How Analytics Help Build this Champion
Posted by Tom Davenport on January 31, 2008 8:54 AM
Last spring, on baseball’s Opening Day, I confidently identified the Boston Red Sox on these very pages as the eventual World Series winner—based in part on their analytical prowess. You may recall that I was correct in that prediction. This Sunday, I will go out on a much more solid limb and pick the Patriots to triumph in the Super Bowl. I’m more of a baseball guy than a football nut, but fortunately both of the Boston teams I cheer for are not only winners of late, but also heavy users of analytical approaches to their respective games (the Celtics aren’t doing badly either, but I think Kevin Garnett is more of a factor in their success than any statistician).
Like the Red Sox (or any analytically-oriented sports team, for that matter) the primary analytical application for the Pats is selecting the best players for the lowest price. This is particularly critical in the NFL, with its stringent salary cap. In-depth analytics helped the team select its players and conserve its dough. (Until last year the team had only a middle-ranking payroll in the National Football League, but now Tom Brady is getting expensive!) The team selects players using its own scouting services rather than the NFL-generic one that other teams employ; Brady, for example, was the 199th pick in 2000. They rate potential draft choices on such nontraditional factors as intelligence and willingness to subsume personal ego for the benefit of the team (though I had my doubts about their fidelity to that variable when they signed the famously mercurial Randy Moss before this season).
The Patriots also make extensive use of analytics for on-the-field decisions. They employ statistics, for example, to decide whether to punt or “go for it” on fourth down, whether to try for one point or two after a touchdown, and whether to throw out the yellow flag and challenge a referee’s ruling. Both its coaches and players (particularly quarterback Tom Brady) are renowned for their extensive study of game video and statistics, and head coach Bill Belichick has been known to peruse articles by academic economists on statistical probabilities of football outcomes—over breakfast cereal, the legend goes.
Off the field, the team uses detailed analytics to assess and improve the “total fan experience.” At every home game, for example, twenty to twenty-five people have specific assignments to make quantitative measurements of the stadium food, parking, personnel, bathroom cleanliness, and other factors. The team prides itself not only on scoring the most points ever this season, but also on having the lowest wait time for women’s restrooms in the NFL. External vendors of services are monitored for contract renewal and have incentives to improve their performance. This won’t help them win the Super Bowl, but it helps fill Gillette Stadium every home game.
Belichick deserves a lot of credit for the analytical emphasis (God knows, he can’t get by on charm), but so do the team’s owners. Just as the Red Sox owner John Henry moved the Sox in an analytical direction, Bob and (especially, I’m told) Jonathan Kraft believed that analytics could make a difference in football. Jonathan is a Harvard Business School alumnus and a former management consultant. In addition to Belichick, they hired Scott Pioli, a former Wall Street investment analyst and now the “player personnel” guru.
The only thing the Patriots lack is an analytical secret weapon equivalent to Bill James, the god of baseball statistics who acts as a “senior adviser” to the Sox. I’m not sure there is a Bill James of football. If there is, the Pats need to hire him (or her). Such a move could keep the Patriots dynasty going for many years to come.
From Harvard Business’ discussionleader.hbsp.com, posted by Tom Davenport on January 31, 2008. URL http://discussionleader.hbsp.com/davenport/2008/01/how_analytics_help_build_this.html
Tim Berry is the founder and chairman of Palo Alto Software and Bplans.com. Follow him on Twitter @Timberry.
