Week 1 / Against the spread

Every bet.
Every reason.

These are the 24 model disagreements that clear the Week 1 evidence floor. Each edge is rebuilt below from its exact contributing signals.

24recommendations
51games with lines
5reasons shown per game
0.0current injury points applied

Week 1 historical replay · 2017–2025

62.7%ATS

185 wins, 110 losses and 1 push.

The Week 1 recommendation policy was replayed across 296 historical selections. It covered 62.7% of the 295 bets with a win or loss. That is encouraging—and it still includes 110 losses. No system is perfect.

See the year-by-year results

How to read the board

The edge has to add up.

The football model starts from the consensus provider spread it was trained to use. An independent roster check can add to or temper that move. A second year-over-year check uses opponent-adjusted offensive-line quality and actual transfer production, but only when that evidence reinforces the same side. The model then estimates competitive possessions from tempo, drive length, efficiency and play style. That pace estimate can make a small change to the displayed forecast in projected non-blowouts, but it cannot create a bet or reverse the side. The board then finds the most favorable named-book spread for that same side. Positive reason points support the recommendation; together they equal the displayed best-book edge.

When both coaching statuses are known and neither team changed head coaches, an intentional model disagreement is enough. If either team changed coaches—or coaching status is unknown—the best-book edge must reach 2.00 points. Coaching can make the model abstain; it never flips the side. Prices refresh automatically, so the named book and line are part of the recommendation—not a timeless pick.

Season by season

The wins did not arrive in a straight line.

Each season was tested using predictions built without that season's final scores. This is historical research, not an untouched live betting record; model development used this general era.

SeasonRecordBetsWin rate
201724–143863.2%
201820–18–13952.6%
201922–163857.9%
20204–04100.0%
202127–174461.4%
202228–184660.9%
202311–31478.6%
202426–83476.5%
202523–163959.0%

How to read the historical matches

Thirty comparisons, not a forecast.

Each recommendation includes the 30 closest prior Week 1 bets based on the spread, model edge, expected total, game setting and the mix of core, roster reasons and the tracked pace profile. The closest available games are not exact copies.

Treat the record as context—not the chance this bet will cover. In a 30-game sample, one result moves the win rate by 3.3 percentage points. These comparisons never change the recommendation, edge or ranking.

Current card

Ranked by best-book edge.

All times Eastern. Best available named-book spreads refreshed Aug 12, 2026, 10:52 PM EDT.

01
Sat, Sep 5, 12:30 PM EDT

Ball State at Ohio State

Best-book edge+8.5
Recommendation · DraftKings

Ohio State −49.5

Best book
DraftKings · −49.5
Consensus line
−50.25
All-Takes line
−58.01

Evidence gate: intentional model departure. Neither team changed head coaches, so an intentional pre-pace production-model departure is sufficient. The pace adjustment changes the forecast, not whether a bet exists.

Core football model+7.0

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check0.0

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production+0.8

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions0.0

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price+0.8

DraftKings has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets22–8

73.3% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2023

Alabama −39.5 vs. Middle Tennessee

+7.6Covered by 9.5
2019

Texas A&M −33.25 vs. Texas State

+7.0Covered by 0.8
2022

Minnesota −36 vs. New Mexico State

+5.1Covered by 2.0
2017

Oklahoma −43.17 vs. UTEP

+3.4Covered by 5.8
2017

Clemson −38.17 vs. Kent State

+3.8Covered by 14.8
2022

Tennessee −36.25 vs. Ball State

+6.2Covered by 12.8
2025

Ole Miss −38.5 vs. Georgia State

+3.1Covered by 17.5
2023

Texas −35 vs. Rice

+6.9Missed by 8.0
2018

Ohio State −40 vs. Oregon State

+3.3Covered by 6.0
2023

Auburn −35 vs. Massachusetts

+3.6Covered by 10.0
2019

Alabama −33.75 vs. Duke

+5.6Covered by 5.3
2024

Alabama −31 vs. Western Kentucky

+3.9Covered by 32.0
2022

Alabama −42.12 vs. Utah State

+1.8Covered by 12.9
2025

Penn State −45.5 vs. Nevada

+1.2Missed by 10.5
2021

Fresno State −27.70 vs. UConn

+4.4Covered by 17.3
2021

Tennessee −37.30 vs. Bowling Green

+6.2Missed by 5.3
2023

Tennessee −28 vs. Virginia

+6.0Covered by 8.0
2019

Michigan −35.62 vs. Middle Tennessee

+2.4Missed by 16.6
2021

Texas A&M −29.5 vs. Kent State

+3.9Covered by 1.5
2024

Oklahoma −42.5 vs. Temple

+0.5Covered by 5.5
2023

Texas A&M −38 vs. New Mexico

+7.2Covered by 4.0
2021

Auburn −37.30 vs. Akron

+5.0Covered by 12.7
2024

Georgia Tech −20.5 vs. Georgia State

+7.5Covered by 2.5
2021

Oklahoma −31.25 vs. Tulane

+2.0Missed by 26.3
2021

Wake Forest −32.5 vs. Old Dominion

+1.0Missed by 0.5
2021

Pittsburgh −37.60 vs. Massachusetts

+0.2Covered by 6.4
2018

Alabama −23.75 vs. Louisville

+4.8Covered by 13.3
2017

Memphis −27.17 vs. UL Monroe

+2.0Missed by 19.2
2022

Maryland −24 vs. Buffalo

+3.6Missed by 3.0
2022

Texas −37.38 vs. UL Monroe

+7.4Covered by 4.6
02
Sat, Sep 5, 7:45 PM EDT

Florida Atlantic at Florida

Best-book edge+8.0
Recommendation · Bovada

Florida Atlantic +26.5

Best book
Bovada · +26.5
Consensus line
+26
All-Takes line
+18.47

Evidence floor: 2.00 points. A 2.00-point evidence floor applies because at least one team changed head coaches. The coaching signal can cause an abstention; it never flips the model's side. The new pace adjustment is shown in the forecast but does not create a bet.

Core football model+5.6

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check+1.9

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production0.0

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions0.0

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price+0.5

Bovada has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets19–11

63.3% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2023

Colorado +20.5 at TCU

+3.9Covered by 23.5
2022

UConn +23.75 at Utah State

+2.9Covered by 12.8
2019

Vanderbilt +22.62 vs. Georgia

+6.3Missed by 1.4
2018

Kent State +18.12 at Illinois

+3.7Covered by 11.1
2019

Florida Atlantic +27.62 at Ohio State

+3.5Covered by 3.6
2019

Kent State +24.5 at Arizona State

+2.8Covered by 1.5
2018

Army +13.88 at Duke

+4.6Missed by 6.1
2024

Western Michigan +24 at Wisconsin

+2.8Covered by 10.0
2025

Jacksonville State +16.5 at UCF

+3.0Covered by 9.5
2023

San José State +30.5 at USC

+3.3Covered by 2.5
2022

Hawai'i +16.5 vs. Western Kentucky

+7.0Missed by 15.5
2017

Rutgers +28 vs. Washington

+4.4Covered by 12.0
2018

Massachusetts +20.38 at Boston College

+2.5Missed by 13.6
2023

Utah State +23.5 at Iowa

+4.4Covered by 13.5
2022

Memphis +17 at Mississippi State

+2.8Missed by 9.0
2025

Western Michigan +18.5 at Michigan State

+2.9Covered by 1.5
2024

Nevada +27 vs. SMU

+3.0Covered by 22.0
2021

New Mexico State +31.20 at San Diego State

+2.2Covered by 13.2
2022

Bowling Green +24 at UCLA

+2.3Missed by 4.0
2019

Louisville +19.25 vs. Notre Dame

+3.1Covered by 1.3
2021

Hawai'i +17.60 at UCLA

+3.1Missed by 16.4
2023

Louisiana Tech +20 at SMU

+2.7Missed by 4.0
2018

Rice +25.88 vs. Houston

+2.3Covered by 7.9
2024

Old Dominion +20 at South Carolina

+2.5Covered by 16.0
2017

San José State +21 vs. South Florida

+3.4Covered by 1.0
2017

Rice +31.17 vs. Stanford

+5.0Missed by 23.8
2019

Louisiana +19 vs. Mississippi State

+3.4Covered by 9.0
2017

Nevada +24 at Northwestern

+4.1Covered by 13.0
2021

Miami +19.5 vs. Alabama

+3.9Missed by 11.5
2019

Massachusetts +16.67 at Rutgers

+2.3Missed by 10.3
03
Sat, Sep 5, 10:30 PM EDT

Western Kentucky at Nevada

Best-book edge+8.6
Recommendation · DraftKings

Nevada +4

Best book
DraftKings · +4
Consensus line
+3.25
All-Takes line
−4.60

Evidence gate: intentional model departure. Neither team changed head coaches, so an intentional pre-pace production-model departure is sufficient. The pace adjustment changes the forecast, not whether a bet exists.

Core football model+7.0

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check+0.0

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production0.0

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions+0.8

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price+0.8

DraftKings has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets16–14

53.3% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2022

Nevada +2 vs. Texas State

+6.9Covered by 26.0
2023

Jacksonville State +1.17 vs. UTEP

+5.9Covered by 4.2
2022

Arizona +6.25 at San Diego State

+4.1Covered by 24.3
2022

Hawai'i +16.5 vs. Western Kentucky

+7.0Missed by 15.5
2024

Georgia Tech +11.5 vs. Florida State

+5.8Covered by 14.5
2019

SMU +1.75 at Arkansas State

+4.0Covered by 8.8
2018

LSU +3.12 vs. Miami

+5.1Covered by 19.1
2025

UTEP +3.5 at Utah State

+3.5Missed by 8.5
2024

West Virginia +10 vs. Penn State

+2.9Missed by 12.0
2023

Northern Illinois +8.5 at Boston College

+3.0Covered by 11.5
2018

Army +13.88 at Duke

+4.6Missed by 6.1
2022

East Carolina +12.5 vs. NC State

+2.8Covered by 11.5
2023

North Texas +6.5 vs. California

+2.1Missed by 30.5
2022

Florida State +4.5 vs. LSU

+3.3Covered by 5.5
2019

Vanderbilt +22.62 vs. Georgia

+6.3Missed by 1.4
2022

Army +1.5 at Coastal Carolina

+3.2Missed by 8.5
2024

Sam Houston +10.5 at Rice

+6.0Covered by 30.5
2018

Northwestern +1 at Purdue

+2.6Covered by 5.0
2022

Temple +9.5 at Duke

+2.4Missed by 20.5
2021

Army +2.5 at Georgia State

+4.0Covered by 35.5
2021

West Virginia −2.5 at Maryland

+4.3Missed by 8.5
2023

Florida International +10 at Louisiana Tech

+5.5Covered by 5.0
2025

Notre Dame −2.5 at Miami

+4.8Missed by 5.5
2019

Louisville +19.25 vs. Notre Dame

+3.1Covered by 1.3
2025

Stanford +1.5 at Hawai'i

+4.0Missed by 1.5
2022

Utah −2.5 at Florida

+3.9Missed by 5.5
2017

California +12.5 at North Carolina

+2.7Covered by 17.5
2017

Florida Atlantic +8.67 vs. Navy

+2.3Missed by 14.3
2025

Sam Houston +13.5 vs. UNLV

+3.8Missed by 3.5
2018

Maryland +12.62 vs. Texas

+2.6Covered by 17.6
04
Mon, Sep 7, 7:30 PM EDT

SMU at Florida State

Best-book edge+8.2
Recommendation · Bovada

Florida State +2.5

Best book
Bovada · +2.5
Consensus line
+2.5
All-Takes line
−5.67

Evidence gate: intentional model departure. Neither team changed head coaches, so an intentional pre-pace production-model departure is sufficient. The pace adjustment changes the forecast, not whether a bet exists.

Core football model+7.0

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check0.0

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production+0.5

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions+0.7

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price0.0

Bovada has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets16–14

53.3% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2022

Nevada +2 vs. Texas State

+6.9Covered by 26.0
2023

Jacksonville State +1.17 vs. UTEP

+5.9Covered by 4.2
2022

Arizona +6.25 at San Diego State

+4.1Covered by 24.3
2018

LSU +3.12 vs. Miami

+5.1Covered by 19.1
2022

Hawai'i +16.5 vs. Western Kentucky

+7.0Missed by 15.5
2024

Georgia Tech +11.5 vs. Florida State

+5.8Covered by 14.5
2019

SMU +1.75 at Arkansas State

+4.0Covered by 8.8
2025

UTEP +3.5 at Utah State

+3.5Missed by 8.5
2024

Sam Houston +10.5 at Rice

+6.0Covered by 30.5
2023

Northern Illinois +8.5 at Boston College

+3.0Covered by 11.5
2024

West Virginia +10 vs. Penn State

+2.9Missed by 12.0
2018

Army +13.88 at Duke

+4.6Missed by 6.1
2022

Army +1.5 at Coastal Carolina

+3.2Missed by 8.5
2021

Army +2.5 at Georgia State

+4.0Covered by 35.5
2022

East Carolina +12.5 vs. NC State

+2.8Covered by 11.5
2023

Florida International +10 at Louisiana Tech

+5.5Covered by 5.0
2023

North Texas +6.5 vs. California

+2.1Missed by 30.5
2022

Florida State +4.5 vs. LSU

+3.3Covered by 5.5
2019

Vanderbilt +22.62 vs. Georgia

+6.3Missed by 1.4
2025

Stanford +1.5 at Hawai'i

+4.0Missed by 1.5
2018

Northwestern +1 at Purdue

+2.6Covered by 5.0
2021

West Virginia −2.5 at Maryland

+4.3Missed by 8.5
2022

Temple +9.5 at Duke

+2.4Missed by 20.5
2025

Notre Dame −2.5 at Miami

+4.8Missed by 5.5
2022

Utah −2.5 at Florida

+3.9Missed by 5.5
2017

Alabama −7.33 vs. Florida State

+7.6Covered by 9.7
2017

California +12.5 at North Carolina

+2.7Covered by 17.5
2019

Louisville +19.25 vs. Notre Dame

+3.1Covered by 1.3
2025

Sam Houston +13.5 vs. UNLV

+3.8Missed by 3.5
2017

Florida Atlantic +8.67 vs. Navy

+2.3Missed by 14.3
05
Sat, Sep 5, 12:45 PM EDT

Kent State at South Carolina

Best-book edge+7.2
Recommendation · DraftKings

South Carolina −34.5

Best book
DraftKings · −34.5
Consensus line
−35.5
All-Takes line
−41.73

Evidence gate: intentional model departure. Neither team changed head coaches, so an intentional pre-pace production-model departure is sufficient. The pace adjustment changes the forecast, not whether a bet exists.

Core football model+5.1

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check0.0

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production+1.2

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions0.0

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price+1.0

DraftKings has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets22–8

73.3% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2022

Minnesota −36 vs. New Mexico State

+5.1Covered by 2.0
2019

Texas A&M −33.25 vs. Texas State

+7.0Covered by 0.8
2023

Alabama −39.5 vs. Middle Tennessee

+7.6Covered by 9.5
2017

Clemson −38.17 vs. Kent State

+3.8Covered by 14.8
2023

Auburn −35 vs. Massachusetts

+3.6Covered by 10.0
2023

Texas −35 vs. Rice

+6.9Missed by 8.0
2024

Alabama −31 vs. Western Kentucky

+3.9Covered by 32.0
2021

Fresno State −27.70 vs. UConn

+4.4Covered by 17.3
2025

Ole Miss −38.5 vs. Georgia State

+3.1Covered by 17.5
2019

Alabama −33.75 vs. Duke

+5.6Covered by 5.3
2022

Tennessee −36.25 vs. Ball State

+6.2Covered by 12.8
2017

Oklahoma −43.17 vs. UTEP

+3.4Covered by 5.8
2018

Ohio State −40 vs. Oregon State

+3.3Covered by 6.0
2023

Tennessee −28 vs. Virginia

+6.0Covered by 8.0
2021

Texas A&M −29.5 vs. Kent State

+3.9Covered by 1.5
2024

Georgia Tech −20.5 vs. Georgia State

+7.5Covered by 2.5
2019

Michigan −35.62 vs. Middle Tennessee

+2.4Missed by 16.6
2021

Tennessee −37.30 vs. Bowling Green

+6.2Missed by 5.3
2019

Texas −19.38 vs. Louisiana Tech

+3.3Covered by 11.6
2021

Auburn −37.30 vs. Akron

+5.0Covered by 12.7
2017

Memphis −27.17 vs. UL Monroe

+2.0Missed by 19.2
2022

Alabama −42.12 vs. Utah State

+1.8Covered by 12.9
2022

Maryland −24 vs. Buffalo

+3.6Missed by 3.0
2018

Alabama −23.75 vs. Louisville

+4.8Covered by 13.3
2022

Michigan State −21.75 vs. Western Michigan

+3.3Covered by 0.3
2023

Texas A&M −38 vs. New Mexico

+7.2Covered by 4.0
2021

Oklahoma −31.25 vs. Tulane

+2.0Missed by 26.3
2025

Purdue −16.5 vs. Ball State

+4.5Covered by 14.5
2021

Wake Forest −32.5 vs. Old Dominion

+1.0Missed by 0.5
2025

Penn State −45.5 vs. Nevada

+1.2Missed by 10.5
06
Sat, Aug 29, 5:30 PM EDT

Jacksonville State at North Dakota State

Best-book edge+7.3
Recommendation · DraftKings

Jacksonville State +10

Best book
DraftKings · +10
Consensus line
+8.5
All-Takes line
+2.68

Evidence floor: 2.00 points. A 2.00-point evidence floor applies because at least one team changed head coaches. The coaching signal can cause an abstention; it never flips the model's side. The new pace adjustment is shown in the forecast but does not create a bet.

Core football model+2.8

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check+2.5

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production0.0

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions+0.4

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price+1.5

DraftKings has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets18–12

60.0% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2018

Army +13.88 at Duke

+4.6Missed by 6.1
2025

Jacksonville State +16.5 at UCF

+3.0Covered by 9.5
2021

East Carolina +9.20 at App State

+2.2Missed by 4.8
2025

Stanford +1.5 at Hawai'i

+4.0Missed by 1.5
2023

Jacksonville State +1.17 vs. UTEP

+5.9Covered by 4.2
2022

North Carolina +3 at App State

+3.1Covered by 5.0
2025

Toledo +10 at Kentucky

+1.5Covered by 2.0
2019

SMU +1.75 at Arkansas State

+4.0Covered by 8.8
2020

South Alabama +12.40 at Southern Miss

+3.0Covered by 23.4
2025

Northwestern +6.5 at Tulane

+1.7Missed by 13.5
2022

Arizona +6.25 at San Diego State

+4.1Covered by 24.3
2025

Western Michigan +18.5 at Michigan State

+2.9Covered by 1.5
2018

Troy +9.12 vs. Boise State

+1.3Missed by 26.9
2022

West Virginia +7.5 at Pittsburgh

+4.3Covered by 0.5
2024

West Virginia +10 vs. Penn State

+2.9Missed by 12.0
2023

Northern Illinois +8.5 at Boston College

+3.0Covered by 11.5
2019

Fresno State +14.12 at USC

+2.9Covered by 6.1
2024

Nevada +15.5 at Troy

+2.1Covered by 17.5
2025

UTEP +3.5 at Utah State

+3.5Missed by 8.5
2022

Nevada +2 vs. Texas State

+6.9Covered by 26.0
2022

Temple +9.5 at Duke

+2.4Missed by 20.5
2018

Kent State +18.12 at Illinois

+3.7Covered by 11.1
2025

Rice +14.5 at Louisiana

+4.4Covered by 16.5
2018

Texas State +16.5 at Rutgers

+1.0Missed by 11.5
2022

East Carolina +12.5 vs. NC State

+2.8Covered by 11.5
2024

Boston College +16.5 at Florida State

+3.7Covered by 31.5
2018

Tennessee +9.88 vs. West Virginia

+3.5Missed by 16.1
2022

Memphis +17 at Mississippi State

+2.8Missed by 9.0
2017

Arkansas State +14.5 at Nebraska

+2.5Covered by 7.5
2022

Army +1.5 at Coastal Carolina

+3.2Missed by 8.5
07
Fri, Sep 4, 8:00 PM EDT

UTEP at Oklahoma

Best-book edge+6.8
Recommendation · Bovada

Oklahoma −40.5

Best book
Bovada · −40.5
Consensus line
−40.5
All-Takes line
−47.28

Evidence gate: intentional model departure. Neither team changed head coaches, so an intentional pre-pace production-model departure is sufficient. The pace adjustment changes the forecast, not whether a bet exists.

Core football model+5.6

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check0.0

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production+1.1

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions0.0

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price0.0

Bovada has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets23–7

76.7% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2022

Minnesota −36 vs. New Mexico State

+5.1Covered by 2.0
2023

Alabama −39.5 vs. Middle Tennessee

+7.6Covered by 9.5
2019

Texas A&M −33.25 vs. Texas State

+7.0Covered by 0.8
2017

Clemson −38.17 vs. Kent State

+3.8Covered by 14.8
2023

Auburn −35 vs. Massachusetts

+3.6Covered by 10.0
2023

Texas −35 vs. Rice

+6.9Missed by 8.0
2024

Alabama −31 vs. Western Kentucky

+3.9Covered by 32.0
2017

Oklahoma −43.17 vs. UTEP

+3.4Covered by 5.8
2025

Ole Miss −38.5 vs. Georgia State

+3.1Covered by 17.5
2019

Alabama −33.75 vs. Duke

+5.6Covered by 5.3
2022

Tennessee −36.25 vs. Ball State

+6.2Covered by 12.8
2018

Ohio State −40 vs. Oregon State

+3.3Covered by 6.0
2021

Fresno State −27.70 vs. UConn

+4.4Covered by 17.3
2023

Tennessee −28 vs. Virginia

+6.0Covered by 8.0
2021

Tennessee −37.30 vs. Bowling Green

+6.2Missed by 5.3
2019

Michigan −35.62 vs. Middle Tennessee

+2.4Missed by 16.6
2023

Texas A&M −38 vs. New Mexico

+7.2Covered by 4.0
2021

Auburn −37.30 vs. Akron

+5.0Covered by 12.7
2025

Penn State −45.5 vs. Nevada

+1.2Missed by 10.5
2022

Alabama −42.12 vs. Utah State

+1.8Covered by 12.9
2024

Georgia Tech −20.5 vs. Georgia State

+7.5Covered by 2.5
2021

Texas A&M −29.5 vs. Kent State

+3.9Covered by 1.5
2024

Oklahoma −42.5 vs. Temple

+0.5Covered by 5.5
2018

Alabama −23.75 vs. Louisville

+4.8Covered by 13.3
2017

Memphis −27.17 vs. UL Monroe

+2.0Missed by 19.2
2021

Wake Forest −32.5 vs. Old Dominion

+1.0Missed by 0.5
2021

Kentucky −31.10 vs. UL Monroe

+5.2Covered by 3.9
2019

Texas −19.38 vs. Louisiana Tech

+3.3Covered by 11.6
2021

Oklahoma −31.25 vs. Tulane

+2.0Missed by 26.3
2022

Michigan State −21.75 vs. Western Michigan

+3.3Covered by 0.3
08
Sat, Aug 29, 7:00 PM EDT

Hawai'i at Stanford

Best-book edge+7.1
Recommendation · Bovada

Hawai'i +5.5

Best book
Bovada · +5.5
Consensus line
+4.25
All-Takes line
−1.64

Evidence floor: 2.00 points. A 2.00-point evidence floor applies because at least one team changed head coaches. The coaching signal can cause an abstention; it never flips the model's side. The new pace adjustment is shown in the forecast but does not create a bet.

Core football model+4.1

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check0.0

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production+1.0

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions+0.7

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price+1.3

Bovada has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets21–9

70.0% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2021

Army +2.5 at Georgia State

+4.0Covered by 35.5
2022

Arizona +6.25 at San Diego State

+4.1Covered by 24.3
2022

Army +1.5 at Coastal Carolina

+3.2Missed by 8.5
2024

Sam Houston +10.5 at Rice

+6.0Covered by 30.5
2023

Northern Illinois +8.5 at Boston College

+3.0Covered by 11.5
2025

Stanford +1.5 at Hawai'i

+4.0Missed by 1.5
2019

SMU +1.75 at Arkansas State

+4.0Covered by 8.8
2025

UTEP +3.5 at Utah State

+3.5Missed by 8.5
2025

Rice +14.5 at Louisiana

+4.4Covered by 16.5
2022

Nevada +2 vs. Texas State

+6.9Covered by 26.0
2022

Temple +9.5 at Duke

+2.4Missed by 20.5
2018

Army +13.88 at Duke

+4.6Missed by 6.1
2025

LSU +3.5 at Clemson

+2.3Covered by 10.5
2023

Florida International +10 at Louisiana Tech

+5.5Covered by 5.0
2025

California +3 at Oregon State

+2.5Covered by 22.0
2022

Rutgers +9 at Boston College

+5.3Covered by 10.0
2018

San Diego State +13.75 at Stanford

+2.6Missed by 7.3
2018

Northwestern +1 at Purdue

+2.6Covered by 5.0
2022

West Virginia +7.5 at Pittsburgh

+4.3Covered by 0.5
2024

Eastern Michigan +2.5 at Massachusetts

+1.9Covered by 16.5
2019

Fresno State +14.12 at USC

+2.9Covered by 6.1
2017

Ball State +5.33 at Illinois

+4.7Covered by 2.3
2017

California +12.5 at North Carolina

+2.7Covered by 17.5
2018

LSU +3.12 vs. Miami

+5.1Covered by 19.1
2017

App State +12.5 at Georgia

+3.4Missed by 8.5
2024

West Virginia +10 vs. Penn State

+2.9Missed by 12.0
2017

Arkansas State +14.5 at Nebraska

+2.5Covered by 7.5
2023

Jacksonville State +1.17 vs. UTEP

+5.9Covered by 4.2
2024

Coastal Carolina +3.5 at Jacksonville State

+1.5Covered by 31.5
2017

Bowling Green +17.17 at Michigan State

+3.8Missed by 7.8
09
Sat, Aug 29, 10:00 PM EDT

Memphis at UNLV

Best-book edge+6.4
Recommendation · Bovada

Memphis +5.5

Best book
Bovada · +5.5
Consensus line
+4.25
All-Takes line
−0.87

Evidence floor: 2.00 points. A 2.00-point evidence floor applies because at least one team changed head coaches. The coaching signal can cause an abstention; it never flips the model's side. The new pace adjustment is shown in the forecast but does not create a bet.

Core football model+4.1

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check0.0

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production+0.8

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions+0.2

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price+1.3

Bovada has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets18–12

60.0% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2025

UTEP +3.5 at Utah State

+3.5Missed by 8.5
2019

SMU +1.75 at Arkansas State

+4.0Covered by 8.8
2022

Army +1.5 at Coastal Carolina

+3.2Missed by 8.5
2023

Northern Illinois +8.5 at Boston College

+3.0Covered by 11.5
2025

LSU +3.5 at Clemson

+2.3Covered by 10.5
2022

Arizona +6.25 at San Diego State

+4.1Covered by 24.3
2021

Army +2.5 at Georgia State

+4.0Covered by 35.5
2017

California +12.5 at North Carolina

+2.7Covered by 17.5
2022

Temple +9.5 at Duke

+2.4Missed by 20.5
2025

Stanford +1.5 at Hawai'i

+4.0Missed by 1.5
2023

Florida International +10 at Louisiana Tech

+5.5Covered by 5.0
2018

SMU +3.5 at North Texas

+3.2Missed by 19.5
2017

Troy +10.5 at Boise State

+1.9Missed by 0.5
2024

Coastal Carolina +3.5 at Jacksonville State

+1.5Covered by 31.5
2019

Florida International +3 at Tulane

+1.4Missed by 25.0
2024

Sam Houston +10.5 at Rice

+6.0Covered by 30.5
2018

Northwestern +1 at Purdue

+2.6Covered by 5.0
2022

Nevada +2 vs. Texas State

+6.9Covered by 26.0
2023

North Texas +6.5 vs. California

+2.1Missed by 30.5
2018

Cincinnati +14.25 at UCLA

+4.5Covered by 23.3
2017

Arkansas State +14.5 at Nebraska

+2.5Covered by 7.5
2024

North Carolina +2.5 at Minnesota

+1.3Covered by 4.5
2018

Kent State +18.12 at Illinois

+3.7Covered by 11.1
2020

South Alabama +12.40 at Southern Miss

+3.0Covered by 23.4
2022

Boise State +1.62 at Oregon State

+1.1Missed by 15.4
2024

West Virginia +10 vs. Penn State

+2.9Missed by 12.0
2018

San Diego State +13.75 at Stanford

+2.6Missed by 7.3
2022

Georgia State +12.5 at South Carolina

+1.6Missed by 8.5
2017

Texas A&M +6.33 at UCLA

+1.0Covered by 5.3
2024

Eastern Michigan +2.5 at Massachusetts

+1.9Covered by 16.5
10
Sat, Sep 5, 3:30 PM EDT · Neutral

Baylor vs. Auburn

Best-book edge+6.4
Recommendation · DraftKings

Auburn −6.5

Best book
DraftKings · −6.5
Consensus line
−6.75
All-Takes line
−12.93

Evidence floor: 2.00 points. A 2.00-point evidence floor applies because at least one team changed head coaches. The coaching signal can cause an abstention; it never flips the model's side. The new pace adjustment is shown in the forecast but does not create a bet.

Core football model+5.6

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check0.0

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production0.0

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions+0.6

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price+0.3

DraftKings has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets21–9

70.0% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2021

Kansas State −3.20 vs. Stanford

+4.2Covered by 13.8
2024

LSU −4 vs. USC

+5.3Missed by 11.0
2025

Tennessee −13.5 vs. Syracuse

+5.6Covered by 5.5
2019

Colorado −12.62 vs. Colorado State

+4.1Covered by 8.4
2022

Georgia −16 vs. Oregon

+7.0Covered by 30.0
2024

Georgia −14 vs. Clemson

+6.3Covered by 17.0
2017

Colorado −3 vs. Colorado State

+3.7Covered by 11.0
2017

Alabama −7.33 vs. Florida State

+7.6Covered by 9.7
2017

LSU −14.67 vs. BYU

+4.4Covered by 12.3
2018

Auburn −1.5 vs. Washington

+4.0Covered by 3.5
2025

Kansas State −3.5 vs. Iowa State

+5.2Missed by 6.5
2018

Alabama −23.75 vs. Louisville

+4.8Covered by 13.3
2021

West Virginia −2.5 at Maryland

+4.3Missed by 8.5
2019

Florida −7 vs. Miami

+3.1Missed by 3.0
2017

Colorado State −3.5 vs. Oregon State

+3.9Covered by 27.5
2022

Utah −2.5 at Florida

+3.9Missed by 5.5
2025

Notre Dame −2.5 at Miami

+4.8Missed by 5.5
2025

South Carolina −8.5 vs. Virginia Tech

+2.4Covered by 4.5
2019

Auburn −3.62 vs. Oregon

+1.0Covered by 2.4
2025

TCU −3 at North Carolina

+3.2Covered by 31.0
2023

Notre Dame −20 vs. Navy

+3.3Covered by 19.0
2024

Georgia Tech −20.5 vs. Georgia State

+7.5Covered by 2.5
2017

Virginia Tech −4.67 vs. West Virginia

+1.9Covered by 2.3
2019

Texas −19.38 vs. Louisiana Tech

+3.3Covered by 11.6
2024

Northwestern −4 vs. Miami (OH)

+2.5Covered by 3.0
2018

Wyoming −5 at New Mexico State

+3.2Covered by 17.0
2021

Notre Dame −7.30 at Florida State

+2.9Missed by 4.3
2021

Duke −6.40 at Charlotte

+2.9Missed by 9.4
2023

Tennessee −28 vs. Virginia

+6.0Covered by 8.0
2022

Florida Atlantic −6 at Ohio

+2.8Missed by 9.0
11
Sat, Aug 29, 12:00 PM EDT · Neutral

North Carolina vs. TCU

Best-book edge+6.5
Recommendation · DraftKings

TCU −6.5

Best book
DraftKings · −6.5
Consensus line
−7
All-Takes line
−12.97

Evidence gate: intentional model departure. Neither team changed head coaches, so an intentional pre-pace production-model departure is sufficient. The pace adjustment changes the forecast, not whether a bet exists.

Core football model+5.3

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check0.0

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production0.0

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions+0.6

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price+0.5

DraftKings has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets22–8

73.3% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2025

Tennessee −13.5 vs. Syracuse

+5.6Covered by 5.5
2024

Georgia −14 vs. Clemson

+6.3Covered by 17.0
2021

Kansas State −3.20 vs. Stanford

+4.2Covered by 13.8
2017

LSU −14.67 vs. BYU

+4.4Covered by 12.3
2019

Colorado −12.62 vs. Colorado State

+4.1Covered by 8.4
2022

Georgia −16 vs. Oregon

+7.0Covered by 30.0
2017

Alabama −7.33 vs. Florida State

+7.6Covered by 9.7
2018

Auburn −1.5 vs. Washington

+4.0Covered by 3.5
2019

Florida −7 vs. Miami

+3.1Missed by 3.0
2025

Kansas State −3.5 vs. Iowa State

+5.2Missed by 6.5
2024

Northwestern −4 vs. Miami (OH)

+2.5Covered by 3.0
2024

LSU −4 vs. USC

+5.3Missed by 11.0
2025

South Carolina −8.5 vs. Virginia Tech

+2.4Covered by 4.5
2023

Notre Dame −20 vs. Navy

+3.3Covered by 19.0
2022

Utah −2.5 at Florida

+3.9Missed by 5.5
2025

Notre Dame −2.5 at Miami

+4.8Missed by 5.5
2018

Wyoming −5 at New Mexico State

+3.2Covered by 17.0
2021

West Virginia −2.5 at Maryland

+4.3Missed by 8.5
2019

Auburn −3.62 vs. Oregon

+1.0Covered by 2.4
2017

Colorado −3 vs. Colorado State

+3.7Covered by 11.0
2024

James Madison −7.5 at Charlotte

+2.6Covered by 15.5
2025

Alabama −13.5 at Florida State

+3.0Missed by 27.5
2018

LSU +3.12 vs. Miami

+5.1Covered by 19.1
2019

Utah −6.12 at BYU

+3.8Covered by 11.9
2017

Virginia Tech −4.67 vs. West Virginia

+1.9Covered by 2.3
2018

Alabama −23.75 vs. Louisville

+4.8Covered by 13.3
2022

Florida Atlantic −6 at Ohio

+2.8Missed by 9.0
2025

TCU −3 at North Carolina

+3.2Covered by 31.0
2022

Illinois −14 vs. Wyoming

+2.9Covered by 18.0
2018

Iowa −9.25 vs. Northern Illinois

+1.8Covered by 16.8
12
Sat, Sep 5, 7:00 PM EDT

Missouri State at Texas A&M

Best-book edge+5.7
Recommendation · DraftKings

Texas A&M −38.5

Best book
DraftKings · −38.5
Consensus line
−39
All-Takes line
−44.15

Evidence floor: 2.00 points. A 2.00-point evidence floor applies because at least one team changed head coaches. The coaching signal can cause an abstention; it never flips the model's side. The new pace adjustment is shown in the forecast but does not create a bet.

Core football model+4.1

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check0.0

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production+1.1

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions0.0

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price+0.5

DraftKings has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets21–9

70.0% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2017

Clemson −38.17 vs. Kent State

+3.8Covered by 14.8
2022

Minnesota −36 vs. New Mexico State

+5.1Covered by 2.0
2023

Auburn −35 vs. Massachusetts

+3.6Covered by 10.0
2024

Alabama −31 vs. Western Kentucky

+3.9Covered by 32.0
2025

Ole Miss −38.5 vs. Georgia State

+3.1Covered by 17.5
2023

Alabama −39.5 vs. Middle Tennessee

+7.6Covered by 9.5
2017

Oklahoma −43.17 vs. UTEP

+3.4Covered by 5.8
2023

Texas −35 vs. Rice

+6.9Missed by 8.0
2018

Ohio State −40 vs. Oregon State

+3.3Covered by 6.0
2019

Texas A&M −33.25 vs. Texas State

+7.0Covered by 0.8
2021

Fresno State −27.70 vs. UConn

+4.4Covered by 17.3
2019

Michigan −35.62 vs. Middle Tennessee

+2.4Missed by 16.6
2019

Alabama −33.75 vs. Duke

+5.6Covered by 5.3
2022

Alabama −42.12 vs. Utah State

+1.8Covered by 12.9
2025

Penn State −45.5 vs. Nevada

+1.2Missed by 10.5
2022

Tennessee −36.25 vs. Ball State

+6.2Covered by 12.8
2021

Texas A&M −29.5 vs. Kent State

+3.9Covered by 1.5
2024

Oklahoma −42.5 vs. Temple

+0.5Covered by 5.5
2021

Wake Forest −32.5 vs. Old Dominion

+1.0Missed by 0.5
2021

Pittsburgh −37.60 vs. Massachusetts

+0.2Covered by 6.4
2021

Auburn −37.30 vs. Akron

+5.0Covered by 12.7
2017

Memphis −27.17 vs. UL Monroe

+2.0Missed by 19.2
2021

Oklahoma −31.25 vs. Tulane

+2.0Missed by 26.3
2017

Wisconsin −27.33 vs. Utah State

+0.9Covered by 21.7
2023

Tennessee −28 vs. Virginia

+6.0Covered by 8.0
2021

Tennessee −37.30 vs. Bowling Green

+6.2Missed by 5.3
2024

Kentucky −25.5 vs. Southern Miss

+0.9Covered by 5.5
2022

Michigan State −21.75 vs. Western Michigan

+3.3Covered by 0.3
2022

Maryland −24 vs. Buffalo

+3.6Missed by 3.0
2019

Army −23 vs. Rice

+1.7Missed by 16.0
13
Sun, Sep 6, 7:30 PM EDT · Neutral

Wisconsin vs. Notre Dame

Best-book edge+5.6
Recommendation · Bovada

Notre Dame −20.5

Best book
Bovada · −20.5
Consensus line
−20.5
All-Takes line
−26.14

Evidence gate: intentional model departure. Neither team changed head coaches, so an intentional pre-pace production-model departure is sufficient. The pace adjustment changes the forecast, not whether a bet exists.

Core football model+5.1

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check+0.5

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production0.0

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions0.0

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price0.0

Bovada has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets26–4

86.7% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2017

LSU −14.67 vs. BYU

+4.4Covered by 12.3
2024

Georgia −14 vs. Clemson

+6.3Covered by 17.0
2025

Tennessee −13.5 vs. Syracuse

+5.6Covered by 5.5
2023

Notre Dame −20 vs. Navy

+3.3Covered by 19.0
2022

Georgia −16 vs. Oregon

+7.0Covered by 30.0
2023

Tennessee −28 vs. Virginia

+6.0Covered by 8.0
2019

Colorado −12.62 vs. Colorado State

+4.1Covered by 8.4
2018

Alabama −23.75 vs. Louisville

+4.8Covered by 13.3
2019

Alabama −33.75 vs. Duke

+5.6Covered by 5.3
2024

Georgia Tech −20.5 vs. Georgia State

+7.5Covered by 2.5
2019

Texas −19.38 vs. Louisiana Tech

+3.3Covered by 11.6
2022

Michigan State −21.75 vs. Western Michigan

+3.3Covered by 0.3
2017

Alabama −7.33 vs. Florida State

+7.6Covered by 9.7
2025

Purdue −16.5 vs. Ball State

+4.5Covered by 14.5
2019

Florida −7 vs. Miami

+3.1Missed by 3.0
2021

Kansas State −3.20 vs. Stanford

+4.2Covered by 13.8
2021

Cincinnati −22.30 vs. Miami (OH)

+2.2Covered by 12.7
2025

Alabama −13.5 at Florida State

+3.0Missed by 27.5
2022

Illinois −14 vs. Wyoming

+2.9Covered by 18.0
2019

Army −23 vs. Rice

+1.7Missed by 16.0
2021

Arkansas −19.60 vs. Rice

+2.8Covered by 1.4
2025

South Carolina −8.5 vs. Virginia Tech

+2.4Covered by 4.5
2024

Maryland −20.5 vs. UConn

+1.4Covered by 22.5
2022

Minnesota −36 vs. New Mexico State

+5.1Covered by 2.0
2024

Syracuse −17.5 vs. Ohio

+2.9Missed by 1.5
2024

Northwestern −4 vs. Miami (OH)

+2.5Covered by 3.0
2019

Texas A&M −33.25 vs. Texas State

+7.0Covered by 0.8
2018

Auburn −1.5 vs. Washington

+4.0Covered by 3.5
2024

Alabama −31 vs. Western Kentucky

+3.9Covered by 32.0
2023

Auburn −35 vs. Massachusetts

+3.6Covered by 10.0
14
Fri, Sep 4, 9:00 PM EDT

Miami at Stanford

Best-book edge+5.4
Recommendation · DraftKings

Miami −21.5

Best book
DraftKings · −21.5
Consensus line
−22.5
All-Takes line
−26.94

Evidence floor: 2.00 points. A 2.00-point evidence floor applies because at least one team changed head coaches. The coaching signal can cause an abstention; it never flips the model's side. The new pace adjustment is shown in the forecast but does not create a bet.

Core football model+2.8

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check0.0

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production+1.7

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions0.0

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price+1.0

DraftKings has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets18–12

60.0% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2025

Purdue −16.5 vs. Ball State

+4.5Covered by 14.5
2024

Syracuse −17.5 vs. Ohio

+2.9Missed by 1.5
2021

Mississippi State −20.90 vs. Louisiana Tech

+2.9Missed by 19.9
2021

Baylor −13.60 at Texas State

+5.3Missed by 4.6
2025

Arizona −13.5 vs. Hawai'i

+3.5Covered by 20.5
2025

Minnesota −16.5 vs. Buffalo

+2.4Missed by 3.5
2025

Kansas −14 vs. Fresno State

+2.3Covered by 10.0
2025

Texas State −14 vs. Eastern Michigan

+4.5Covered by 11.0
2021

NC State −19.70 vs. South Florida

+2.7Covered by 25.3
2025

Wisconsin −17.5 vs. Miami (OH)

+2.0Missed by 0.5
2022

Illinois −14 vs. Wyoming

+2.9Covered by 18.0
2024

Virginia Tech −13.5 at Vanderbilt

+1.7Missed by 20.5
2025

Alabama −13.5 at Florida State

+3.0Missed by 27.5
2024

Michigan State −11.5 vs. Florida Atlantic

+2.6Missed by 5.5
2019

Texas −19.38 vs. Louisiana Tech

+3.3Covered by 11.6
2024

Pittsburgh −23.5 vs. Kent State

+1.2Covered by 7.5
2025

Rutgers −11.5 vs. Ohio

+2.9Missed by 8.5
2021

Fresno State −27.70 vs. UConn

+4.4Covered by 17.3
2019

Army −23 vs. Rice

+1.7Missed by 16.0
2025

Washington −20.5 vs. Colorado State

+0.9Missed by 3.5
2025

Mississippi State −11.5 at Southern Miss

+2.2Covered by 5.5
2021

Cincinnati −22.30 vs. Miami (OH)

+2.2Covered by 12.7
2022

Kentucky −15.5 vs. Miami (OH)

+2.4Covered by 8.5
2022

Michigan State −21.75 vs. Western Michigan

+3.3Covered by 0.3
2024

Maryland −20.5 vs. UConn

+1.4Covered by 22.5
2019

Utah −6.12 at BYU

+3.8Covered by 11.9
2024

Indiana −20.5 vs. Florida International

+5.0Covered by 3.5
2024

Alabama −31 vs. Western Kentucky

+3.9Covered by 32.0
2024

Kentucky −25.5 vs. Southern Miss

+0.9Covered by 5.5
2025

Texas A&M −21.5 vs. UTSA

+4.8Missed by 3.5
15
Sat, Sep 5, 3:30 PM EDT

Boise State at Oregon

Best-book edge+5.1
Recommendation · Bovada

Boise State +24.5

Best book
Bovada · +24.5
Consensus line
+24.5
All-Takes line
+19.36

Evidence gate: intentional model departure. Neither team changed head coaches, so an intentional pre-pace production-model departure is sufficient. The pace adjustment changes the forecast, not whether a bet exists.

Core football model+2.4

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check+2.2

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production+0.5

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions0.0

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price0.0

Bovada has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets23–7

76.7% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2022

UConn +23.75 at Utah State

+2.9Covered by 12.8
2025

Jacksonville State +16.5 at UCF

+3.0Covered by 9.5
2025

Western Michigan +18.5 at Michigan State

+2.9Covered by 1.5
2021

New Mexico State +31.20 at San Diego State

+2.2Covered by 13.2
2018

Kent State +18.12 at Illinois

+3.7Covered by 11.1
2022

Bowling Green +24 at UCLA

+2.3Missed by 4.0
2019

Kent State +24.5 at Arizona State

+2.8Covered by 1.5
2024

Kennesaw State +24 at UTSA

+2.2Covered by 12.0
2018

Army +13.88 at Duke

+4.6Missed by 6.1
2023

Colorado +20.5 at TCU

+3.9Covered by 23.5
2019

Florida Atlantic +27.62 at Ohio State

+3.5Covered by 3.6
2018

New Mexico State +21.5 at Minnesota

+1.5Missed by 16.5
2024

Boston College +16.5 at Florida State

+3.7Covered by 31.5
2020

South Alabama +12.40 at Southern Miss

+3.0Covered by 23.4
2022

Memphis +17 at Mississippi State

+2.8Missed by 9.0
2024

Western Michigan +24 at Wisconsin

+2.8Covered by 10.0
2018

Massachusetts +20.38 at Boston College

+2.5Missed by 13.6
2024

Old Dominion +20 at South Carolina

+2.5Covered by 16.0
2019

Fresno State +14.12 at USC

+2.9Covered by 6.1
2018

Utah State +23.5 at Michigan State

+2.6Covered by 16.5
2024

Nevada +15.5 at Troy

+2.1Covered by 17.5
2018

Hawai'i +17.12 at Colorado State

+3.2Covered by 26.1
2018

Rice +25.88 vs. Houston

+2.3Covered by 7.9
2018

Texas State +16.5 at Rutgers

+1.0Missed by 11.5
2023

Utah State +23.5 at Iowa

+4.4Covered by 13.5
2019

Miami (OH) +24.25 at Iowa

+1.2Covered by 0.3
2019

Georgia Southern +27.12 at LSU

+2.1Missed by 24.9
2017

Rutgers +28 vs. Washington

+4.4Covered by 12.0
2018

Central Michigan +17.5 at Kentucky

+1.3Covered by 2.5
2025

Rice +14.5 at Louisiana

+4.4Covered by 16.5
16
Sat, Sep 5, 6:00 PM EDT

Wyoming at Colorado State

Best-book edge+5.1
Recommendation · DraftKings

Wyoming +4

Best book
DraftKings · +4
Consensus line
+3.75
All-Takes line
−1.09

Evidence floor: 2.00 points. A 2.00-point evidence floor applies because at least one team changed head coaches. The coaching signal can cause an abstention; it never flips the model's side. The new pace adjustment is shown in the forecast but does not create a bet.

Core football model+3.1

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check+1.2

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production0.0

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions+0.6

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price+0.3

DraftKings has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets18–12

60.0% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2022

Arizona +6.25 at San Diego State

+4.1Covered by 24.3
2019

SMU +1.75 at Arkansas State

+4.0Covered by 8.8
2018

Army +13.88 at Duke

+4.6Missed by 6.1
2025

Stanford +1.5 at Hawai'i

+4.0Missed by 1.5
2023

Northern Illinois +8.5 at Boston College

+3.0Covered by 11.5
2025

UTEP +3.5 at Utah State

+3.5Missed by 8.5
2025

Northwestern +6.5 at Tulane

+1.7Missed by 13.5
2022

Temple +9.5 at Duke

+2.4Missed by 20.5
2022

Army +1.5 at Coastal Carolina

+3.2Missed by 8.5
2023

Jacksonville State +1.17 vs. UTEP

+5.9Covered by 4.2
2025

Toledo +10 at Kentucky

+1.5Covered by 2.0
2024

West Virginia +10 vs. Penn State

+2.9Missed by 12.0
2021

East Carolina +9.20 at App State

+2.2Missed by 4.8
2018

Northwestern +1 at Purdue

+2.6Covered by 5.0
2024

North Carolina +2.5 at Minnesota

+1.3Covered by 4.5
2025

Jacksonville State +16.5 at UCF

+3.0Covered by 9.5
2022

Nevada +2 vs. Texas State

+6.9Covered by 26.0
2022

North Carolina +3 at App State

+3.1Covered by 5.0
2022

East Carolina +12.5 vs. NC State

+2.8Covered by 11.5
2021

Army +2.5 at Georgia State

+4.0Covered by 35.5
2023

North Texas +6.5 vs. California

+2.1Missed by 30.5
2020

South Alabama +12.40 at Southern Miss

+3.0Covered by 23.4
2022

Boise State +1.62 at Oregon State

+1.1Missed by 15.4
2017

Wyoming +12.17 at Iowa

+1.6Missed by 8.8
2019

Fresno State +14.12 at USC

+2.9Covered by 6.1
2025

LSU +3.5 at Clemson

+2.3Covered by 10.5
2017

Arkansas State +14.5 at Nebraska

+2.5Covered by 7.5
2025

California +3 at Oregon State

+2.5Covered by 22.0
2018

Troy +9.12 vs. Boise State

+1.3Missed by 26.9
2024

Coastal Carolina +3.5 at Jacksonville State

+1.5Covered by 31.5
17
Fri, Sep 4, 8:00 PM EDT

Toledo at Michigan State

Best-book edge+4.4
Recommendation · Bovada

Michigan State −10.5

Best book
Bovada · −10.5
Consensus line
−10.5
All-Takes line
−14.89

Evidence floor: 2.00 points. A 2.00-point evidence floor applies because at least one team changed head coaches. The coaching signal can cause an abstention; it never flips the model's side. The new pace adjustment is shown in the forecast but does not create a bet.

Core football model+3.5

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check+0.9

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production0.0

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions0.0

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price0.0

Bovada has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets20–10

66.7% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2018

Iowa −9.25 vs. Northern Illinois

+1.8Covered by 16.8
2025

Alabama −13.5 at Florida State

+3.0Missed by 27.5
2022

Florida Atlantic −6 at Ohio

+2.8Missed by 9.0
2025

Rutgers −11.5 vs. Ohio

+2.9Missed by 8.5
2025

Notre Dame −2.5 at Miami

+4.8Missed by 5.5
2022

Illinois −14 vs. Wyoming

+2.9Covered by 18.0
2024

James Madison −7.5 at Charlotte

+2.6Covered by 15.5
2018

Wyoming −5 at New Mexico State

+3.2Covered by 17.0
2019

Texas −19.38 vs. Louisiana Tech

+3.3Covered by 11.6
2025

Purdue −16.5 vs. Ball State

+4.5Covered by 14.5
2021

Rutgers −14.20 vs. Temple

+1.5Covered by 32.8
2021

Arkansas −19.60 vs. Rice

+2.8Covered by 1.4
2022

Michigan State −21.75 vs. Western Michigan

+3.3Covered by 0.3
2019

Stanford −6.5 vs. Northwestern

+1.1Covered by 3.5
2022

Utah −2.5 at Florida

+3.9Missed by 5.5
2019

Utah −6.12 at BYU

+3.8Covered by 11.9
2025

TCU −3 at North Carolina

+3.2Covered by 31.0
2022

Arkansas −6.62 vs. Cincinnati

+1.5Covered by 0.4
2022

Vanderbilt −9.25 at Hawai'i

+2.0Covered by 43.8
2017

Colorado State −3.5 vs. Oregon State

+3.9Covered by 27.5
2025

Wyoming −8.5 at Akron

+1.6Covered by 1.5
2025

Ohio State −3 vs. Texas

+1.4Covered by 4.0
2022

Kentucky −15.5 vs. Miami (OH)

+2.4Covered by 8.5
2021

Duke −6.40 at Charlotte

+2.9Missed by 9.4
2021

Notre Dame −7.30 at Florida State

+2.9Missed by 4.3
2024

Virginia Tech −13.5 at Vanderbilt

+1.7Missed by 20.5
2021

Cincinnati −22.30 vs. Miami (OH)

+2.2Covered by 12.7
2024

Michigan State −11.5 vs. Florida Atlantic

+2.6Missed by 5.5
2021

West Virginia −2.5 at Maryland

+4.3Missed by 8.5
2025

Utah −6.5 at UCLA

+1.3Covered by 26.5
18
Sat, Sep 5, 12:00 PM EDT

North Texas at Indiana

Best-book edge+4.3
Recommendation · Bovada

North Texas +41

Best book
Bovada · +41
Consensus line
+40.75
All-Takes line
+36.72

Evidence floor: 2.00 points. A 2.00-point evidence floor applies because at least one team changed head coaches. The coaching signal can cause an abstention; it never flips the model's side. The new pace adjustment is shown in the forecast but does not create a bet.

Core football model+3.3

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check+0.8

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production0.0

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions0.0

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price+0.3

Bovada has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets21–9

70.0% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2024

Akron +50.5 at Ohio State

+2.8Covered by 4.5
2017

Georgia Southern +34.33 at Auburn

+1.8Covered by 0.3
2024

Colorado State +35.5 at Texas

+1.4Missed by 16.5
2021

New Mexico State +31.20 at San Diego State

+2.2Covered by 13.2
2023

San José State +30.5 at USC

+3.3Covered by 2.5
2019

New Mexico State +32.88 at Washington State

+1.5Missed by 18.1
2018

Coastal Carolina +30.75 at South Carolina

+2.4Missed by 3.3
2019

South Alabama +35.5 at Nebraska

+1.1Covered by 21.5
2024

Nevada +27 vs. SMU

+3.0Covered by 22.0
2019

Florida Atlantic +27.62 at Ohio State

+3.5Covered by 3.6
2017

Rutgers +28 vs. Washington

+4.4Covered by 12.0
2024

Western Michigan +24 at Wisconsin

+2.8Covered by 10.0
2018

Rice +25.88 vs. Houston

+2.3Covered by 7.9
2019

Kent State +24.5 at Arizona State

+2.8Covered by 1.5
2022

UConn +23.75 at Utah State

+2.9Covered by 12.8
2017

Akron +29.83 at Penn State

+2.6Missed by 22.2
2019

Georgia Southern +27.12 at LSU

+2.1Missed by 24.9
2022

Bowling Green +24 at UCLA

+2.3Missed by 4.0
2017

Rice +31.17 vs. Stanford

+5.0Missed by 23.8
2024

Kennesaw State +24 at UTSA

+2.2Covered by 12.0
2019

Miami (OH) +24.25 at Iowa

+1.2Covered by 0.3
2023

Colorado +20.5 at TCU

+3.9Covered by 23.5
2023

Utah State +23.5 at Iowa

+4.4Covered by 13.5
2022

Colorado State +31 at Michigan

+4.9Missed by 13.0
2018

Utah State +23.5 at Michigan State

+2.6Covered by 16.5
2025

Old Dominion +23.5 at Indiana

+0.5Covered by 10.5
2018

Massachusetts +20.38 at Boston College

+2.5Missed by 13.6
2018

UNLV +24.62 at USC

+1.7Covered by 2.6
2024

Old Dominion +20 at South Carolina

+2.5Covered by 16.0
2021

Florida Atlantic +23.5 at Florida

+0.2Covered by 2.5
19
Sat, Sep 5, 3:30 PM EDT

Texas State at Texas

Best-book edge+3.8
Recommendation · Bovada

Texas −30.5

Best book
Bovada · −30.5
Consensus line
−30.5
All-Takes line
−34.33

Evidence gate: intentional model departure. Neither team changed head coaches, so an intentional pre-pace production-model departure is sufficient. The pace adjustment changes the forecast, not whether a bet exists.

Core football model+3.6

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check0.0

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production+0.3

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions0.0

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price0.0

Bovada has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets22–8

73.3% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2024

Alabama −31 vs. Western Kentucky

+3.9Covered by 32.0
2021

Fresno State −27.70 vs. UConn

+4.4Covered by 17.3
2017

Memphis −27.17 vs. UL Monroe

+2.0Missed by 19.2
2023

Auburn −35 vs. Massachusetts

+3.6Covered by 10.0
2025

Ole Miss −38.5 vs. Georgia State

+3.1Covered by 17.5
2021

Texas A&M −29.5 vs. Kent State

+3.9Covered by 1.5
2022

Maryland −24 vs. Buffalo

+3.6Missed by 3.0
2022

Minnesota −36 vs. New Mexico State

+5.1Covered by 2.0
2017

Clemson −38.17 vs. Kent State

+3.8Covered by 14.8
2021

Oklahoma −31.25 vs. Tulane

+2.0Missed by 26.3
2018

Ohio State −40 vs. Oregon State

+3.3Covered by 6.0
2021

Wake Forest −32.5 vs. Old Dominion

+1.0Missed by 0.5
2019

Texas −19.38 vs. Louisiana Tech

+3.3Covered by 11.6
2022

Washington −23.62 vs. Kent State

+2.3Covered by 1.4
2019

Michigan −35.62 vs. Middle Tennessee

+2.4Missed by 16.6
2017

Oklahoma −43.17 vs. UTEP

+3.4Covered by 5.8
2022

Michigan State −21.75 vs. Western Michigan

+3.3Covered by 0.3
2019

Texas A&M −33.25 vs. Texas State

+7.0Covered by 0.8
2017

Wisconsin −27.33 vs. Utah State

+0.9Covered by 21.7
2022

Tennessee −36.25 vs. Ball State

+6.2Covered by 12.8
2022

Missouri −20.38 vs. Louisiana Tech

+2.8Covered by 7.6
2022

Alabama −42.12 vs. Utah State

+1.8Covered by 12.9
2024

Pittsburgh −23.5 vs. Kent State

+1.2Covered by 7.5
2018

Penn State −24.5 vs. App State

+1.3Missed by 17.5
2024

Kentucky −25.5 vs. Southern Miss

+0.9Covered by 5.5
2018

Alabama −23.75 vs. Louisville

+4.8Covered by 13.3
2019

Alabama −33.75 vs. Duke

+5.6Covered by 5.3
2021

Cincinnati −22.30 vs. Miami (OH)

+2.2Covered by 12.7
2019

Army −23 vs. Rice

+1.7Missed by 16.0
2022

Ohio State −17 vs. Notre Dame

+1.5Missed by 6.0
20
Thu, Sep 3, 8:00 PM EDT

Colorado at Georgia Tech

Best-book edge+4.1
Recommendation · Bovada

Colorado +7

Best book
Bovada · +7
Consensus line
+7
All-Takes line
+2.86

Evidence gate: intentional model departure. Neither team changed head coaches, so an intentional pre-pace production-model departure is sufficient. The pace adjustment changes the forecast, not whether a bet exists.

Core football model+3.7

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check0.0

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production0.0

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions+0.4

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price0.0

Bovada has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets14–16

46.7% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2023

Northern Illinois +8.5 at Boston College

+3.0Covered by 11.5
2022

Arizona +6.25 at San Diego State

+4.1Covered by 24.3
2022

Temple +9.5 at Duke

+2.4Missed by 20.5
2025

UTEP +3.5 at Utah State

+3.5Missed by 8.5
2019

SMU +1.75 at Arkansas State

+4.0Covered by 8.8
2018

Army +13.88 at Duke

+4.6Missed by 6.1
2024

West Virginia +10 vs. Penn State

+2.9Missed by 12.0
2017

California +12.5 at North Carolina

+2.7Covered by 17.5
2017

Arkansas State +14.5 at Nebraska

+2.5Covered by 7.5
2017

Wyoming +12.17 at Iowa

+1.6Missed by 8.8
2022

East Carolina +12.5 vs. NC State

+2.8Covered by 11.5
2024

North Carolina +2.5 at Minnesota

+1.3Covered by 4.5
2018

Northwestern +1 at Purdue

+2.6Covered by 5.0
2017

Troy +10.5 at Boise State

+1.9Missed by 0.5
2022

Army +1.5 at Coastal Carolina

+3.2Missed by 8.5
2023

North Texas +6.5 vs. California

+2.1Missed by 30.5
2019

Massachusetts +16.67 at Rutgers

+2.3Missed by 10.3
2025

Northwestern +6.5 at Tulane

+1.7Missed by 13.5
2019

Florida International +3 at Tulane

+1.4Missed by 25.0
2022

Georgia State +12.5 at South Carolina

+1.6Missed by 8.5
2018

San Diego State +13.75 at Stanford

+2.6Missed by 7.3
2025

LSU +3.5 at Clemson

+2.3Covered by 10.5
2022

Boise State +1.62 at Oregon State

+1.1Missed by 15.4
2022

Memphis +17 at Mississippi State

+2.8Missed by 9.0
2025

Toledo +10 at Kentucky

+1.5Covered by 2.0
2024

Coastal Carolina +3.5 at Jacksonville State

+1.5Covered by 31.5
2021

Central Michigan +13.80 at Missouri

+1.7Covered by 3.8
2025

Jacksonville State +16.5 at UCF

+3.0Covered by 9.5
2017

Charlotte +14 at Eastern Michigan

+1.4Missed by 3.0
2025

Western Michigan +18.5 at Michigan State

+2.9Covered by 1.5
21
Sat, Sep 5, 7:30 PM EDT

Western Michigan at Michigan

Best-book edge+3.7
Recommendation · Bovada

Michigan −26.5

Best book
Bovada · −26.5
Consensus line
−26.5
All-Takes line
−30.21

Evidence floor: 2.00 points. A 2.00-point evidence floor applies because at least one team changed head coaches. The coaching signal can cause an abstention; it never flips the model's side. The new pace adjustment is shown in the forecast but does not create a bet.

Core football model+3.1

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check0.0

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production+0.6

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions0.0

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price0.0

Bovada has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets18–12

60.0% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2019

Army −23 vs. Rice

+1.7Missed by 16.0
2023

Auburn −35 vs. Massachusetts

+3.6Covered by 10.0
2021

Cincinnati −22.30 vs. Miami (OH)

+2.2Covered by 12.7
2019

Texas −19.38 vs. Louisiana Tech

+3.3Covered by 11.6
2024

Kentucky −25.5 vs. Southern Miss

+0.9Covered by 5.5
2024

Maryland −20.5 vs. UConn

+1.4Covered by 22.5
2017

Wisconsin −27.33 vs. Utah State

+0.9Covered by 21.7
2024

Syracuse −17.5 vs. Ohio

+2.9Missed by 1.5
2022

Michigan State −21.75 vs. Western Michigan

+3.3Covered by 0.3
2024

Alabama −31 vs. Western Kentucky

+3.9Covered by 32.0
2024

Pittsburgh −23.5 vs. Kent State

+1.2Covered by 7.5
2025

Purdue −16.5 vs. Ball State

+4.5Covered by 14.5
2017

Clemson −38.17 vs. Kent State

+3.8Covered by 14.8
2022

Minnesota −36 vs. New Mexico State

+5.1Covered by 2.0
2025

Washington −20.5 vs. Colorado State

+0.9Missed by 3.5
2018

Penn State −24.5 vs. App State

+1.3Missed by 17.5
2022

Illinois −14 vs. Wyoming

+2.9Covered by 18.0
2019

Michigan State −23.12 vs. Tulsa

+1.4Missed by 2.1
2025

Minnesota −16.5 vs. Buffalo

+2.4Missed by 3.5
2021

Fresno State −27.70 vs. UConn

+4.4Covered by 17.3
2025

Wisconsin −17.5 vs. Miami (OH)

+2.0Missed by 0.5
2021

Arkansas −19.60 vs. Rice

+2.8Covered by 1.4
2017

Memphis −27.17 vs. UL Monroe

+2.0Missed by 19.2
2021

Mississippi State −20.90 vs. Louisiana Tech

+2.9Missed by 19.9
2025

Kansas −14 vs. Fresno State

+2.3Covered by 10.0
2021

NC State −19.70 vs. South Florida

+2.7Covered by 25.3
2019

Michigan −35.62 vs. Middle Tennessee

+2.4Missed by 16.6
2025

Alabama −13.5 at Florida State

+3.0Missed by 27.5
2022

Ohio State −17 vs. Notre Dame

+1.5Missed by 6.0
2022

Washington −23.62 vs. Kent State

+2.3Covered by 1.4
22
Sat, Sep 5, 12:00 PM EDT

Ohio at Nebraska

Best-book edge+3.5
Recommendation · DraftKings

Nebraska −23.5

Best book
DraftKings · −23.5
Consensus line
−23.75
All-Takes line
−26.97

Evidence gate: intentional model departure. Neither team changed head coaches, so an intentional pre-pace production-model departure is sufficient. The pace adjustment changes the forecast, not whether a bet exists.

Core football model+3.2

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check0.0

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production0.0

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions0.0

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price+0.3

DraftKings has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets19–11

63.3% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2019

Army −23 vs. Rice

+1.7Missed by 16.0
2024

Maryland −20.5 vs. UConn

+1.4Covered by 22.5
2019

Texas −19.38 vs. Louisiana Tech

+3.3Covered by 11.6
2021

Cincinnati −22.30 vs. Miami (OH)

+2.2Covered by 12.7
2024

Kentucky −25.5 vs. Southern Miss

+0.9Covered by 5.5
2022

Michigan State −21.75 vs. Western Michigan

+3.3Covered by 0.3
2017

Wisconsin −27.33 vs. Utah State

+0.9Covered by 21.7
2018

Penn State −24.5 vs. App State

+1.3Missed by 17.5
2025

Washington −20.5 vs. Colorado State

+0.9Missed by 3.5
2023

Auburn −35 vs. Massachusetts

+3.6Covered by 10.0
2022

Illinois −14 vs. Wyoming

+2.9Covered by 18.0
2019

Michigan State −23.12 vs. Tulsa

+1.4Missed by 2.1
2024

Pittsburgh −23.5 vs. Kent State

+1.2Covered by 7.5
2024

Syracuse −17.5 vs. Ohio

+2.9Missed by 1.5
2021

Arkansas −19.60 vs. Rice

+2.8Covered by 1.4
2024

Alabama −31 vs. Western Kentucky

+3.9Covered by 32.0
2025

Alabama −13.5 at Florida State

+3.0Missed by 27.5
2022

Ohio State −17 vs. Notre Dame

+1.5Missed by 6.0
2025

Purdue −16.5 vs. Ball State

+4.5Covered by 14.5
2025

Minnesota −16.5 vs. Buffalo

+2.4Missed by 3.5
2022

Minnesota −36 vs. New Mexico State

+5.1Covered by 2.0
2017

Memphis −27.17 vs. UL Monroe

+2.0Missed by 19.2
2023

Notre Dame −20 vs. Navy

+3.3Covered by 19.0
2017

Clemson −38.17 vs. Kent State

+3.8Covered by 14.8
2025

Wisconsin −17.5 vs. Miami (OH)

+2.0Missed by 0.5
2022

Washington −23.62 vs. Kent State

+2.3Covered by 1.4
2025

Kansas −14 vs. Fresno State

+2.3Covered by 10.0
2021

Rutgers −14.20 vs. Temple

+1.5Covered by 32.8
2018

Iowa −9.25 vs. Northern Illinois

+1.8Covered by 16.8
2024

Virginia Tech −13.5 at Vanderbilt

+1.7Missed by 20.5
23
Sat, Sep 5, 7:00 PM EDT

Florida International at South Florida

Best-book edge+3.1
Recommendation · Bovada

Florida International +12.5

Best book
Bovada · +12.5
Consensus line
+12.5
All-Takes line
+9.38

Evidence floor: 2.00 points. A 2.00-point evidence floor applies because at least one team changed head coaches. The coaching signal can cause an abstention; it never flips the model's side. The new pace adjustment is shown in the forecast but does not create a bet.

Core football model+3.0

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check+0.0

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production0.0

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions+0.1

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price0.0

Bovada has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets13–16–1

44.8% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2019

Massachusetts +16.67 at Rutgers

+2.3Missed by 10.3
2022

Temple +9.5 at Duke

+2.4Missed by 20.5
2017

California +12.5 at North Carolina

+2.7Covered by 17.5
2023

Northern Illinois +8.5 at Boston College

+3.0Covered by 11.5
2022

Memphis +17 at Mississippi State

+2.8Missed by 9.0
2017

Arkansas State +14.5 at Nebraska

+2.5Covered by 7.5
2017

Troy +10.5 at Boise State

+1.9Missed by 0.5
2017

Wyoming +12.17 at Iowa

+1.6Missed by 8.8
2021

Central Michigan +13.80 at Missouri

+1.7Covered by 3.8
2017

Charlotte +14 at Eastern Michigan

+1.4Missed by 3.0
2022

Georgia State +12.5 at South Carolina

+1.6Missed by 8.5
2018

Kent State +18.12 at Illinois

+3.7Covered by 11.1
2024

Old Dominion +20 at South Carolina

+2.5Covered by 16.0
2022

East Carolina +12.5 vs. NC State

+2.8Covered by 11.5
2025

Western Michigan +18.5 at Michigan State

+2.9Covered by 1.5
2025

Jacksonville State +16.5 at UCF

+3.0Covered by 9.5
2025

UTEP +3.5 at Utah State

+3.5Missed by 8.5
2021

San José State +14 at USC

+0.9Missed by 9.0
2024

West Virginia +10 vs. Penn State

+2.9Missed by 12.0
2022

Arizona +6.25 at San Diego State

+4.1Covered by 24.3
2023

Colorado +20.5 at TCU

+3.9Covered by 23.5
2019

Louisville +19.25 vs. Notre Dame

+3.1Covered by 1.3
2018

San Diego State +13.75 at Stanford

+2.6Missed by 7.3
2025

Coastal Carolina +12.5 at Virginia

+0.3Missed by 28.5
2018

Army +13.88 at Duke

+4.6Missed by 6.1
2023

North Texas +6.5 vs. California

+2.1Missed by 30.5
2018

North Carolina +7 at California

+0.7Push
2018

Central Michigan +17.5 at Kentucky

+1.3Covered by 2.5
2018

Massachusetts +20.38 at Boston College

+2.5Missed by 13.6
2021

Hawai'i +17.60 at UCLA

+3.1Missed by 16.4
24
Thu, Sep 3, 9:00 PM EDT

UAB at Illinois

Best-book edge+2.6
Recommendation · Bovada

Illinois −27.5

Best book
Bovada · −27.5
Consensus line
−27.5
All-Takes line
−30.15

Evidence gate: intentional model departure. Neither team changed head coaches, so an intentional pre-pace production-model departure is sufficient. The pace adjustment changes the forecast, not whether a bet exists.

Core football model+2.6

Team strength and matchup evidence that survives the production model's market cross-check.

2026 roster check0.0

An independently estimated roster-change signal is used only when it agrees with the core edge.

Year-over-year production0.0

Opponent-adjusted offensive-line quality and actual transfer production are added only when they reinforce the current model side.

Expected possessions0.0

Opponent-adjusted clock, drive, efficiency and play-style forecasts make a small pace adjustment in projected non-blowouts. Pace cannot choose a side by itself.

Best-book price0.0

Bovada has the most favorable current spread for the model side. This execution improvement is separate from the football model.

Verified injuriesPending

No current-game injury adjustment is applied until verified pregame availability is supplied.

30 closest historical Week 1 bets19–11

63.3% of decided matches covered. Useful context, not a predicted probability.

Inspect all 30 historical matches

Ordered from most to least similar. “Closest” means the best available matches in the historical pool; it does not mean the games are identical.

YearHistorical recommendationEdgeResult
2017

Memphis −27.17 vs. UL Monroe

+2.0Missed by 19.2
2024

Alabama −31 vs. Western Kentucky

+3.9Covered by 32.0
2017

Wisconsin −27.33 vs. Utah State

+0.9Covered by 21.7
2019

Texas −19.38 vs. Louisiana Tech

+3.3Covered by 11.6
2021

Wake Forest −32.5 vs. Old Dominion

+1.0Missed by 0.5
2024

Kentucky −25.5 vs. Southern Miss

+0.9Covered by 5.5
2023

Auburn −35 vs. Massachusetts

+3.6Covered by 10.0
2022

Washington −23.62 vs. Kent State

+2.3Covered by 1.4
2018

Penn State −24.5 vs. App State

+1.3Missed by 17.5
2024

Pittsburgh −23.5 vs. Kent State

+1.2Covered by 7.5
2021

Fresno State −27.70 vs. UConn

+4.4Covered by 17.3
2022

Michigan State −21.75 vs. Western Michigan

+3.3Covered by 0.3
2022

Maryland −24 vs. Buffalo

+3.6Missed by 3.0
2019

Army −23 vs. Rice

+1.7Missed by 16.0
2021

Cincinnati −22.30 vs. Miami (OH)

+2.2Covered by 12.7
2021

Oklahoma −31.25 vs. Tulane

+2.0Missed by 26.3
2021

Texas A&M −29.5 vs. Kent State

+3.9Covered by 1.5
2022

Ohio State −17 vs. Notre Dame

+1.5Missed by 6.0
2025

Ole Miss −38.5 vs. Georgia State

+3.1Covered by 17.5
2022

Missouri −20.38 vs. Louisiana Tech

+2.8Covered by 7.6
2025

Washington −20.5 vs. Colorado State

+0.9Missed by 3.5
2019

Michigan −35.62 vs. Middle Tennessee

+2.4Missed by 16.6
2017

Clemson −38.17 vs. Kent State

+3.8Covered by 14.8
2024

Maryland −20.5 vs. UConn

+1.4Covered by 22.5
2022

Minnesota −36 vs. New Mexico State

+5.1Covered by 2.0
2019

Michigan State −23.12 vs. Tulsa

+1.4Missed by 2.1
2018

Ohio State −40 vs. Oregon State

+3.3Covered by 6.0
2021

Pittsburgh −37.60 vs. Massachusetts

+0.2Covered by 6.4
2017

Arizona State −24.67 vs. New Mexico State

+0.2Missed by 18.7
2021

NC State −19.70 vs. South Florida

+2.7Covered by 25.3

Before kickoff

The card is supposed to change.

Lines, quarterbacks and skill-player availability will move. The production card should be rebuilt with the latest verified inputs; these recommendations are a dated model snapshot, not guarantees.

See the team ratings behind the card