Nightly self-improvement · slate 2026-08-30
Generated 2026-08-31T17:54:01.226752-04:00 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
2026-08-30
status completed
Winners
7/14
Brier 0.2592
Learned or changed
0
of 5 stages
Blocked
0
wanted to act and could not
Release
v4.10.0
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 | skipped_open_rollback_monitor: |
| Auto-apply | nothing to learn | no_promotable_candidate |
| Code agent | nothing to learn | disabled_by_flag: code changes are made on the development host, never in the container. |
Games
| Game | Predicted | Actual | Home win prob | Result | Main driver | Starters |
|---|---|---|---|---|---|---|
| BAL@OAK | 8-6 | 8-5 | 0.449 | hit | Chris Bassitt: 15 proj / 17 real outs; Jeffrey Springs: 12 proj / 14 real outs | |
| BOS@NYY | 5-4 | 1-16 | 0.470 | miss | Starting-pitcher projection was wrong | Ranger Suarez: 15 proj / 16 real outs; Will Warren: 15 proj / 21 real outs |
| CHW@MIN | 7-4 | 1-5 | 0.313 | miss | Run environment far lower than projected (over-dispersed) | Jordan Hicks: 13 proj / 2 real outs; Zebby Matthews: 17 proj / 25 real outs |
| CIN@CHC | 4-7 | 7-5 | 0.560 | miss | Backed the wrong side with real conviction | Chase Burns: 17 proj / 17 real outs; Shota Imanaga: 16 proj / 15 real outs |
| COL@ATL | 3-6 | 2-3 | 0.580 | hit | Mason Adams: 11 proj / 15 real outs; Tyler Mahle: 17 proj / 20 real outs | |
| HOU@NYM | 4-5 | 6-3 | 0.542 | miss | Small edge on the losing side | Ethan Pecko: 11 proj / 15 real outs; Zac Thornton: 17 proj / 15 real outs |
| KC@CLE | 5-6 | 1-11 | 0.530 | hit | Seth Lugo: 16 proj / 11 real outs; Parker Messick: 18 proj / 18 real outs | |
| LAD@DET | 4-3 | 6-1 | 0.451 | hit | Tyler Glasnow: 17 proj / 21 real outs; Framber Valdez: 15 proj / 18 real outs | |
| MIA@WAS | 3-4 | 6-2 | 0.575 | miss | Backed the wrong side with real conviction | Janson Junk: 15 proj / 14 real outs; Andrew Alvarez: 15 proj / 18 real outs |
| PHI@LAA | 5-2 | 5-2 | 0.335 | hit | Zack Wheeler: 16 proj / 18 real outs; Yusei Kikuchi: 13 proj / 21 real outs | |
| PIT@STL | 5-4 | 5-4 | 0.440 | hit | Braxton Ashcraft: 19 proj / 15 real outs; Matthew Liberatore: 15 proj / 17 real outs | |
| SD@TB | 3-4 | 4-5 | 0.534 | hit | Robbie Ray: 17 proj / 18 real outs; Drew Rasmussen: 18 proj / 18 real outs | |
| SEA@TOR | 5-4 | 0-7 | 0.470 | miss | Starting-pitcher projection was wrong | Logan Gilbert: 18 proj / 15 real outs; Max Scherzer: 14 proj / 21 real outs |
| TEX@MIL | 4-5 | 7-4 | 0.577 | miss | Backed the wrong side with real conviction | Kumar Rocker: 15 proj / 9 real outs; Dustin May: 14 proj / 13 real outs |
Day over day
Rolling evidence window against the prior night. Associations, not causal proof.
| Metric | Prior | Current | Change | Result |
|---|---|---|---|---|
| brier_score | 0.2338 | 0.2385 | 0.0047 | regressed |
| accuracy | 0.6348 | 0.6083 | -0.0265 | regressed |
| team_run_mae | 2.4261 | 2.5000 | 0.0739 | regressed |
| total_run_mae | 3.7391 | 3.6000 | -0.1391 | improved |
| calibration_error | 0.1236 | 0.1106 | -0.0130 | improved |
| starter_er_mae | 1.7917 | 1.7563 | -0.0354 | improved |
| starter_outs_mae | 3.1042 | 3.1639 | 0.0597 | 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 ## Sources received Six of seven manifest entries were loaded: 1. **Prediction review** — `/app/output/reports/prediction_review_20260830.json`, 349,718 bytes; official outcomes and scoring errors. 2. **Forensic deep audit** — `/app/output/reports/PREDICTION_SYSTEM_DEEP_AUDIT_20260830.json`, 226,049 bytes; probability, provenance, starter, and decision contracts. 3. **Continuous learning** — `/app/output/learning/continuous_learning_daily_20260830.json`, 2,264,904 bytes; pipeline-stage quality, calibration, lineage, score distribution, and starter comparisons. 4. **Detailed report** — `/app/output/reports/detailed_20260830.html`, 331,974 bytes. 5. **Ultra-detailed player audit** — `/app/output/reports/20260830_ultra_detailed_player_audit.html`, 2,061,684 bytes. 6. **Structural audit** — `/app/output/reports/audit_report_20260830.html`, 138,926 bytes. The following manifest item is explicitly **missing**: 7. **Integrity revision audit** — `/app/output/reports/audit_report_20260830_integrity_revision.html`; intended to provide post-remediation proof for immutable revision artifacts. ## Slate coverage All 14 expected games were reviewed: - BAL@OAK - BOS@NYY - CHW@MIN - CIN@CHC - COL@ATL - HOU@NYM - KC@CLE - LAD@DET - MIA@WAS - PHI@LAA - PIT@STL - SD@TB - SEA@TOR - TEX@MIL Coverage is structurally complete: 14/14 canonical artifacts loaded, 14/14 continuous-learning games, 14/14 forensic deep audits, 14/14 starter comparisons, and 14/14 str