Nightly self-improvement · slate 2026-08-29
Generated 2026-08-30T02:00:00.004351-04:00 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
2026-08-29
status completed
Winners
16/17
Brier 0.1975
Learned or changed
0
of 5 stages
Blocked
1
wanted to act and could not
Release
v4.9.1
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@SF | 7-6 | 7-1 | 0.473 | hit | Merrill Kelly: 15 proj / 13 real outs; Matt Wilkinson: 9 proj / 14 real outs | |
| ARI@SF | 5-6 | 2-7 | 0.519 | hit | ||
| BAL@OAK | 7-6 | 5-3 | 0.445 | hit | Shane Baz: 18 proj / 17 real outs; Jack Perkins: 12 proj / 15 real outs | |
| BOS@NYY | 5-4 | 6-0 | 0.470 | hit | Jake Bennett: 17 proj / 4 real outs; Carlos Rodón: 17 proj / 11 real outs | |
| BOS@NYY | 3-4 | 2-9 | 0.534 | hit | ||
| CHW@MIN | 7-5 | 3-2 | 0.429 | hit | Erick Fedde: 15 proj / 15 real outs; Bailey Ober: 16 proj / 15 real outs | |
| CIN@CHC | 5-7 | 5-17 | 0.560 | hit | Andrew Abbott: 15 proj / 9 real outs; Kevin Gausman: 17 proj / 15 real outs | |
| COL@ATL | 2-7 | 1-2 | 0.734 | hit | Zach Agnos: 13 proj / 8 real outs; Martín Pérez: 17 proj / 17 real outs | |
| HOU@NYM | 4-5 | 2-6 | 0.530 | hit | Bennett Sousa: 12 proj / 2 real outs; Nolan McLean: 17 proj / 21 real outs | |
| KC@CLE | 5-4 | 8-3 | 0.448 | hit | Daniel Lynch IV: 15 proj / 10 real outs; Foster Griffin: 17 proj / 15 real outs | |
| LAD@DET | 4-2 | 1-2 | 0.335 | miss | Backed the wrong side with real conviction | Blake Snell: 16 proj / 15 real outs; Keider Montero: 18 proj / 16 real outs |
| MIA@WAS | 2-3 | 4-5 | 0.560 | hit | Sandy Alcantara: 19 proj / 18 real outs; Cade Cavalli: 18 proj / 12 real outs | |
| PHI@LAA | 6-3 | 4-2 | 0.335 | hit | Cristopher Sánchez: 18 proj / 21 real outs; Ryan Johnson: 16 proj / 15 real outs | |
| PIT@STL | 4-3 | 6-2 | 0.470 | hit | Lake Bachar: 15 proj / 7 real outs; Kyle Leahy: 17 proj / 9 real outs | |
| SD@TB | 4-5 | 6-7 | 0.621 | hit | Walker Buehler: 15 proj / 15 real outs; Nick Martinez: 18 proj / 18 real outs | |
| SEA@TOR | 4-5 | 3-4 | 0.540 | hit | Kade Anderson: 15 proj / 13 real outs; José Soriano: 16 proj / 21 real outs | |
| TEX@MIL | 3-7 | 3-5 | 0.734 | hit | Cal Quantrill: 17 proj / 18 real outs; Shane Drohan: 16 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 | 0.2418 | 0.2338 | -0.0080 | improved |
| accuracy | 0.5833 | 0.6348 | 0.0515 | improved |
| team_run_mae | 2.4306 | 2.4261 | -0.0045 | improved |
| total_run_mae | 3.7130 | 3.7391 | 0.0261 | regressed |
| calibration_error | 0.1004 | 0.1236 | 0.0232 | regressed |
| starter_er_mae | 1.7854 | 1.7917 | 0.0063 | regressed |
| starter_outs_mae | 3.1073 | 3.1042 | -0.0031 | improved |
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.