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=tennis_wta
| Predicted-prob bucket | N | Mean predicted | Actual win rate | Gap (actual − predicted) | Brier |
|---|---|---|---|---|---|
| 50-55% | 113 | 52.5% | 57.5% | +5.1pp | 0.244 |
| 55-60% | 142 | 56.8% | 56.3% | -0.5pp | 0.245 |
| 60-65% | 53 | 62.0% | 50.9% | -11.1pp | 0.262 |
| 65-70% | 65 | 67.5% | 70.8% | +3.3pp | 0.208 |
| 70-75% | 29 | 73.0% | 75.9% | +2.9pp | 0.181 |
| 75-80% | 168 | 75.8% | 73.2% | -2.6pp | 0.198 |
| 80-90% | 71 | 88.0% | 91.5% | +3.5pp | 0.079 |
| 90%+ | 27 | 93.3% | 96.3% | +3.0pp | 0.035 |