Idiots Write About Sports

Perhaps you laughed when I dared to suggest that the Miss World Cup beauty pageant might be a leading indicator for how the soccer tourney itself might play out. (“Laughed.” “Shook your head sadly and vowed never to take me seriously again.” It’s all semantics really.) I don’t like to pat myself on the back, but it appears that, once again, I have been proven to be a pioneer in merging the twin arts of sports analysis and staring at attractive women.

Throughout the course of the World Cup, Deadspin has been providing match-by-match live blogs of each contest. (Gaze at the one for the England-Sweden confrontation for a better idea of what I’m talking about.) Accompanying each live blog are two pictures, each representing the competing nations. One picture usually features a face-painting SuperFan; the other, a ridiculously beautiful woman.

Well, a Deadspin reader did the math. And it will surprise no one who read my World Cup preview — or indeed, anyone who lived through junior high — that beauty trumps dorkery nearly every time. Or, to turn the floor over to this Bill James-esque innovator in the study of babe-ology:

while “Babe” pictures have coincided with only 25.0% of the teams involved in the 20 games, they have accounted for 35.3% of the wins and, perhaps not coincidentally, 35.1% of points. I’m not going to get in p factors and two-tailed tests and such here, but it appears that having a Babe picture involved correlates with an increased likelihood of a result of in the area of 50% over teams with a picture of a Dude (the Kid group serves as an effective control group, as it is about 1/8th across the board).

There are only two possible hypotheses: 1) Better teams attract more babes, making it easier for you to find pictures to attach to the live blogs; and 2) the pictures you attach to the live blog posts have some influence on the games. Initially, it would seem that the first hypothesis is more persuasive. However, I must point out that “Babe Teams” have not scored a disproportionate number of goals — 27.3%, which is only 2.3% higher than the expected 25.0% and, as such, well within the standard error. As such, there must be some other larger, subtler, less obviously discernible force that is influencing these outcomes.

I believe if the statistical analysis community were to embrace this brave pioneer’s theories, we would see a wellspring of renewed interest in sabermetrics. If nothing else, it would make the center-spread of each month’s SABR newsletter a far more intriguing proposition.

Post a comment