For every bucket of predicted probability, what did the model actually hit? A well-calibrated model has actual win rate ≈ predicted prob. A negative gap means the model is overconfident in that bucket (dangerous); positive gap means it's underpredicting (safe). Rows turn red when |gap| > 5pp AND N ≥ 10.
Combined "all sports" is rarely meaningful — sports differ in market efficiency, signal availability, and base rates. Use the chips to drill into a single sport. N<50 (red) means the calibration is brittle; N≥200 (green) is trustworthy.
Filter: window=90d · sport=mlb
| Predicted-prob bucket | N | Mean predicted | Actual win rate | Gap (actual − predicted) | Brier |
|---|---|---|---|---|---|
| <50% | 565 | 45.9% | 40.9% | -5.0pp | 0.245 |
| 50-55% | 928 | 52.1% | 48.7% | -3.4pp | 0.251 |
| 55-60% | 572 | 57.4% | 56.8% | -0.6pp | 0.245 |
| 60-65% | 263 | 62.3% | 59.3% | -2.9pp | 0.241 |
| 65-70% | 135 | 66.8% | 60.0% | -6.8pp | 0.244 |
| 70-75% | 17 | 71.5% | 82.4% | +10.8pp | 0.155 |
| 75-80% | 4 | 75.9% | 75.0% | -0.9pp | 0.195 |
| 80-90% | 1 | 85.7% | 100.0% | +14.3pp | 0.020 |
| 90%+ | 1 | 90.9% | 100.0% | +9.1pp | 0.008 |