Model Predictions
| Model | Rider | Canisius | Spread | Result |
|---|---|---|---|---|
| Elo | 42.1% | 57.9% | - | โ |
| Bradley-Terry | 66.5% | 33.5% | - | โ |
| Bradley-Terry Recency | 64.9% | 35.1% | - | โ |
| Margin | 44.3% | 55.7% | - | โ |
| Margin Recency | 43.8% | 56.2% | - | โ |
| Pythagorean | 60.8% | 39.2% | - | โ |
| Adjusted Efficiency | 59.0% | 41.0% | - | โ |
| Log Adjusted | 58.8% | 41.2% | - | โ |
| Points Off/Def | 45.7% | 54.3% | - | โ |
| Points Off/Def Recency | 45.0% | 55.0% | - | โ |
| Core Ensemble | 54.7% | 45.3% | - | โ |
| Recency Ensemble | 54.6% | 45.4% | - | โ |
| Dynamic Bradley-Terry | 61.9% | 38.1% | - | โ |
| Adjusted Context Blend | 48.8% | 51.2% | - | โ |
Actual: Canisius -3 (75-78). Win probabilities: out-of-sample forecasts from the weekly fit before the game, the same forecasts as the model pages.