Denver Broncos vs San Francisco 49ers — Win Probability

Model Underestimated SF Dominance in 10-Point Win Over Denver

San Francisco's 24-14 victory over Denver was not an upset—the model favored the 49ers pregame at 59% win probability, matching the market's 3-point spread. But the margin of victory revealed a significant gap between projection and result.

The model expected San Francisco to win by 3 points, with a combined score of 48.4 (SF 25.7, Denver 22.7). The actual game saw the 49ers dominate by 10 points with a final score of 24-14. That 7-point differential between expected margin and actual margin represents a substantial miss on the intensity of San Francisco's performance.

The market and the model aligned on direction but diverged on magnitude. Both pegged this as a close contest; the 3-point spread and 59% win probability both suggested a competitive game decided by possession. Instead, Denver's offense was held to 14 points—8.7 below projection—while San Francisco scored just 1.7 points above expectation. The underperformance came primarily from Denver's side of the ball.

For a model that correctly identified the winner, this represents a calibration issue on game control rather than pick accuracy. The 49ers' defense was more disruptive than the pregame data suggested, or Denver's offense regressed beyond what was forecasted. The 7-point margin miss is material in a model whose brand depends on precision—this was not a nail-biter that played out as expected, but a convincing road win that looked closer on paper than it did on the field.

The result validates the directional call but exposes a blind spot in estimating how thoroughly one team would dominate the other.