Philadelphia Eagles vs Chicago Bears — Win Probability

Chicago's 20-Point Upset Over Philadelphia Exposes Model's Pregame Miscalibration

The Chicago Bears defeated Philadelphia 27-7 on Sunday, a 20-point margin that inverted the model's entire read on the game. Before kickoff, the model projected a Philadelphia win by 3.5 points, with the Eagles favored at 40% win probability for Chicago. The market agreed: Philadelphia was a 3.5-point favorite. Neither the model nor the betting line saw what actually happened.

The model's pregame expectation was Philadelphia 22.5, Chicago 19. The actual result: Chicago 27, Philadelphia 7. That's a 24-point swing in Chicago's direction from the projection.

This was a genuine upset. Chicago entered with only a 40% implied win probability in the model's simulation, meaning the Bears were treated as the underdog despite playing at home. Philadelphia was the clear favorite on both the model's internal odds and the published spread. The model's expected margin of -3.5 (favoring Philadelphia) became a +20 actual margin (favoring Chicago), a gap of 23.5 points between what was modeled and what occurred.

For a projection system built on historical tendencies and measurable inputs, this represents a substantial miss. The model did not anticipate a Bears team that would outscore Philadelphia by 20 in a head-to-head matchup. Whether Chicago's dominant performance reflected variables the model underweighted—offensive execution, defensive pressure, or Philadelphia's unpreparedness—or whether randomness simply broke decisively one direction, the numbers show a clear divergence between pregame expectation and Sunday's result.

Both teams enter Week 4 at 2-1. The model's projection accuracy will depend on whether this performance signals a systematic misread of these teams' true strength or represents the normal variance that occurs across a 17-week season.