PlayDecoded
Menu

Model Accuracy

We believe in showing our work. Here's exactly how accurate our win probability model has been across every game we've analyzed.

Overall Performance

55.4%
Overall accuracy across 364 games
201 correct
162 incorrect
1 ties
nfl2025
59.3%
54/91 decided games
nfl2024
54.0%
147/272 decided games

Calibration by Confidence

When the model says 70%, does the favored team actually win ~70% of the time? Points above the dashed line outperform; below underperform.

50-55%: predicted ~52%, actual 50.0% (88 games)55-60%: predicted ~58%, actual 56.4% (266 games)60-65%: predicted ~62%, actual 83.3% (6 games)65-70%: predicted ~67%, actual 100.0% (2 games)70-75%: predicted ~75%, actual 0.0% (1 games)50%75%100%50%75%100%Perfect calibration
Model ConfidenceGamesCorrectActual Accuracy
50-55%884450.0%
55-60%26615056.4%
60-65%6583.3%
65-70%22100.0%
70-75%100.0%

Week-over-Week Trend

How the model has performed over recent weeks. The dotted line marks 50% (coin flip).

50%
W18
W3
W6
W9
W13
W16

How We Measure Accuracy

A prediction is correct when the team our model gives > 50% win probability actually wins the game. Ties are excluded from accuracy calculations.

Calibration measures whether our confidence levels are meaningful. If we say a team has a 70% chance of winning, they should win roughly 70% of those games -- not 90%, not 50%.

For context, a coin flip gives 50% accuracy. Our model targets outright winners: predicting which team wins the game.