Sports keep getting smarter.
The seventh annual MIT Sloan Sports Analytics Conference kicks off on Friday, and this yearly gathering of stats geeks, sports execs, and companies looking to make a buck in the burgeoning world of sports data continues to grow in prominence. (A few years ago, Grantland founder Bill Simmons astutely labeled the conference “dorkapalooza.”) Around 2,700 people signed up to attend this year’s conference, a 23% increase over last year: representatives from 80 teams in Major League Baseball, the NBA, NFL, NHL, Major League Soccer and the English Premier League will be on hand. In its infancy, the conference attracted 2 to 4 sponsors: there are now 13 sponsors, including Under Armour, SAP, ESPN, TicketMaster, and StubHub. Nate Silver is making an appearance; he’ll be treated like Bono at this thing. “The reality is that big data works in sports,” says Houston Rockets general manager Daryl Morey, a former MIT business student who founded the conference.
If you follow sports and have a Twitter account, be prepared for a barrage of missives about “spatio-temporal data” and other esoteric topics. The research papers presented at the conference are always worth a look. In a paper entitled “Going For Three: Predicting the Likelihood of Field Goal Success with Logistic Regression,” three PhD students from MIT’s Department of Aeronautics and Astronautics write that, “through the development of a binomial logistic regression model, we found that situational (psychological) factors have no statistical effect on the outcome of [NFL] field goal attempts, despite what fans, coaches, and the media believe. On the other hand, environmental factors – temperature, precipitation, wind, field surface, and altitude – do have major impacts upon the difficulty of a field goal attempt.”
Upside: maybe this research will convince coaches to stop trying to “ice” the kicker at the end of games by a timeout before the snap, since psychological tricks don’t have a statistically significant impact. Downside: does it bode well for America’s aeronautic and astronautic future that these guys are spending their time studying field goals?
Another hot topic at this year’s conference: advanced “fanalytics.” Over the past few years, cameras have been honed in on the performers on the field, to capture their every movement and tendency, to help them improve future performance — and to help their opponents scout them out. A group of NYU computer scientists are putting the focus on the fans: their paper, “Beyond the Kiss Cam: Measuring the Fan With Computer Vision Analytics,” presents a software program they’ve developed that can automatically identify different fan behaviors, in real-time, during a game. What Jumbotron clips get the fans most fired up? If a team knows this information, it can pick the right highlights to play at crunch time, to help motivate the home team. If you’re an advertiser, you can measure what kind of message creates the most positive response, in different stadiums and arenas. “We’ve been putting this big X-ray on the players,” says Chris Bregler, one of the co-authors of the paper. “But the fans are the ones spending all the money. That where there’s bigger commercial potential.”
Is all this just the beginning of the sports analytics movement? “I don’t know where we are on the merry-go-round,” says Morey. “Sports was pretty late to the party. Financial services companies, consumer products companies – they’ve all been analyzing data to make decisions for years.” Every year at the Sloan conference, the message becomes clearer. If you want to make it in sports — off the field — start crunching numbers.