Cincinnati Bengals vs Houston Texans — Win Probability

Cincinnati upsets Houston as model underestimated Bengals by 16.5 points

Cincinnati's 20-6 victory over Houston in Week 2 was a clear upset, and the model's pregame projection fell substantially short of capturing how decisively the Bengals would play.

Before kickoff, the model gave Cincinnati a 43% win probability and projected Houston to win by 2.5 points—an expected score of 24.5 to 22. That aligned closely with the market line, which favored Houston by 2.5 points. The model was effectively calling this a toss-up with a slight Houston lean.

What happened was different in both direction and magnitude. Cincinnati won 20-6, a 14-point margin. The Bengals scored 2 fewer points than the model expected them to (20 vs. 22), but Houston scored 18.5 fewer (6 vs. 24.5). The gap between the projected margin (-2.5 for Houston) and the actual result (+14 for Cincinnati) was 16.5 points—a substantial miss.

This wasn't a close call that went the other way. The model projected a competitive game decided by a field goal, and instead saw a dominant Bengals performance where Houston's offense nearly shut down. Cincinnati's defense was the primary story the numbers didn't anticipate; the model's projection of 24.5 points for Houston reflected an expectation of offensive execution that didn't materialize.

For Cincinnati, the win improved them to 2-0. Houston fell to 0-2. The model correctly identified this as a game Houston was favored in—the upset was real—but underestimated both Cincinnati's ability to control the game and Houston's vulnerability. The 16.5-point swing suggests the pregame assessment of Houston's strength was too generous, or Cincinnati's defensive capability was undervalued.