Deep slate diagnosis — 20260919

• Games: 15 · hits 11 · misses 4 · winner accuracy 0.7333
• Mean total-run bias on losses: 5.0 (negative = under-projected)
• Mean team-run error on losses: 6.5 · mean Brier on losses: 0.405

Loss primary drivers

• Run environment far lower than projected (over-dispersed): 2
• Starting-pitcher projection was wrong: 1
• Backed the wrong side with real conviction: 1

High-level causal issue register

[CRITICAL] Published probability differs materially without a recorded transformation

• Domain: probability_lineage · affected: ATL@HOU, SEA@COL, SF@LAD
• Evidence: 3 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] An intermediate stage materially degraded winner direction

• Domain: pipeline_stages · affected: ATL@HOU, MIA@SD, WAS@STL
• Evidence: historical_ml → variance_applied changed winner accuracy by -0.2000 (-3 hits)
• Action: List the exact games whose winner flipped at this stage, and keep both the before-stage and after-stage picks in the trace.
• Do not infer: Even if a later stage recovers the pick, this unstable step is still a risk; do not treat it as harmless.

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

• Domain: calibration · affected: 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
• Evidence: weighted absolute calibration error 0.3458 across 15 games
• Action: Watch the same confidence buckets over the full rolling window before changing any calibration weight.
• Do not infer: A single slate has far too few games to establish a stable calibration curve.

[HIGH] Validated run models explain very little holdout variance

• Domain: model_validation · affected: 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
• Evidence: home/away/total R² = 0.0082/0.0273/0.0092
• 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.

[MEDIUM] Most high-raw picks lost before/after shrinkage

• Domain: calibration · affected: NYY@ARI, SEA@COL, SF@LAD
• Evidence: 1/3 games with raw pick probability ≥0.65 won
• Action: Track how often high-raw-probability picks lose across multiple slates, and keep the raw to calibrated to published lineage.
• Do not infer: This one slate is not enough to justify a new per-bucket shrinkage rule.

[INFO] Every settled game received structural QA

• Domain: coverage · affected: 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
• 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: 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
• 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: 20260909, 20260910, 20260911, 20260912, 20260913, 20260914, 20260915, 20260916, 20260917, 20260918, 20260919 (10 prior slates loaded)
• Highest-confidence pick loss streak: 1 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: improving · series [0.6667, 0.6, 0.6667, 0.5333, 0.6, 0.7, 0.4, 0.6667, 0.4444, 0.6667, 0.7333]
• weighted_calibration_error: regressing · series [0.1069, 0.1171, 0.2091, 0.1711, 0.0827, 0.2812, 0.2795, 0.1661, 0.1016, 0.1247, 0.3458]
• margin_bias: improving · series [-2.133, -1.0, -2.933, -2.667, -1.8, -0.4, -1.667, -1.133, -2.111, -2.0, -0.667]
• mean_team_run_error_on_losses: regressing · series [5.6, 6.5, 5.0, 6.143, 5.0, 6.333, 5.556, 6.4, 3.6, 5.2, 6.5]

Pipeline stage quality

• historical_ml: winner 10/15 (0.6667), team-run MAE 2.3021, margin MAE 2.7024
• variance_applied: winner 7/15 (0.4667), team-run MAE 2.3333, margin MAE 2.6667
• enhanced_prediction: winner 11/15 (0.7333), team-run MAE 2.3667, margin MAE 2.4667
• embedded_savant_run_environment: winner 11/15 (0.7333), team-run MAE 2.4, margin MAE 2.4
• fundamental_edge_target_review: winner 11/15 (0.7333), team-run MAE 2.4, margin MAE 2.4
• capped_target: winner 11/15 (0.7333), team-run MAE 2.4, margin MAE 2.4
• simulation_result: winner 11/15 (0.7333), team-run MAE 2.4, margin MAE 2.4
• published_final: winner 11/15 (0.7333), team-run MAE 2.4, margin MAE 2.4

Probability lineage

• ATL@HOU: raw 0.5609 → calibrated 0.5984 → published 0.562; no recorded transformation explains the publication delta
• BOS@TB: raw 0.5686 → calibrated 0.5404 → published 0.5292; score coherence followed by matchup-recency guardrail
• CHC@CIN: raw 0.6443 → calibrated 0.6837 → published 0.56; score coherence followed by matchup-recency guardrail
• DET@CHW: raw 0.5703 → calibrated 0.6277 → published 0.56; score coherence followed by matchup-recency guardrail
• KC@PIT: raw 0.5 → calibrated 0.5482 → published 0.541; no recorded transformation explains the publication delta
• MIA@SD: raw 0.6431 → calibrated 0.6938 → published 0.58; score coherence followed by matchup-recency guardrail
• MIL@BAL: raw 0.5795 → calibrated 0.5462 → published 0.53; score coherence followed by matchup-recency guardrail
• MIN@LAA: raw 0.6036 → calibrated 0.5752 → published 0.5544; score coherence followed by matchup-recency guardrail
• NYY@ARI: raw 0.7133 → calibrated 0.7112 → published 0.7112; published probability matches the calibrated/governed value
• OAK@CLE: raw 0.6147 → calibrated 0.6689 → published 0.58; score coherence followed by matchup-recency guardrail
• PHI@NYM: raw 0.5684 → calibrated 0.5403 → published 0.5291; score coherence followed by matchup-recency guardrail
• SEA@COL: raw 0.7647 → calibrated 0.7642 → published 0.7213; no recorded transformation explains the publication delta
• SF@LAD: raw 0.8181 → calibrated 0.6271 → published 0.5651; no recorded transformation explains the publication delta
• TOR@TEX: raw 0.542 → calibrated 0.611 → published 0.611; published probability matches the calibrated/governed value
• WAS@STL: raw 0.5218 → calibrated 0.5597 → published 0.5431; score coherence followed by matchup-recency guardrail

Every game, in detail

ATL@HOU — ✅ hit

• Predicted ATL 4-3 (conf -0.12, moderate) · actual ATL 6-3
• Root cause: strong_correct_call
• Errors: winner correct · Brier 0.191831 · team-run err 2.0 · total bias -2

BOS@TB — ✅ hit

• Predicted TB 4-5 (conf +0.06, lean) · actual TB 1-2
• Root cause: right_winner_wrong_score
• Errors: winner correct · Brier 0.221655 · team-run err 6.0 · total bias +6
• Contributing factors:
• total runs over-projected by 6 (9 vs 3)
• BOS offense over-projected by 3
• TB offense over-projected by 3

CHC@CIN — ✅ hit

• Predicted CHC 7-4 (conf -0.12, moderate) · actual CHC 5-2
• Root cause: solid_correct_call
• Errors: winner correct · Brier 0.1936 · team-run err 4.0 · total bias +4
• Contributing factors:
• total runs over-projected by 4 (11 vs 7)
• Matthew Boyd (CHC) allowed 3 fewer ER than projected (1 vs 4)

DET@CHW — ❌ miss

• Predicted DET 6-4 (conf -0.12, moderate) · actual CHW 1-3
• Root cause: Run environment far lower than projected (over-dispersed)
• Secondary cause: Backed the wrong side with real conviction
• Errors: winner WRONG · Brier 0.3136 · team-run err 6.0 · total bias +6
• Contributing factors:
• total runs over-projected by 6 (10 vs 4)
• DET offense over-projected by 5
• committed to DET at 0.12 confidence and it lost

KC@PIT — ✅ hit

• Predicted PIT 4-5 (conf +0.08, lean) · actual PIT 5-6
• Root cause: solid_correct_call
• Errors: winner correct · Brier 0.210699 · team-run err 2.0 · total bias -2
• Contributing factors:
• Noah Cameron (KC) allowed 3 more ER than projected (4 vs 1)

MIA@SD — ✅ hit

• Predicted SD 5-8 (conf +0.16, moderate) · actual SD 6-7
• Root cause: strong_correct_call
• Errors: winner correct · Brier 0.1764 · team-run err 2.0 · total bias +0

MIL@BAL — ❌ miss

• Predicted BAL 4-5 (conf +0.06, lean) · actual MIL 1-0
• Root cause: Run environment far lower than projected (over-dispersed)
• Errors: winner WRONG · Brier 0.2809 · team-run err 8.0 · total bias +8
• Contributing factors:
• total runs over-projected by 8 (9 vs 1)
• MIL offense over-projected by 3
• BAL offense over-projected by 5
• Robert Gasser (MIL) allowed 3 fewer ER than projected (0 vs 3)

MIN@LAA — ✅ hit

• Predicted LAA 3-4 (conf +0.11, moderate) · actual LAA 5-6
• Root cause: solid_correct_call
• Errors: winner correct · Brier 0.19859 · team-run err 4.0 · total bias -4
• Contributing factors:
• total runs under-projected by 4 (7 vs 11)

NYY@ARI — ❌ miss

• Predicted NYY 5-2 (conf -0.42, strong) · actual ARI 3-5
• Root cause: Backed the wrong side with real conviction
• Secondary cause: Small edge on the losing side
• Errors: winner WRONG · Brier 0.505875 · team-run err 5.0 · total bias -1
• Contributing factors:
• ARI offense under-projected by 3
• committed to NYY at 0.42 confidence and it lost

OAK@CLE — ✅ hit

• Predicted CLE 4-6 (conf +0.16, moderate) · actual CLE 6-12
• Root cause: right_winner_wrong_score
• Errors: winner correct · Brier 0.1764 · team-run err 8.0 · total bias -8
• Contributing factors:
• total runs under-projected by 8 (10 vs 18)
• CLE offense under-projected by 6
• Jacob Lopez (OAK) allowed 6 more ER than projected (10 vs 4)
• Tanner Bibee (CLE) allowed 5 more ER than projected (6 vs 1)

PHI@NYM — ✅ hit

• Predicted NYM 5-6 (conf +0.06, lean) · actual NYM 3-10
• Root cause: right_winner_wrong_score
• Errors: winner correct · Brier 0.22174 · team-run err 6.0 · total bias -2
• Contributing factors:
• NYM offense under-projected by 4

SEA@COL — ❌ miss

• Predicted SEA 10-5 (conf -0.44, strong) · actual COL 3-5
• Root cause: Starting-pitcher projection was wrong
• Secondary cause: Run environment far lower than projected (over-dispersed)
• Errors: winner WRONG · Brier 0.520279 · team-run err 7.0 · total bias +7
• Contributing factors:
• total runs over-projected by 7 (15 vs 8)
• SEA offense over-projected by 7
• Bryce Miller (SEA) allowed 5 fewer ER than projected (0 vs 5)
• committed to SEA at 0.44 confidence and it lost

SF@LAD — ✅ hit

• Predicted LAD 1-7 (conf +0.13, moderate) · actual LAD 4-10
• Root cause: right_winner_wrong_score
• Errors: winner correct · Brier 0.189111 · team-run err 6.0 · total bias -6
• Contributing factors:
• total runs under-projected by 6 (8 vs 14)
• SF offense under-projected by 3
• LAD offense under-projected by 3
• Tarik Skubal (LAD) allowed 4 more ER than projected (4 vs 0)

TOR@TEX — ✅ hit

• Predicted TEX 2-4 (conf +0.22, strong) · actual TEX 2-6
• Root cause: strong_correct_call
• Errors: winner correct · Brier 0.151319 · team-run err 2.0 · total bias -2

WAS@STL — ✅ hit

• Predicted WAS 5-4 (conf -0.09, lean) · actual WAS 8-5
• Root cause: solid_correct_call
• Errors: winner correct · Brier 0.208754 · team-run err 4.0 · total bias -4
• Contributing factors:
• total runs under-projected by 4 (9 vs 13)
• WAS 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: 3 helped vs 1 hurt
• Would help:
• DET@CHW (miss): lost a 0.12-confidence pick; wider tails lower overconfidence
• NYY@ARI (miss): lost a 0.42-confidence pick; wider tails lower overconfidence
• SEA@COL (miss): lost a 0.44-confidence pick; wider tails lower overconfidence
• Would hurt / risk:
• TOR@TEX (hit): strong correct call at 0.22; widening softens a good edge

player.pitcher.v4 form weights / player.pitcher_prior_games

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