Four business lessons from the MIT Sloan Sports Analytics Conference
Takeaways from the Oakland A’s, Boston Red Sox, and, of course, Nate Silver
By Brian Eastwood |
March 15, 2016
Moneyball author Michael Lewis, statistician Bill James, Oakland Athletics executive Billy Beane, and ESPN’s Jackie MacMullan
Amid the discussions about jump shots, penalty kicks, and rabid fans at the MIT Sloan Sports Analytics Conference, held March 11-12 in Boston, executives and data scientists alike shared lessons just as relevant in the boardroom as in the locker room. Here are four of the best.
Seek out unrecognized potential. On the face of it, the 2002 Oakland Athletics were a ragtag bunch of young men who didn’t even look like Major League Baseball players. However, general manager Billy Beane and assistant general manager Paul DePodesta—central figures of the Michael Lewis book Moneyball—had assembled a group of players that others teams had overlooked because they were short, overweight, or in the wrong role.
“We were trying to find value where it wasn’t readily apparent,” said DePodesta, now the chief strategy officer for the National Football League’s Cleveland Browns. One such player, Scott Hatteberg, had been a backup catcher for the Boston Red Sox but became an everyday first baseman for the Athletics, who valued his ability to get on base and were willing to teach him how to pay a new position.
Baseball statistician Bill James, whose work influenced the personnel decisions Beane and DePodesta made, said judging players by what they cannot do is “not relevant,” adding, “You only win ballgames with things that people can do. You don’t win ballgames with things that people can’t do.”
Build a model and stick with it. James’ approach to objective analytics, known as sabermetrics, applies predictive modeling in a variety of ways in an effort to avoid bias when evaluating talent.
Doing it right, though, means defining the rules and evidence for the model ahead of time and then letting the model run, said Nate Silver, founder of FiveThirtyEight and author of The Signal and the Noise. Avoid the temptation to change the rules “midstream” if the model is not producing the desired results, as this can introduce bias to the model.
“There are strong incentives to stick to the process and model and get it right,” Silver said.
Always look for new opportunities. Today’s sports ownership groups have diversified their holdings. Fenway Sports Management owns an English soccer team (Liverpool), a NASCAR racing team (Roush Fenway Racing), and an 80 percent stake in a television station (New England Sports Network)—in addition to the Boston Red Sox and Fenway Park.
Sam Kennedy, president of the Red Sox, said it’s important to identify “blue chip properties” that will help a company get through lean times. As an example, he pointed to Fenway Park itself. The park was built in 1912, and the previous owners of the Red Sox had considered tearing it down, but the current ownership group saw the park as untapped revenue opportunity as well as a competitive advantage.
After all, Kennedy said, “The Red Sox aren’t the Red Sox without Fenway.”
Don’t underestimate the human element. For all the discussion of analytics, conference speakers encouraged the audience of sports industry professionals, researchers, and students to remember the power of intuition and soft skills.
Analytics may help the Red Sox identify the companies most likely to buy an advertisement on top of the Green Monster, Fenway Park’s 37-foot leftfield wall, Kennedy said—but those companies only agree to pay for signs because the Red Sox take the time to build a relationship with them and understand their needs.
As Brian Lafemina, the NFL’s senior vice president of club business development, put it, “Everyone should be analytical, but not everyone needs to be an analyst.”