Well, that was quite a game. Despite a 10-point fourth quarter comeback, the New England Patriots looked like they were headed for another gut-wrenching defeat at the hands of a circus catch (this time with the Seattle Seahawks’ Jermaine Kearse playing the role of David Tyree), before undrafted rookie Malcolm Butler stepped in front of a Russell Wilson pass and sealed the Patriots’ fourth Super Bowl crown since 2001.
Much will be made of the Seahawks’ decision not to run the ball from the Patriots’ 1-yard line, instead opting for the fateful pass that Butler intercepted. During the 2014 season, 57.5 percent of all rushing plays from the opponent’s 1 yard-line ended in touchdowns — and, as Harvard Sports Analytics Collective pointed out Sunday night, you could make a strong case for Seattle’s probability of scoring being even higher against New England’s defense:
But I’ll leave most of the second-guessing to others for now. In the moment, I was mainly interested in where this game ranked among all classic Super Bowls in terms of excitement. (After all, it featured an impressive comeback and some wild swings in win probability late in the game.) To quantify how thrilling the game was, we once again turn to the Excitement Index, which we wrote about in our Super Bowl preview Friday. The Excitement Index measures the sum of the absolute changes in win probability throughout a game, which theoretically captures how many extreme ups and downs there were in a given Super Bowl.
Perhaps surprisingly, though, this year’s game only ranks 12th all-time if you use the data from NumberFire’s in-game Win Probability tool:
(Consider this an unofficial early return; the numbers we used last week were from Pro-Football-Reference, which won’t post Super Bowl win probability until Monday. It’s worth noting that some of the NumberFire probabilities don’t completely match what Brian Burke’s model lists.)
Certainly, there were some plays that moved the probability needle dramatically. According to NumberFire’s model, Wilson’s interception dropped Seattle’s chances of winning from 64 percent to essentially zero. The bobbled completion to Kearse several snaps earlier raised the Seahawks’ probability by 30.8 percentage points. But most of the changes before that sequence were more gradual. The Patriots built a modest cushion for most of the first half before the Seahawks tied the game; likewise, the Seahawks rise to 91 percent win probability early in the fourth quarter was a slow march.
At the same time, models like the Excitement Index are, well, just numerical representations of football reality. No index can really capture the craziness and excitement of what just transpired on the field — and we wouldn’t have it any other way.