CFB Betting Edge

2025 · Week 15 · FBS · backtest — what the model predicted vs what happened, and vs the market
Week 15 backtest · FBS · 9 games — what the model predicted vs what happened
Straight up
5–456%
ATS vs close
3–633%
Model margin MAE
13.1
Market margin MAE
10.3
→ market was closer (model can't beat the close ATS)
Results — model read vs the final score + how the closing-line lean held up
Model
pred James Madison 20.8 · actual James Madison 17
winner James Madison
ATS vs close
leaned Troy +24.5Cover ✓
margin err: model 3.8 · mkt 7.5 · closer
Model
pred Jacksonville State 10.8 · actual Kennesaw State 4
winner Jacksonville State
ATS vs close
leaned Jacksonville State +3.0Miss ✗
margin err: model 14.8 · mkt 1.0 · mkt closer
Duke @ VirginiaFinal 2720
Model
pred Virginia 6.8 · actual Duke 7
winner Virginia
ATS vs close
leaned Virginia -3.5Miss ✗
margin err: model 13.8 · mkt 10.5 · mkt closer
Model
pred Ohio State 6.7 · actual Indiana 3
winner Ohio State
ATS vs close
leaned Ohio State -3.5Miss ✗
margin err: model 9.7 · mkt 6.5 · mkt closer
Model
pred Western Michigan 5.4 · actual Western Michigan 10
winner Western Michigan
ATS vs close
leaned Western Michigan -2.5Cover ✓
margin err: model 4.6 · mkt 7.5 · closer
UNLV @ Boise StateFinal 2138
Model
pred Boise State 5.4 · actual Boise State 17
winner Boise State
ATS vs close
leaned UNLV +6.0Miss ✗
margin err: model 11.6 · mkt 11.0 · mkt closer
BYU @ Texas TechFinal 734
Model
pred Texas Tech 2.6 · actual Texas Tech 27
winner Texas Tech
ATS vs close
leaned BYU +12.5Miss ✗
margin err: model 24.4 · mkt 14.5 · mkt closer
Georgia @ AlabamaFinal 287
Model
pred Georgia 0.3 · actual Georgia 21
winner Georgia
ATS vs close
leaned Alabama +1.5Miss ✗
margin err: model 20.7 · mkt 19.5 · mkt closer
Model
pred North Texas 1.1 · actual Tulane 13
winner North Texas
ATS vs close
leaned Tulane +1.5Cover ✓
margin err: model 14.1 · mkt 14.5 · closer
The model beats SP+ but not the closing line (ATS ≈ break-even). “Model−Mkt” is the gap between our projected home margin and the market's implied margin — a read on disagreement, not a profitable edge. Upcoming games have no closing line yet, so the market number marked ~ is the consensus midpoint of the best posted spreads across books.