Model Insights
Technical validation metrics and predictive stability.
Brier Score
0.2431
Lower is better (0 = perfect)
Log Loss
0.6792
Lower is better
Total Sample
4136
NHL games tested
Rolling 30-Game Win Rate
Win Rate by Day of Week
Margin of Victory (ML Wins)
Metric Interpretation
Brier Score
0.243
Coin Flip Brier
0.2500
Calibration
Fair
Model beats coin flip by
2.8%
Injury-Adjusted Predictions
Injury Model
Active
Player Stats Used
xGF% / CF%
Time Decay
Enabled
Our model weights each injury by player value using advanced analytics and adjusts impact over time as teams adapt to absences.
Injury Impact Analysis
How skater injuries affect pick accuracy. Based on 426 live picks with tracked injury data.
Win Rate by Opponent Injury Count
Win Rate by Injury Advantage
Win Rate by Picked Team's Injury Count
Injury Breakdown Detail
| Category | Scenario | Wins | Losses | Win Rate | Games |
|---|---|---|---|---|---|
| Opponent Injuries | 0 injuries | 102 | 73 | 58.3% | 175 |
| 1-2 injuries | 124 | 81 | 60.5% | 205 | |
| 3+ injuries | 29 | 17 | 63.0% | 46 | |
| Picked Team Injuries | 0 injuries | 152 | 90 | 62.8% | 242 |
| 1-2 injuries | 86 | 66 | 56.6% | 152 | |
| 3+ injuries | 17 | 15 | 53.1% | 32 | |
| Injury Edge | WP has injury edge | 119 | 64 | 65.0% | 183 |
| Even injuries | 75 | 62 | 54.7% | 137 | |
| Opponent healthier | 61 | 45 | 57.5% | 106 |