Army lucky 21 
Temple unlucky 17 Stolen · Army wins 9% of replays Army won on an off day 53% to win a fresh game, 9% with this day’s form: defense far below its usual level. College football, with the luck taken out.
The scoreboard says who won. We replay every FBS game 4,000 times with the way both teams actually played, and show who earned it and who got lucky: a postgame win expectancy for every game, and luck rankings for every team.
It opens the site with your team and marks it everywhere. Saved in this browser only.
Week 4, 2026: 4 of 71 winners were outplayed
The line on each card fills with how often the loser wins the replay: the fuller, the more luck decided it. Outlined: the winner was outplayed.
Army lucky 21 Temple unlucky 17 Stolen · Army wins 9% of replays Army won on an off day 53% to win a fresh game, 9% with this day’s form: defense far below its usual level. 
Iowa lucky 20 
Michigan unlucky 19 Stolen · Iowa wins 13% of replays 
Navy unlucky 20 
UAB lucky 24 Winner outplayed · UAB wins 32% of replays 
Wake Forest lucky 30 
Louisville unlucky 27 Winner outplayed · Wake Forest wins 38% of replays 
Minnesota 27 
Washington 24 Coin flip · Minnesota wins 40% of replays 
Air Force 36 
Nevada 33 Coin flip · Air Force wins 46% of replays 
Nebraska 31 
Michigan State 13 Coin flip · Nebraska wins 48% of replays 
App State 31 
NC State 41 Coin flip · NC State wins 53% of replays NC State won on an off day 84% to win a fresh game, 53% with this day’s form: far below what the matchup promised. 
Oregon 41 
USC 27 Coin flip · Oregon wins 53% of replays 
Southern Miss 21 
Tulane 24 Coin flip · Tulane wins 55% of replays 
Kennesaw State 14 
Arkansas State 17 Coin flip · Arkansas State wins 56% of replays 
Georgia Tech 27 
Stanford 34 Coin flip · Stanford wins 59% of replays Luckiest so far
South Florida4–0+1.6 wins
Marshall3–1+1.6 wins
Florida Atlantic3–1+1.1 wins- Iowa4–0+1.1 wins
North Dakota State4–0+1.0 wins
Texas Tech4–0+1.0 wins
Fresno State3–1+0.9 wins
Texas4–0+0.9 wins
Clemson3–1+0.9 wins
Tulsa3–1+0.9 wins
Unluckiest so far
- Georgia Tech1–3−1.9 wins
Rutgers1–3−1.6 wins- Temple1–3−1.5 wins
- Nevada1–3−1.2 wins
UL Monroe0–4−1.2 wins
Utah State1–3−1.2 wins
Sam Houston1–3−1.2 wins
Northern Illinois0–4−1.2 wins
Charlotte0–4−1.1 wins
Rice1–3−1.0 wins
All 138 FBS teams, after Week 4Standings, luck taken outBy conferenceEvery 2026 upsetExpected winsOne-score gamesTurnover luckPower rankingsEvery team
Every week of 2026
| Week | Dates | Winners outplayed | Biggest heist |
|---|---|---|---|
| Week 4 | Sep 24, 2026–Sep 26, 2026 | 4 of 71 | Army 21–17 Temple · won 9% of replays |
| Week 3 | Sep 17, 2026–Sep 19, 2026 | 7 of 75 | Arizona State 24–17 Kansas · won 8% of replays |
| Week 2 | Sep 10, 2026–Sep 12, 2026 | 12 of 86 | Illinois State 28–24 Northern Illinois · won 17% of replays |
| Week 1 | Aug 29, 2026–Sep 7, 2026 | 4 of 99 | South Alabama 39–14 SE Louisiana · won 12% of replays |
The biggest heists of 2026 so far
- Arizona State 24, Kansas 17 · Week 3 · Arizona State wins 8% of replays
- Army 21, Temple 17 · Week 4 · Army wins 9% of replays
- South Alabama 39, SE Louisiana 14 · Week 1 · South Alabama wins 12% of replays
- Iowa 20, Michigan 19 · Week 4 · Iowa wins 13% of replays
- Idaho State 29, Utah State 17 · Week 1 · Idaho State wins 16% of replays
How it works
Every play of every game goes back into the pot. Each team’s own plays from that game are drawn at random, by down and distance, into new drives and a new game, 4,000 times. Kicks, fumble bounces and turnovers happen at the rates you would expect, not the ones that happened. How often a team wins those replays is what its performance was worth that day: a postgame win expectancy, built from the plays themselves.
The difference between that and the result is luck. A loss in a game a team wins 9 times in 10 scores −90 on our luck index. Kicks, fumbles, conversions and referee calls are valued one by one, each kick against the kicker’s own record, so you can see where it came from.
College play-by-play comes merged from several stats providers; we check it against the final scores and rebuild what doesn’t hold up. Methodology and validation