Nightly self-improvement · slate 2026-09-02
Generated 2026-09-03T02:00:00.008251-04:00 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
2026-09-02
status completed
Winners
5/15
Brier 0.2708
Learned or changed
0
of 5 stages
Blocked
0
wanted to act and could not
Release
v4.11.4
consistent
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 | unavailable: |
| 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 |
|---|---|---|---|---|---|---|
| ATL@WAS | 5-4 | 9-0 | 0.470 | hit | Grant Holmes: 15 proj / 9 real outs; Brad Lord: 15 proj / 10 real outs | |
| BAL@COL | 7-5 | 5-6 | 0.443 | miss | Backed the wrong side with real conviction | Trevor Rogers: 17 proj / 20 real outs; Tomoyuki Sugano: 14 proj / 18 real outs |
| CHW@HOU | 5-6 | 0-2 | 0.530 | hit | Davis Martin: 15 proj / 17 real outs; Hayden Wesneski: 16 proj / 21 real outs | |
| DET@MIN | 4-5 | 11-6 | 0.559 | miss | Run environment far higher than projected (under-dispersed) | Drew Anderson: 13 proj / 17 real outs; Dean Kremer: 15 proj / 12 real outs |
| MIA@KC | 5-4 | 9-6 | 0.444 | hit | Eury Pérez: 16 proj / 14 real outs; Noah Cameron: 19 proj / 15 real outs | |
| MIL@CHC | 6-4 | 9-5 | 0.460 | hit | Jacob Misiorowski: 18 proj / 12 real outs; David Peterson: 12 proj / 11 real outs | |
| NYM@TB | 3-4 | 10-4 | 0.549 | miss | Run environment far higher than projected (under-dispersed) | Justin Hagenman: 9 proj / 9 real outs; Griffin Jax: 15 proj / 8 real outs |
| NYY@LAA | 4-2 | 6-3 | 0.470 | hit | Cam Schlittler: 19 proj / 24 real outs; Reid Detmers: 18 proj / 21 real outs | |
| OAK@TEX | 4-5 | 9-2 | 0.560 | miss | Backed the wrong side with real conviction | Jacob Lopez: 15 proj / 18 real outs; Cody Bradford: 15 proj / 12 real outs |
| PHI@ARI | 5-4 | 0-1 | 0.470 | miss | Starting-pitcher projection was wrong | Andrew Painter: 15 proj / 21 real outs; Taylor Clarke: 18 proj / 5 real outs |
| SD@CIN | 5-4 | 3-7 | 0.470 | miss | Small edge on the losing side | Casey Mize: 15 proj / 12 real outs; Brandon Williamson: 13 proj / 16 real outs |
| SEA@BOS | 4-5 | 8-3 | 0.520 | miss | Genuine toss-up that flipped | Bryce Miller: 16 proj / 14 real outs; Patrick Sandoval: 15 proj / 12 real outs |
| SF@PIT | 5-6 | 5-4 | 0.546 | miss | Several errors of similar size | Landen Roupp: 16 proj / 15 real outs; Bubba Chandler: 15 proj / 15 real outs |
| STL@LAD | 4-5 | 8-6 | 0.530 | miss | Run environment far higher than projected (under-dispersed) | Brycen Mautz: 9 proj / 3 real outs; Yoshinobu Yamamoto: 20 proj / 21 real outs |
| TOR@CLE | 2-3 | 11-0 | 0.584 | miss | Run environment far higher than projected (under-dispersed) | Dylan Cease: 18 proj / 18 real outs; Joey Cantillo: 15 proj / 16 real outs |
Day over day
Rolling evidence window against the prior night. Associations, not causal proof.
| Metric | Prior | Current | Change | Result |
|---|---|---|---|---|
| brier_score | 0.2519 | 0.2535 | 0.0016 | regressed |
| accuracy | 0.5667 | 0.5250 | -0.0417 | regressed |
| team_run_mae | 2.7458 | 2.7542 | 0.0084 | regressed |
| total_run_mae | 3.8083 | 3.8750 | 0.0667 | regressed |
| calibration_error | 0.1128 | 0.1009 | -0.0119 | improved |
| starter_er_mae | 1.9483 | 1.9463 | -0.0020 | improved |
| starter_outs_mae | 2.9267 | 3.0537 | 0.1270 | regressed |
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 seven manifest entries were reviewed as evidence records: | Source | Status | Size | Purpose | |---|---:|---:|---| | `prediction review` — `/app/output/reports/prediction_review_20260902.json` | Loaded | 372,146 bytes | Official outcomes and scoring errors | | `forensic deep audit` — `/app/output/reports/PREDICTION_SYSTEM_DEEP_AUDIT_20260902.json` | Loaded | 240,269 bytes | Probability, provenance, starter, and decision contracts | | `continuous learning` — `/app/output/learning/continuous_learning_daily_20260902.json` | Loaded | 1,925,532 bytes | Pipeline-stage quality, calibration, probability lineage, score distribution, starter comparisons | | `detailed report` — `/app/output/reports/detailed_20260902.html` | Loaded | 352,626 bytes | Full game and simulation rendering | | `ultra detailed player audit` — `/app/output/reports/20260902_ultra_detailed_player_audit.html` | Loaded | 30,625,934 bytes | Player-level prediction trace | | `structural audit` — `/app/output/reports/audit_report_20260902.html` | Loaded | 1,878,790 bytes | Invariant findings | | `integrity revision audit` — `/app/output/reports/audit_report_20260902_integrity_revision.html` | **Missing** | Not supplied | Post-Luna remediation proof for immutable revision artifacts | The missing integrity-revision audit is a genuine evidence gap. It prevents independent confirmation that any post-generation remediation or immutable revision process was correctly