Bootstrap by game: no MLB market bin's gap interval excludes zero (10 of 10 within noise)
mlb/model: 3/10 eligible bins have a calibration-gap 95% CI that excludes 0; 7 within noise. mlb/market: 0/10 eligible bins have a calibration-gap 95% CI that excludes 0; 10 within noise. soccer_intl/model: 8/9 eligible bins have a calib...
Data integrity
These MLB/soccer numbers are revision 2, computed on the segment-clean corpus (2026-09-16). Revision 1 values are withdrawn and kept in the regeneration receipt. Segmentation excluded 51,635 of the 78,986 MLB ticks (65.4 percent); the specifically identified label mismatches were a smaller 27,076 (34.3 percent).
- calibration_stability
How to read this figure
Each bin shows a gap (predicted minus actual) with a 95% confidence interval; if the interval excludes 0, the miss is real rather than noise.
Published coverage
The committed artifact is partial; only its published coverage is shown here.
| Population | Status | Games | Usable buckets | Published reason |
|---|---|---|---|---|
| MLB | unknown | - | - | - |
| SOCCER_INTL | unknown | - | - | - |
What it means
A cluster bootstrap (resampling game_id, 1000 draws) asks which calibration gaps survive sampling noise. For the market, none do -- no MLB market bin's gap interval excludes zero at this sample. The model has three real misses: it over-forecasts long shots in the 0.1-0.2 and 0.2-0.3 bands and under-forecasts the 0.8-0.9 band. The top bin is no longer a significant miss: 0.9516-forecasts win 0.9798, a gap of +0.0282 whose interval includes zero.
Caveats and confounds
Soccer is much noisier -- 8 of 9 eligible model bins and 7 of 9 market bins deviate significantly on just 27 games, so the small corpus limits any firm conclusion there.