Chicago Bears vs Green Bay Packers — Win Probability

GB's 24-Point Win Over CHI Defied Model and Market Alike

Green Bay's 31-7 victory over Chicago on Sunday was an upset by both analytical and betting measures—and the magnitude of the gap between projection and result was substantial.

Pre-kickoff, Matrix's model favored Chicago slightly, projecting a 1-point Bears win (CHI 24.2, GB 23.2). The market agreed, installing Chicago as a 1-point favorite. Green Bay's win probability sat at just 47%, meaning the model gave the Packers less than an even coin flip to win.

What happened instead: Green Bay won by 24, a margin that swung 25 points from the model's expected outcome. The Packers' actual 31 points nearly matched their 23.2 projection, but Chicago's collapse—scoring just 7—represented a 17-point underperformance from the projected 24.2.

This was a clean upset. Green Bay wasn't favored and delivered a dominant performance that bore no resemblance to the close game the model anticipated. The Bears' defense and offensive execution both fell far short of expectation, while Green Bay's defense operated at a level the pregame simulation hadn't fully accounted for.

The size of the miss—25 points worse than expected for Chicago—isn't explainable by normal variance. This falls into the category of a scenario where either team composition, preparation, or execution deviated significantly from what the model's inputs suggested. Both teams finished the week at 3-2, so the loss didn't derail Chicago's season trajectory, but the performance gap between projection and reality was among the larger discrepancies a single game can produce.

For a model that emphasizes factual calibration, this serves as a reminder that even balanced pregame assessments can face high-impact surprises. Green Bay's win was real. So was the model's miss.