Archive

A history of our predictions. A model is only worth what its predictions are, so all of ours are made public after games end.

Were the model's chances right?

Every game is sorted by the chance that the Field Theory model gave each team to win. Each bar shows the amount of teams in that prediction range that actually won. When the model's chances are honest, the bar boundary ends up inside the shaded band.

How often the model's favorites won, by the chance it gave them
Model's chance for the favoriteFavorites wonUnderdogs won
90% to 100% (underdog 0% to 10%)96%4%
80% to 90% (underdog 10% to 20%)82%18%
70% to 80% (underdog 20% to 30%)76%24%
60% to 70% (underdog 30% to 40%)63%37%
50% to 60% (underdog 40% to 50%)50%50%
76%Winners picked1148 of 1508 games, not including toss-ups
12.7Average missPoints between projected and final margin among FBS matchups
1522GamesFinal with a pregame prediction, all seasons

Click here to see Field Theory's advanced statistics.

Week 16

1 of 1 winners picked (100%) · average miss 2.9 pts

Sat, Dec 13 · Neutral siteCorrect pick
TeamFair moneylineWinFinal
Army+14441%16
Navy−14459%17
Spread
Navy −4
Projected margin
Navy by 3.9
Off by2.9

Our Top 25 and archived predictions are free to use and distribute.