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 graded755755 with a closing line
Model MAE12.58points
Market MAE11.97points
Gap to market+0.62points (lower is better)
Model closer48.9%
80% coverage79%target 80%
By prediction source
Live season accuracy by prediction source
Source
Games
MAE
RMSE
Bias
Total MAE
50 / 80 / 90
Gap
Closer
Pure model
755
12.80
16.42
-1.14
12.96
52% / 80% / 90%
+0.83
47.2%
Published (market-informed)
755
12.58
16.17
-1.58
13.69
50% / 79% / 87%
+0.62
48.9%
Closing line
755
11.97
15.18
-1.11
12.50
–
–
–
Published is the market-informed line, total and scores the site shows: the pure margin mixed with the pregame market spread the model saw when it published. Pure model is the independent projection behind it. Closing line is the benchmark graded on its own. Coverage is the share of results inside each source's 50, 80, and 90 percent intervals: the pure model's frozen ones, and for the published line the dispersion picks are priced with. The published total is graded only for games published since it existed. Home win probability from the published pregame line scores Brier 0.148 and log loss 0.450 over 755 games.
By week
By week
Segment
Games
Model MAE
Market MAE
Gap
Model closer
Bias
Week 1
171
15.49
12.75
+2.74
41.5%
-5.71
Week 2
120
11.89
11.53
+0.36
49.2%
-1.90
Week 3
119
10.29
10.51
-0.22
54.6%
-0.53
Week 4
120
11.97
11.65
+0.32
47.5%
+1.01
Week 5
114
13.01
13.29
-0.28
51.8%
-0.26
Week 6
104
11.65
11.92
-0.27
53.8%
+0.16
Week 7thin
By opponent classification
By opponent classification
Segment
Games
Model MAE
Market MAE
Gap
Model closer
Bias
FBS vs FBS
336
11.58
11.46
+0.11
48.8%
-0.39
FBS vs FCS
119
15.42
12.58
+2.84
41.2%
-8.59
FCS vs FCS
300
12.58
12.29
+0.30
52.0%
-0.14
Data and model diagnostics
These explain forecast error, not betting records. Missing-input counts include unavailable injury data and other gaps in preseason sources. Projected winning margin is the absolute model margin: under 7 means the model expects either team to win by fewer than 7 points, regardless of the sportsbook line. Favorite, underdog, over and under pick records are in Recommendations.
By input completeness
By input completeness
Segment
Games
Model MAE
Market MAE
Gap
Model closer
Bias
1 to 3
319
11.71
11.58
+0.13
48.3%
-0.17
4 or more
436
13.22
12.25
+0.97
49.3%
-2.62
By projected winning margin (points)
By projected winning margin (points)
Segment
Games
Model MAE
Market MAE
Gap
Model closer
Bias
under 7
259
11.79
11.35
+0.44
49.8%
-0.21
7 to 14
181
12.84
Segments grade the published line. Market columns use only the games in that segment with a closing spread. A segment marked thin has fewer than 30 games.
Loading historical comparison…
7
10.17
9.79
+0.38
28.6%
-4.80
12.06
+0.78
48.6%
-1.44
14 to 21
125
12.16
11.38
+0.78
44.8%
-2.80
21 or more
190
13.70
13.10
+0.60
50.5%
-2.79
Historical walk-forward backtest (pure model)
Frozen, 2021–2025
Every prediction below was made walking forward through each season with only the data available at the time, then frozen. They come from the pure model with no market blend, so they are historical stand-ins for the model's own skill, not for the published line. 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
Season
Games
Model MAE
Market MAE
Gap
Model closer
Bias
2021
760
13.60
12.49
+1.10
43.9%
-0.28
2022
757
13.86
12.28
+1.58
40.8%
+0.29
2023
771
13.26
12.00
+1.25
43.3%
-0.24
2024
773
13.46
12.02
+1.44
43.6%
-0.76
2025
792
12.84
11.93
+0.91
45.8%
-0.32
All seasons
3853
13.40
12.14
+1.26
43.5%
-0.27
Bias is the mean signed error of the model's home margin: positive means the model leans toward home teams. The published line and the in-game projections lean on the market precisely because the closing line remains the better pregame forecast.
