CFB Betting Edge

2024 · Week 4 · Conference USA · backtest — what the model predicted vs what happened, and vs the market
Week 4 backtest · Conference USA · 7 games — what the model predicted vs what happened
Straight up
5–271%
ATS vs close
4–357%
Model margin MAE
10.5
Market margin MAE
9.6
→ 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 Liberty 13.0 · actual Liberty 11
winner Liberty
ATS vs close
leaned Liberty -7.5Cover ✓
margin err: model 2.0 · mkt 3.5 · closer
Model
pred Jacksonville State 10.4 · actual Jacksonville State 37
winner Jacksonville State
ATS vs close
leaned Jacksonville State -6.0Cover ✓
margin err: model 26.6 · mkt 31.0 · closer
Model
pred Sam Houston 7.2 · actual Sam Houston 20
winner Sam Houston
ATS vs close
leaned New Mexico State +15.0Miss ✗
margin err: model 12.8 · mkt 5.0 · mkt closer
Model
pred Louisiana Tech 1.5 · actual Tulsa 3
winner Louisiana Tech
ATS vs close
leaned Tulsa +2.5Cover ✓
margin err: model 4.5 · mkt 5.5 · closer
Model
pred Colorado State 1.4 · actual Colorado State 10
winner Colorado State
ATS vs close
leaned UTEP +8.5Miss ✗
margin err: model 8.6 · mkt 1.5 · mkt closer
Model
pred Toledo 10.6 · actual Western Kentucky 5
winner Toledo
ATS vs close
leaned Toledo -2.5Miss ✗
margin err: model 15.6 · mkt 7.5 · mkt closer
Model
pred Duke 24.5 · actual Duke 28
winner Duke
ATS vs close
leaned Duke -14.5Cover ✓
margin err: model 3.5 · mkt 13.5 · 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.