Nightly self-improvement · slate 2026-09-07
Generated 2026-09-08T05:00:00.005655-04:00 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
2026-09-07
status completed
Winners
6/11
Brier 0.2780
Learned or changed
0
of 5 stages
Blocked
0
wanted to act and could not
Release
v4.26.0
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_config: code changes are made on the development host, never in the container. |
Games
| Game | Predicted | Actual | Home win prob | Result | Main driver | Starters |
|---|---|---|---|---|---|---|
| ARI@KC | 3-5 | 5-4 | 0.580 | miss | Backed the wrong side with real conviction | Jose Cabrera: 14 proj / 6 real outs; Noah Cameron: 18 proj / 18 real outs |
| ATL@PHI | 3-4 | 0-1 | 0.530 | hit | Grant Holmes: 14 proj / 18 real outs; Jesús Luzardo: 19 proj / 27 real outs | |
| CHC@MIL | 5-7 | 3-4 | 0.507 | hit | Matthew Boyd: 17 proj / 21 real outs; Robert Gasser: 15 proj / 14 real outs | |
| CIN@LAD | 5-6 | 3-6 | 0.533 | hit | Chase Burns: 17 proj / 9 real outs; Emmet Sheehan: 15 proj / 17 real outs | |
| CLE@BAL | 3-4 | 4-6 | 0.508 | hit | Joey Cantillo: 14 proj / 7 real outs; Trevor Rogers: 17 proj / 22 real outs | |
| LAA@BOS | 3-5 | 2-5 | 0.623 | hit | Grayson Rodriguez: 15 proj / 12 real outs; Brayan Bello: 13 proj / 15 real outs | |
| MIN@DET | 5-4 | 4-5 | 0.470 | miss | Several errors of similar size | Joe Ryan: 17 proj / 12 real outs; Troy Melton: 18 proj / 17 real outs |
| NYM@MIA | 4-5 | 9-4 | 0.521 | miss | Small edge on the losing side | Jonah Tong: 10 proj / 15 real outs; Eury Pérez: 16 proj / 12 real outs |
| STL@SF | 7-5 | 4-5 | 0.373 | miss | Backed the wrong side with real conviction | Michael McGreevy: 16 proj / 20 real outs; Logan Webb: 16 proj / 18 real outs |
| TOR@OAK | 6-2 | 5-6 | 0.300 | miss | Backed the wrong side with real conviction | Dylan Cease: 18 proj / 15 real outs; Jacob Lopez: 15 proj / 18 real outs |
| WAS@SD | 5-6 | 2-3 | 0.530 | hit | Jake Irvin: 14 proj / 17 real outs; Nick Pivetta: 15 proj / 15 real outs |
Day over day
Rolling evidence window against the prior night. Associations, not causal proof.
| Metric | Prior | Current | Change | Result |
|---|---|---|---|---|
| brier_score | 0.2509 | 0.2572 | 0.0063 | regressed |
| accuracy | 0.5583 | 0.5333 | -0.0250 | regressed |
| team_run_mae | 2.6083 | 2.5958 | -0.0125 | improved |
| total_run_mae | 3.6500 | 3.6750 | 0.0250 | regressed |
| calibration_error | 0.0640 | 0.0900 | 0.0260 | regressed |
| starter_er_mae | 1.9795 | 1.9958 | 0.0163 | regressed |
| starter_outs_mae | 3.0123 | 3.0586 | 0.0463 | 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 All seven `source_manifest` entries were checked against the supplied evidence: | Source | Status | Purpose | |---|---:|---| | `prediction_review_20260907.json` | Loaded | Official outcomes and scoring errors | | `PREDICTION_SYSTEM_DEEP_AUDIT_20260907.json` | Loaded | Probability, provenance, starter, and decision contracts | | `continuous_learning_daily_20260907.json` | Loaded | Pipeline-stage quality, calibration, probability lineage, score distribution, and starter comparisons | | `detailed_20260907.html` | Loaded | Full game and simulation rendering | | `20260907_ultra_detailed_player_audit.html` | Loaded | Player-level prediction trace | | `audit_report_20260907.html` | Loaded | Invariant findings | | `audit_report_20260907_integrity_revision.html` | **Missing** | Post-Luna remediation proof for immutable revision artifacts | The missing integrity-revision audit is the only explicitly missing source. It prevents independent confirmation of immutable revision-artifact remediation, although the supplied structural and per-game audits report no critical or error findings. ## Slate coverage All 11 expected games were reviewed: `ARI@KC`, `ATL@PHI`, `CHC@MIL`, `CIN@LAD`, `CLE@BAL`, `LAA@BOS`, `MIN@DET`, `NYM@MIA`, `STL@SF`, `TOR@OAK`, and `WAS@SD`. Coverage is internally complete for the principal slate artifacts: - 11 expected games - 11 canonical artifacts loaded - 11 continuous-learning games - 11 forensic deep-aud