Nightly self-improvement · slate 2026-09-09
Generated 2026-09-10T05:00:00.007116-04:00 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
2026-09-09
status completed
Winners
10/15
Brier 0.2236
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@KC | 5-7 | 2-5 | 0.530 | hit | Zac Gallen: 15 proj / 10 real outs; Daniel Lynch IV: 15 proj / 15 real outs | |
| CHC@MIL | 5-6 | 6-8 | 0.524 | hit | Kevin Gausman: 17 proj / 12 real outs; Logan Henderson: 19 proj / 15 real outs | |
| CIN@LAD | 5-8 | 1-14 | 0.580 | hit | Rhett Lowder: 14 proj / 10 real outs; Yoshinobu Yamamoto: 19 proj / 21 real outs | |
| CLE@BAL | 4-3 | 5-9 | 0.442 | miss | Starting-pitcher projection was wrong | Foster Griffin: 17 proj / 7 real outs; Shane Baz: 17 proj / 17 real outs |
| COL@NYY | 4-9 | 1-6 | 0.784 | hit | Tomoyuki Sugano: 14 proj / 18 real outs; Will Warren: 15 proj / 18 real outs | |
| HOU@PHI | 3-4 | 7-11 | 0.520 | hit | Hunter Brown: 17 proj / 15 real outs; Cristopher Sánchez: 19 proj / 18 real outs | |
| LAA@BOS | 4-5 | 6-4 | 0.529 | miss | Several errors of similar size | Ryan Johnson: 16 proj / 18 real outs; Jake Bennett: 17 proj / 16 real outs |
| MIN@DET | 4-5 | 2-7 | 0.524 | hit | Zebby Matthews: 18 proj / 7 real outs; Keider Montero: 17 proj / 12 real outs | |
| NYM@MIA | 7-4 | 15-14 | 0.458 | hit | Robert Stock: 13 proj / 12 real outs; Janson Junk: 15 proj / 9 real outs | |
| PIT@CHW | 5-4 | 4-2 | 0.402 | hit | Lake Bachar: 15 proj / 5 real outs; Davis Martin: 15 proj / 15 real outs | |
| STL@SF | 5-4 | 6-7 | 0.479 | miss | Small edge on the losing side | Andre Pallante: 17 proj / 15 real outs; Blade Tidwell: 15 proj / 12 real outs |
| TB@ATL | 4-3 | 7-2 | 0.423 | hit | Griffin Jax: 15 proj / 15 real outs; Reynaldo López: 15 proj / 14 real outs | |
| TEX@SEA | 6-4 | 2-3 | 0.451 | miss | Run environment far lower than projected (over-dispersed) | Cody Bradford: 15 proj / 12 real outs; Kade Anderson: 15 proj / 18 real outs |
| TOR@OAK | 6-5 | 0-2 | 0.464 | miss | Run environment far lower than projected (over-dispersed) | Braydon Fisher: 15 proj / 3 real outs; Brady Basso: 13 proj / 12 real outs |
| WAS@SD | 4-5 | 2-9 | 0.524 | hit | Jackson Kent: 14 proj / 9 real outs; Walker Buehler: 16 proj / 7 real outs |
Day over day
Rolling evidence window against the prior night. Associations, not causal proof.
| Metric | Prior | Current | Change | Result |
|---|---|---|---|---|
| brier_score | 0.2542 | 0.2444 | -0.0097 | improved |
| accuracy | 0.5500 | 0.5833 | 0.0333 | improved |
| team_run_mae | 2.5792 | 2.5500 | -0.0292 | improved |
| total_run_mae | 3.8917 | 4.0000 | 0.1083 | regressed |
| calibration_error | 0.0641 | 0.0690 | 0.0049 | regressed |
| starter_er_mae | 2.0207 | 1.9879 | -0.0328 | improved |
| starter_outs_mae | 3.0083 | 3.1498 | 0.1415 | 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.