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=americanfootball_nfl
| Predicted-prob bucket | N | Mean predicted | Actual win rate | Gap (actual − predicted) | Brier |
|---|---|---|---|---|---|
| <50% | 48 | 48.6% | 45.8% | -2.8pp | 0.249 |
| 50-55% | 145 | 51.5% | 46.9% | -4.7pp | 0.251 |
| 55-60% | 45 | 57.7% | 62.2% | +4.6pp | 0.236 |
| 60-65% | 31 | 62.2% | 64.5% | +2.3pp | 0.229 |
| 65-70% | 14 | 66.9% | 71.4% | +4.5pp | 0.206 |
| 70-75% | 10 | 73.1% | 80.0% | +6.9pp | 0.162 |
| 75-80% | 14 | 77.3% | 64.3% | -13.1pp | 0.254 |
| 80-90% | 6 | 85.0% | 66.7% | -18.4pp | 0.252 |