Seattle Seahawks vs Denver Broncos — Win Probability

Broncos' 3-Point Upset Over Seahawks Defies Model's 2.4-Point Seattle Projection

Denver's 17-14 victory over Seattle on Sunday was a straightforward upset by the numbers: the model projected a 2.4-point Seahawks win before kickoff, yet the Broncos won by 3. That 5.4-point swing exceeded the pregame expectation significantly.

The model rated Denver at just 43% win probability entering the game, a clear underdog read. The market agreed, favoring Seattle by 2.5 points—nearly identical to the model's expected margin. Both the simulation and the betting line expected a close game, but both got the direction wrong.

The model's pregame scoring projection was Seattle 21.5, Denver 19.1. The Broncos finished with 17 points, 2.1 below projection. Seattle scored 14, 7.5 fewer than expected. This was not a game where one team dramatically outperformed—both fell short of simulated output, though Seattle's shortfall was steeper. The actual margin (3 points) made for a closer result than the model anticipated, but in Denver's favor rather than Seattle's.

For a model calibrated on historical patterns, missing the winner by 5.4 points registers as a notable gap. However, the game itself remained tight—a field goal decided it. The miss was not a blowout projection failure; it was a direction error in a low-probability upset scenario. The model assigned Denver less than even odds, which proved pessimistic in hindsight.

Denver improved to 4-2 with the win. Seattle fell to 3-3. The Broncos' upset victory represents the kind of outcome that occurs outside historical distributions—uncommon enough to warrant a projection miss, common enough that it happens.