I had 'em all out there steppin' and fetchin' like their heads were on fire and their asses was catchin'. "The Ballad of Uneasy Rider" - Charlie Daniels

Model Performance

Every FINAL game is scored against each model's pre-game prediction, built only from rating snapshots dated before tip-off. The Book Closing Line row is the sportsbook benchmark. Model predictions are for analytical purposes only and do not constitute betting advice.

Spread predicted home margin vs. actual

Model Games MAE RMSE Winner %
Baseline (ML) 5,531 9.37 11.91 70.6%
Baseline Plus (ML) 5,425 9.29 11.79 70.6%
Model 3 (ML) 5,531 9.41 12.00 70.9%
Model 4 (ML) 5,293 9.18 11.66 70.6%
Massey 5,531 9.53 12.10 70.4%
Adjusted Efficiency 5,531 9.52 12.10 70.3%
Book Closing Line 5,705 8.92 11.23 72.7%

Total predicted combined score vs. actual

Model Games MAE RMSE Bias
Baseline (ML) 5,531 14.17 17.97 +4.38
Baseline Plus (ML) 5,425 13.90 17.63 +3.68
Model 3 (ML) 5,531 14.27 18.12 +5.07
Model 4 (ML) 5,293 13.95 17.69 +3.90
Massey Totals 5,531 14.10 17.74 -1.28
Adjusted Efficiency 5,531 14.13 17.77 -1.30
Book Closing Line 5,734 13.37 16.89 +0.60

Win Probability predicted home-win chance vs. outcome

Model Games Log Loss Brier Accuracy
Baseline (ML) 5,531 0.5541 0.1881 70.8%
Baseline Plus (ML) 5,425 0.5538 0.1880 70.6%
Model 3 (ML) 5,531 0.5553 0.1885 70.6%
Model 4 (ML) 5,293 0.5543 0.1883 70.9%
Massey 5,531 0.5584 0.1887 70.4%
Bradley-Terry 5,531 0.6246 0.2076 68.8%
Weighted Bradley-Terry 5,531 0.6835 0.2157 68.7%
Adjusted Efficiency 5,531 0.5581 0.1887 70.3%
Book Closing Line 5,729 0.5129 0.1732 73.0%

vs. Closing Line identical games only — the honest head-to-head

Model Games Model MAE Book MAE Δ ATS O/U CLV
Baseline (ML) 5,487 9.37 8.89 +0.47 50.6% (5,487) 50.2% (5,513)
Baseline Plus (ML) 5,381 9.29 8.86 +0.43 50.5% (5,381) 50.4% (5,407)
Model 3 (ML) 5,487 9.41 8.89 +0.51 50.3% (5,487) 50.8% (5,513)
Model 4 (ML) 5,249 9.18 8.87 +0.31 50.2% (5,249) 50.4% (5,275)
Massey 5,487 9.53 8.89 +0.64 50.3% (5,487)
Massey Totals 0 49.6% (5,513)
Adjusted Efficiency 5,487 9.52 8.89 +0.63 50.2% (5,487) 49.5% (5,513)

All columns pair each model with the book closing line on the same games, so coverage differences can't skew the comparison. ATS/O-U count only games where the model disagrees with the closing number (pushes excluded) — 52.4% is breakeven at standard −110 vig. CLV counts games where the line moved off the opener and the model had taken a side against it; above 50% means the market tended to move toward the model. Not betting advice.

Month by Month do the models close the gap to the book as the season develops?

Monthly accuracy over the full season (the window selector does not apply here). Lower is better for both metrics; the dashed line is the book benchmark. Early-season months reflect ratings that are still converging on limited games.

By Conference where does the model beat the book?

Conference Games Model MAE Book MAE Δ vs Book Winner %
Big South Conference 164 9.17 9.50 -0.33 77.4%
Horizon League 188 8.99 8.99 -0.00 69.1%
Conference USA 204 9.02 8.99 +0.03 62.7%
Southeastern Conference 354 10.08 9.94 +0.14 75.4%
Southland Conference 195 8.39 8.25 +0.14 70.8%
Patriot League 183 8.83 8.62 +0.21 67.8%
Atlantic Sun Conference 221 9.23 9.00 +0.23 71.0%
Southern Conference 179 8.75 8.51 +0.23 70.4%
Big Sky Conference 175 8.87 8.59 +0.29 70.3%
Sun Belt Conference 257 9.21 8.90 +0.31 66.1%
Atlantic 10 Conference 277 9.00 8.62 +0.38 68.2%
Southwestern Athletic Conference 222 9.68 9.29 +0.39 67.1%
Mid-American Conference 235 9.29 8.90 +0.39 73.2%
Coastal Athletic Association 240 9.01 8.62 +0.40 70.0%
Big East Conference 221 8.94 8.54 +0.40 74.2%
Missouri Valley Conference 207 9.87 9.47 +0.40 70.0%
Big 12 Conference 343 9.78 9.35 +0.42 76.1%
Big West Conference 193 9.37 8.90 +0.47 67.9%
Mountain West Conference 230 9.74 9.27 +0.47 74.8%
Big Ten Conference 368 9.98 9.48 +0.50 76.9%
Summit League 159 10.20 9.68 +0.52 69.2%
Ivy League 136 9.63 9.10 +0.54 66.9%
West Coast Conference 233 9.69 9.15 +0.54 76.0%
Northeast Conference 179 8.82 8.23 +0.60 74.9%
American Conference 251 8.72 8.11 +0.60 72.5%
Atlantic Coast Conference 373 9.05 8.44 +0.62 74.8%
Western Athletic Conference 125 8.80 8.16 +0.64 72.0%
America East Conference 167 9.91 9.24 +0.66 74.3%
Metro Atlantic Athletic Conference 231 9.34 8.65 +0.69 69.7%
Ohio Valley Conference 179 10.19 9.44 +0.75 66.5%
Mid-Eastern Athletic Conference 149 9.24 8.10 +1.14 72.5%

Paired comparison: only games where both the selected model and the book closing line made a spread prediction, sorted best-for-the-model first. Cross-conference games count toward both teams' conferences. Small samples swing wildly — read the Games column before drawing conclusions.

Calibration when a model says 70%, does the home team win 70% of the time?

Predictions are grouped into deciles of predicted home-win probability; each point compares the decile's average prediction (x) with the actual home-win rate (y). A perfectly calibrated model follows the dashed diagonal.