2025-2026 NCAAW Model Performance Analysis
Comparing prediction accuracy across 6031 games using multiple rating models.
Model Performance Leaderboard
Models ranked by AUC on 2025-2026 games. Hover over column headers for explanations.
| # | Model | AUC | Acc | Brier | LogLoss | n | AUC 7d | Acc 7d | Brier 7d | n 7d |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Margin Recency | 0.893 | 79.9% | 0.140 | 0.434 | 5395 | 0.800 | 71.4% | 0.180 | 14 |
| 2 | Dynamic Bradley-Terry | 0.887 | 79.2% | 0.143 | 0.445 | 5395 | 0.733 | 78.6% | 0.204 | 14 |
| 3 | Margin Regression | 0.878 | 79.4% | 0.148 | 0.455 | 5048 | 0.711 | 71.4% | 0.202 | 14 |
| 4 | Core Ensemble | 0.877 | 80.7% | 0.189 | 0.562 | 5395 | 0.756 | 71.4% | 0.194 | 14 |
| 5 | Bradley-Terry | 0.876 | 78.8% | 0.149 | 0.462 | 5048 | 0.800 | 57.1% | 0.212 | 14 |
| 6 | Recency Ensemble | 0.876 | 80.4% | 0.187 | 0.560 | 5395 | 0.756 | 71.4% | 0.191 | 14 |
| 7 | Pythagorean Adjusted | 0.871 | 78.0% | 0.151 | 0.466 | 5048 | 0.711 | 71.4% | 0.208 | 14 |
| 8 | Pythagorean Log | 0.870 | 78.0% | 0.151 | 0.468 | 5048 | 0.711 | 71.4% | 0.204 | 14 |
| 9 | Points O/D | 0.870 | 78.1% | 0.153 | 0.469 | 5048 | 0.711 | 78.6% | 0.197 | 14 |
| 10 | Points O/D Recency | 0.867 | 77.7% | 0.158 | 0.486 | 5395 | 0.756 | 78.6% | 0.180 | 14 |
| 11 | Adjusted Context Blend | 0.857 | 76.9% | 0.194 | 0.567 | 5395 | 0.667 | 64.3% | 0.216 | 14 |
| 12 | Avg Margin (Baseline) | 0.843 | 75.9% | 0.166 | 0.502 | 5048 | 0.644 | 50.0% | 0.226 | 14 |
| 13 | Pure ELO | 0.841 | 75.4% | 0.168 | 0.511 | 5048 | 0.689 | 64.3% | 0.215 | 14 |
| 14 | Pythagorean Efficiency | 0.840 | 75.4% | 0.170 | 0.528 | 5048 | - | - | - | 0 |
| 15 | Bradley-Terry Recency | 0.836 | 74.3% | 0.183 | 0.548 | 5395 | 0.689 | 57.1% | 0.224 | 14 |
| 16 | Pythagorean Raw | 0.794 | 73.6% | 0.189 | 0.563 | 5048 | 0.600 | 50.0% | 0.226 | 14 |
| 17 | Home Team (Baseline) | 0.595 | 59.6% | 0.241 | 0.675 | 5048 | 0.633 | 64.3% | 0.231 | 14 |
Methodology
ELO / Bradley-Terry
- ELO: Iterative updates, K=64, HCA=100
- BT: Static logistic regression on all games
- Both model win probability, not margin
- ELO updates after each game; BT fits once
Pythagorean Models
- Raw: Bill James formula (pts scored/allowed)
- Efficiency: Pace-adjusted (pts per possession)
- Adjusted: KenPom-style opponent adjustment
- Log: Log-linear multiplicative scale
Margin Regression
- Team-level ridge regression on point margin
- Linear Bradley-Terry (margin target)
- Alpha=0.05 (CV-tuned)
- Learns home advantage from data (~6 pts)
Baselines
- Home Team: Always predict home wins (60%)
- Avg Margin: Higher average margin wins
- Models should beat these to add value