Deep slate diagnosis — 20261001

• Games: 1 · hits 1 · misses 0 · winner accuracy 1.0
• Mean total-run bias on losses: None (negative = under-projected)
• Mean team-run error on losses: None · mean Brier on losses: None

Loss primary drivers

High-level causal issue register

[HIGH] Published margins are compressed relative to realized margins

• Domain: score_distribution · affected: PHI@ATL
• Evidence: 1/1 predictions were ≤2 runs; 1/1 actual margins were ≥4; mean predicted margin 1.00 vs actual 4.00
• Action: Check whether predicted margins are too narrow across several recent out-of-sample slates (measuring tail coverage and margin MAE) before widening the score spread.
• Do not infer: A single blowout game is not a reason to widen every score prediction.

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

• Domain: calibration · affected: PHI@ATL
• Evidence: weighted absolute calibration error 0.4832 across 1 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: PHI@ATL
• Evidence: home/away/total R² = 0.0021/0.0309/0.0024
• 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: PHI@ATL
• Evidence: 4/1 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: PHI@ATL
• 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: 20260921, 20260922, 20260923, 20260924, 20260925, 20260926, 20260927, , 20260929, 20260930, 20261001 (10 prior slates loaded)
• Highest-confidence pick loss streak: 0 slate(s)
• Actionable pattern detected: True

Escalated persistent patterns

• score_margin_compression — 3 consecutive slates (5 in window)
• Target lever(s): score.variance_multiplier, score.variance_floor
• Direction: widen run distributions
• Guardrail: score_margin_compression has persisted 3 consecutive slates; this is no longer single-slate noise. Treat it as an actionable pattern and propose a bounded shadow experiment against a target lever.
• 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.
• single_slate_calibration_instability — 3 consecutive slates (7 in window)
• Target lever(s): confidence.empirical_max_weight, confidence.base_cap, confidence.moderate_cap
• Direction: increase empirical shrinkage
• Guardrail: single_slate_calibration_instability has persisted 3 consecutive slates; this is no longer single-slate noise. Treat it as an actionable pattern and propose a bounded shadow experiment against a target lever.
• weak_run_model_signal — 3 consecutive slates (10 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 3 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 [0.3333, 0.4, 0.5, 0.6667, 0.6471, 0.5385, 0.6429, None, 1.0, 0.75, 1.0]
• weighted_calibration_error: regressing · series [0.2028, 0.1881, 0.1108, 0.1701, 0.1302, 0.1524, 0.1031, None, 0.4792, 0.2317, 0.4832]
• margin_bias: regressing · series [-2.667, -1.867, -1.188, -1.083, -1.706, -3.0, -1.571, None, -4.25, -2.75, -3.0]
• mean_team_run_error_on_losses: improving · series [3.0, 6.222, 5.75, 3.5, 5.333, 6.167, 4.8, None, None, 2.0, None]

Pipeline stage quality

• historical_ml: winner 1/1 (1.0), team-run MAE 1.6753, margin MAE 3.3506
• variance_applied: winner 1/1 (1.0), team-run MAE 1.5, margin MAE 3.0
• enhanced_prediction: winner 1/1 (1.0), team-run MAE 1.5, margin MAE 3.0
• embedded_savant_run_environment: winner 1/1 (1.0), team-run MAE 1.5, margin MAE 3.0
• fundamental_edge_target_review: winner 1/1 (1.0), team-run MAE 1.5, margin MAE 3.0
• capped_target: winner 1/1 (1.0), team-run MAE 1.5, margin MAE 3.0
• simulation_result: winner 1/1 (1.0), team-run MAE 1.5, margin MAE 3.0
• published_final: winner 1/1 (1.0), team-run MAE 1.5, margin MAE 3.0

Probability lineage

• PHI@ATL: raw 0.5453 → calibrated 0.5278 → published 0.5168; score coherence followed by matchup-recency guardrail

Every game, in detail

PHI@ATL — ✅ hit

• Predicted ATL 3-4 (conf +0.03, toss_up) · actual ATL 2-6
• Root cause: toss_up_went_our_way
• Errors: winner correct · Brier 0.233465 · team-run err 3.0 · total bias -1

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: 0 helped vs 0 hurt

player.pitcher.v4 form weights / player.pitcher_prior_games

• Targets: starter_er_mae, starter_outs_mae
• Net: 0 helped vs 1 hurt
• Would hurt / risk:
• PHI@ATL (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: 0 helped vs 0 hurt