CFB Betting Edge

2025 · Week 4 · Sun Belt · backtest — what the model predicted vs what happened, and vs the market
Week 4 backtest · Sun Belt · 9 games — what the model predicted vs what happened
Straight up
3–633%
ATS vs close
2–722%
Model margin MAE
17.9
Market margin MAE
13.4
→ 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 Vanderbilt 32.3 · actual Vanderbilt 49
winner Vanderbilt
ATS vs close
leaned Vanderbilt -27.5Cover ✓
margin err: model 16.7 · mkt 21.5 · closer
Model
pred South Alabama 23.1 · actual Coastal Carolina 18
winner South Alabama
ATS vs close
leaned South Alabama -14.5Miss ✗
margin err: model 41.1 · mkt 32.5 · mkt closer
UL Monroe @ UTEPFinal 3125
Model
pred UTEP 21.7 · actual UL Monroe 6
winner UTEP
ATS vs close
leaned UTEP -5.5Miss ✗
margin err: model 27.7 · mkt 11.5 · mkt closer
Model
pred Louisiana Tech 10.8 · actual Louisiana Tech 10
winner Louisiana Tech
ATS vs close
leaned Louisiana Tech -3.0Cover ✓
margin err: model 0.8 · mkt 7.0 · closer
Model
pred Liberty 10.4 · actual James Madison 18
winner Liberty
ATS vs close
leaned Liberty +7.5Miss ✗
margin err: model 28.4 · mkt 10.5 · mkt closer
Troy @ BuffaloFinal 2117
Model
pred Buffalo 9.3 · actual Troy 4
winner Buffalo
ATS vs close
leaned Buffalo -5.5Miss ✗
margin err: model 13.3 · mkt 9.5 · mkt closer
Model
pred Marshall 0.8 · actual Marshall 14
winner Marshall
ATS vs close
leaned Middle Tennessee +2.5Miss ✗
margin err: model 13.2 · mkt 11.5 · mkt closer
Model
pred Louisiana 4.7 · actual Eastern Michigan 3
winner Louisiana
ATS vs close
leaned Louisiana -2.5Miss ✗
margin err: model 7.7 · mkt 5.5 · mkt closer
Model
pred Arkansas State 5.0 · actual Kennesaw State 7
winner Arkansas State
ATS vs close
leaned Arkansas State -4.5Miss ✗
margin err: model 12.0 · mkt 11.5 · mkt 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.