Tennessee Titans vs Indianapolis Colts — Win Probability

Tennessee Upsets Indianapolis as Model Projected Colts Dominance

Tennessee's 26-20 victory over Indianapolis on Sunday was a clear upset, flipping the script on what the pregame model anticipated across nearly every dimension.

Before kickoff, the model projected Indianapolis to win by 6.5 points, with an expected final score of 25-18 in the Colts' favor. That projection aligned closely with the market, which had Indianapolis favored by 6.5 points. The model gave Tennessee only a 31 percent win probability—a 69 percent lean toward Indianapolis.

What actually happened was a 6-point Tennessee win, creating a 12.5-point swing from the model's expected margin. The real final margin (6 points, Tennessee) stood in stark contrast to the projected margin (6.5 points, Indianapolis). Tennessee scored 26 points versus the model's projection of 18.5; Indianapolis scored 20 versus the projected 25.

The Titans' victory qualifies as a genuine upset. Tennessee not only won a game the model favored them to lose by nearly a touchdown, but they did so by flipping the sign of the margin entirely. The actual outcome fell well outside the range of the model's pregame expectation.

Neither team's scoring output fell close to projection. Tennessee exceeded its expected 18.5 points by 7.5 points. Indianapolis fell short of its projected 25 by 5 points. The combination created a larger-than-expected Tennessee advantage.

This gap between projection and result does not necessarily invalidate the model's reasoning—Indianapolis entered as the stronger team on paper, and the Colts were favored by the market for a reason. But Sunday's result was a meaningful miss, with the Titans playing well enough to overcome a significant pre-game probability disadvantage.