Indianapolis Colts vs Kansas City Chiefs — Win Probability

KC's Close Call: Model Projected 5.7-Point Win, Got 3 Instead

Kansas City's 33-30 victory over Indianapolis on Sunday was the outcome the model expected—a Chiefs win with 67% win probability pregame—but the margin fell short of the projection by 2.7 points.

The model simulated a Kansas City win by 5.7 points, with the Chiefs scoring 26.1 and Indianapolis 20.4. The market's betting line was nearly identical, setting KC as a 5.8-point favorite. Both projection and line aligned closely, suggesting a confident read on Kansas City's edge.

What actually happened: Kansas City won 33-30, a 3-point margin. The Chiefs scored 6.9 points above their expected output; Indianapolis scored 9.6 points above theirs. That combination compressed the gap between the teams from the projected 5.7 points down to 3.

This was not an upset. The model got the winner right and the direction right. The 2.7-point miss in margin is modest relative to the volatility inherent in football simulation. Kansas City's expected score was conservative relative to their actual output, while Indianapolis performed significantly better than the model anticipated.

The model's 67% win probability for Kansas City proved sound—it assigned them favorites' odds and they won. The spread miss (5.8 market vs. 3 actual) reflects how both the model and the betting market slightly underestimated Indianapolis's ability to stay close, but not by a degree that would trigger significant model recalibration. Kansas City improved to 2-0; Indianapolis fell to 0-2.