2018-2019 NCAAMD2 Model Performance
Season
Every ranked model predicted the same 3571 games (2018-11-25 to 2019-03-30, 18 weekly windows) out of sample: each week's games were predicted by models fit only on the season's earlier games. 54 of 3625 games are left out because a model had no rating for a team yet.
Model Performance Leaderboard
Ranked by full-season AUC. The final week is the season's last 7 played days (7 games, 2019-03-24 to 2019-03-30), predicted from every earlier game: a small sample that shows the latest form.
| # | Model | AUC | Acc | Brier | LogLoss | AUC final wk | Acc final wk |
|---|---|---|---|---|---|---|---|
| 1 | Points Off/Def Recency Points Off/Def Recency Off/def points regression with exponential recency weights. More โ | 0.801 | 72.4% | 0.185 | 0.550 | 0.750 | 57.1% |
| 2 | Points Off/Def Points Off/Def Raw points regression with separate offensive and defensive team parameters. More โ | 0.801 | 72.4% | 0.186 | 0.551 | 0.750 | 57.1% |
| 3 | Margin Recency Margin Recency Margin regression with exponential recency weights on newer games. More โ | 0.801 | 72.3% | 0.186 | 0.551 | 0.750 | 71.4% |
| 4 | Margin Margin Linear team-strength model fit on point differential instead of binary wins. More โ | 0.800 | 72.1% | 0.186 | 0.552 | 0.750 | 71.4% |
| 5 | Core Ensemble Core Ensemble Equal-logit blend of Elo, recency BT, recency margin, log-adjusted pyth, and points off/def. More โ | 0.793 | 71.9% | 0.187 | 0.553 | 0.750 | 71.4% |
| 6 | Recency Ensemble Recency Ensemble Equal-logit blend of Elo, recency BT, recency margin, log-adjusted pyth, and recency points off/def. More โ | 0.793 | 71.9% | 0.187 | 0.553 | 0.750 | 71.4% |
| 7 | Adjusted Context Blend Adjusted Context Blend Chronologically trained component blend with rest and signed venue context. More โ | 0.785 | 71.4% | 0.196 | 0.586 | 0.750 | 57.1% |
| 8 | Adjusted Efficiency Adjusted Efficiency Opponent-adjusted efficiency model with separate offensive and defensive components. More โ | 0.783 | 70.7% | 0.190 | 0.560 | 0.917 | 71.4% |
| 9 | Log Adjusted Log Adjusted Log-scale adjusted efficiency model that downweights blowout leverage. More โ | 0.783 | 70.8% | 0.190 | 0.560 | 0.917 | 71.4% |
| 10 | Dynamic Bradley-Terry Dynamic Bradley-Terry Time-evolving paired-comparison model with latent team strength drift. More โ | 0.767 | 70.2% | 0.197 | 0.577 | 0.750 | 57.1% |
| 11 | Bradley-Terry Recency Bradley-Terry Recency Static Bradley-Terry with exponential recency weights on newer games. More โ | 0.763 | 69.6% | 0.198 | 0.579 | 0.750 | 57.1% |
| 12 | Bradley-Terry Bradley-Terry Static logistic paired-comparison model with one team strength parameter. More โ | 0.763 | 69.5% | 0.198 | 0.579 | 0.750 | 57.1% |
| 13 | Avg Margin Baseline Avg Margin Baseline Predict from simple average scoring margin in the training window. More โ | 0.763 | 69.6% | 0.198 | 0.580 | 0.917 | 71.4% |
| 14 | Pythagorean Pythagorean Pythagorean win expectation from raw points scored and allowed. More โ | 0.760 | 69.4% | 0.205 | 0.596 | 0.917 | 71.4% |
| 15 | Elo Elo Streaming paired-comparison rating with recency baked into sequential updates. More โ | 0.754 | 67.2% | 0.203 | 0.590 | 0.750 | 57.1% |
| 16 | Home Team Baseline Home Team Baseline Always favor the home team with a fixed prior. More โ | 0.594 | 59.4% | 0.241 | 0.675 | 0.542 | 57.1% |
Methodology
Out-of-sample windows
- The season is split into 7-day windows ending on its last played day
- Each window is predicted by models refit on all earlier games of the season
- The first window starts after 21 days of games
- Every model is scored on the same games; a model that cannot predict most of them is listed as not compared
Elo / Bradley-Terry
- Elo: the published league Elo (preseason priors and carryover where configured), frozen at the window start
- Bradley-Terry: static logistic fit on the earlier games
- Recency variants weight games with a 120-day half-life in NCAA leagues (keeping the non-conference games that link conferences) and a 30-day half-life in pro leagues
Pythagorean and margin models
- Raw, adjusted and log-adjusted Pythagorean from scores; efficiency from box-score possessions
- Margin: ridge regression on point margin with a learned home advantage
- Ensembles average the log-odds of five families
Baselines
- Home Team: the home team wins 60% (50% at neutral sites)
- Avg Margin: the higher average margin so far wins
- Models should beat these to add value