By Barrett Hansen
Super Bowl XLVIII between the Seattle Seahawks and Denver Broncos was one of the most hyped championship matchups in recent memory: Peyton Manning and the highest-scoring offense in NFL history were pitted against one of the best passing defenses of all time in what figured to be an epic battle. The line hovered right around DEN -2.5 up through kickoff. However, for many experts and football insiders, the game was essentially a toss-up, with not much to distinguish between the two powerhouses, and a fantastic finish nearly a guarantee.
The game played out much differently than anticipated, with the Seahawks thoroughly dismantling a Denver squad ravaged by injuries. Such an outcome seemed so bizarre considering the perceived balance before kickoff, and prompts the question: How improbable was Seattle’s 43-8 blowout?
To answer that, I looked at NFL spread data going back to 1978. I found that of the 516 games in which the line was 2.5 points at kickoff, the favorite won about 53% of the games by an average of 0.82 points, indicating these matchups are relatively even. The distribution of margins of victories centered on that mean with a standard deviation of 13.6. Consequently, the Seahawks’ 35-point victory as underdogs gave this past Super Bowl a Z Score of -2.6, meaning that the likelihood of a win at least that lopsided was a mere 0.43%.
Here is a graph illustrating the distribution of these games. The solid red line signifies the mean, with the dotted red lines indicating the first and second standard deviations. The blue line represents the Seahawks victory, and is negative because they were technically the underdog.
One flaw with this method is that big games like the Super Bowl—which feature a disproportionate number of casual fans—could produce lines that don’t truly reflect game probability. For example, this game’s spread was probably skewed by Americans’ blind love of Peyton Manning. To account for that, I also examined games with a zero or one point spread in toss-up games, reflecting the uncertainty most industry experts had going into the game.
Favorites won 260 of the 514 games with a spread at one or below, good for a 50.3% clip. The average margin of victory was -0.35, meaning underdogs actually outscored the favorites overall, with a standard deviation of 13.78. Calling the game a pick ’em gives the 35 point victory an average Z Score of 2.54 (since it’s arbitrary who’s dubbed the “favorite” in a pick ’em game I averaged the Z Scores of a +35 and -35 margin of victory), which boosts the probability modestly to 0.56%.
Given the extreme margin of victory, perhaps the line didn’t properly value the Seahawks. But given the line, the probability that either team would get blown out by 35 or more was about one in a hundred, leaving us to forever speculate about the Super Bowl epic that could have been.