Deep slate diagnosis — 20260829

• Games: 17 · hits 16 · misses 1 · winner accuracy 0.9412
• Mean total-run bias on losses: 3.0 (negative = under-projected)
• Mean team-run error on losses: 3.0 · mean Brier on losses: 0.443

Loss primary drivers

• Backed the wrong side with real conviction: 1

High-level causal issue register

[HIGH] Blocking canonical artifact error

• Domain: artifact_integrity · affected: 20260829_ARI@SF_2
• Evidence: structural audit totals: {'critical': 0, 'error': 2, 'warning': 1, 'info': 125}
• Action: quarantine the slate from parameter learning and repair the generation boundary
• Do not infer: do not attribute prediction accuracy changes to a malformed artifact

[HIGH] Published confidence buckets were poorly aligned on this slate

• Domain: calibration · affected: ARI@SF, ARI@SF, BAL@OAK, BOS@NYY, BOS@NYY, CHW@MIN, CIN@CHC, COL@ATL, HOU@NYM, KC@CLE, LAD@DET, MIA@WAS, PHI@LAA, PIT@STL, SD@TB, SEA@TOR, TEX@MIL
• Evidence: weighted absolute calibration error 0.3961 across 17 games
• Action: monitor the same buckets over the registered rolling window before changing calibration weights
• Do not infer: seven games cannot establish a stable calibration curve

[HIGH] Validated run models explain very little holdout variance

• Domain: model_validation · affected: ARI@SF, ARI@SF, BAL@OAK, BOS@NYY, BOS@NYY, CHW@MIN, CIN@CHC, COL@ATL, HOU@NYM, KC@CLE, LAD@DET, MIA@WAS, PHI@LAA, PIT@STL, SD@TB, SEA@TOR, TEX@MIL
• Evidence: home/away/total R² = 0.0452/0.0220/0.0224
• Action: treat exact scores as low-confidence estimates and investigate feature/target signal before tuning tails
• Do not infer: positive R² means better than the mean baseline, not that exact scores are reliable

[INFO] Every settled game received structural QA

• Domain: coverage · affected: ARI@SF, ARI@SF, BAL@OAK, BOS@NYY, BOS@NYY, CHW@MIN, CIN@CHC, COL@ATL, HOU@NYM, KC@CLE, LAD@DET, MIA@WAS, PHI@LAA, PIT@STL, SD@TB, SEA@TOR, TEX@MIL
• Evidence: 17/17 games structurally audited
• Action: retain the per-game receipt in every report
• Do not infer: complete coverage does not erase findings inside the audited games

[INFO] Material calibrated-to-published movements have recorded lineage

• Domain: probability_lineage · affected: BOS@NYY, BOS@NYY, CIN@CHC, COL@ATL, HOU@NYM, MIA@WAS, PIT@STL, SEA@TOR, TEX@MIL
• Evidence: 9 material movement(s), all explained by named score-coherence and/or recency transformations
• Action: keep displaying each probability layer and transformation
• Do not infer: a higher published value is not automatically a governor bypass

[INFO] Simulation produced no winner-direction intervention on this slate

• Domain: simulation · affected: ARI@SF, ARI@SF, BAL@OAK, BOS@NYY, BOS@NYY, CHW@MIN, CIN@CHC, COL@ATL, HOU@NYM, KC@CLE, LAD@DET, MIA@WAS, PHI@LAA, PIT@STL, SD@TB, SEA@TOR, TEX@MIL
• Evidence: flip candidates 0; published flips 0
• Action: measure simulation contribution over rolling slates and distinguish score realization from winner flipping
• Do not infer: zero flips is not itself a defect and does not mean simulation did no score-reconciliation work

Pipeline stage quality

• historical_ml: winner 15/17 (0.8824), team-run MAE 2.1119, margin MAE 3.0082
• variance_applied: winner 13/17 (0.7647), team-run MAE 2.1765, margin MAE 2.9412
• enhanced_prediction: winner 16/17 (0.9412), team-run MAE 2.1765, margin MAE 2.9412
• embedded_savant_run_environment: winner 15/17 (0.8824), team-run MAE 2.2647, margin MAE 3.1176
• fundamental_edge_target_review: winner 16/17 (0.9412), team-run MAE 2.2353, margin MAE 3.0588
• capped_target: winner 16/17 (0.9412), team-run MAE 2.2353, margin MAE 3.0588
• simulation_result: winner 16/17 (0.9412), team-run MAE 2.2353, margin MAE 3.0588
• published_final: winner 16/17 (0.9412), team-run MAE 2.2353, margin MAE 3.0588

Probability lineage

• ARI@SF: raw 0.5085 → calibrated 0.5082 → published 0.5269; score coherence followed by matchup-recency guardrail
• ARI@SF: raw 0.5085 → calibrated 0.5082 → published 0.5185; score coherence followed by matchup-recency guardrail
• BAL@OAK: raw 0.568 → calibrated 0.5661 → published 0.5546; score coherence followed by matchup-recency guardrail
• BOS@NYY: raw 0.5085 → calibrated 0.5082 → published 0.53; score coherence followed by matchup-recency guardrail
• BOS@NYY: raw 0.5085 → calibrated 0.5082 → published 0.5337; score coherence followed by matchup-recency guardrail
• CHW@MIN: raw 0.585 → calibrated 0.5814 → published 0.5707; score coherence followed by matchup-recency guardrail
• CIN@CHC: raw 0.6815 → calibrated 0.6705 → published 0.56; score coherence followed by matchup-recency guardrail
• COL@ATL: raw 0.8613 → calibrated 0.6551 → published 0.7345; score_coherence aligned probability to the published score margin
• HOU@NYM: raw 0.5085 → calibrated 0.5082 → published 0.53; score coherence followed by matchup-recency guardrail
• KC@CLE: raw 0.56 → calibrated 0.5584 → published 0.5518; score coherence followed by matchup-recency guardrail
• LAD@DET: raw 0.67 → calibrated 0.6608 → published 0.6654; score_coherence aligned probability to the published score margin
• MIA@WAS: raw 0.585 → calibrated 0.5814 → published 0.5601; score coherence followed by matchup-recency guardrail
• PHI@LAA: raw 0.67 → calibrated 0.6608 → published 0.6654; score_coherence aligned probability to the published score margin
• PIT@STL: raw 0.5085 → calibrated 0.5082 → published 0.53; score coherence followed by matchup-recency guardrail
• SD@TB: raw 0.6199 → calibrated 0.6157 → published 0.6214; no recorded transformation explains the publication delta
• SEA@TOR: raw 0.624 → calibrated 0.6193 → published 0.54; score coherence followed by matchup-recency guardrail
• TEX@MIL: raw 0.8613 → calibrated 0.6551 → published 0.7345; score_coherence aligned probability to the published score margin

Every game, in detail

ARI@SF — ✅ hit

• Predicted ARI 7-6 (conf -0.05, lean) · actual ARI 7-1
• Root cause: right_winner_wrong_score
• Errors: winner correct · Brier 0.223787 · team-run err 5.0 · total bias +5
• Contributing factors:
• total runs over-projected by 5 (13 vs 8)
• SF offense over-projected by 5
• Matt Wilkinson (SF) allowed 5 fewer ER than projected (1 vs 6)

ARI@SF — ✅ hit

• Predicted SF 5-6 (conf +0.04, toss_up) · actual SF 2-7
• Root cause: toss_up_went_our_way
• Errors: winner correct · Brier 0.231822 · team-run err 4.0 · total bias +2
• Contributing factors:
• ARI offense over-projected by 3

BAL@OAK — ✅ hit

• Predicted BAL 7-6 (conf -0.11, moderate) · actual BAL 5-3
• Root cause: right_winner_wrong_score
• Errors: winner correct · Brier 0.19839 · team-run err 5.0 · total bias +5
• Contributing factors:
• total runs over-projected by 5 (13 vs 8)
• OAK offense over-projected by 3
• Jack Perkins (OAK) allowed 5 fewer ER than projected (1 vs 6)

BOS@NYY — ✅ hit

• Predicted BOS 5-4 (conf -0.06, lean) · actual BOS 6-0
• Root cause: right_winner_wrong_score
• Errors: winner correct · Brier 0.2209 · team-run err 5.0 · total bias +3
• Contributing factors:
• total runs over-projected by 3 (9 vs 6)
• NYY offense over-projected by 4
• Jake Bennett (BOS) allowed 3 fewer ER than projected (0 vs 3)

BOS@NYY — ✅ hit

• Predicted NYY 3-4 (conf +0.07, lean) · actual NYY 2-9
• Root cause: right_winner_wrong_score
• Errors: winner correct · Brier 0.21746 · team-run err 6.0 · total bias -4
• Contributing factors:
• total runs under-projected by 4 (7 vs 11)
• NYY offense under-projected by 5

CHW@MIN — ✅ hit

• Predicted CHW 7-5 (conf -0.14, moderate) · actual CHW 3-2
• Root cause: right_winner_wrong_score
• Errors: winner correct · Brier 0.184274 · team-run err 7.0 · total bias +7
• Contributing factors:
• total runs over-projected by 7 (12 vs 5)
• CHW offense over-projected by 4
• MIN offense over-projected by 3
• Bailey Ober (MIN) allowed 4 fewer ER than projected (1 vs 5)

CIN@CHC — ✅ hit

• Predicted CHC 5-7 (conf +0.12, moderate) · actual CHC 5-17
• Root cause: right_winner_wrong_score
• Errors: winner correct · Brier 0.1936 · team-run err 10.0 · total bias -10
• Contributing factors:
• total runs under-projected by 10 (12 vs 22)
• CHC offense under-projected by 10
• Andrew Abbott (CIN) allowed 4 more ER than projected (7 vs 3)
• Kevin Gausman (CHC) allowed 3 more ER than projected (5 vs 2)

COL@ATL — ✅ hit

• Predicted ATL 2-7 (conf +0.47, strong) · actual ATL 1-2
• Root cause: right_winner_wrong_score
• Errors: winner correct · Brier 0.070514 · team-run err 6.0 · total bias +6
• Contributing factors:
• total runs over-projected by 6 (9 vs 3)
• ATL offense over-projected by 5
• Zach Agnos (COL) allowed 4 fewer ER than projected (0 vs 4)

HOU@NYM — ✅ hit

• Predicted NYM 4-5 (conf +0.06, lean) · actual NYM 2-6
• Root cause: solid_correct_call
• Errors: winner correct · Brier 0.2209 · team-run err 3.0 · total bias +1
• Contributing factors:
• Bennett Sousa (HOU) allowed 4 fewer ER than projected (1 vs 5)

KC@CLE — ✅ hit

• Predicted KC 5-4 (conf -0.10, moderate) · actual KC 8-3
• Root cause: solid_correct_call
• Errors: winner correct · Brier 0.200878 · team-run err 4.0 · total bias -2
• Contributing factors:
• KC offense under-projected by 3

LAD@DET — ❌ miss

• Predicted LAD 4-2 (conf -0.33, strong) · actual DET 1-2
• Root cause: Backed the wrong side with real conviction
• Errors: winner WRONG · Brier 0.442752 · team-run err 3.0 · total bias +3
• Contributing factors:
• total runs over-projected by 3 (6 vs 3)
• LAD offense over-projected by 3
• committed to LAD at 0.33 confidence and it lost

MIA@WAS — ✅ hit

• Predicted WAS 2-3 (conf +0.12, moderate) · actual WAS 4-5
• Root cause: solid_correct_call
• Errors: winner correct · Brier 0.193494 · team-run err 4.0 · total bias -4
• Contributing factors:
• total runs under-projected by 4 (5 vs 9)

PHI@LAA — ✅ hit

• Predicted PHI 6-3 (conf -0.33, strong) · actual PHI 4-2
• Root cause: strong_correct_call
• Errors: winner correct · Brier 0.11196 · team-run err 3.0 · total bias +3
• Contributing factors:
• total runs over-projected by 3 (9 vs 6)

PIT@STL — ✅ hit

• Predicted PIT 4-3 (conf -0.06, lean) · actual PIT 6-2
• Root cause: solid_correct_call
• Errors: winner correct · Brier 0.2209 · team-run err 3.0 · total bias -1

SD@TB — ✅ hit

• Predicted TB 4-5 (conf +0.24, strong) · actual TB 6-7
• Root cause: solid_correct_call
• Errors: winner correct · Brier 0.143333 · team-run err 4.0 · total bias -4
• Contributing factors:
• total runs under-projected by 4 (9 vs 13)

SEA@TOR — ✅ hit

• Predicted TOR 4-5 (conf +0.08, lean) · actual TOR 3-4
• Root cause: solid_correct_call
• Errors: winner correct · Brier 0.2116 · team-run err 2.0 · total bias +2

TEX@MIL — ✅ hit

• Predicted MIL 3-7 (conf +0.47, strong) · actual MIL 3-5
• Root cause: strong_correct_call
• Errors: winner correct · Brier 0.070514 · team-run err 2.0 · total bias +2

Cross-game lever tension

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

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

• Targets: brier_score, calibration_error
• Net: 1 helped vs 4 hurt
• Would help:
• LAD@DET (miss): lost a 0.33-confidence pick; wider tails lower overconfidence
• Would hurt / risk:
• COL@ATL (hit): strong correct call at 0.47; widening softens a good edge
• PHI@LAA (hit): strong correct call at 0.33; widening softens a good edge
• SD@TB (hit): strong correct call at 0.24; widening softens a good edge
• TEX@MIL (hit): strong correct call at 0.47; widening softens a good edge

player.pitcher.v4 form weights / player.pitcher_prior_games

• Targets: starter_er_mae, starter_outs_mae
• Net: 7 helped vs 5 hurt
• Would help:
• ARI@SF (hit): starter ER off by 5
• BAL@OAK (hit): starter ER off by 5
• BOS@NYY (hit): starter ER off by 3
• CHW@MIN (hit): starter ER off by 4
• CIN@CHC (hit): starter ER off by 4
• COL@ATL (hit): starter ER off by 4
• HOU@NYM (hit): starter ER off by 4
• Would hurt / risk:
• KC@CLE (hit): starter projection was accurate (ER off 1); reweight risks disturbing it
• LAD@DET (miss): starter projection was accurate (ER off 1); reweight risks disturbing it
• SD@TB (hit): starter projection was accurate (ER off 1); reweight risks disturbing it
• SEA@TOR (hit): starter projection was accurate (ER off 0); reweight risks disturbing it
• TEX@MIL (hit): starter projection was accurate (ER off 1); reweight risks disturbing it

recent. / season. run-mean weighting

• Targets: team_run_mae, total_run_mae, accuracy
• Net: 11 helped vs 2 hurt
• Would help:
• ARI@SF (hit): team runs off by 5
• ARI@SF (hit): team runs off by 4
• BAL@OAK (hit): team runs off by 5
• BOS@NYY (hit): team runs off by 5
• BOS@NYY (hit): team runs off by 6
• CHW@MIN (hit): team runs off by 7
• CIN@CHC (hit): team runs off by 10
• COL@ATL (hit): team runs off by 6
• KC@CLE (hit): team runs off by 4
• MIA@WAS (hit): team runs off by 4
• SD@TB (hit): team runs off by 4
• Would hurt / risk:
• SEA@TOR (hit): team runs already close (off 2); reweight risks disturbing it
• TEX@MIL (hit): team runs already close (off 2); reweight risks disturbing it