Skip to content

Model Performance

Live grading runs after each Monday refresh
Last graded: Sep 1, 3:03 PM EDT

2026 season, graded as played

Every row is a frozen record: the projection published before kickoff, the closing spread from the CFBD lines feed (median across providers, the same definition as the backtest), and the final score. Nothing is re-fit or re-priced after a result is known. The closing line is the bar the model is measured against, not an input to it.

Games graded5353 with a closing line
Model MAE12.86points
Market MAE11.80points
Gap to market+1.05points (lower is better)
Model closer49.1%
80% coverage85%target 80%

By prediction source

Live season accuracy by prediction source
SourceGamesMAERMSEBiasTotal MAE50 / 80 / 90GapCloser
Pure model5312.8615.58-0.2512.6453% / 85% / 96%+1.0549.1%
Market-informed blend5312.8615.55-0.23+1.0647.2%
Closing line5311.8014.03-0.2011.56

Pure model is the independent projection. The market-informed blend mixes the pure margin with the pregame market spread the model saw when it published, so it is a product line, not model skill. Closing line is the benchmark graded on its own. Coverage is the share of results inside the pure model's frozen 50, 80, and 90 percent intervals. Home win probability, taken from the frozen pregame margin distribution before kickoff, scores Brier 0.207 and log loss 0.600 over 53 games.

By week

By week
SegmentGamesModel MAEMarket MAEGapModel closerBias
Week 15312.8611.80+1.0549.1%-0.25

By opponent classification

By opponent classification
SegmentGamesModel MAEMarket MAEGapModel closerBias
FBS vs FBSthin813.6813.44+0.2450.0%+1.04
FCS vs FCS4512.7111.51+1.2048.9%-0.48

By model favorite size

By model favorite size
SegmentGamesModel MAEMarket MAEGapModel closerBias
under 7thin2510.2010.88-0.6856.0%-2.94
7 to 14thin1115.6311.36+4.2627.3%+3.32
14 to 21thin1211.1110.71+0.4050.0%-2.60
21 or morethin524.2520.00+4.2560.0%+10.95

By missing preseason inputs

By missing preseason inputs
SegmentGamesModel MAEMarket MAEGapModel closerBias
1 to 3thin614.8714.33+0.5450.0%+4.75
4 or more4712.6011.48+1.1248.9%-0.89

Segments are pure model rows. Market columns use only the games in that segment with a closing spread. A segment marked thin has fewer than 30 games.

Historical walk-forward backtest

Frozen, 20212025

Every prediction below was made walking forward through each season with only the data available at the time, then frozen. These are historical stand-ins for live performance, not live results. The market benchmark is the closing spread: the strongest public forecast of a game's margin. Beating it consistently is rare, and the model is measured against it, not against a naive baseline.

Games graded3853
Model MAE13.40points
Market MAE12.14points
Gap to market+1.26points (lower is better)
Model closer43.5%
Backtest accuracy by season
SeasonGamesModel MAEMarket MAEGapModel closerBias
202176013.6012.49+1.1043.9%-0.28
202275713.8612.28+1.5840.8%+0.29
202377113.2612.00+1.2543.3%-0.24
202477313.4612.02+1.4443.6%-0.76
202579212.8411.93+0.9145.8%-0.32
All seasons385313.4012.14+1.2643.5%-0.27

Bias is the mean signed error of the model's home margin: positive means the model leans toward home teams. In-game projections anchor on the market closing line precisely because the closing line remains the better pregame forecast.