MLB Prediction System

KPI Slate Report

2026-10-01 ยท executive view

slate 2026-10-01games 4kpi applied 0/4model 20261005T125039Z_f5114dc
Slate at a glance
KPIs registered
40
feeding 44 model features
Games on card
4
KPI layer applied to 0
Batters resolved
18
of 662 in the league
Starters resolved
2
of 866 in the league
Mean run movement
-
combined home plus away
What the KPI layer is worth

Measured out of sample, not on this slate

Frozen evaluation over 91 untouched holdout games. Coefficients use development dates and weights use selection dates before the holdout is scored.

TargetWeightBaselineWith KPIGainMovedMean moveVerdict
home_runs0.442.72532.6529+0.0723-/910.000inconclusive
away_runs0.782.61542.5301+0.0853-/910.000inconclusive
home_win0.000.24400.2440+0.0000-/910.000inconclusive
Catalog use

Registered versus consumed

34 of 40 registered KPIs feed the joint game model. Legacy per-KPI applied weights do not represent the active application path.

KPIConsumedStatusInterpretation
batter_barrel_rate_vs_offspeedyesjoint_model_inputskill measurement
batter_ground_ball_rate_vs_slideryesjoint_model_inputskill measurement
batter_hard_hit_rate_vs_breakingyesjoint_model_inputskill measurement
batter_whiff_rate_vs_fastballyesjoint_model_inputskill measurement
batter_whiff_rate_vs_breakingyesjoint_model_inputskill measurement
batter_whiff_rate_vs_offspeedyesjoint_model_inputskill measurement
batter_barrel_rate_vs_fastballyesjoint_model_inputskill measurement
batter_barrel_rate_vs_breakingyesjoint_model_inputskill measurement
batter_hard_hit_rate_vs_fastballyesjoint_model_inputskill measurement
batter_contact_xwoba_vs_fastballyesjoint_model_inputskill measurement
batter_contact_xwoba_vs_breakingyesjoint_model_inputskill measurement
batter_exit_velocity_vs_fastballyesjoint_model_inputskill measurement
batter_ground_ball_rate_vs_fastballyesjoint_model_inputskill measurement
batter_contact_rate_vs_breakingyesjoint_model_inputskill measurement
batter_whiff_rate_vs_rhpyesjoint_model_inputskill measurement
batter_whiff_rate_vs_lhpyesjoint_model_inputskill measurement
batter_barrel_rate_vs_rhpyesjoint_model_inputskill measurement
batter_barrel_rate_vs_lhpyesjoint_model_inputskill measurement
batter_strikeout_rateyesjoint_model_inputskill measurement
batter_walk_rateyesjoint_model_inputskill measurement
batter_home_run_rateyesjoint_model_inputskill measurement
batter_contact_xwobayesjoint_model_inputskill measurement
batter_bat_speedyesjoint_model_inputskill measurement
pitcher_barrel_rate_allowedyesjoint_model_inputskill measurement
pitcher_whiff_rateyesjoint_model_inputskill measurement
pitcher_whiff_rate_fastballyesjoint_model_inputskill measurement
pitcher_whiff_rate_breakingyesjoint_model_inputskill measurement
pitcher_hard_hit_rate_allowedyesjoint_model_inputskill measurement
pitcher_ground_ball_rate_inducedyesjoint_model_inputskill measurement
pitcher_contact_xwoba_allowedyesjoint_model_inputskill measurement
pitcher_strikeout_rateyesjoint_model_inputskill measurement
pitcher_walk_rateyesjoint_model_inputskill measurement
pitcher_home_run_rate_allowedyesjoint_model_inputskill measurement
pitcher_release_speedyesjoint_model_inputpitch-mix velocity, not overall pitcher quality
pitcher_whiff_rate_vs_rhhnoresearch_onlyskill measurement
pitcher_whiff_rate_vs_lhhnoresearch_onlyskill measurement
pitcher_barrel_rate_vs_rhhnoresearch_onlyskill measurement
pitcher_barrel_rate_vs_lhhnoresearch_onlyskill measurement
matchup_barrel_rate_archetypenoresearch_onlyskill measurement
matchup_whiff_rate_archetypenoresearch_onlyskill measurement
Game by game

What the layer did to this card

GameBaselineKPI onlyMove AMove HPrior publishedThis runActualError gainState
BOS @ NYY0.00-0.000.00-0.00--4-5 NYY4-5 NYY---
CHC @ SD0.00-0.000.00-0.00--4-5 SD4-5 SD---
CHW @ HOU0.00-0.000.00-0.00--6-5 CHW6-5 CHW---
PHI @ ATL0.00-0.000.00-0.00--3-4 ATL3-4 ATL2-6+0-
Settled scoring across 1 games, against the prior card. Total run error 3.000 before, 3.000 after, a gain of +0.000 runs per game. Winners 1/1 before, 1/1 after. This regenerated slate is diagnostic. Intermediate baseline-to-final movement does not isolate KPI impact without paired final outputs.
Luna matchup review

State on this slate

absent: 4. Per-game Luna matchup review status as recorded in each artifact.
Luna KPI review

KPIs Luna created

Latest review 20260929: 0 KPI-level faults, 2 player-level faults, Luna received; report 20260929_kpi_luna_review.html. A Luna KPI moves hitter lines only after 400+ rows over 5+ slates after its proposal with a gain interval above zero.
KPIstatustargetproposedprospective evidencegain interval
luna_pitcher_release_speed_vs_righthanded_startersshadowSO20260923122 rows / 4 slates[-0.00163, 0.00176]
luna_pitcher_barrel_rate_vs_rhh_righthanded_startersshadowHR20260923122 rows / 4 slates[0.00032, 0.00237]
luna_batter_contact_rate_vs_breaking_righthanded_startersshadowH20260923122 rows / 4 slates[-0.00707, 0.0092]
luna_batter_walk_rate_vs_righthanded_pitchers_for_home_runsretiredHR202609240 rows / 0 slatesNone
luna_batter_contact_rate_vs_breaking_left_handed_starters_for_strikeoutsshadowSO202609240 rows / 0 slatesNone
luna_pitcher_contact_xwoba_allowed_vs_righthanded_starters_for_strikeoutsshadowSO20260924122 rows / 4 slates[-0.00051, 3e-05]
luna_pitcher_whiff_rate_breaking_vs_left_handed_hitters_for_hitsshadowH20260924122 rows / 4 slates[-6e-05, 0.00153]
luna_pitcher_home_run_rate_vs_righthanded_startersshadowHR20260925122 rows / 4 slates[0.0001, 0.00189]
luna_batter_ground_ball_rate_vs_fastball_vs_lhp_for_home_runsshadowHR202609260 rows / 0 slatesNone
luna_pitcher_whiff_rate_vs_right_handed_starters_for_batter_strikeoutsshadowSO2026092951 rows / 2 slatesNone
luna_batter_whiff_rate_vs_breaking_right_handed_starters_for_hitsshadowH2026092951 rows / 2 slatesNone
generated 2026-10-06T13:59:18.981926+00:00 · kpi_slate_report_v3
model 20261005T125039Z_f5114dc0 · bundle 20261004T081701Z_6064881e
sources: output/games/revisions · output/learning/kpi/{ledger,models,values} · statsapi.mlb.com official schedule