Nightly self-improvement · slate 2026-09-28
Generated 2026-09-29T05:00:00.006046-04:00 ET from official finals (mlb-api). Every learning stage is listed with what it did. "Blocked" means it wanted to act and a safety rule stopped it; "nothing to learn" means it had no eligible evidence.
Slate
2026-09-28
status completed
Winners
0/0
Brier –
Learned or changed
0
of 5 stages
Blocked
0
wanted to act and could not
Release
v4.31.0
DRIFT
What each learning stage did
| Stage | Outcome | Detail |
|---|---|---|
| KPI settlement | failed | did not run |
| Bullpen usage model | nothing to learn | No pregame bullpen forecasts to settle. Refit: not attempted. |
| Adaptive parameter proposal (Luna) | nothing to learn | skipped_open_rollback_monitor: |
| Auto-apply | nothing to learn | no_promotable_candidate |
| Code agent | nothing to learn | disabled_by_config: code changes are made on the development host, never in the container. |
Games
| Game | Predicted | Actual | Home win prob | Result | Main driver | Starters |
|---|---|---|---|---|---|---|
| No per-game diagnosis on disk. | ||||||
Day over day
Rolling evidence window against the prior night. Associations, not causal proof.
| Metric | Prior | Current | Change | Result |
|---|---|---|---|---|
| brier_score | 0.2462 | 0.2462 | 0.0000 | flat |
| accuracy | 0.5726 | 0.5726 | 0.0000 | flat |
| team_run_mae | 2.5684 | 2.5684 | 0.0000 | flat |
| total_run_mae | 3.8034 | 3.8034 | 0.0000 | flat |
| calibration_error | 0.0862 | 0.0862 | 0.0000 | flat |
| starter_er_mae | 1.7965 | 1.7965 | 0.0000 | flat |
| starter_outs_mae | 4.0442 | 4.0442 | 0.0000 | flat |
Calibration produced for the next slate
| Metric | Scale used | Measured | Rows | Status |
|---|---|---|---|---|
| No calibration produced. | ||||
Luna KPI candidates and their promotion gate
Promotion needs 400+ prospective rows over 5+ slates after the proposal, with a gain interval above zero. Only verified pregame forecasts count; local replays never do.
| KPI | Status | Target | Proposed | Rows | Slates | Gain interval | Progress |
|---|---|---|---|---|---|---|---|
| luna_pitcher_release_speed_vs_righthanded_starters | shadow | SO | 20260923 | 330/400 | 5/5 | [-0.0007, 0.0007] | |
| luna_pitcher_barrel_rate_vs_rhh_righthanded_starters | shadow | HR | 20260923 | 330/400 | 5/5 | [-0.0004, 0.0049] | |
| luna_batter_contact_rate_vs_breaking_righthanded_starters | shadow | H | 20260923 | 330/400 | 5/5 | [-0.0045, 0.0044] | |
| luna_batter_contact_rate_vs_breaking_left_handed_starters_for_strikeouts | shadow | SO | 20260924 | 122/400 | 4/5 | [-0.0027, 0.0021] | |
| luna_pitcher_contact_xwoba_allowed_vs_righthanded_starters_for_strikeouts | shadow | SO | 20260924 | 122/400 | 4/5 | [-0.0001, 0.0000] | |
| luna_pitcher_whiff_rate_breaking_vs_left_handed_hitters_for_hits | shadow | H | 20260924 | 122/400 | 4/5 | [-0.0056, -0.0011] | |
| luna_pitcher_home_run_rate_vs_righthanded_starters | shadow | HR | 20260925 | 122/400 | 4/5 | [-0.0010, 0.0056] | |
| luna_batter_ground_ball_rate_vs_fastball_vs_lhp_for_home_runs | shadow | HR | 20260926 | 86/400 | 3/5 | [-0.0001, -0.0000] | |
| luna_pitcher_whiff_rate_vs_right_handed_starters_for_batter_strikeouts | shadow | SO | 20260929 | 51/400 | 2/5 | [–, –] | |
| luna_batter_whiff_rate_vs_breaking_right_handed_starters_for_hits | shadow | H | 20260929 | 51/400 | 2/5 | [–, –] | |
| luna_batter_walk_rate_vs_righthanded_pitchers_for_home_runs | retired | HR | 20260924 | 0/400 | 0/5 | [–, –] |
Operator actions the cycle asked for (0)
Changes to code or a model recipe are never made by the nightly cycle. These are the measured findings that need one.
| Priority | Owner | Finding | Action |
|---|---|---|---|
| No operator actions requested. | |||
Luna post-mortem
Luna (gpt-5.6-luna) post-mortem, status received. Full analysis.
## 1. Evidence receipt and gaps ### Source manifest All eight manifest entries were reviewed as untrusted evidence metadata: | Source | Status | Evidence | |---|---|---| | Prediction review | Loaded | 4,003 bytes; intended to contain official outcomes and scoring errors | | Forensic deep audit | Loaded | 15,888 bytes; intended to contain probability, provenance, starter, and decision contracts | | Continuous learning | Loaded | 6,809 bytes; intended to contain stage quality, calibration, lineage, score distribution, and starter comparisons | | Detailed report | Missing | `/app/output/reports/detailed_20260928.html` | | Ultra-detailed player audit | Missing | `/app/output/reports/20260928_ultra_detailed_player_audit.html` | | Structural audit | Missing | `/app/output/reports/audit_report_20260928.html` | | Integrity revision audit | Missing | `/app/output/reports/audit_report_20260928_integrity_revision.html` | ### Game and slate coverage No games were supplied: - `date`: blank - `games_reviewed`: `0` - `hits`: `0` - `misses`: `0` - `continuous_learning_evidence.summary.games`: `0` - `settled`: `0` - `coverage_matrix.expected_games`: `0` - `canonical_artifacts_loaded`: `0` - `structural_audits_completed`: `0` - `forensic_deep_audit_games`: `0` - `continuous_learning_games`: `0` - `games_with_starter_comparisons`: `0` - `all_games_structurally_audited`: `false` Thus, there is no game-level evidence from which to calculate winner accuracy, Brier score, run error, starter