CFB Betting Edge

2025 · Week 4 · ACC · backtest — what the model predicted vs what happened, and vs the market
Week 4 backtest · ACC · 10 games — what the model predicted vs what happened
Straight up
7–370%
ATS vs close
6–460%
Model margin MAE
17.3
Market margin MAE
15.0
→ 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 Florida State 34.0 · actual Florida State 56
winner Florida State
ATS vs close
leaned Kent State +45.0Miss ✗
margin err: model 22.0 · mkt 11.0 · mkt closer
Model
pred Louisville 21.0 · actual Louisville 23
winner Louisville
ATS vs close
leaned Bowling Green +26.5Cover ✓
margin err: model 2.0 · mkt 3.5 · closer
Syracuse @ ClemsonFinal 3421
Model
pred Clemson 17.8 · actual Syracuse 13
winner Clemson
ATS vs close
leaned Clemson -17.5Miss ✗
margin err: model 30.8 · mkt 30.5 · mkt closer
Stanford @ VirginiaFinal 2048
Model
pred Virginia 17.2 · actual Virginia 28
winner Virginia
ATS vs close
leaned Virginia -16.5Cover ✓
margin err: model 10.8 · mkt 11.5 · closer
Model
pred Georgia Tech 14.9 · actual Georgia Tech 21
winner Georgia Tech
ATS vs close
leaned Temple +24.5Cover ✓
margin err: model 6.1 · mkt 3.5 · mkt closer
SMU @ TCUFinal 2435
Model
pred TCU 13.7 · actual TCU 11
winner TCU
ATS vs close
leaned TCU -6.5Cover ✓
margin err: model 2.7 · mkt 4.5 · closer
Florida @ MiamiFinal 726
Model
pred Miami 10.7 · actual Miami 19
winner Miami
ATS vs close
leaned Miami -8.5Cover ✓
margin err: model 8.3 · mkt 10.5 · closer
Model
pred UCF 9.6 · actual UCF 25
winner UCF
ATS vs close
leaned UCF -7.0Cover ✓
margin err: model 15.4 · mkt 18.0 · closer
NC State @ DukeFinal 3345
Model
pred NC State 12.3 · actual Duke 12
winner NC State
ATS vs close
leaned NC State +3.0Miss ✗
margin err: model 24.3 · mkt 9.0 · mkt closer
Model
pred California 16.7 · actual San Diego State 34
winner California
ATS vs close
leaned California -14.0Miss ✗
margin err: model 50.7 · mkt 48.0 · 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.