In 2007, its inaugural year, the MIT Sloan Sports Analytics Conference drew 100 attendees, most of them students. Two years later nearly 500 sports executives, fans and "quant jocks" showed up for a daylong session that included Celtics guard Ray Allen as a featured speaker. This year's incarnation, held last weekend at the Boston Convention & Exhibition Center, attracted a sold-out crowd of 1,000—and a waiting list of 400. The hope is that your math skills allow you to deduce that the conference has grown dramatically. Otherwise you might have had trouble following, say, the session titled Improved NBA Adjusted +/- Using Regularization and Out-of-Sample Testing.
This is an article from the March 15, 2010 issue
Christened Dorkapalooza by ESPN columnist Bill Simmons, the conference is, quite literally, the brainchild of Houston Rockets general manager Daryl Morey. An MIT alum, Morey was one of the first NBA executives to value analytics—crassly: the use of data to tell, confirm or deny a story—in his decision-making. "You get tunnel vision in your sport," says Morey. "This is a way to pick up on [analysis] done in other sports that might be useful."
This year's event attracted heavy hitters, including Mavericks owner Mark Cuban, Colts president Bill Polian and writer Michael (Moneyball) Lewis. Topics included new metrics to measure defense in baseball and using risk analysis to determine whether to punt on fourth down. (This writer co-presented a paper on basketball officiating.) John Huizenga, a Chicago Business School professor who doubles as Yao Ming's agent, and Sandy Weil, a statistics consultant, offered empirical proof that Tim Duncan's 152 blocked shots in '07--08 were more valuable than Dwight Howard's 176. (Duncan, for instance, was better at keeping the ball in bounds and got called for goaltending less often.) Says Cuban, "This is the cutting of edge of analysis. I recommend no other NBA teams attend."
The keynote panel was an exercise in self-examination. Moderated by Lewis, What Geeks Don't Get: the Limits of Moneyball addressed the shortcomings of analytics and the necessity of mixing science with art. Still, the takeaway from Dorkapalooza was clear: A data-driven front office doesn't guarantee success, but it sure beats the alternative.