Nightly self-improvement · slate 2026-09-25
Generated 2026-09-26T05:00:00.010328-04:00 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
2026-09-25
status completed
Winners
11/17
Brier 0.2228
Learned or changed
0
of 5 stages
Blocked
1
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) | blocked | quarantined_invalid_artifact: slate quarantined from new parameter proposals until critical/error artifact findings are corrected |
| 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@SD | 6-5 | 11-4 | 0.495 | hit | Brandon Pfaadt: 18 proj / 23 real outs; Casey Mize: 14 proj / 12 real outs | |
| ATL@MIA | 5-4 | 0-3 | 0.466 | miss | Run environment far lower than projected (over-dispersed) | Ray Kerr: 15 proj / 9 real outs; Eury Pérez: 15 proj / 27 real outs |
| BAL@NYY | 5-6 | 10-2 | 0.505 | miss | Genuine toss-up that flipped | Trevor Rogers: 19 proj / 15 real outs; Brendan Beck: 9 proj / 7 real outs |
| BAL@NYY | 5-6 | 3-6 | 0.512 | hit | ||
| CHC@BOS | 5-4 | 3-4 | 0.495 | miss | Genuine toss-up that flipped | Clay Holmes: 17 proj / 15 real outs; Alec Gamboa: 15 proj / 12 real outs |
| CHC@BOS | 3-4 | 0-2 | 0.557 | hit | ||
| CIN@TOR | 3-5 | 5-6 | 0.636 | hit | Nick Lodolo: 15 proj / 15 real outs; José Soriano: 18 proj / 9 real outs | |
| CLE@KC | 4-3 | 12-9 | 0.440 | hit | Gavin Williams: 17 proj / 10 real outs; Noah Cameron: 17 proj / 15 real outs | |
| COL@CHW | 4-9 | 1-6 | 0.723 | hit | Tomoyuki Sugano: 15 proj / 20 real outs; Sean Burke: 15 proj / 16 real outs | |
| HOU@OAK | 5-4 | 5-6 | 0.461 | miss | Several errors of similar size | Hunter Brown: 16 proj / 16 real outs; Jacob Lopez: 13 proj / 18 real outs |
| LAA@SEA | 4-3 | 7-5 | 0.482 | hit | Reid Detmers: 18 proj / 18 real outs; Bryce Miller: 16 proj / 18 real outs | |
| LAD@SF | 6-2 | 2-0 | 0.434 | hit | Tarik Skubal: 19 proj / 21 real outs; Yunior Marte: 14 proj / 15 real outs | |
| NYM@WAS | 4-5 | 6-7 | 0.513 | hit | Jefry Yan: 12 proj / 5 real outs; Andrew Alvarez: 17 proj / 14 real outs | |
| PIT@DET | 5-4 | 7-8 | 0.495 | miss | Run environment far higher than projected (under-dispersed) | Bubba Chandler: 15 proj / 8 real outs; Jackson Jobe: 16 proj / 16 real outs |
| STL@MIL | 4-5 | 4-8 | 0.542 | hit | Michael McGreevy: 16 proj / 14 real outs; Robert Gasser: 15 proj / 18 real outs | |
| TB@PHI | 4-5 | 2-0 | 0.535 | miss | Starting-pitcher projection was wrong | Freddy Peralta: 17 proj / 16 real outs; Cristopher Sánchez: 19 proj / 21 real outs |
| TEX@MIN | 4-5 | 2-10 | 0.542 | hit | Jacob deGrom: 17 proj / 12 real outs; Joe Ryan: 15 proj / 18 real outs |
Day over day
Rolling evidence window against the prior night. Associations, not causal proof.
| Metric | Prior | Current | Change | Result |
|---|---|---|---|---|
| brier_score | 0.2454 | 0.2407 | -0.0047 | improved |
| accuracy | 0.5630 | 0.5882 | 0.0252 | improved |
| team_run_mae | 2.4832 | 2.5168 | 0.0336 | regressed |
| total_run_mae | 3.3529 | 3.6387 | 0.2858 | regressed |
| calibration_error | 0.0535 | 0.0736 | 0.0201 | regressed |
| starter_er_mae | 1.7323 | 1.7400 | 0.0077 | regressed |
| starter_outs_mae | 3.5984 | 3.6480 | 0.0496 | 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
No Luna post-mortem this night.