Deep slate diagnosis — 20260913

• Games: 15 · hits 9 · misses 6 · winner accuracy 0.6
• Mean total-run bias on losses: -0.333 (negative = under-projected)
• Mean team-run error on losses: 5.0 · mean Brier on losses: 0.305

Loss primary drivers

• Small edge on the losing side: 2
• Run environment far lower than projected (over-dispersed): 1
• Genuine toss-up that flipped: 1
• Backed the wrong side with real conviction: 2

High-level causal issue register

[CRITICAL] Published probability differs materially without a recorded transformation

• Domain: probability_lineage · affected: COL@DET, LAA@WAS
• Evidence: 2 game(s) have an unexplained ≥2pp calibrated-to-published delta
• Action: Block any promotion until the missing probability transformation is recorded and reconciled.
• Do not infer: Do not assume the probability governor was bypassed until you have traced the pick-side math.

[HIGH] Canonical artifact warning requires reconciliation

• Domain: artifact_integrity · affected: slate-wide
• Evidence: structural audit totals: {'critical': 0, 'error': 0, 'warning': 1, 'info': 110}
• Action: Fix the flagged stat or presentation contract before treating this slate as clean evidence.
• Do not infer: A warning flags a display or contract issue; it does not prove the model picked the wrong side.

[HIGH] Validated run models explain very little holdout variance

• Domain: model_validation · affected: BAL@TOR, CHW@STL, CIN@MIL, CLE@MIN, COL@DET, HOU@TB, KC@BOS, LAA@WAS, LAD@MIA, NYM@NYY, PHI@ATL, PIT@CHC, SD@SF, SEA@OAK, TEX@ARI
• Evidence: home/away/total R² = 0.0118/0.0210/0.0207
• Action: Treat exact run scores as low-confidence estimates, and look for stronger features or a better target before tuning the score tails.
• Do not infer: A small positive R² only means the model beats guessing the average; it does not make the exact scores reliable.

[INFO] Every settled game received structural QA

• Domain: coverage · affected: BAL@TOR, CHW@STL, CIN@MIL, CLE@MIN, COL@DET, HOU@TB, KC@BOS, LAA@WAS, LAD@MIA, NYM@NYY, PHI@ATL, PIT@CHC, SD@SF, SEA@OAK, TEX@ARI
• Evidence: 15/15 games structurally audited
• Action: Keep the per-game audit receipt in every report.
• Do not infer: Full audit coverage does not cancel the findings raised inside the audited games.

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

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

Cross-slate trends (persistent-pattern memory)

• Window: 20260903, 20260904, 20260905, 20260906, 20260907, 20260908, 20260909, 20260910, 20260911, 20260912, 20260913 (10 prior slates loaded)
• Highest-confidence pick loss streak: 0 slate(s)
• Actionable pattern detected: True

Escalated persistent patterns

• simulation_inert_on_slate — 11 consecutive slates (11 in window)
• Target lever(s): simulation.allow_score
• Direction: confirm the simulation module is running, not silently no-oping
• Guardrail: 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 slates (11 in window)
• Target lever(s): training.xgb.n_estimators, training.xgb.max_depth, training.xgb.learning_rate
• Direction: investigate features/target via isolated training backtest (manual)
• Guardrail: 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.

Metric direction across the window

• winner_accuracy: flat · series [0.5556, 0.625, 0.5333, 0.6, 0.5455, 0.7333, 0.6667, 0.6, 0.6667, 0.5333, 0.6]
• weighted_calibration_error: improving · series [0.1474, 0.1316, 0.1, 0.0937, 0.2611, 0.1825, 0.1069, 0.1171, 0.2091, 0.1711, 0.0827]
• margin_bias: regressing · series [-1.778, -0.688, -1.467, -1.467, -0.182, -0.867, -2.133, -1.0, -2.933, -2.667, -1.8]
• mean_team_run_error_on_losses: improving · series [6.75, 6.0, 5.286, 5.333, 3.8, 6.75, 5.6, 6.5, 5.0, 6.143, 5.0]

Pipeline stage quality

• historical_ml: winner 9/15 (0.6), team-run MAE 2.3252, margin MAE 3.2574
• variance_applied: winner 8/15 (0.5333), team-run MAE 2.3667, margin MAE 3.2667
• enhanced_prediction: winner 9/15 (0.6), team-run MAE 2.5333, margin MAE 3.4667
• embedded_savant_run_environment: winner 9/15 (0.6), team-run MAE 2.4333, margin MAE 3.2667
• fundamental_edge_target_review: winner 9/15 (0.6), team-run MAE 2.4333, margin MAE 3.2667
• capped_target: winner 9/15 (0.6), team-run MAE 2.4333, margin MAE 3.2667
• simulation_result: winner 9/15 (0.6), team-run MAE 2.4333, margin MAE 3.2667
• published_final: winner 9/15 (0.6), team-run MAE 2.4333, margin MAE 3.2667

Probability lineage

• BAL@TOR: raw 0.5579 → calibrated 0.5348 → published 0.5296; score coherence followed by matchup-recency guardrail
• CHW@STL: raw 0.5407 → calibrated 0.5255 → published 0.5085; score coherence followed by matchup-recency guardrail
• CIN@MIL: raw 0.587 → calibrated 0.7356 → published 0.54; score coherence followed by matchup-recency guardrail
• CLE@MIN: raw 0.6131 → calibrated 0.5637 → published 0.5461; score coherence followed by matchup-recency guardrail
• COL@DET: raw 0.6515 → calibrated 0.7168 → published 0.6464; no recorded transformation explains the publication delta
• HOU@TB: raw 0.5595 → calibrated 0.5357 → published 0.5202; score coherence followed by matchup-recency guardrail
• KC@BOS: raw 0.6344 → calibrated 0.5918 → published 0.56; score coherence followed by matchup-recency guardrail
• LAA@WAS: raw 0.6486 → calibrated 0.6294 → published 0.5711; no recorded transformation explains the publication delta
• LAD@MIA: raw 0.6331 → calibrated 0.7075 → published 0.6; score coherence followed by matchup-recency guardrail
• NYM@NYY: raw 0.6318 → calibrated 0.5973 → published 0.5639; score coherence followed by matchup-recency guardrail
• PHI@ATL: raw 0.5713 → calibrated 0.5596 → published 0.543; score coherence followed by matchup-recency guardrail
• PIT@CHC: raw 0.5418 → calibrated 0.5261 → published 0.5189; score coherence followed by matchup-recency guardrail
• SD@SF: raw 0.5995 → calibrated 0.5739 → published 0.539; score coherence followed by matchup-recency guardrail
• SEA@OAK: raw 0.6045 → calibrated 0.5982 → published 0.56; score coherence followed by matchup-recency guardrail
• TEX@ARI: raw 0.5 → calibrated 0.5078 → published 0.5057; published probability matches the calibrated/governed value

Every game, in detail

BAL@TOR — ✅ hit

• Predicted TOR 2-4 (conf +0.06, lean) · actual TOR 1-8
• Root cause: right_winner_wrong_score
• Errors: winner correct · Brier 0.221256 · team-run err 5.0 · total bias -3
• Contributing factors:
• total runs under-projected by 3 (6 vs 9)
• TOR offense under-projected by 4

CHW@STL — ✅ hit

• Predicted STL 5-6 (conf +0.02, toss_up) · actual STL 1-3
• Root cause: toss_up_went_our_way
• Errors: winner correct · Brier 0.241616 · team-run err 7.0 · total bias +7
• Contributing factors:
• total runs over-projected by 7 (11 vs 4)
• CHW offense over-projected by 4
• STL offense over-projected by 3
• David Sandlin (CHW) allowed 6 fewer ER than projected (0 vs 6)

CIN@MIL — ❌ miss

• Predicted MIL 4-8 (conf +0.08, lean) · actual CIN 4-3
• Root cause: Run environment far lower than projected (over-dispersed)
• Errors: winner WRONG · Brier 0.2916 · team-run err 5.0 · total bias +5
• Contributing factors:
• total runs over-projected by 5 (12 vs 7)
• MIL offense over-projected by 5

CLE@MIN — ❌ miss

• Predicted MIN 4-5 (conf +0.09, lean) · actual CLE 9-2
• Root cause: Small edge on the losing side
• Errors: winner WRONG · Brier 0.298172 · team-run err 8.0 · total bias -2
• Contributing factors:
• CLE offense under-projected by 5
• MIN offense over-projected by 3
• Tanner Bibee (CLE) allowed 3 fewer ER than projected (2 vs 5)

COL@DET — ✅ hit

• Predicted DET 4-7 (conf +0.29, strong) · actual DET 1-8
• Root cause: solid_correct_call
• Errors: winner correct · Brier 0.125063 · team-run err 4.0 · total bias +2
• Contributing factors:
• COL offense over-projected by 3

HOU@TB — ✅ hit

• Predicted TB 3-4 (conf +0.04, lean) · actual TB 4-14
• Root cause: right_winner_wrong_score
• Errors: winner correct · Brier 0.230215 · team-run err 11.0 · total bias -11
• Contributing factors:
• total runs under-projected by 11 (7 vs 18)
• TB offense under-projected by 10

KC@BOS — ✅ hit

• Predicted BOS 2-4 (conf +0.12, moderate) · actual BOS 1-4
• Root cause: strong_correct_call
• Errors: winner correct · Brier 0.1936 · team-run err 1.0 · total bias +1
• Contributing factors:
• Noah Cameron (KC) allowed 3 fewer ER than projected (1 vs 4)

LAA@WAS — ✅ hit

• Predicted WAS 5-6 (conf +0.14, moderate) · actual WAS 5-6
• Root cause: strong_correct_call
• Errors: winner correct · Brier 0.183936 · team-run err 0.0 · total bias +0

LAD@MIA — ❌ miss

• Predicted LAD 8-5 (conf -0.20, strong) · actual MIA 4-6
• Root cause: Backed the wrong side with real conviction
• Secondary cause: Small edge on the losing side
• Errors: winner WRONG · Brier 0.36 · team-run err 5.0 · total bias +3
• Contributing factors:
• total runs over-projected by 3 (13 vs 10)
• LAD offense over-projected by 4
• committed to LAD at 0.20 confidence and it lost

NYM@NYY — ✅ hit

• Predicted NYY 5-6 (conf +0.13, moderate) · actual NYY 0-2
• Root cause: right_winner_wrong_score
• Errors: winner correct · Brier 0.190144 · team-run err 9.0 · total bias +9
• Contributing factors:
• total runs over-projected by 9 (11 vs 2)
• NYM offense over-projected by 5
• NYY offense over-projected by 4

PHI@ATL — ❌ miss

• Predicted ATL 4-5 (conf +0.09, lean) · actual PHI 9-4
• Root cause: Small edge on the losing side
• Errors: winner WRONG · Brier 0.294891 · team-run err 6.0 · total bias -4
• Contributing factors:
• total runs under-projected by 4 (9 vs 13)
• PHI offense under-projected by 5

PIT@CHC — ❌ miss

• Predicted CHC 3-4 (conf +0.04, toss_up) · actual PIT 4-3
• Root cause: Genuine toss-up that flipped
• Errors: winner WRONG · Brier 0.269242 · team-run err 2.0 · total bias +0

SD@SF — ✅ hit

• Predicted SD 6-4 (conf -0.08, lean) · actual SD 6-4
• Root cause: solid_correct_call
• Errors: winner correct · Brier 0.212503 · team-run err 0.0 · total bias +0
• Contributing factors:
• Logan Webb (SF) allowed 5 more ER than projected (6 vs 1)

SEA@OAK — ❌ miss

• Predicted SEA 6-5 (conf -0.12, moderate) · actual OAK 7-8
• Root cause: Backed the wrong side with real conviction
• Secondary cause: Small edge on the losing side
• Errors: winner WRONG · Brier 0.3136 · team-run err 4.0 · total bias -4
• Contributing factors:
• total runs under-projected by 4 (11 vs 15)
• OAK offense under-projected by 3
• committed to SEA at 0.12 confidence and it lost

TEX@ARI — ✅ hit

• Predicted TEX 4-3 (conf -0.01, toss_up) · actual TEX 7-6
• Root cause: toss_up_went_our_way
• Errors: winner correct · Brier 0.244381 · team-run err 6.0 · total bias -6
• Contributing factors:
• total runs under-projected by 6 (7 vs 13)
• TEX offense under-projected by 3
• ARI offense under-projected by 3

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: 2 helped vs 1 hurt
• Would help:
• LAD@MIA (miss): lost a 0.20-confidence pick; wider tails lower overconfidence
• SEA@OAK (miss): lost a 0.12-confidence pick; wider tails lower overconfidence
• Would hurt / risk:
• COL@DET (hit): strong correct call at 0.29; widening softens a good edge

player.pitcher.v4 form weights / player.pitcher_prior_games

• Targets: starter_er_mae, starter_outs_mae
• Net: 4 helped vs 7 hurt
• Would help:
• CHW@STL (hit): starter ER off by 6
• CLE@MIN (miss): starter ER off by 3
• KC@BOS (hit): starter ER off by 3
• SD@SF (hit): starter ER off by 5
• Would hurt / risk:
• BAL@TOR (hit): starter projection was accurate (ER off 0); reweight risks disturbing it
• CIN@MIL (miss): starter projection was accurate (ER off 1); reweight risks disturbing it
• COL@DET (hit): starter projection was accurate (ER off 1); reweight risks disturbing it
• HOU@TB (hit): starter projection was accurate (ER off 1); reweight risks disturbing it
• NYM@NYY (hit): starter projection was accurate (ER off 1); reweight risks disturbing it
• PHI@ATL (miss): starter projection was accurate (ER off 1); reweight risks disturbing it
• TEX@ARI (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 4 hurt
• Would help:
• BAL@TOR (hit): team runs off by 5
• CHW@STL (hit): team runs off by 7
• CIN@MIL (miss): team runs off by 5
• CLE@MIN (miss): team runs off by 8
• COL@DET (hit): team runs off by 4
• HOU@TB (hit): team runs off by 11
• LAD@MIA (miss): team runs off by 5
• NYM@NYY (hit): team runs off by 9
• PHI@ATL (miss): team runs off by 6
• SEA@OAK (miss): team runs off by 4
• TEX@ARI (hit): team runs off by 6
• Would hurt / risk:
• KC@BOS (hit): team runs already close (off 1); reweight risks disturbing it
• LAA@WAS (hit): team runs already close (off 0); reweight risks disturbing it
• PIT@CHC (miss): team runs already close (off 2); reweight risks disturbing it
• SD@SF (hit): team runs already close (off 0); reweight risks disturbing it