B current=0 box=1, SO current=1 box=2; Spencer Torkelson: H current=0 box=2, 2B current=0 box=1, RBI current=1 box=0, R current=0 box=1; Kevin McGonigle: H current=1 box=0, R current=1 box=0, SO current=0 box=2; Hao-Yu Lee: AB current=3 box=4, HR current=0 box=1, RBI current=1 box=3, R current=0 box=1; Eduardo Valencia: AB current=4 box=3, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0; Max Clark: AB current=3 box=4, 2B current=0 box=1, RBI current=0 box=2	DET
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

KC@PITHIT

Predicted: PIT 3-4 (conf +0.17, moderate)Actual: PIT 5-8Brier 0.171total bias -6

Root cause: Right winner, wrong score

Evidence-backed diagnosis

published system edge 0.17 (moderate); this is not the half-scale distance from 50%
team-run absolute error 6; total bias -6
pregame home-minus-away evidence: season +0.661, recent +0.256, pitcher +0.006, composite +0.622
confidence pipeline reached 0.627 before later publication/recency controls; published home probability was 0.586
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 (7 vs 13)
PIT offense under-projected by 4
Paul Skenes (PIT) allowed 3 more ER than projected (5 vs 2)

Team	Starter	Projected	Actual	ER err
KC	Randy Dobnak	3 ER / 18 outs	4 ER / 11 outs	1
PIT	Paul Skenes	2 ER / 16 outs	5 ER / 18 outs	3

Decision internals: run factors home×0.989 / away×0.983; 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:

Carter Jensen (KC C): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.73]
Henry Davis (PIT C): 1-for-3, 0 HR, 1 RBI (projected 0.0 H, 0.0 HR) [run impact 1.11]
Konnor Griffin (PIT SS): 4-for-5, 0 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.1]
Vinnie Pasquantino (KC 1B): 2-for-4, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.07]
Oneil Cruz (PIT CF): 3-for-5, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact 0.92]

Starter miss: Randy Dobnak (KC) posted 9.82 ERA vs projected 4.50 (4 ER in 11 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

Randy Dobnak (KC) threw 11 outs, 4 ER = 9.82 ERA against a projected 4.5 ERA (5.32 ERA error)

Bullpen behind him: 3 ER over 3 outs.

The HOME lineup he faced produced 8 runs on 15 hits and 1 HR, against a projected 5 runs / 8 hits (3 runs vs projection).

Who did the damage to him:

Henry Davis: 1-for-3, 0 HR, 1 RBI [+1.11 runs vs projection] season xwOBA 0.292, below average - a good day, not a model error
Konnor Griffin: 4-for-5, 0 HR, 2 RBI [+1.1 runs vs projection] season xwOBA 0.302, below average - a good day, not a model error
Oneil Cruz: 3-for-5, 0 HR, 0 RBI [+0.92 runs vs projection] season xwOBA 0.352, elite - real projection error

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

HOME pitching

Paul Skenes (PIT) threw 18 outs, 5 ER = 7.5 ERA against a projected 3.38 ERA (4.12 ERA error)

Bullpen behind him: 0 ER over 3 outs.

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

Who did the damage to him:

Vinnie Pasquantino: 2-for-4, 1 HR, 2 RBI [+1.07 runs vs projection] season xwOBA 0.328, above average - real projection error
Nick Loftin: 1-for-3, 0 HR, 0 RBI [+0.64 runs vs projection] season xwOBA 0.291, below average - a good day, not a model error
Michael Massey: 1-for-4, 1 HR, 1 RBI [+0.62 runs vs projection] season xwOBA 0.308, 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)
KC	Randy Dobnak (Starter)	11 / 18	9.82 / 4.5 (+5.32)	4 / 3	7 / 2	1 / 0	2 / 0	1.47	1 / 7	3.57	venue away: 1.84 vs 3.86 ERA (ok) · opp run factor 0.99 (average) · 13 prior appearances this season
KC	Nate Pearson (Setup Man)	3 / 3	27 / 9 (+18.00)	3 / 1	3 / 4	0 / 0	1 / 0	0.42	2 / 0	0.86
PIT	Paul Skenes (Starter)	18 / 16	7.5 / 3.38 (+4.12)	5 / 2	6 / 3	2 / 1	1 / 2	1.72	7 / 5	4.75	venue home: 3.7 vs 3.87 ERA (ok) · opp run factor 0.98 (weak) · 30 prior appearances this season
PIT	Gregory Soto (Middle Reliever (High Leverage))	3 / 4	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0.56	2 / 1	1.23

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
Carter Jensen (C)	KC	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	3 / 1	-1.73	0.314 average	no signal
Henry Davis (C)	PIT	3 / 3	1 / 0	1 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 0	1 / 1	1.11	0.292 below average	good day
Konnor Griffin (SS)	PIT	5 / 4	4 / 1	0 / 1	0 / 0	0 / 0	2 / 1	1 / 1	0 / 0	0 / 1	1.1	0.302 below average	good day
Vinnie Pasquantino (1B)	KC	4 / 4	2 / 1	0 / 0	0 / 0	1 / 0	2 / 1	2 / 0	0 / 1	0 / 1	1.07	0.328 above average	projection too low
Oneil Cruz (CF)	PIT	5 / 4	3 / 1	1 / 0	0 / 0	0 / 0	0 / 1	2 / 1	0 / 1	1 / 1	0.92	0.352 elite	projection too low
Bobby Witt Jr. (SS)	KC	4 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	0 / 1	-0.64	0.378 elite	bad day
Nick Loftin (RF)	KC	3 / 3	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0.64	0.291 below average	good day
Michael Massey (2B)	KC	4 / 4	1 / 1	0 / 1	0 / 0	1 / 0	1 / 0	1 / 0	0 / 0	0 / 1	0.62	0.308 average	no signal
Kyle Isbel (CF)	KC	2 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 1	0.47	0.258 weak	—
Salvador Perez (DH)	KC	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	2 / 1	-0.47	0.311 average	—
Spencer Horwitz (1B)	PIT	3 / 4	2 / 1	0 / 0	0 / 0	0 / 0	2 / 0	2 / 0	1 / 1	0 / 0	0.47	0.309 average	—
Bryan Reynolds (LF)	PIT	5 / 4	2 / 1	2 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	1 / 1	0.45	0.355 elite	—
John Rave (LF)	KC	2 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	1 / 1	0.33	0.32 average	—
Maikel Garcia (3B)	KC	4 / 4	1 / 1	1 / 0	0 / 0	0 / 0	2 / 1	0 / 1	0 / 0	1 / 0	0.31	0.315 average	—
Ronny Simon (RF)	PIT	4 / 4	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0 / 0	1 / 1	0.31	0.33 above average	—
Ryan O'Hearn (DH)	PIT	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 0	1 / 0	1 / 1	-0.14	0.318 average	—
Brandon Lowe (2B)	PIT	5 / 4	1 / 1	0 / 0	0 / 0	1 / 1	2 / 1	1 / 1	0 / 0	2 / 1	0	0.325 above average	—
Nick Gonzales (3B)	PIT	3 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 1	0	0.326 above average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	0.661	positive favors home; negative favors away
Recent form	0.256	positive favors home; negative favors away
Starting-pitcher matchup	0.006	positive favors home; negative favors away
Composite	0.622	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.27-5.81	n/a	Base ML historical model prediction
variance_applied	4-6	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	3-4	0.586	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run)
embedded_savant_run_environment	3-4	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	3-4	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	3-4	n/a	Capped at max 12 realistic runs
simulation_result	3-4	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.27-5.81	4-6	0.456	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-6	3-4	-1	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run)
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
KC	Bobby Witt Jr.	batter / SS	worse	predicted_AB -2.55
KC	Carter Jensen	batter / C	worse	predicted_SO 1.871
PIT	Nick Gonzales	batter / 3B	worse	predicted_AB -1.53
KC	Michael Massey	batter / 2B	worse	predicted_SO 1.417
KC	Salvador Perez	batter / DH	worse	predicted_AB -1.35
PIT	Oneil Cruz	batter / CF	worse	predicted_H -1.35
PIT	Brandon Lowe	batter / 2B	worse	predicted_H -1.13
PIT	Konnor Griffin	batter / SS	worse	predicted_SO -1.1

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 pitcher projections differ from final simulation pitching box score (H: current=6, sim=7; BB: current=0, sim=1).	KC
info	CURRENT_FINALIZED_MATCH_BOX	KC model-prediction batting totals differ from final simulation box score (AB: current=33, sim=31; H: current=8, sim=4; 2B: current=2, sim=1; RBI: current=4, sim=3; R: current=4, sim=3; BB: current=3, sim=6; 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: Maikel Garcia: 2B current=0 box=1, SO current=0 box=1; Bobby Witt Jr.: AB current=4 box=2, 2B current=1 box=0, HR current=0 box=1, RBI current=0 box=1; Carter Jensen: H current=1 box=0, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0; John Rave: H current=1 box=0, R current=1 box=0, BB current=0 box=1; Nick Loftin: H current=1 box=0, BB current=0 box=1; Vinnie Pasquantino: R current=0 box=1, BB current=1 box=0, SO current=1 box=0; Michael Massey: H current=1 box=0, 2B current=1 box=0, SO current=1 box=2; Kyle Isbel: AB current=3 box=4, H current=0 box=1	KC
info	CURRENT_FINALIZED_MATCH_BOX	PIT model-prediction batting totals differ from final simulation box score (AB: current=35, sim=32; H: current=8, sim=7; 2B: current=2, sim=1; HR: current=1, sim=0; RBI: current=5, sim=4; R: current=5, sim=4; BB: current=3, sim=1; SF: current=0, sim=1).	PIT
info	CURRENT_PLAYER_MATCH_BOX	PIT current finalized hitter lines are model predictions and differ from the simulation box score: Oneil Cruz: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, BB current=1 box=0; Konnor Griffin: 2B current=1 box=0, SO current=1 box=0; Ryan O'Hearn: AB current=4 box=3, 2B current=0 box=1, RBI current=1 box=0, SO current=1 box=0; Ronny Simon: H current=1 box=2, R current=1 box=0; Brandon Lowe: H current=1 box=0, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0; Bryan Reynolds: 2B current=1 box=0, BB current=1 box=0; Nick Gonzales: AB current=4 box=3, BB current=0 box=1; Henry Davis: H 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 4-6 (conf +0.16, moderate)Actual: SD 2-8Brier 0.176total bias +0

Root cause: Solid correct call

Evidence-backed diagnosis

published system edge 0.16 (moderate); this is not the half-scale distance from 50%
team-run absolute error 4; total bias +0
pregame home-minus-away evidence: season +0.385, recent +1.038, pitcher -0.051, composite +0.719
confidence pipeline reached 0.660 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

No large single-factor errors.

Team	Starter	Projected	Actual	ER err
MIA	Tyler Phillips	0 ER / 15 outs	0 ER / 11 outs	0
SD	Nick Pivetta	3 ER / 15 outs	1 ER / 15 outs	-2

Decision internals: run factors home×0.998 / away×1.049; 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:

Agustín Ramírez (MIA DH): 3-for-4, 1 HR, 1 RBI (projected 0.0 H, 0.0 HR) [run impact 2.34]
Manny Machado (SD DH): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.4]
Ty France (SD 1B): 2-for-4, 1 HR, 3 RBI (projected 1.0 H, 0.0 HR) [run impact 1.4]
Heriberto Hernández (MIA LF): 1-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -0.93]
Jackson Merrill (SD CF): 1-for-4, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 0.93]

Starter miss: Nick Pivetta (SD) posted 1.80 ERA vs projected 5.40 (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

Tyler Phillips (MIA) threw 11 outs, 0 ER = 0 ERA against a projected 0 ERA (0 ERA error)

Bullpen behind him: 4 ER over 3 outs.

The HOME lineup he faced produced 8 runs on 12 hits and 2 HR, against a projected 5 runs / 7 hits (3 runs vs projection).

Who did the damage to him:

Ty France: 2-for-4, 1 HR, 3 RBI [+1.4 runs vs projection] season xwOBA 0.333, above average - real projection error
Jackson Merrill: 1-for-4, 1 HR, 1 RBI [+0.93 runs vs projection] season xwOBA 0.349, elite - real projection error
Xander Bogaerts: 2-for-4, 0 HR, 0 RBI [+0.92 runs vs projection] season xwOBA 0.323, above average - real projection error

Verdict: offense: the lineup scored well below its projection

HOME pitching

Nick Pivetta (SD) threw 15 outs, 1 ER = 1.8 ERA against a projected 5.4 ERA (-3.6 ERA error) only 6 prior appearances in the covered season - the projection had little of his own record to stand on

Bullpen behind him: 0 ER over 1 outs.

The AWAY lineup he faced produced 2 runs on 10 hits and 1 HR, against a projected 5 runs / 7 hits (-3 runs vs projection).

Who did the damage to him:

Agustín Ramírez: 3-for-4, 1 HR, 1 RBI [+2.34 runs vs projection] season xwOBA 0.279, weak - a good day, not a model error
Otto Lopez: 2-for-4, 0 HR, 0 RBI [+0.78 runs vs projection] season xwOBA 0.327, above average - real projection error
Javier Sanoja: 2-for-4, 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

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
MIA	Tyler Phillips (Starter)	11 / 15	0 / 0	0 / 0	2 / 0	0 / 0	1 / 0	1.89	3 / 6	4	venue away: 5.82 vs 3.17 ERA (worse here) · opp run factor 1 (average) · 35 prior appearances this season
MIA	Tyler Zuber (Setup Man)	3 / 1	36 / 54 (-18.00)	4 / 2	4 / 3	1 / 1	0 / 0	0.37	0 / 1	1.05
SD	Nick Pivetta (Starter)	15 / 15	1.8 / 5.4 (-3.60)	1 / 3	4 / 1	1 / 0	1 / 3	0.78	4 / 6	5.18	venue home: 3.38 vs 4 ERA (ok) · opp run factor 1.05 (average) · 6 prior appearances this season
SD	Adrian Morejon (Setup Man)	1 / 3	0 / 9 (-9.00)	0 / 1	0 / 1	0 / 0	0 / 0	0.23	1 / 1	1.18

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
Agustín Ramírez (DH)	MIA	4 / 3	3 / 0	0 / 0	0 / 0	1 / 0	1 / 0	1 / 0	0 / 0	0 / 1	2.34	0.279 weak	good day
Manny Machado (DH)	SD	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	1 / 1	2 / 1	-1.4	0.333 above average	bad day
Ty France (1B)	SD	4 / 4	2 / 1	0 / 0	0 / 0	1 / 0	3 / 1	1 / 0	0 / 0	1 / 1	1.4	0.333 above average	projection too low
Heriberto Hernández (LF)	MIA	4 / 4	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 0	2 / 1	-0.93	0.35 elite	bad day
Jackson Merrill (CF)	SD	4 / 4	1 / 1	0 / 0	0 / 0	1 / 0	1 / 1	1 / 1	0 / 0	2 / 1	0.93	0.349 elite	projection too low
Xander Bogaerts (SS)	SD	4 / 4	2 / 0	1 / 0	0 / 0	0 / 0	0 / 0	2 / 0	0 / 1	1 / 0	0.92	0.323 above average	projection too low
Kyle Stowers (1B)	MIA	5 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	0 / 2	-0.8	0.324 above average	bad day
Samad Taylor (LF)	SD	3 / 3	2 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	1 / 0	0 / 1	0.8	0.306 average	no signal
Otto Lopez (SS)	MIA	4 / 4	2 / 1	1 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0 / 0	0 / 1	0.78	0.327 above average	projection too low
Griffin Conine (RF)	MIA	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	1 / 1	-0.47	0.361 elite	—
Javier Sanoja (3B)	MIA	4 / 4	2 / 1	1 / 1	0 / 0	0 / 0	1 / 1	0 / 1	0 / 0	0 / 0	0.47	0.292 below average	—
Dustin Harris (RF)	SD	5 / 4	2 / 1	0 / 0	0 / 0	0 / 0	2 / 0	1 / 1	0 / 0	2 / 0	0.47	0.336 above average	—
Jake Cronenworth (3B)	SD	3 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0 / 1	0.47	0.271 weak	—
Xavier Edwards (2B)	MIA	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	-0.33	0.312 average	—
Fernando Tatis Jr. (2B)	SD	4 / 4	1 / 1	0 / 1	0 / 0	0 / 0	1 / 1	1 / 1	1 / 1	2 / 1	-0.31	0.377 elite	—
Jakob Marsee (CF)	MIA	3 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	1 / 1	0	0.311 average	—
Joe Mack (C)	MIA	2 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0	0.287 below average	—
Ethan Salas (C)	SD	4 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0	0.178 weak	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	0.385	positive favors home; negative favors away
Recent form	1.038	positive favors home; negative favors away
Starting-pitcher matchup	-0.051	positive favors home; negative favors away
Composite	0.719	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	3.12-4.75	n/a	Base ML historical model prediction
variance_applied	3-5	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	3.12-4.75	3-5	0.371	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	3-5	4-6	0	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
MIA	Heriberto Hernández	batter / LF	worse	predicted_RBI 3.45
SD	Ty France	batter / 1B	better	predicted_RBI 2.41
SD	Fernando Tatis Jr.	batter / 2B	better	predicted_AB -1.993
SD	Jackson Merrill	batter / CF	worse	predicted_RBI 1.91
MIA	Xavier Edwards	batter / 2B	worse	predicted_AB -1.82
MIA	Otto Lopez	batter / SS	worse	predicted_H -1.52
MIA	Kyle Stowers	batter / 1B	worse	predicted_SO -1.365
MIA	Javier Sanoja	batter / 3B	worse	predicted_AB -1.33

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-prediction batting totals differ from final simulation box score (AB: current=34, sim=29; H: current=7, sim=2; HR: current=1, sim=0; R: current=5, sim=4; HBP: current=0, sim=2).	MIA
info	CURRENT_PLAYER_MATCH_BOX	MIA current finalized hitter lines are model predictions and differ from the simulation box score: Otto Lopez: H current=1 box=0, R current=1 box=0; Griffin Conine: H current=1 box=0, RBI current=1 box=0, R current=1 box=0; Kyle Stowers: AB current=4 box=3, H current=1 box=0, RBI current=1 box=0, R current=0 box=1; Heriberto Hernández: 2B current=0 box=1, HR current=1 box=0, RBI current=1 box=4, SO current=1 box=2; Javier Sanoja: AB current=4 box=3, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Jakob Marsee: AB current=4 box=3, H current=1 box=0, R current=1 box=0, BB current=1 box=0; Xavier Edwards: AB current=4 box=2, H current=1 box=0, R current=0 box=1, BB current=1 box=2; Agustín Ramírez: H current=0 box=1, R current=0 box=1	MIA
info	CURRENT_FINALIZED_MATCH_BOX	SD model-prediction batting totals differ from final simulation box score (AB: current=34, sim=29; H: current=7, sim=5; HR: current=1, sim=2; RBI: current=4, sim=6; R: current=5, sim=6; BB: current=4, sim=2; SO: current=7, sim=9; HBP: current=0, sim=2).	SD
info	CURRENT_PLAYER_MATCH_BOX	SD current finalized hitter lines are model predictions and differ from the simulation box score: Fernando Tatis Jr.: AB current=4 box=2, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Dustin Harris: AB current=4 box=3, H current=1 box=0, BB current=0 box=1; Manny Machado: HR current=1 box=0, RBI current=1 box=0, BB current=1 box=0; Jake Cronenworth: AB current=4 box=3, H current=1 box=0, BB current=1 box=0, SO current=1 box=2; Samad Taylor: H current=1 box=0, R current=1 box=0; Jackson Merrill: HR current=0 box=1, RBI current=1 box=3, SO current=1 box=2; Ty France: AB current=4 box=3, H current=1 box=2, 2B current=0 box=1, HR current=0 box=1; Xander Bogaerts: H current=0 box=1, R current=0 box=1, BB current=1 box=0, SO current=0 box=1	SD
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

MIL@BALHIT

Predicted: MIL 7-4 (conf -0.33, strong)Actual: MIL 6-5Brier 0.112total bias +0

Root cause: Strong correct call

Evidence-backed diagnosis

published system edge 0.33 (strong); this is not the half-scale distance from 50%
team-run absolute error 2; total bias +0
pregame home-minus-away evidence: season -1.247, recent -0.492, pitcher -0.062, composite -1.486
confidence pipeline reached 0.738 before later publication/recency controls; published home probability was 0.334
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
MIL	Dustin May	4 ER / 14 outs	3 ER / 7 outs	-1
BAL	Cade Povich	2 ER / 13 outs	3 ER / 11 outs	1

Decision internals: run factors home×0.979 / away×1.041; savant lineup adjustment applied

Player-level breakdown (projected vs actual)

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

Top deviations that drove the game:

Coby Mayo (BAL 3B): 3-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact 2.2]
Jackson Chourio (MIL LF): 2-for-5, 1 HR, 3 RBI (projected 1.0 H, 1.0 HR) [run impact 0.8]
Garrett Mitchell (MIL CF): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.8]
Gunnar Henderson (BAL SS): 3-for-5, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact 0.61]
Colton Cowser (BAL CF): 2-for-3, 0 HR, 0 RBI (projected 0.0 H, 0.0 HR) [run impact 0.61]

Starter miss: Dustin May (MIL) posted 11.57 ERA vs projected 7.71 (3 ER in 7 outs); he is materially worse pitching away (5.50 ERA there vs 3.87 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

Dustin May (MIL) threw 7 outs, 3 ER = 11.57 ERA against a projected 7.71 ERA (3.86 ERA error)

Bullpen behind him: 2 ER over 9 outs.

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

Who did the damage to him:

Coby Mayo: 3-for-4, 0 HR, 0 RBI [+2.2 runs vs projection] season xwOBA 0.328, above average - real projection error
Gunnar Henderson: 3-for-5, 0 HR, 0 RBI [+0.61 runs vs projection] season xwOBA 0.318, average - a good day, not a model error
Colton Cowser: 2-for-3, 0 HR, 0 RBI [+0.61 runs vs projection] season xwOBA 0.315, average - a good day, not a model error

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

HOME pitching

Cade Povich (BAL) threw 11 outs, 3 ER = 7.36 ERA against a projected 4.15 ERA (3.21 ERA error) only 7 prior appearances in the covered season - the projection had little of his own record to stand on

Bullpen behind him: 2 ER over 9 outs.

The AWAY lineup he faced produced 6 runs on 9 hits and 1 HR, against a projected 6 runs / 9 hits (0 runs vs projection).

Who did the damage to him:

Jackson Chourio: 2-for-5, 1 HR, 3 RBI [+0.8 runs vs projection] season xwOBA 0.338, above average - real projection error
Joey Ortiz: 1-for-4, 0 HR, 0 RBI [+0.33 runs vs projection] season xwOBA 0.301, below average - a good day, not a model error
Cooper Pratt: 1-for-3, 0 HR, 0 RBI [+0.33 runs vs projection] season xwOBA 0.316, average - a good day, not a model error

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

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
MIL	Dustin May (Starter)	7 / 14	11.57 / 7.71 (+3.86)	3 / 4	6 / 6	1 / 1	1 / 2	1.72	4 / 4	4.5	venue away: 5.5 vs 3.87 ERA (worse here) · opp run factor 0.98 (average) · 28 prior appearances this season
MIL	Abner Uribe (Setup Man)	3 / 3	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0.33	3 / 0	0.97
MIL	Antonio Senzatela (Middle Reliever (High Leverage))	3 / 3	0 / 0	0 / 0	2 / 0	0 / 0	0 / 0	0.44	0 / 0	0.87
MIL	Trevor Megill (Closer)	3 / 3	18 / 0 (+18.00)	2 / 0	2 / 0	0 / 0	0 / 0	0.27	1 / 0	1.17
BAL	Cade Povich (Starter)	11 / 13	7.36 / 4.15 (+3.21)	3 / 2	5 / 5	1 / 2	2 / 2	1.83	3 / 7	4.2	venue home: 0.71 vs 7.2 ERA (ok) · opp run factor 1.04 (strong) · 7 prior appearances this season
BAL	Andrew Kittredge (Closer)	2 / 3	0 / 9 (-9.00)	0 / 1	0 / 2	0 / 0	0 / 0	0.19	1 / 1	0.91
BAL	Josh Walker (Middle Reliever (High Leverage))	2 / 4	0 / 6.75 (-6.75)	0 / 1	0 / 1	0 / 1	2 / 0	0.22	1 / 0	0.96
BAL	Rico Garcia (Setup Man)	2 / 5	27 / 5.4 (+21.60)	2 / 1	2 / 2	0 / 0	1 / 0	0.29	1 / 1	1.09
BAL	Yennier Cano (Middle Reliever (High Leverage))	3 / 2	0 / 13.5 (-13.50)	0 / 1	0 / 1	0 / 1	1 / 0	0.16	1 / 1	0.67

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

Batter	Team	AB	H	2B	3B	HR	RBI	R	BB	SO	Run impact	Season xwOBA	Signal
Coby Mayo (3B)	BAL	4 / 4	3 / 1	3 / 0	0 / 0	0 / 0	0 / 1	1 / 1	1 / 0	0 / 2	2.2	0.328 above average	projection too low
Jackson Chourio (LF)	MIL	5 / 4	2 / 1	0 / 0	0 / 0	1 / 1	3 / 1	1 / 1	1 / 0	1 / 1	0.8	0.338 above average	projection too low
Garrett Mitchell (CF)	MIL	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	2 / 1	-0.8	0.357 elite	bad day
Gunnar Henderson (SS)	BAL	5 / 4	3 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	1 / 1	0.61	0.318 average	no signal
Colton Cowser (CF)	BAL	3 / 3	2 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	0.61	0.315 average	no signal
Andrew Vaughn (1B)	MIL	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	-0.47	0.332 above average	—
Christian Encarnacion-Strand (DH)	BAL	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	0 / 0	2 / 1	-0.47	0.342 elite	—
Jeremiah Jackson (2B)	BAL	5 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	3 / 1	-0.47	0.291 below average	—
Joey Ortiz (3B)	MIL	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 1	0.33	0.301 below average	—
Christian Yelich (DH)	MIL	5 / 4	1 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	0 / 1	0 / 1	-0.33	0.312 average	—
Cooper Pratt (SS)	MIL	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	1 / 0	0 / 1	0.33	0.316 average	—
Brice Turang (2B)	MIL	5 / 4	2 / 1	0 / 1	0 / 0	0 / 0	0 / 1	1 / 1	0 / 1	3 / 1	-0.17	0.347 elite	—
Dylan Beavers (LF)	BAL	5 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 1	1 / 0	1 / 0	1 / 1	0.02	0.296 below average	—
William Contreras (C)	MIL	4 / 4	1 / 1	1 / 1	0 / 0	0 / 0	1 / 1	1 / 1	1 / 1	1 / 1	0	0.335 above average	—
Luis Lara (RF)	MIL	3 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	0 / 0	0 / 1	0	0.298 below average	—
Pete Alonso (1B)	BAL	4 / 4	1 / 1	0 / 0	0 / 0	1 / 1	3 / 1	1 / 1	1 / 1	2 / 1	0	0.394 elite	—
Leody Taveras (RF)	BAL	5 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 1	0	0.269 weak	—
Samuel Basallo (C)	BAL	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	1 / 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	-1.247	positive favors home; negative favors away
Recent form	-0.492	positive favors home; negative favors away
Starting-pitcher matchup	-0.062	positive favors home; negative favors away
Composite	-1.486	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.57-4.31	n/a	Base ML historical model prediction
variance_applied	5-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	7-4	0.666	Enhanced ML with recent form, H2H, low-score analysis
embedded_savant_run_environment	7-4	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	7-4	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	7-4	n/a	Capped at max 12 realistic runs
simulation_result	7-4	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.57-4.31	5-4	0.739	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-4	7-4	2	Enhanced ML with recent form, H2H, low-score analysis
embedded_savant_run_environment	7-4	7-4	-0	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	7-4	7-4	-0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	7-4	7-4	-0	Capped at max 12 realistic runs
simulation_result	7-4	7-4	-0	Final inning-by-inning simulation result

Largest player-level adjustments from the ultra-detailed trace

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

Team	Player	Role	Net effect	Largest change
MIL	Brice Turang	batter / 2B	better	predicted_RBI 2.18
MIL	Jackson Chourio	batter / LF	better	predicted_H 1.71
BAL	Leody Taveras	batter / RF	worse	predicted_RBI 1.568
MIL	Garrett Mitchell	batter / CF	better	predicted_H 1.52
MIL	William Contreras	batter / C	worse	predicted_SO 1.365
MIL	Cooper Pratt	batter / SS	worse	predicted_SO 1.355
BAL	Colton Cowser	batter / CF	worse	predicted_AB 1.345
BAL	Christian Encarnacion-Strand	batter / DH	better	predicted_RBI 1.327

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 pitcher projections differ from final simulation pitching box score (ER: current=6, sim=7).	BAL
info	CURRENT_FINALIZED_MATCH_BOX	BAL model-prediction batting totals differ from final simulation box score (H: current=8, sim=6; BB: current=3, sim=2; SO: current=10, sim=5).	BAL
info	CURRENT_PLAYER_MATCH_BOX	BAL current finalized hitter lines are model predictions and differ from the simulation box score: Dylan Beavers: AB current=4 box=3, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Pete Alonso: HR current=1 box=0, RBI current=1 box=0, BB current=1 box=0, SO current=1 box=0; Christian Encarnacion-Strand: H current=1 box=2, HR current=0 box=1, RBI current=1 box=2, SO current=1 box=0; Gunnar Henderson: H current=1 box=0, R current=1 box=0, BB current=1 box=0; Coby Mayo: RBI current=1 box=0, SO current=2 box=1; Leody Taveras: 2B current=0 box=1, RBI current=0 box=2, R current=0 box=1, SO current=1 box=0; Samuel Basallo: AB current=4 box=3, H current=1 box=0, BB current=0 box=1; Colton Cowser: AB current=3 box=4, BB current=1 box=0, SO current=1 box=0	BAL
info	CURRENT_FINALIZED_MATCH_BOX	MIL model-prediction batting totals differ from final simulation box score (AB: current=33, sim=38; H: current=9, sim=11; 2B: current=2, sim=3; HR: current=1, sim=4; RBI: current=5, sim=7; R: current=6, sim=7; BB: current=4, 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: Andrew Vaughn: AB current=3 box=4, H current=1 box=0, R current=0 box=1, BB current=0 box=1; Brice Turang: H current=1 box=3, HR current=0 box=1, RBI current=1 box=3, SO current=1 box=0; William Contreras: AB current=4 box=5, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Joey Ortiz: HR current=0 box=1, RBI current=0 box=1, R current=0 box=1; Luis Lara: R current=0 box=1; Jackson Chourio: AB current=4 box=5, H current=1 box=3, 2B current=0 box=1; Christian Yelich: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, BB current=1 box=0; Cooper Pratt: AB current=3 box=4, SO current=1 box=2	MIL
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

MIN@LAAMISS

Predicted: LAA 4-5 (conf +0.07, lean)Actual: MIN 3-0Brier 0.286total bias +6

Root cause: Starting-pitcher projection was wrong

Evidence-backed diagnosis

published system edge 0.07 (lean); this is not the half-scale distance from 50%
team-run absolute error 6; total bias +6
pregame home-minus-away evidence: season +0.107, recent +0.560, pitcher -0.057, composite +0.073
confidence pipeline reached 0.559 before later publication/recency controls; published home probability was 0.534
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)
LAA offense over-projected by 5
Connor Prielipp (MIN) allowed 4 fewer ER than projected (0 vs 4)

Team	Starter	Projected	Actual	ER err
MIN	Connor Prielipp	4 ER / 15 outs	0 ER / 21 outs	-4
LAA	Grayson Rodriguez	3 ER / 15 outs	3 ER / 20 outs	0

Decision internals: run factors home×0.931 / away×0.988; savant lineup adjustment applied

Player-level breakdown (projected vs actual)

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

Top deviations that drove the game:

Walker Jenkins (MIN CF): 3-for-4, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 1.54]
Kody Clemens (MIN 2B): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.4]
Zach Neto (LAA DH): 0-for-2, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.07]
Mike Trout (LAA CF): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.8]
Ryan Jeffers (MIN C): 1-for-3, 0 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact -0.64]

Starter miss: Connor Prielipp (MIN) posted 0.00 ERA vs projected 7.20 (0 ER in 21 outs); he is materially worse pitching away (6.40 ERA there vs 4.85 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

Connor Prielipp (MIN) threw 21 outs, 0 ER = 0 ERA against a projected 7.2 ERA (-7.2 ERA error)

Bullpen behind him: 0 ER over 3 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:

Jose Siri: 2-for-3, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.3, below average - a good day, not a model error

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

HOME pitching

Grayson Rodriguez (LAA) threw 20 outs, 3 ER = 4.05 ERA against a projected 5.4 ERA (-1.35 ERA error)

Bullpen behind him: 0 ER over 0 outs.

The AWAY lineup he faced produced 3 runs on 9 hits and 1 HR, against a projected 4 runs / 8 hits (-1 runs vs projection).

Who did the damage to him:

Walker Jenkins: 3-for-4, 1 HR, 1 RBI [+1.54 runs vs projection] season xwOBA 0.273, weak - a good day, not a model error
Brooks Lee: 2-for-4, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.275, weak - a good day, not a model error

Verdict: offense: the lineup scored well below its projection

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
MIN	Connor Prielipp (Starter)	21 / 15	0 / 7.2 (-7.20)	0 / 4	2 / 2	0 / 2	2 / 3	1.8	8 / 6	5.17	venue away: 6.4 vs 4.85 ERA (worse here) · opp run factor 0.93 (weak) · 22 prior appearances this season
MIN	Tommy Nance (Reliever)	3 / 0	0 / n/a	0 / 1	2 / 1	0 / 0	1 / 0	n/a	2 / 0	n/a
LAA	Grayson Rodriguez (Starter)	20 / 15	4.05 / 5.4	3 / 3	8 / 3	1 / 1	0 / 2	2.06	4 / 4	5.48	venue home: 7.54 vs 4.54 ERA (worse here) · opp run factor 0.99 (average) · 18 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
Walker Jenkins (CF)	MIN	4 / 3	3 / 1	0 / 0	0 / 0	1 / 0	1 / 0	1 / 0	0 / 1	0 / 1	1.54	0.273 weak	good day
Kody Clemens (2B)	MIN	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	0 / 0	0 / 1	-1.4	0.336 above average	bad day
Zach Neto (DH)	LAA	2 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	2 / 1	1 / 2	-1.07	0.311 average	no signal
Mike Trout (CF)	LAA	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	2 / 1	-0.8	0.379 elite	bad day
Ryan Jeffers (C)	MIN	3 / 4	1 / 1	0 / 1	0 / 0	0 / 0	1 / 1	0 / 1	0 / 1	0 / 1	-0.64	0.35 elite	bad day
Emmanuel Rodriguez (LF)	MIN	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	3 / 1	-0.47	0.424 elite	—
Brooks Lee (3B)	MIN	4 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	0 / 0	0 / 1	0.47	0.275 weak	—
Denzer Guzman (3B)	LAA	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	-0.47	0.265 weak	—
Jose Siri (LF)	LAA	3 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	1 / 2	0.47	0.3 below average	—
Vaughn Grissom (1B)	LAA	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 0	1 / 0	1 / 1	-0.45	0.306 average	—
Luke Keaschall (RF)	MIN	3 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	-0.33	0.317 average	—
Wade Meckler (RF)	LAA	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	-0.33	0.311 average	—
Josh Bell (DH)	MIN	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 0	0 / 0	0 / 1	0	0.333 above average	—
Royce Lewis (1B)	MIN	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0	0.308 average	—
Ben Ross (SS)	MIN	4 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	3 / 1	0	0.17 weak	—
Oswald Peraza (SS)	LAA	4 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0	0.275 weak	—
Travis d'Arnaud (C)	LAA	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0	0.244 weak	—
Christian Moore (2B)	LAA	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	0	0.222 weak	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	0.107	positive favors home; negative favors away
Recent form	0.56	positive favors home; negative favors away
Starting-pitcher matchup	-0.057	positive favors home; negative favors away
Composite	0.073	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.52-3.54	n/a	Base ML historical model prediction
variance_applied	5-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	3-4	0.534	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by winner_starter_is_below_average, projected_margin_is_one_run, matchup_recency_guardrail)
embedded_savant_run_environment	4-4	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	4-5	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	4-5	n/a	Capped at max 12 realistic runs
simulation_result	4-5	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.52-3.54	5-4	-0.02	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-4	3-4	2	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by winner_starter_is_below_average, projected_margin_is_one_run, matchup_recency_guardrail)
embedded_savant_run_environment	3-4	4-4	-1	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	4-4	4-5	1	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	4-5	4-5	0	Capped at max 12 realistic runs
simulation_result	4-5	4-5	0	Final inning-by-inning simulation result

Largest player-level adjustments from the ultra-detailed trace

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

Team	Player	Role	Net effect	Largest change
LAA	Mike Trout	batter / CF	better	predicted_RBI 3.57
MIN	Luke Keaschall	batter / RF	better	predicted_BB 1.608
MIN	Ben Ross	batter / SS	better	predicted_AB 1.5
LAA	Ben Joyce	pitcher / Closer	better	predicted_IP 1.1
LAA	Jose Siri	batter / LF	worse	predicted_SO 1.099
MIN	Josh Bell	batter / DH	worse	predicted_H -1.06
LAA	Oswald Peraza	batter / SS	worse	predicted_SO -1.019
MIN	Tommy Nance	pitcher / Reliever	worse	predicted_ER 1

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=32, sim=34; H: current=7, sim=4; 2B: current=1, sim=0; HR: current=1, sim=2; RBI: current=3, sim=5; R: current=4, sim=5; BB: current=3, sim=4; 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: Zach Neto: H current=1 box=0, HR current=1 box=0, RBI current=1 box=0; Vaughn Grissom: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, SO current=1 box=0; Oswald Peraza: AB current=3 box=4, SO current=1 box=0; Travis d'Arnaud: H current=0 box=1, R current=0 box=1, BB current=0 box=1, SO current=1 box=0; Mike Trout: H current=1 box=2, HR current=0 box=2, RBI current=0 box=4; Wade Meckler: AB current=3 box=4, RBI current=0 box=1, BB current=1 box=0, SO current=1 box=0; Denzer Guzman: H current=1 box=0, R current=1 box=0; Jose Siri: AB current=4 box=3, H current=1 box=0, RBI current=1 box=0, R current=0 box=1	LAA
info	CURRENT_FINALIZED_MATCH_BOX	MIN model-prediction batting totals differ from final simulation box score (H: current=8, sim=4; 2B: current=1, sim=0; BB: current=3, sim=4; SO: current=8, sim=7).	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=3, RBI current=0 box=1, BB current=1 box=2; Ryan Jeffers: AB current=4 box=3, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Emmanuel Rodriguez: H current=1 box=0, SO current=1 box=2; Brooks Lee: AB current=4 box=5, R current=0 box=1; Kody Clemens: HR current=1 box=0, SO current=1 box=0; Josh Bell: H current=1 box=0, RBI current=1 box=0; Royce Lewis: H current=1 box=0; Ben Ross: AB current=3 box=4, H current=0 box=1, HR current=0 box=1, RBI current=0 box=1	MIN
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

NYY@ARIHIT

Predicted: NYY 5-4 (conf -0.01, toss_up)Actual: NYY 9-2Brier 0.245total bias -2

Root cause: Toss-up that went our way

Evidence-backed diagnosis

published system edge 0.01 (toss_up); this is not the half-scale distance from 50%
team-run absolute error 6; total bias -2
pregame home-minus-away evidence: season -0.786, recent -0.428, pitcher +0.073, composite -0.560
confidence pipeline reached 0.505 before later publication/recency controls; published home probability was 0.495
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

NYY offense under-projected by 4
Eduardo Rodriguez (ARI) allowed 3 more ER than projected (7 vs 4)

Team	Starter	Projected	Actual	ER err
NYY	Gerrit Cole	2 ER / 17 outs	2 ER / 18 outs	0
ARI	Eduardo Rodriguez	4 ER / 18 outs	7 ER / 16 outs	3

Decision internals: run factors home×1.028 / away×0.955; savant lineup adjustment applied

Player-level breakdown (projected vs actual)

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

Top deviations that drove the game:

Ben Rice (NYY DH): 2-for-3, 1 HR, 2 RBI (projected 1.0 H, 1.0 HR) [run impact 1.42]
George Lombard Jr. (NYY SS): 2-for-4, 1 HR, 3 RBI (projected 1.0 H, 0.0 HR) [run impact 1.4]
Corbin Carroll (ARI RF): 2-for-4, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.38]
Anthony Volpe (NYY 2B): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.8]
Austin Wells (NYY C): 1-for-3, 0 HR, 0 RBI (projected 0.0 H, 0.0 HR) [run impact 0.8]

Pitcher versus batter: what actually broke this game

AWAY pitching

Gerrit Cole (NYY) threw 18 outs, 2 ER = 3 ERA against a projected 3.18 ERA (-0.18 ERA error)

Bullpen behind him: 0 ER over 0 outs.

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

Who did the damage to him:

Corbin Carroll: 2-for-4, 1 HR, 2 RBI [+1.38 runs vs projection] season xwOBA 0.341, elite - real projection error
Geraldo Perdomo: 2-for-3, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.327, above average - a good day, not a model error

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

HOME pitching

No settled pitching line for this club. The forecast named Eduardo Rodriguez (ARI) at a projected 6 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 9 runs on 8 hits and 2 HR, against a projected 4 runs / 8 hits (5 runs vs projection).

Who did the damage to him:

Ben Rice: 2-for-3, 1 HR, 2 RBI [+1.42 runs vs projection] season xwOBA 0.376, elite - real projection error
George Lombard Jr.: 2-for-4, 1 HR, 3 RBI [+1.4 runs vs projection] season xwOBA 0.276, weak - a good day, not a model error
Austin Wells: 1-for-3, 0 HR, 0 RBI [+0.8 runs vs projection] season xwOBA 0.287, below average - a good day, not a model error

Verdict: offense: the lineup scored well below its projection

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
NYY	Gerrit Cole (Starter)	18 / 17	3 / 3.18	2 / 2	4 / 4	1 / 1	0 / 5	0.77	6 / 4	5.87	venue away: 3.45 vs 3.93 ERA (ok) · opp run factor 1.03 (average) · 20 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
Ben Rice (DH)	NYY	3 / 4	2 / 1	0 / 0	1 / 0	1 / 1	2 / 1	2 / 1	2 / 1	0 / 1	1.42	0.376 elite	projection too low
George Lombard Jr. (SS)	NYY	4 / 4	2 / 1	0 / 0	0 / 0	1 / 0	3 / 0	1 / 1	0 / 0	0 / 1	1.4	0.276 weak	good day
Corbin Carroll (RF)	ARI	4 / 4	2 / 1	1 / 0	0 / 0	1 / 0	2 / 0	1 / 1	0 / 1	0 / 1	1.38	0.341 elite	projection too low
Anthony Volpe (2B)	NYY	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	2 / 0	-0.8	0.299 below average	projection too high
Austin Wells (C)	NYY	3 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0.8	0.287 below average	good day
Lars Nootbaar (LF)	ARI	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	2 / 1	-0.8	0.346 elite	bad day
Gabriel Moreno (C)	ARI	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	1 / 0	-0.78	0.363 elite	bad day
Ketel Marte (2B)	ARI	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	1 / 0	-0.78	0.347 elite	bad day
Cody Bellinger (LF)	NYY	5 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 1	1 / 1	0 / 1	0 / 0	-0.64	0.342 elite	bad day
Spencer Jones (CF)	NYY	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 0	0 / 0	2 / 1	-0.47	0.329 above average	—
Pavin Smith (1B)	ARI	2 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	-0.47	0.271 weak	—
Geraldo Perdomo (SS)	ARI	3 / 4	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0 / 0	0.47	0.327 above average	—
Tim Tawa (CF)	ARI	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	2 / 1	-0.47	0.314 average	—
Nolan Arenado (3B)	ARI	4 / 4	2 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 0	0 / 1	-0.46	0.293 below average	—
Paul Goldschmidt (1B)	NYY	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	2 / 0	1 / 0	0 / 1	0.33	0.316 average	—
Heliot Ramos (RF)	NYY	3 / 3	0 / 1	0 / 1	0 / 0	0 / 0	0 / 0	2 / 1	2 / 0	1 / 1	-0.12	0.328 above average	—
Amed Rosario (3B)	NYY	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 0	0 / 1	0	0.312 average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-0.786	positive favors home; negative favors away
Recent form	-0.428	positive favors home; negative favors away
Starting-pitcher matchup	0.073	positive favors home; negative favors away
Composite	-0.56	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.25-4.48	n/a	Base ML historical model prediction
variance_applied	4-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-4	0.505	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run)
embedded_savant_run_environment	5-4	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	5-4	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	5-4	n/a	Capped at max 12 realistic runs
simulation_result	5-4	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.25-4.48	4-4	0.224	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)
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
ARI	Geraldo Perdomo	batter / SS	better	predicted_RBI 2.7
NYY	Amed Rosario	batter / 3B	worse	predicted_AB 2.22
ARI	Gabriel Moreno	batter / C	worse	predicted_AB -2.18
ARI	Pavin Smith	batter / 1B	worse	predicted_AB 2.097
NYY	Paul Goldschmidt	batter / 1B	worse	predicted_RBI 1.7
NYY	Heliot Ramos	batter / RF	better	predicted_AB 1.396
NYY	Ben Rice	batter / DH	better	predicted_RBI 1.32
ARI	Ketel Marte	batter / 2B	worse	predicted_H -1.29

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 pitcher projections differ from final simulation pitching box score (H: current=9, sim=11).	ARI
info	CURRENT_FINALIZED_MATCH_BOX	ARI model-prediction batting totals differ from final simulation box score (AB: current=34, sim=33; H: current=8, sim=6; 2B: current=2, sim=1; HR: current=1, sim=2; BB: current=3, sim=7; SO: current=6, sim=7; HBP: current=0, sim=1).	ARI
info	CURRENT_PLAYER_MATCH_BOX	ARI current finalized hitter lines are model predictions and differ from the simulation box score: Lars Nootbaar: H current=1 box=0; Gabriel Moreno: AB current=4 box=3, 2B current=1 box=0, RBI current=1 box=0, BB current=0 box=2; Pavin Smith: AB current=4 box=5, H current=1 box=0; Geraldo Perdomo: H current=1 box=2, 2B current=0 box=1, HR current=0 box=1, RBI current=0 box=3; Jesús Sánchez: AB current=3 box=4; Tim Tawa: AB current=3 box=2, RBI current=1 box=0, BB current=0 box=1, HBP current=0 box=1; Ketel Marte: AB current=4 box=5, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Nolan Arenado: AB current=4 box=3, BB current=0 box=1, SO current=1 box=0	ARI
info	CURRENT_FINALIZED_MATCH_BOX	NYY model-prediction batting totals differ from final simulation box score (AB: current=32, sim=38; H: current=8, sim=11; 3B: current=0, sim=1; RBI: current=4, sim=5; R: current=4, sim=5; SO: current=7, sim=6; 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=5, H current=1 box=2; Cody Bellinger: H current=1 box=2, 2B current=1 box=0, RBI current=1 box=0; Ben Rice: 2B current=0 box=1, HR current=1 box=0, RBI current=1 box=2, SO current=1 box=0; Amed Rosario: AB current=3 box=5, 3B current=0 box=1; Paul Goldschmidt: AB current=4 box=5, HR current=0 box=1, RBI current=1 box=2, R current=0 box=1; Anthony Volpe: AB current=3 box=4, H current=1 box=2, RBI current=0 box=1, R current=0 box=1; George Lombard Jr.: H current=1 box=0, R current=1 box=0; Spencer Jones: RBI current=1 box=0, BB current=0 box=1, SO current=1 box=0	NYY
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

OAK@CLEHIT

Predicted: CLE 3-7 (conf +0.16, moderate)Actual: CLE 3-5Brier 0.176total bias +2

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 +2
pregame home-minus-away evidence: season +1.242, recent +0.845, pitcher +0.031, composite +1.682
confidence pipeline reached 0.698 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
OAK	Mason Barnett	3 ER / 11 outs	3 ER / 10 outs	0
CLE	Daniel Espino	1 ER / 9 outs	3 ER / 3 outs	2

Decision internals: run factors home×1.025 / away×0.994; 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:

Shea Langeliers (OAK DH): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.73]
Brayan Rocchio (CLE SS): 2-for-3, 0 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.11]
Zack Gelof (OAK LF): 1-for-4, 1 HR, 3 RBI (projected 1.0 H, 0.0 HR) [run impact 0.93]
Jo Adell (CLE DH): 1-for-4, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -0.93]
Travis Bazzana (CLE 2B): 1-for-3, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 0.93]

Starter miss: Daniel Espino (CLE) posted 27.00 ERA vs projected 3.00 (3 ER in 3 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

Mason Barnett (OAK) threw 10 outs, 3 ER = 8.1 ERA against a projected 7.36 ERA (0.74 ERA error)

Bullpen behind him: 1 ER over 3 outs.

The HOME lineup he faced produced 5 runs on 10 hits and 2 HR, against a projected 4 runs / 8 hits (1 runs vs projection).

Who did the damage to him:

Brayan Rocchio: 2-for-3, 0 HR, 2 RBI [+1.11 runs vs projection] season xwOBA 0.293, below average - a good day, not a model error
Travis Bazzana: 1-for-3, 1 HR, 1 RBI [+0.93 runs vs projection] season xwOBA 0.303, below average - a good day, not a model error
José Ramírez: 2-for-5, 1 HR, 2 RBI [+0.76 runs vs projection] season xwOBA 0.328, above average - real projection error

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

HOME pitching

Daniel Espino (CLE) threw 3 outs, 3 ER = 27 ERA against a projected 3 ERA (24 ERA error) only 11 prior appearances in the covered season - the projection had little of his own record to stand on

Bullpen behind him: 0 ER over 6 outs.

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

Who did the damage to him:

Zack Gelof: 1-for-4, 1 HR, 3 RBI [+0.93 runs vs projection] season xwOBA 0.312, average - a good day, not a model error

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

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
OAK	Mason Barnett (Starter)	10 / 11	8.1 / 7.36	3 / 3	7 / 6	0 / 0	4 / 3	2.58	4 / 4	3.97	venue away: 14.14 vs 8 ERA (ok) · opp run factor 1.03 (weak) · 18 prior appearances this season
OAK	Seth Johnson (Reliever)	3 / 7	9 / 0 (+9.00)	1 / 0	2 / 0	1 / 0	0 / 0	n/a	1 / 0	n/a
CLE	Daniel Espino (Starter)	3 / 9	27 / 3 (+24.00)	3 / 1	1 / 2	1 / 0	3 / 2	0	1 / 6	3.96	opp run factor 0.99 (weak) · 11 prior appearances this season
CLE	Cade Smith (Closer)	3 / 4	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0.35	2 / 1	1.61
CLE	Hunter Gaddis (Setup Man)	3 / 4	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0.31	0 / 1	0.82

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
Shea Langeliers (DH)	OAK	4 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 1	1 / 1	-1.73	0.362 elite	bad day
Brayan Rocchio (SS)	CLE	3 / 3	2 / 1	1 / 0	0 / 0	0 / 0	2 / 0	1 / 0	1 / 0	0 / 0	1.11	0.293 below average	good day
Zack Gelof (LF)	OAK	4 / 4	1 / 1	0 / 0	0 / 0	1 / 0	3 / 1	1 / 1	0 / 0	2 / 1	0.93	0.312 average	no signal
Jo Adell (DH)	CLE	4 / 4	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 0	2 / 1	-0.93	0.328 above average	bad day
Travis Bazzana (2B)	CLE	3 / 3	1 / 1	0 / 0	0 / 0	1 / 0	1 / 0	1 / 1	1 / 1	0 / 1	0.93	0.303 below average	good day
Jonah Heim (C)	OAK	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	-0.78	0.287 below average	projection too high
José Ramírez (3B)	CLE	5 / 4	2 / 1	0 / 1	0 / 0	1 / 0	2 / 1	1 / 1	0 / 1	1 / 1	0.76	0.328 above average	projection too low
Henry Bolte (CF)	OAK	2 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	1 / 1	1 / 1	-0.47	0.318 average	—
Petey Halpin (CF)	CLE	4 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 1	0.47	0.29 below average	—
Jeff McNeil (1B)	OAK	4 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0 / 1	-0.31	0.302 below average	—
Chase DeLauter (RF)	CLE	3 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	0 / 1	-0.31	0.344 elite	—
Tommy White (3B)	OAK	2 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	-0.14	0.292 below average	—
Lawrence Butler (RF)	OAK	2 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	1 / 1	2 / 1	1 / 1	-0.14	0.323 above average	—
Steven Kwan (LF)	CLE	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	2 / 1	1 / 0	-0.14	0.324 above average	—
Alika Williams (SS)	OAK	2 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0	0.242 weak	—
Michael Stefanic (2B)	OAK	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	0	0.344 elite	—
Patrick Bailey (C)	CLE	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	3 / 1	0	0.307 average	—
Nathaniel Lowe (1B)	CLE	3 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	1 / 1	0	0.337 above average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	1.242	positive favors home; negative favors away
Recent form	0.845	positive favors home; negative favors away
Starting-pitcher matchup	0.031	positive favors home; negative favors away
Composite	1.682	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	3.07-6.91	n/a	Base ML historical model prediction
variance_applied	3-7	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-7	0.58	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by matchup_recency_guardrail)
embedded_savant_run_environment	3-7	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	3-7	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	3-7	n/a	Capped at max 12 realistic runs
simulation_result	3-7	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	3.07-6.91	3-7	0.163	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	3-7	4-7	-1	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by matchup_recency_guardrail)
embedded_savant_run_environment	4-7	3-7	1	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	3-7	3-7	0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	3-7	3-7	0	Capped at max 12 realistic runs
simulation_result	3-7	3-7	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
OAK	Mason Barnett	pitcher / Starter	worse	predicted_IP -4.8
OAK	Chris Roycroft	pitcher / Reliever	worse	predicted_H 4
CLE	Nathaniel Lowe	batter / 1B	better	predicted_RBI 2.72
CLE	Jo Adell	batter / DH	better	predicted_RBI 2.389
OAK	Seth Johnson	pitcher / Reliever	better	predicted_IP 2
CLE	Daniel Espino	pitcher / Starter	worse	predicted_IP -1.667
OAK	Zack Gelof	batter / LF	worse	predicted_RBI 1.531
OAK	Michael Stefanic	batter / 2B	worse	predicted_AB 1.5

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=34; H: current=8, sim=10; 2B: current=2, sim=0; RBI: current=4, sim=7; R: current=4, sim=7; BB: current=4, sim=5; SO: current=7, sim=5).	CLE
info	CURRENT_PLAYER_MATCH_BOX	CLE current finalized hitter lines are model predictions and differ from the simulation box score: Jo Adell: H current=1 box=3, HR current=1 box=0, RBI current=1 box=3, SO current=1 box=0; Travis Bazzana: AB current=3 box=4, RBI current=0 box=1, BB current=1 box=0, SO current=1 box=2; José Ramírez: 2B current=1 box=0, RBI current=1 box=0, SO current=1 box=2; Chase DeLauter: 2B current=1 box=0, RBI current=1 box=0, R current=0 box=1, SO current=1 box=0; Steven Kwan: H current=1 box=0, R current=1 box=0; Patrick Bailey: AB current=3 box=4, H current=1 box=0, SO current=1 box=0; Nathaniel Lowe: AB current=4 box=3, H current=1 box=2, HR current=0 box=1, RBI current=1 box=3; Petey Halpin: H current=0 box=1, R current=0 box=1, BB current=0 box=1, SO current=1 box=0	CLE
info	CURRENT_FINALIZED_MATCH_BOX	OAK model pitcher projections differ from final simulation pitching box score (ER: current=6, sim=7).	OAK
info	CURRENT_FINALIZED_MATCH_BOX	OAK model-prediction batting totals differ from final simulation box score (AB: current=34, sim=33; H: current=8, sim=6; 2B: current=2, sim=1; HR: current=1, sim=0; RBI: current=4, sim=3; R: current=4, sim=3; BB: current=3, sim=4; SO: current=9, sim=11).	OAK
info	CURRENT_PLAYER_MATCH_BOX	OAK current finalized hitter lines are model predictions and differ from the simulation box score: Shea Langeliers: AB current=4 box=3, HR current=1 box=0, RBI current=1 box=0, SO current=1 box=2; Henry Bolte: AB current=4 box=5, R current=1 box=0, BB current=1 box=0, SO current=1 box=2; Zack Gelof: AB current=4 box=3, RBI current=1 box=2, BB current=0 box=1; Jeff McNeil: AB current=4 box=3, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Jonah Heim: H current=1 box=0, 2B current=1 box=0, RBI current=0 box=1, R current=0 box=1; Alika Williams: AB current=3 box=4, 2B current=0 box=1; Lawrence Butler: AB current=4 box=3, H current=1 box=0, RBI current=1 box=0, R current=1 box=0; Michael Stefanic: AB current=3 box=4, H current=0 box=1, SO current=1 box=0	OAK
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

PHI@NYMMISS

Predicted: PHI 5-4 (conf -0.02, toss_up)Actual: NYM 3-6Brier 0.262total bias +0

Root cause: Genuine toss-up that flipped

Evidence-backed diagnosis

published system edge 0.02 (toss_up); this is not the half-scale distance from 50%
team-run absolute error 4; total bias +0
pregame home-minus-away evidence: season -0.453, recent -0.102, pitcher +0.045, composite -0.355
confidence pipeline reached 0.516 before later publication/recency controls; published home probability was 0.488
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
PHI	Tim Mayza	2 ER / 15 outs	3 ER / 1 outs	1
NYM	Zac Thornton	1 ER / 17 outs	2 ER / 18 outs	1

Decision internals: run factors home×0.998 / away×0.953; 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:

Edmundo Sosa (PHI LF): 3-for-3, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 2.18]
Juan Soto (NYM LF): 3-for-5, 1 HR, 3 RBI (projected 1.0 H, 0.0 HR) [run impact 1.85]
Kyle Schwarber (PHI DH): 0-for-2, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -1.4]
Derek Hill (PHI CF): 1-for-3, 1 HR, 1 RBI (projected 0.0 H, 0.0 HR) [run impact 1.4]
Bryce Harper (PHI RF): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.8]

Starter miss: Tim Mayza (PHI) posted 81.00 ERA vs projected 3.60 (3 ER in 1 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

Tim Mayza (PHI) threw 1 outs, 3 ER = 81 ERA against a projected 3.6 ERA (77.4 ERA error)

Bullpen behind him: 0 ER over 0 outs.

The HOME lineup he faced produced 6 runs on 13 hits and 1 HR, against a projected 4 runs / 8 hits (2 runs vs projection).

Who did the damage to him:

Juan Soto: 3-for-5, 1 HR, 3 RBI [+1.85 runs vs projection] season xwOBA 0.413, elite - real projection error
Brett Baty: 1-for-3, 0 HR, 0 RBI [+0.8 runs vs projection] season xwOBA 0.311, average - a good day, not a model error
Carson Benge: 2-for-3, 0 HR, 1 RBI [+0.47 runs vs projection] season xwOBA 0.337, 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

Zac Thornton (NYM) threw 18 outs, 2 ER = 3 ERA against a projected 1.59 ERA (1.41 ERA error)

Bullpen behind him: 1 ER over 5 outs.

The AWAY lineup he faced produced 3 runs on 6 hits and 2 HR, against a projected 4 runs / 8 hits (-1 runs vs projection).

Who did the damage to him:

Edmundo Sosa: 3-for-3, 1 HR, 1 RBI [+2.18 runs vs projection] season xwOBA 0.328, above average - real projection error
Derek Hill: 1-for-3, 1 HR, 1 RBI [+1.4 runs vs projection] season xwOBA 0.288, 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	Tim Mayza (Starter)	1 / 15	81 / 3.6 (+77.40)	3 / 2	3 / 2	0 / 0	1 / 2	0.41	0 / 10	4.06	venue away: 1.69 vs 0 ERA (ok) · opp run factor 1 (average) · 56 prior appearances this season
NYM	Zac Thornton (Starter)	18 / 17	3 / 1.59	2 / 1	5 / 3	1 / 0	1 / 2	0.85	4 / 4	4.64	venue home: 3.8 vs 2.86 ERA (worse here) · opp run factor 0.95 (average) · 13 prior appearances this season
NYM	Nathan Lavender (Setup Man)	5 / 3	5.4 / 0 (+5.40)	1 / 0	1 / 0	1 / 0	0 / 0	0.23	1 / 1	1.34

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
Edmundo Sosa (LF)	PHI	3 / 3	3 / 1	1 / 0	0 / 0	1 / 0	1 / 1	2 / 0	0 / 0	0 / 1	2.18	0.328 above average	projection too low
Juan Soto (LF)	NYM	5 / 4	3 / 1	1 / 0	0 / 0	1 / 0	3 / 1	1 / 1	0 / 1	0 / 0	1.85	0.413 elite	projection too low
Kyle Schwarber (DH)	PHI	2 / 4	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	1 / 1	1 / 1	-1.4	0.364 elite	bad day
Derek Hill (CF)	PHI	3 / 3	1 / 0	0 / 0	0 / 0	1 / 0	1 / 0	1 / 0	0 / 0	0 / 1	1.4	0.288 below average	good day
Bryce Harper (RF)	PHI	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	3 / 1	-0.8	0.385 elite	bad day
Brett Baty (1B)	NYM	3 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0.8	0.311 average	no signal
Francisco Lindor (SS)	NYM	5 / 4	2 / 1	1 / 0	0 / 0	0 / 1	0 / 1	1 / 1	0 / 1	2 / 1	-0.48	0.33 above average	—
Alec Bohm (1B)	PHI	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0 / 0	-0.47	0.327 above average	—
Luis Arraez (2B)	PHI	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	-0.47	0.307 average	—
J.T. Realmuto (C)	PHI	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	-0.47	0.323 above average	—
Carson Benge (RF)	NYM	3 / 4	2 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	0 / 0	0 / 1	0.47	0.337 above average	—
Mark Vientos (DH)	NYM	3 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 0	0 / 1	0.33	0.313 average	—
Marcus Semien (2B)	NYM	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	0 / 1	-0.33	0.304 average	—
Trea Turner (SS)	PHI	4 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	-0.31	0.291 below average	—
Bo Bichette (3B)	NYM	5 / 4	1 / 1	0 / 1	0 / 0	0 / 0	0 / 1	1 / 0	0 / 0	1 / 1	-0.31	0.325 above average	—
Bryson Stott (3B)	PHI	4 / 3	1 / 1	1 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 1	0 / 1	-0.02	0.326 above average	—
A.J. Ewing (CF)	NYM	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	2 / 0	0 / 1	0 / 0	2 / 1	0	0.32 average	—
Luis Torrens (C)	NYM	4 / 3	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	2 / 1	0	0.295 below average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-0.453	positive favors home; negative favors away
Recent form	-0.102	positive favors home; negative favors away
Starting-pitcher matchup	0.045	positive favors home; negative favors away
Composite	-0.355	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.46-3.74	n/a	Base ML historical model prediction
variance_applied	4-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-4	0.512	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	4.46-3.74	4-4	-0.72	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
NYM	Bo Bichette	batter / 3B	worse	predicted_SO 2.197
NYM	Francisco Lindor	batter / SS	worse	predicted_SO 2.096
PHI	Luis Arraez	batter / 2B	worse	predicted_RBI 1.51
NYM	Mark Vientos	batter / DH	better	predicted_RBI 1.506
PHI	Alec Bohm	batter / 1B	worse	predicted_SO 1.413
NYM	Luis Torrens	batter / C	better	predicted_H 1.3
PHI	Kyle Schwarber	batter / DH	better	predicted_H 1.17
PHI	Edmundo Sosa	batter / LF	worse	predicted_AB 1.169

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=32, sim=34; H: current=8, sim=7; 3B: current=0, sim=1; HR: current=1, sim=0; BB: current=3, sim=4; SO: current=8, sim=12).	NYM
info	CURRENT_PLAYER_MATCH_BOX	NYM current finalized hitter lines are model predictions and differ from the simulation box score: Carson Benge: H current=1 box=0, R current=1 box=0, SO current=1 box=0; Mark Vientos: AB current=3 box=4, RBI current=1 box=2, R current=0 box=1, SO current=1 box=2; A.J. Ewing: AB current=4 box=3, R current=1 box=0, BB current=0 box=1; Marcus Semien: AB current=3 box=4, BB current=1 box=0; Brett Baty: AB current=3 box=4, H current=0 box=1, 3B current=0 box=1, RBI current=0 box=1; Francisco Lindor: H current=1 box=0, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0; Bo Bichette: AB current=4 box=3, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Luis Torrens: AB current=3 box=4, H current=1 box=2, 2B current=0 box=1, R current=0 box=1	NYM
info	CURRENT_FINALIZED_MATCH_BOX	PHI model-prediction batting totals differ from final simulation box score (AB: current=32, sim=35; H: current=8, sim=9; HR: current=1, sim=0; RBI: current=4, sim=5; R: current=4, sim=5; SO: current=7, sim=6; SF: current=0, sim=1; HBP: current=0, sim=1).	PHI
info	CURRENT_PLAYER_MATCH_BOX	PHI current finalized hitter lines are model predictions and differ from the simulation box score: Bryce Harper: H current=1 box=2; Alec Bohm: H current=1 box=0, RBI current=1 box=0, SO current=0 box=2; Edmundo Sosa: AB current=3 box=4, H current=1 box=0, RBI current=1 box=0; Bryson Stott: AB current=3 box=4, R current=0 box=1, BB current=1 box=0, SO current=1 box=0; Kyle Schwarber: AB current=4 box=5, H current=1 box=2, 2B current=0 box=1, HR current=1 box=0; Luis Arraez: H current=1 box=2, RBI current=0 box=2, R current=0 box=1, SF current=0 box=1; Trea Turner: 2B current=1 box=0, RBI current=0 box=1, BB current=0 box=1, SO current=1 box=0; J.T. Realmuto: H current=1 box=0, R current=1 box=0, BB current=0 box=1	PHI
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

SEA@COLHIT

Predicted: SEA 9-5 (conf -0.52, strong)Actual: SEA 5-4Brier 0.057total bias +5

Root cause: Right winner, wrong score

Evidence-backed diagnosis

published system edge 0.52 (strong); this is not the half-scale distance from 50%
team-run absolute error 5; total bias +5
pregame home-minus-away evidence: season -1.003, recent -1.124, pitcher -0.073, composite -1.713
confidence pipeline reached 0.760 before later publication/recency controls; published home probability was 0.240
score-volatility actions: supported_blowout_margin_expansion
forecast-file integrity: clean (0 defects, 8 informational note(s)); the saved artifact is not why the pick missed

Quantified contributing factors

total runs over-projected by 5 (14 vs 9)
SEA offense over-projected by 4
Gabriel Hughes (COL) allowed 3 fewer ER than projected (3 vs 6)

Team	Starter	Projected	Actual	ER err
SEA	Bryan Woo	3 ER / 18 outs	3 ER / 16 outs	0
COL	Gabriel Hughes	6 ER / 12 outs	3 ER / 14 outs	-3

Decision internals: run factors home×0.959 / away×1.031; 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:

Connor Norby (COL DH): 3-for-4, 0 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 1.27]
Jake McCarthy (COL LF): 3-for-5, 0 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.25]
Lazaro Montes (SEA RF): 1-for-4, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 0.93]
Hunter Goodman (COL C): 1-for-5, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -0.93]
Cal Raleigh (SEA C): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.8]

Starter miss: Gabriel Hughes (COL) posted 5.79 ERA vs projected 13.50 (3 ER in 14 outs); no credible venue, opponent-strength or experience signal explains it — an outlier start (variance).

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

Pitcher versus batter: what actually broke this game

AWAY pitching

Bryan Woo (SEA) threw 16 outs, 3 ER = 5.06 ERA against a projected 4.5 ERA (0.56 ERA error)

Bullpen behind him: 1 ER over 6 outs.

The HOME lineup he faced produced 4 runs on 15 hits and 0 HR, against a projected 5 runs / 9 hits (-1 runs vs projection).

Who did the damage to him:

Connor Norby: 3-for-4, 0 HR, 1 RBI [+1.27 runs vs projection] season xwOBA 0.306, average - a good day, not a model error
Jake McCarthy: 3-for-5, 0 HR, 2 RBI [+1.25 runs vs projection] season xwOBA 0.326, above average - real projection error
Ezequiel Tovar: 2-for-4, 0 HR, 0 RBI [+0.78 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

HOME pitching

Gabriel Hughes (COL) threw 14 outs, 3 ER = 5.79 ERA against a projected 13.5 ERA (-7.71 ERA error)

Bullpen behind him: 2 ER over 9 outs.

The AWAY lineup he faced produced 4 runs on 8 hits and 2 HR, against a projected 5 runs / 9 hits (-1 runs vs projection).

Who did the damage to him:

Lazaro Montes: 1-for-4, 1 HR, 1 RBI [+0.93 runs vs projection] season xwOBA 0.346, elite - real projection error
Randy Arozarena: 3-for-4, 0 HR, 0 RBI [+0.61 runs vs projection] season xwOBA 0.346, elite - real projection error
J.P. Crawford: 1-for-4, 1 HR, 1 RBI [+0.6 runs vs projection] season xwOBA 0.323, above average - 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)
SEA	Bryan Woo (Starter)	16 / 18	5.06 / 4.5	3 / 3	10 / 5	0 / 1	1 / 3	1.42	4 / 6	6.26	venue away: 5.23 vs 2.36 ERA (worse here) · opp run factor 0.96 (weak) · 28 prior appearances this season
SEA	Andrés Muñoz (Closer)	3 / 3	0 / 9 (-9.00)	0 / 1	1 / 1	0 / 0	0 / 0	0.46	2 / 1	1.48
SEA	Gabe Speier (Middle Reliever (High Leverage))	1 / 2	27 / 0 (+27.00)	1 / 0	2 / 0	0 / 0	0 / 0	0.34	0 / 2	1.09
SEA	José A. Ferrer (Setup Man)	2 / 2	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0.24	1 / 1	0.6
COL	Gabriel Hughes (Starter)	14 / 12	5.79 / 13.5 (-7.71)	3 / 6	5 / 7	1 / 1	1 / 2	1.82	6 / 2	3.62	venue home: 4.26 vs 8.71 ERA (ok) · opp run factor 1.03 (weak) · 13 prior appearances this season
COL	Brennan Bernardino (Setup Man)	3 / 3	9 / 0 (+9.00)	1 / 0	1 / 0	0 / 0	0 / 2	0.16	1 / 0	0.79
COL	Jaden Hill (Middle Reliever (High Leverage))	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0.59	0 / 2	0.98
COL	Jordan Romano (Closer)	3 / 3	9 / 0 (+9.00)	1 / 0	2 / 0	1 / 0	0 / 0	0.58	1 / 1	1.14

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
Connor Norby (DH)	COL	4 / 4	3 / 1	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	1 / 0	0 / 1	1.27	0.306 average	no signal
Jake McCarthy (LF)	COL	5 / 4	3 / 1	0 / 1	1 / 0	0 / 0	2 / 1	1 / 1	0 / 0	1 / 1	1.25	0.326 above average	projection too low
Lazaro Montes (RF)	SEA	4 / 3	1 / 1	0 / 0	0 / 0	1 / 0	1 / 1	1 / 0	0 / 0	1 / 1	0.93	0.346 elite	projection too low
Hunter Goodman (C)	COL	5 / 4	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 0	0 / 1	-0.93	0.321 above average	bad day
Cal Raleigh (C)	SEA	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	4 / 1	-0.8	0.313 average	no signal
Julio Rodríguez (CF)	SEA	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	1 / 1	-0.78	0.335 above average	bad day
Ezequiel Tovar (SS)	COL	4 / 3	2 / 1	1 / 0	0 / 0	0 / 0	0 / 0	2 / 0	0 / 0	1 / 1	0.78	0.28 weak	good day
Cole Carrigg (CF)	COL	5 / 4	3 / 1	0 / 1	0 / 0	0 / 0	1 / 1	0 / 1	0 / 0	1 / 1	0.63	0.311 average	no signal
Mickey Moniak (RF)	COL	5 / 4	1 / 1	0 / 0	1 / 0	0 / 0	0 / 1	0 / 0	0 / 0	2 / 1	0.62	0.282 weak	good day
Randy Arozarena (LF)	SEA	4 / 4	3 / 1	1 / 1	0 / 0	0 / 0	0 / 1	1 / 1	0 / 1	0 / 1	0.61	0.346 elite	projection too low
J.P. Crawford (SS)	SEA	4 / 4	1 / 1	0 / 0	0 / 0	1 / 0	1 / 0	1 / 1	0 / 1	1 / 1	0.6	0.323 above average	projection too low
Adael Amador (2B)	COL	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	0 / 1	-0.47	0.264 weak	—
Kyle Karros (3B)	COL	5 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	2 / 1	-0.33	0.34 elite	—
TJ Rumfield (1B)	COL	5 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	2 / 0	-0.33	0.337 above average	—
Cole Young (2B)	SEA	4 / 4	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0.31	0.314 average	—
Dominic Canzone (DH)	SEA	3 / 4	1 / 1	1 / 0	0 / 0	0 / 1	1 / 1	1 / 1	1 / 0	0 / 1	-0.29	0.353 elite	—
Brock Rodden (3B)	SEA	1 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 1	-0.14	0.239 weak	—
Josh Naylor (1B)	SEA	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	2 / 0	0 / 0	0 / 0	0 / 0	0	0.313 average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-1.003	positive favors home; negative favors away
Recent form	-1.124	positive favors home; negative favors away
Starting-pitcher matchup	-0.073	positive favors home; negative favors away
Composite	-1.713	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	5.99-3.97	n/a	Base ML historical model prediction
variance_applied	6-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	9-5	0.76	Enhanced ML with recent form, H2H, low-score analysis
embedded_savant_run_environment	9-5	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	9-5	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	9-5	n/a	Capped at max 12 realistic runs
simulation_result	9-5	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	5.99-3.97	6-4	-0.019	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	9-5	-0	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	9-5	9-5	-0	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	9-5	9-5	-0	Capped at max 12 realistic runs
simulation_result	9-5	9-5	-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
COL	Mickey Moniak	batter / RF	worse	predicted_RBI 3.402
SEA	J.P. Crawford	batter / SS	better	predicted_RBI 2.54
SEA	Bryan Woo	pitcher / Starter	better	predicted_SO 2
COL	Gabriel Hughes	pitcher / Starter	worse	predicted_H 2
COL	Nick Frasso	pitcher / Reliever	worse	predicted_ER 2
SEA	Brock Rodden	batter / 3B	better	predicted_RBI 1.81
COL	Hunter Goodman	batter / C	worse	predicted_SO 1.682
SEA	Dominic Canzone	batter / DH	deboosted	predicted_RBI 1.39

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 (H: current=9, sim=10; ER: current=8, sim=9).	COL
info	CURRENT_FINALIZED_MATCH_BOX	COL model-prediction batting totals differ from final simulation box score (H: current=9, sim=7; 2B: current=2, sim=0; HR: current=1, sim=2; BB: current=2, sim=3; SO: current=8, sim=11).	COL
info	CURRENT_PLAYER_MATCH_BOX	COL current finalized hitter lines are model predictions and differ from the simulation box score: Cole Carrigg: 2B current=1 box=0, RBI current=1 box=0; Connor Norby: AB current=4 box=3, H current=1 box=0, BB current=0 box=1; Kyle Karros: AB current=4 box=3, H current=1 box=0, R current=1 box=0, SO current=1 box=0; Mickey Moniak: HR current=0 box=1, RBI current=1 box=4, R current=0 box=1, SO current=1 box=2; Jake McCarthy: AB current=4 box=5, H current=1 box=2, 2B current=1 box=0, RBI current=1 box=0; Hunter Goodman: H current=1 box=0, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0; TJ Rumfield: H current=1 box=2, R current=0 box=1, BB current=1 box=0, SO current=0 box=1; Adael Amador: H current=1 box=0, RBI current=1 box=0, R current=1 box=0, BB current=0 box=1	COL
info	CURRENT_FINALIZED_MATCH_BOX	SEA model-prediction batting totals differ from final simulation box score (AB: current=34, sim=37; H: current=9, sim=10; 2B: current=2, sim=1; RBI: current=5, sim=9; R: current=5, sim=9; BB: current=3, sim=5; SO: current=7, sim=6; HBP: current=0, sim=1).	SEA
info	CURRENT_PLAYER_MATCH_BOX	SEA current finalized hitter lines are model predictions and differ from the simulation box score: Randy Arozarena: H current=1 box=2, 2B current=1 box=0, HR current=0 box=1, RBI current=1 box=2; J.P. Crawford: 2B current=0 box=1, RBI current=0 box=3, SO current=1 box=0; Dominic Canzone: HR current=1 box=0, RBI current=1 box=2, BB current=0 box=1, SO current=1 box=0; Cal Raleigh: AB current=4 box=5, H current=1 box=0, RBI current=1 box=0, BB current=1 box=0; Julio Rodríguez: AB current=4 box=5, H current=1 box=2, 2B current=1 box=0, RBI current=1 box=0; Brock Rodden: AB current=3 box=2, RBI current=0 box=2, R current=0 box=1, BB current=0 box=1; Lazaro Montes: RBI current=1 box=0, R current=0 box=1, BB current=0 box=1, SO current=1 box=2; Josh Naylor: AB current=4 box=5, R current=0 box=1	SEA
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

SF@LADHIT

Predicted: LAD 2-4 (conf +0.08, lean)Actual: LAD 2-8Brier 0.213total bias -4

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 4; total bias -4
pregame home-minus-away evidence: season +1.422, recent +0.582, pitcher +0.078, composite +1.459
confidence pipeline reached 0.600 before later publication/recency controls; published home probability was 0.538
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 (6 vs 10)
LAD offense under-projected by 4

Team	Starter	Projected	Actual	ER err
SF	Cesar Perdomo	4 ER / 15 outs	2 ER / 14 outs	-2
LAD	Tyler Glasnow	2 ER / 17 outs	2 ER / 18 outs	0

Decision internals: run factors home×1.002 / away×0.925; 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): 3-for-4, 2 HR, 3 RBI (projected 1.0 H, 1.0 HR) [run impact 2.2]
Josue De Paula (LAD DH): 2-for-3, 0 HR, 1 RBI (projected 0.0 H, 0.0 HR) [run impact 1.58]
Kyle Tucker (LAD RF): 2-for-3, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 1.4]
Will Smith (LAD C): 1-for-4, 1 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 0.93]
Drew Cavanaugh (SF C): 0-for-2, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.8]

Starter miss: Cesar Perdomo (SF) posted 3.86 ERA vs projected 7.20 (2 ER in 14 outs). Thin prior record (3 prior appearances this season): high projection uncertainty.

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

Pitcher versus batter: what actually broke this game

AWAY pitching

Cesar Perdomo (SF) threw 14 outs, 2 ER = 3.86 ERA against a projected 7.2 ERA (-3.34 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 3 outs.

The HOME lineup he faced produced 7 runs on 10 hits and 4 HR, against a projected 4 runs / 8 hits (3 runs vs projection).

Who did the damage to him:

Mookie Betts: 3-for-4, 2 HR, 3 RBI [+2.2 runs vs projection] season xwOBA 0.336, above average - real projection error
Josue De Paula: 2-for-3, 0 HR, 1 RBI [+1.58 runs vs projection] season xwOBA 0.417, elite - real projection error
Kyle Tucker: 2-for-3, 1 HR, 1 RBI [+1.4 runs vs projection] season xwOBA 0.332, above average - real projection error

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

HOME pitching

Tyler Glasnow (LAD) threw 18 outs, 2 ER = 3 ERA against a projected 3.18 ERA (-0.18 ERA error) only 11 prior appearances in the covered season - the projection had little of his own record to stand on

Bullpen behind him: 0 ER over 3 outs.

The AWAY lineup he faced produced 2 runs on 3 hits and 1 HR, against a projected 2 runs / 6 hits (0 runs vs projection).

Who did the damage to him:

Jung Hoo Lee: 1-for-4, 1 HR, 1 RBI [+0.62 runs vs projection] season xwOBA 0.315, average - a good day, not a model error
Scott Bandura: 1-for-2, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.136, 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)
SF	Cesar Perdomo (Starter)	14 / 15	3.86 / 7.2 (-3.34)	2 / 4	6 / 5	1 / 0	3 / 3	1.36	1 / 4	3.34	venue away: 0 vs 9.64 ERA (ok) · opp run factor 1 (strong) · thin sample: 3 prior appearances this season (high uncertainty)
SF	Carson Seymour (Middle Reliever (High Leverage))	3 / 2	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	0.75	2 / 1	0.72
LAD	Tyler Glasnow (Starter)	18 / 17	3 / 3.18	2 / 2	3 / 2	1 / 1	2 / 1	1.92	10 / 6	5.85	venue home: 4.56 vs 2.84 ERA (worse here) · opp run factor 0.92 (weak) · 11 prior appearances this season
LAD	Tanner Scott (Closer)	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0.2	1 / 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
Mookie Betts (SS)	LAD	4 / 4	3 / 1	0 / 0	0 / 0	2 / 1	3 / 1	2 / 1	1 / 0	0 / 0	2.2	0.336 above average	projection too low
Josue De Paula (DH)	LAD	3 / 3	2 / 0	1 / 0	0 / 0	0 / 0	1 / 0	2 / 0	1 / 0	0 / 1	1.58	0.417 elite	projection too low
Kyle Tucker (RF)	LAD	3 / 3	2 / 1	0 / 0	0 / 0	1 / 0	1 / 1	2 / 1	1 / 1	0 / 1	1.4	0.332 above average	projection too low
Will Smith (C)	LAD	4 / 4	1 / 1	0 / 0	0 / 0	1 / 0	1 / 1	1 / 0	1 / 1	0 / 1	0.93	0.377 elite	projection too low
Drew Cavanaugh (C)	SF	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	2 / 1	-0.8	0.311 average	no signal
Freddie Freeman (1B)	LAD	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	2 / 1	-0.78	0.368 elite	bad day
Jung Hoo Lee (RF)	SF	4 / 4	1 / 1	0 / 1	0 / 0	1 / 0	1 / 0	1 / 0	0 / 0	0 / 1	0.62	0.315 average	no signal
Turner Hill (LF)	SF	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	1 / 1	-0.47	0.328 above average	—
Scott Bandura (DH)	SF	2 / 3	1 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 1	0.47	0.136 weak	—
Shay Whitcomb (3B)	SF	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	3 / 1	-0.47	0.268 weak	—
Miguel Rojas (2B)	LAD	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	-0.47	0.311 average	—
Enrique Hernández (3B)	LAD	4 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 1	-0.47	0.282 weak	—
Drew Gilbert (CF)	SF	2 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 0	0 / 1	-0.14	0.33 above average	—
Teoscar Hernández (LF)	LAD	5 / 4	2 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	0 / 1	1 / 1	0.14	0.334 above average	—
Alex Call (CF)	LAD	1 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 1	-0.14	0.276 weak	—
Jonah Cox (2B)	SF	3 / 4	1 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	0 / 0	2 / 2	0	0.273 weak	—
Christian Koss (SS)	SF	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	2 / 1	0	0.257 weak	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	1.422	positive favors home; negative favors away
Recent form	0.582	positive favors home; negative favors away
Starting-pitcher matchup	0.078	positive favors home; negative favors away
Composite	1.459	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	3.37-5.06	n/a	Base ML historical model prediction
variance_applied	3-5	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	3-4	0.538	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by matchup_recency_guardrail)
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	3.37-5.06	3-5	0.312	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	3-5	3-4	-1	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by matchup_recency_guardrail)
embedded_savant_run_environment	3-4	2-4	1	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
LAD	Freddie Freeman	batter / 1B	better	predicted_RBI 2.61
LAD	Enrique Hernández	batter / 3B	worse	predicted_SO 2.155
LAD	Tyler Glasnow	pitcher / Starter	neutral	predicted_SO 2
LAD	Tanner Scott	pitcher / Closer	neutral	predicted_SO -2
SF	Jonah Cox	batter / 2B	better	predicted_RBI 1.69
LAD	Mookie Betts	batter / SS	worse	predicted_AB -1.53
SF	Scott Bandura	batter / DH	worse	predicted_AB 1.5
LAD	Miguel Rojas	batter / 2B	worse	predicted_AB 1.5

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=31, sim=30; H: current=8, sim=6; 2B: current=1, sim=4; HR: current=1, sim=0; BB: current=3, sim=4; SO: current=7, sim=8).	LAD
info	CURRENT_PLAYER_MATCH_BOX	LAD current finalized hitter lines are model predictions and differ from the simulation box score: Mookie Betts: AB current=4 box=3, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0; Will Smith: AB current=4 box=3, H current=1 box=0, RBI current=1 box=0, SO current=1 box=0; Teoscar Hernández: AB current=4 box=3, H current=1 box=2, 2B current=0 box=1; Freddie Freeman: RBI current=0 box=3; Alex Call: AB current=3 box=2, 2B current=0 box=1, R current=0 box=1, BB current=0 box=1; Josue De Paula: H current=0 box=1, 2B current=0 box=1, SO current=1 box=0; Miguel Rojas: AB current=3 box=4, R current=0 box=1, SO current=0 box=1; Kyle Tucker: AB current=3 box=4, H current=1 box=0, RBI current=1 box=0, R current=1 box=0	LAD
info	CURRENT_FINALIZED_MATCH_BOX	SF model-prediction batting totals differ from final simulation box score (AB: current=33, sim=31; H: current=7, sim=4; RBI: current=4, sim=2; R: current=3, sim=2; SO: current=10, sim=7).	SF
info	CURRENT_PLAYER_MATCH_BOX	SF current finalized hitter lines are model predictions and differ from the simulation box score: Bryce Eldridge: AB current=4 box=3, 2B current=0 box=1, HR current=1 box=0, RBI current=1 box=0; Jonah Cox: H current=1 box=2, HR current=0 box=1, RBI current=1 box=2, SO current=2 box=0; Turner Hill: H current=1 box=0, RBI current=1 box=0, BB current=1 box=0, SO current=1 box=0; Scott Bandura: AB current=3 box=4, SO current=1 box=2; Jung Hoo Lee: 2B current=1 box=0; Drew Cavanaugh: AB current=3 box=2, H current=1 box=0, SO current=1 box=0; Shay Whitcomb: AB current=4 box=3, H current=1 box=0, BB current=0 box=1; Drew Gilbert: H current=1 box=0, RBI current=1 box=0	SF
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

TOR@TEXHIT

Predicted: TEX 4-5 (conf +0.01, toss_up)Actual: TEX 1-7Brier 0.245total bias +1

Root cause: Toss-up that went our way

Evidence-backed diagnosis

published system edge 0.01 (toss_up); this is not the half-scale distance from 50%
team-run absolute error 5; total bias +1
pregame home-minus-away evidence: season +0.123, recent +0.726, pitcher +0.062, composite +0.385
confidence pipeline reached 0.505 before later publication/recency controls; published home probability was 0.505
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

TOR offense over-projected by 3

Team	Starter	Projected	Actual	ER err
TOR	Dylan Cease	3 ER / 18 outs	1 ER / 8 outs	-2
TEX	Kumar Rocker	1 ER / 15 outs	1 ER / 13 outs	0

Decision internals: run factors home×0.998 / away×0.936; 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:

Wyatt Langford (TEX LF): 2-for-2, 0 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 1.13]
George Springer (TOR DH): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -1.11]
Jake Burger (TEX 1B): 1-for-2, 0 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 0.97]
Brett Bateman (TOR CF): 0-for-4, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -0.8]
Myles Straw (TOR LF): 2-for-3, 0 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact 0.78]

Pitcher versus batter: what actually broke this game

AWAY pitching

Dylan Cease (TOR) threw 8 outs, 1 ER = 3.38 ERA against a projected 4.5 ERA (-1.12 ERA error)

Bullpen behind him: 0 ER over 0 outs.

The HOME lineup he faced produced 7 runs on 8 hits and 0 HR, against a projected 4 runs / 7 hits (3 runs vs projection).

Who did the damage to him:

Wyatt Langford: 2-for-2, 0 HR, 1 RBI [+1.13 runs vs projection] season xwOBA 0.319, average - a good day, not a model error
Jake Burger: 1-for-2, 0 HR, 2 RBI [+0.97 runs vs projection] season xwOBA 0.315, average - a good day, not a model error
Nicky Lopez: 2-for-5, 0 HR, 0 RBI [+0.47 runs vs projection] season xwOBA 0.258, weak - a good day, not a model error

Verdict: offense: the lineup scored well below its projection

HOME pitching

Kumar Rocker (TEX) threw 13 outs, 1 ER = 2.08 ERA against a projected 1.8 ERA (0.28 ERA error)

Bullpen behind him: 0 ER over 0 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:

Myles Straw: 2-for-3, 0 HR, 1 RBI [+0.78 runs vs projection] season xwOBA 0.297, below average - a good day, not a model error
Josh Smith: 0-for-3, 0 HR, 0 RBI [+0.33 runs vs projection] season xwOBA 0.284, 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)
TOR	Dylan Cease (Starter)	8 / 18	3.38 / 4.5	1 / 3	3 / 6	0 / 1	6 / 2	2.31	3 / 9	5.9	venue away: 1.89 vs 2.85 ERA (ok) · opp run factor 1 (average) · 28 prior appearances this season
TEX	Kumar Rocker (Starter)	13 / 15	2.08 / 1.8	1 / 1	3 / 4	0 / 0	2 / 2	2.44	2 / 3	5.06	venue home: 4.76 vs 4.57 ERA (ok) · opp run factor 0.94 (weak) · 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
Wyatt Langford (LF)	TEX	2 / 3	2 / 1	0 / 0	0 / 0	0 / 0	1 / 0	2 / 1	3 / 1	0 / 1	1.13	0.319 average	no signal
George Springer (DH)	TOR	4 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	1 / 1	-1.11	0.308 average	no signal
Jake Burger (1B)	TEX	2 / 3	1 / 1	1 / 0	0 / 0	0 / 0	2 / 1	1 / 1	2 / 0	1 / 1	0.97	0.315 average	no signal
Brett Bateman (CF)	TOR	4 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	1 / 1	-0.8	0.306 average	no signal
Myles Straw (LF)	TOR	3 / 3	2 / 1	1 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0 / 0	0.78	0.297 below average	good day
Corey Seager (SS)	TEX	4 / 4	1 / 1	1 / 0	0 / 0	0 / 1	1 / 1	0 / 1	1 / 1	1 / 1	-0.62	0.351 elite	bad day
Kazuma Okamoto (3B)	TOR	3 / 3	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	1 / 0	1 / 1	-0.6	0.33 above average	bad day
Nathan Lukes (RF)	TOR	2 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	-0.47	0.284 below average	—
Ernie Clement (SS)	TOR	3 / 4	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	-0.47	0.26 weak	—
Nicky Lopez (2B)	TEX	5 / 3	2 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	1 / 1	0.47	0.258 weak	—
Alejandro Kirk (C)	TOR	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	0 / 1	1 / 0	-0.33	0.312 average	—
Josh Smith (2B)	TOR	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	1 / 1	0.33	0.284 below average	—
Josh Jung (DH)	TEX	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	1 / 1	0.33	0.349 elite	—
Evan Carter (CF)	TEX	4 / 3	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	1 / 0	0 / 1	2 / 1	-0.33	0.278 weak	—
Elias Díaz (C)	TEX	3 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 0	2 / 1	0.33	0.267 weak	—
Ezequiel Duran (3B)	TEX	4 / 4	1 / 1	1 / 0	0 / 0	0 / 0	1 / 1	1 / 0	0 / 0	0 / 2	0.31	0.304 average	—
Brandon Nimmo (RF)	TEX	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 0	1 / 0	2 / 0	1 / 1	-0.12	0.343 elite	—
Vladimir Guerrero Jr. (1B)	TOR	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	0 / 0	1 / 1	0	0.336 above average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	0.123	positive favors home; negative favors away
Recent form	0.726	positive favors home; negative favors away
Starting-pitcher matchup	0.062	positive favors home; negative favors away
Composite	0.385	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.79-2.53	n/a	Base ML historical model prediction
variance_applied	5-3	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-5	0.505	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by starter_quality_favours_the_other_side, projected_margin_is_one_run)
embedded_savant_run_environment	4-4	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	4-5	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	4-5	n/a	Capped at max 12 realistic runs
simulation_result	4-5	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.79-2.53	5-3	0.259	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	5-3	4-5	3	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by starter_quality_favours_the_other_side, projected_margin_is_one_run)
embedded_savant_run_environment	4-5	4-4	-1	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	4-4	4-5	1	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
TOR	Josh Smith	batter / 2B	worse	predicted_BB 2.647
TOR	George Springer	batter / DH	worse	predicted_SO 2.241
TOR	Dylan Cease	pitcher / Starter	better	predicted_SO 2
TEX	Brandon Nimmo	batter / RF	worse	predicted_H 1.92
TOR	Brett Bateman	batter / CF	worse	predicted_RBI 1.62
TEX	Nicky Lopez	batter / 2B	worse	predicted_SO 1.512
TOR	Myles Straw	batter / LF	worse	predicted_AB 1.5
TEX	Elias Díaz	batter / C	worse	predicted_RBI 1.49

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-prediction batting totals differ from final simulation box score (AB: current=31, sim=33; H: current=7, sim=10; 2B: current=1, sim=5; RBI: current=4, sim=5; R: current=4, sim=5; BB: current=3, sim=4; SO: current=10, sim=12; SF: current=0, sim=1).	TEX
info	CURRENT_PLAYER_MATCH_BOX	TEX current finalized hitter lines are model predictions and differ from the simulation box score: Ezequiel Duran: AB current=4 box=3, 2B current=0 box=1, RBI current=1 box=0, BB current=0 box=1; Josh Jung: 2B current=0 box=1, RBI current=0 box=1, BB current=0 box=1; Brandon Nimmo: H current=1 box=3, RBI current=0 box=1, R current=0 box=1; Jake Burger: AB current=3 box=4, H current=1 box=2, 2B current=0 box=1, RBI current=1 box=0; Wyatt Langford: AB current=3 box=4, H current=1 box=0, R current=1 box=0, BB current=1 box=0; Corey Seager: SO current=1 box=2; Nicky Lopez: BB current=0 box=1, SO current=1 box=2; Evan Carter: AB current=3 box=4, BB current=1 box=0, SO current=1 box=2	TEX
info	CURRENT_FINALIZED_MATCH_BOX	TOR model pitcher projections differ from final simulation pitching box score (H: current=8, sim=10; BB: current=3, sim=4).	TOR
info	CURRENT_FINALIZED_MATCH_BOX	TOR 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=0).	TOR
info	CURRENT_PLAYER_MATCH_BOX	TOR current finalized hitter lines are model predictions and differ from the simulation box score: Brett Bateman: AB current=4 box=5, H current=1 box=2, RBI current=0 box=2, BB current=1 box=0; Vladimir Guerrero Jr.: 2B current=0 box=1, SO current=1 box=0; Nathan Lukes: H current=1 box=2, RBI current=0 box=1, R current=0 box=1, SO current=1 box=0; George Springer: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, R current=1 box=0; Ernie Clement: H current=1 box=0; Myles Straw: AB current=3 box=4, 2B current=0 box=1; Alejandro Kirk: RBI current=1 box=0, R current=0 box=1, BB current=1 box=0, SO current=0 box=1; Kazuma Okamoto: AB current=3 box=4, H current=1 box=0, HR current=1 box=0, RBI current=1 box=0	TOR
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

WAS@STLMISS

Predicted: STL 3-4 (conf +0.02, toss_up)Actual: WAS 9-1Brier 0.259total bias -3

Root cause: Starting-pitcher projection was wrong

Evidence-backed diagnosis

published system edge 0.02 (toss_up); this is not the half-scale distance from 50%
team-run absolute error 9; total bias -3
pregame home-minus-away evidence: season -0.013, recent +0.213, pitcher +0.077, composite +0.395
confidence pipeline reached 0.511 before later publication/recency controls; published home probability was 0.509
score-volatility actions: no_change
forecast-file integrity: clean (0 defects, 8 informational note(s)); the saved artifact is not why the pick missed

Quantified contributing factors

total runs under-projected by 3 (7 vs 10)
WAS offense under-projected by 6
STL offense over-projected by 3
Cade Cavalli (WAS) allowed 3 fewer ER than projected (0 vs 3)
Kyle Leahy (STL) allowed 5 more ER than projected (6 vs 1)

Team	Starter	Projected	Actual	ER err
WAS	Cade Cavalli	3 ER / 17 outs	0 ER / 15 outs	-3
STL	Kyle Leahy	1 ER / 15 outs	6 ER / 9 outs	5

Decision internals: run factors home×1.01 / away×0.934; 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:

Dylan Crews (WAS RF): 2-for-5, 1 HR, 2 RBI (projected 1.0 H, 0.0 HR) [run impact 1.4]
Jacob Young (WAS CF): 1-for-3, 0 HR, 0 RBI (projected 0.0 H, 0.0 HR) [run impact 1.11]
Alec Burleson (STL 1B): 0-for-3, 0 HR, 0 RBI (projected 1.0 H, 0.0 HR) [run impact -1.11]
Jordan Walker (STL RF): 1-for-3, 0 HR, 0 RBI (projected 1.0 H, 1.0 HR) [run impact -0.93]
CJ Abrams (WAS SS): 1-for-4, 0 HR, 1 RBI (projected 1.0 H, 0.0 HR) [run impact -0.64]

Starter miss: Kyle Leahy (STL) posted 18.00 ERA vs projected 1.80 (6 ER in 9 outs); no credible venue, opponent-strength or experience signal explains it — an outlier start (variance).

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

Pitcher versus batter: what actually broke this game

AWAY pitching

Cade Cavalli (WAS) threw 15 outs, 0 ER = 0 ERA against a projected 4.76 ERA (-4.76 ERA error)

Bullpen behind him: 0 ER over 6 outs.

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

Who did the damage to him:

Leonardo Bernal: 3-for-4, 0 HR, 0 RBI [+0.63 runs vs projection] season xwOBA 0.378, elite - real projection error
Nathan Church: 1-for-3, 0 HR, 0 RBI [+0.31 runs vs projection] season xwOBA 0.289, below average - a good day, not a model error

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

HOME pitching

Kyle Leahy (STL) threw 9 outs, 6 ER = 18 ERA against a projected 1.8 ERA (16.2 ERA error)

Bullpen behind him: 0 ER over 0 outs.

The AWAY lineup he faced produced 9 runs on 7 hits and 1 HR, against a projected 5 runs / 8 hits (4 runs vs projection).

Who did the damage to him:

Dylan Crews: 2-for-5, 1 HR, 2 RBI [+1.4 runs vs projection] season xwOBA 0.312, average - a good day, not a model error
Jacob Young: 1-for-3, 0 HR, 0 RBI [+1.11 runs vs projection] season xwOBA 0.314, average - a good day, not a model error
Brady House: 1-for-4, 0 HR, 1 RBI [+0.31 runs vs projection] season xwOBA 0.337, above average - a good day, not a model error

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

Pitcher	Team	Outs	ERA	ER	H	HR	BB	exp BB	SO	exp SO	Context (as-of)
WAS	Cade Cavalli (Starter)	15 / 17	0 / 4.76 (-4.76)	0 / 3	4 / 5	0 / 1	0 / 3	1.46	6 / 6	7.63	venue away: 2.94 vs 3.25 ERA (ok) · opp run factor 1.01 (average) · 31 prior appearances this season
WAS	Brad Lord (Setup Man)	6 / 4	0 / 6.75 (-6.75)	0 / 1	1 / 1	0 / 0	0 / 0	0.63	3 / 0	1.4
STL	Kyle Leahy (Starter)	9 / 15	18 / 1.8 (+16.20)	6 / 1	5 / 2	1 / 0	3 / 2	0.98	1 / 6	4.37	venue home: 2.79 vs 3.94 ERA (ok) · opp run factor 0.93 (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
Dylan Crews (RF)	WAS	5 / 4	2 / 1	0 / 0	0 / 0	1 / 0	2 / 0	2 / 0	0 / 0	1 / 1	1.4	0.312 average	no signal
Jacob Young (CF)	WAS	3 / 3	1 / 0	1 / 0	0 / 0	0 / 0	0 / 0	2 / 0	1 / 0	0 / 1	1.11	0.314 average	no signal
Alec Burleson (1B)	STL	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	1 / 1	-1.11	0.376 elite	bad day
Jordan Walker (RF)	STL	3 / 4	1 / 1	0 / 0	0 / 0	0 / 1	0 / 1	0 / 1	0 / 0	1 / 1	-0.93	0.344 elite	bad day
CJ Abrams (SS)	WAS	4 / 4	1 / 1	0 / 1	0 / 0	0 / 0	1 / 1	1 / 1	0 / 1	1 / 1	-0.64	0.33 above average	bad day
Leonardo Bernal (C)	STL	4 / 4	3 / 1	0 / 1	0 / 0	0 / 0	0 / 1	0 / 0	0 / 0	1 / 1	0.63	0.378 elite	projection too low
James Wood (DH)	WAS	3 / 4	1 / 1	1 / 0	0 / 0	0 / 1	1 / 1	0 / 1	1 / 1	0 / 1	-0.62	0.412 elite	bad day
Thomas Saggese (3B)	STL	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 0	1 / 1	-0.47	0.319 average	—
Masyn Winn (SS)	STL	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	-0.47	0.299 below average	—
Daylen Lile (LF)	WAS	3 / 4	0 / 1	0 / 1	0 / 0	0 / 0	1 / 1	2 / 1	1 / 0	0 / 1	-0.45	0.327 above average	—
JJ Wetherholt (2B)	STL	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	0 / 1	2 / 1	-0.33	0.35 elite	—
Iván Herrera (DH)	STL	4 / 4	1 / 1	0 / 0	0 / 0	0 / 0	1 / 1	0 / 1	0 / 1	1 / 1	-0.33	0.332 above average	—
Brady House (3B)	WAS	4 / 4	1 / 1	1 / 0	0 / 0	0 / 0	1 / 1	1 / 0	0 / 0	1 / 1	0.31	0.337 above average	—
Yohandy Morales (1B)	WAS	4 / 3	1 / 1	1 / 0	0 / 0	0 / 0	2 / 0	0 / 1	0 / 0	1 / 1	0.31	0.336 above average	—
Nathan Church (CF)	STL	3 / 3	1 / 1	1 / 0	0 / 0	0 / 0	0 / 0	1 / 0	0 / 0	0 / 0	0.31	0.289 below average	—
Jorbit Vivas (2B)	WAS	2 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	1 / 0	2 / 1	0 / 0	-0.14	0.292 below average	—
Harry Ford (C)	WAS	3 / 3	0 / 1	0 / 0	0 / 0	0 / 0	0 / 0	0 / 1	1 / 0	2 / 1	-0.14	0.29 below average	—
Bryan Torres (LF)	STL	4 / 3	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	0 / 0	2 / 1	0	0.298 below average	—

Full model trace, player evidence, and audit findings

Fundamental evidence entering the pick

Signal	Home-away edge	Meaning
Season quality	-0.013	positive favors home; negative favors away
Recent form	0.213	positive favors home; negative favors away
Starting-pitcher matchup	0.077	positive favors home; negative favors away
Composite	0.395	positive favors home; negative favors away

Score and confidence evolution

Stage	Away-home runs	Confidence	What changed
historical_ml	4.43-4.04	n/a	Base ML historical model prediction
variance_applied	4-4	n/a	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	3-4	0.509	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run)
embedded_savant_run_environment	3-4	n/a	Target score adjusted by temporal Savant lineup quality versus opposing starter quality
fundamental_edge_target_review	3-4	n/a	Refreshed score target and win probability after Savant/pitcher/team-strength adjustments
capped_target	3-4	n/a	Capped at max 12 realistic runs
simulation_result	3-4	n/a	Final inning-by-inning simulation result

Official postgame stage attribution

Stage	Before	After	Toward final pick	What changed
variance_applied	4.43-4.04	4-4	0.384	Deterministic baseline target (variance factor 1.0)
enhanced_prediction	4-4	3-4	1	Enhanced ML with recent form, H2H, low-score analysis (confidence shrunk by projected_margin_is_one_run)
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
WAS	Cade Cavalli	pitcher / Starter	better	predicted_SO 4
WAS	Brad Lord	pitcher / Setup Man	worse	predicted_SO -3
WAS	Brady House	batter / 3B	worse	predicted_SO 2
STL	Bryan Torres	batter / LF	better	predicted_RBI 1.705
WAS	CJ Abrams	batter / SS	worse	predicted_AB -1.6
STL	Leonardo Bernal	batter / C	worse	predicted_AB -1.54
STL	Nathan Church	batter / CF	worse	predicted_AB -1.47
WAS	James Wood	batter / DH	worse	predicted_SO -1.317

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 pitcher projections differ from final simulation pitching box score (H: current=3, sim=2).	STL
info	CURRENT_FINALIZED_MATCH_BOX	STL model-prediction batting totals differ from final simulation box score (AB: current=32, sim=30; H: current=8, sim=6; 3B: current=0, sim=1; R: current=5, sim=4; SO: current=8, sim=6).	STL
info	CURRENT_PLAYER_MATCH_BOX	STL current finalized hitter lines are model predictions and differ from the simulation box score: JJ Wetherholt: AB current=4 box=3, H current=1 box=0, R current=1 box=0, SO current=1 box=0; Iván Herrera: BB current=1 box=0, SO current=1 box=0; Thomas Saggese: H current=1 box=0, R current=1 box=0, BB current=0 box=1; Jordan Walker: HR current=1 box=0; Bryan Torres: H current=0 box=2, 3B current=0 box=1, HR current=0 box=1, RBI current=0 box=2; Leonardo Bernal: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Alec Burleson: RBI current=1 box=0, BB current=1 box=0, SO current=1 box=2; Nathan Church: AB current=3 box=2, H current=1 box=0, BB current=0 box=1	STL
info	CURRENT_FINALIZED_MATCH_BOX	WAS model-prediction batting totals differ from final simulation box score (AB: current=32, sim=29; H: current=8, sim=2; HR: current=1, sim=0; RBI: current=4, sim=3; R: current=5, sim=3; SO: current=8, sim=10).	WAS
info	CURRENT_PLAYER_MATCH_BOX	WAS current finalized hitter lines are model predictions and differ from the simulation box score: James Wood: H current=1 box=0, HR current=1 box=0, RBI current=1 box=0, R current=1 box=0; Brady House: H current=1 box=0, RBI current=1 box=0, SO current=1 box=3; Jorbit Vivas: AB current=3 box=2, 2B current=0 box=1, RBI current=0 box=1, R current=0 box=1; Daylen Lile: H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0, R current=1 box=0; Dylan Crews: H current=1 box=0, SO current=1 box=2; CJ Abrams: AB current=4 box=3, H current=1 box=0, 2B current=1 box=0, RBI current=1 box=0; Yohandy Morales: 2B current=0 box=1, RBI current=0 box=1, SO current=1 box=0; Harry Ford: AB current=3 box=2, H current=1 box=0, R current=1 box=0, BB current=0 box=1	WAS
info	RECONCILIATION_PROPAGATION	No score_target_reconciliation step was used for this game.

Cross-game lever tension

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

Day over day: did it improve or worsen

Compared against packet 20260917. 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	0.2425	0.2413	-0.001171	improved
accuracy	0.5917	0.5833	-0.0084	regressed
team_run_mae	2.5333	2.4333	-0.1	improved
total_run_mae	3.5333	3.2333	-0.3	improved
calibration_error	0.0728	0.0572	-0.0156	improved
starter_er_mae	1.7326	1.686	-0.0466	improved
starter_outs_mae	3.2984	3.3721	+0.0737	regressed

Deep slate report — 20260918

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@TBMISS

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@CINMISS

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@CHWHIT

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@BALHIT

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@LAAMISS

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@ARIHIT

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)