Indianapolis Colts vs Pittsburgh Steelers — Win Probability

Model Underestimated Pittsburgh in Comfortable Win Over Indianapolis

The Pittsburgh Steelers' 34-22 victory over Indianapolis on Sunday was more decisive than pregame projections suggested, though the model correctly identified Pittsburgh as the favorite.

Before kickoff, the model expected Pittsburgh to win by 3 points, with a final score of 22.3 to 19.3. The market agreed on the spread, setting Pittsburgh as a 3-point favorite. Pittsburgh's win probability stood at 59%, reflecting a modest but clear edge.

The actual result differed meaningfully from that projection. Pittsburgh scored 34 points and Indianapolis 22, producing a 12-point margin—four times larger than the expected 3-point differential. The Steelers' actual output exceeded the model's expectation by 11.7 points, while the Colts fell short by 2.7 points.

This was not an upset—Pittsburgh was favored and won—but it was a more dominant performance than the pregame simulation anticipated. The model's 59% win probability for Pittsburgh proved directionally correct, yet it underestimated the magnitude of the Steelers' advantage. A 3-point expected margin closing as a 12-point actual margin represents a substantial gap between projection and reality.

For Pittsburgh (3-2), the win confirmed the pregame lean but exceeded it in scale. For Indianapolis (2-3), the result landed worse than the neutral expectation would have suggested. The model's framework correctly identified Pittsburgh as the more likely winner, but the simulation underestimated Pittsburgh's offensive output and the defensive dominance that produced the final spread. This gap suggests either Pittsburgh played significantly above its pregame baseline, Indianapolis underperformed its projection, or both.