Ben Brostoff, a columnist for The Chronicle’s opinion section, recently got the chance to attend MIT’s Sloan Sports Conference, a one-day event focused on the increasing role of analytics (stats) in sports. At the forum, Brostoff rubbed shoulders with sports luminaries like writer Bill Simmons, Dallas Mavericks GM Mark Cuban and legendary 3-point assassin Steve Kerr, among others. Brostoff wrote this exclusively for The Chronicle Sports Blog.
Take a look at the attendees list for the MIT Sloan Sports Conference when you have a chance.
On March 6, the Boston Convention & Exhibition Center housed more influential minds at the tops of their respective sports than probably any previous gathering in human history. There were over 1,000 people at Sloan from sports management, media and fandom, plus an additional 400 on the waiting list. Panel discussions generated lines bigger than the ones at Alpine on a Monday morning: you had a better chance of finding an empty seat at Cameron that same night. Welcome to the big show. Everywhere you walked, there was a recognizable face. Is that Bill Simmons talking to Brian Kenney about Sugar Ray Leonard? Are Adam Silver, Daryl Morey and Mark Cuban really humoring three Harvard kids waving resumes? Wait… that can’t be… Steve Kerr? And a ragtag team of representative from MLB, NFL, NHL, FIFA, ESPN, Reebok, Nike, Bloomberg, EA Sport and Black Rock? In short, if you were anyone of consequence in sports, you were at Sloan.
You would figure, then, that Duke athletics would have a few representatives. Sports analytics should in fact be synonymous with Duke itself. This is how Duke has marketed itself for two decades: sports with an analytical, cerebral bend. Shane Battier, as portrayed in Chris Ballard’s “The Art of a Beautiful Game” and this must-read NY Times feature, perhaps best embodies this distinct Duke quality. Battier’s knowledge of analytics—for instance, Kobe Bryant’s effective field goal percentage from different zones on the floor, Manu Ginobli’s statistical tendencies—render him the unique player who combines athletic prowess and the scientific method. Battier’s four years at Duke weren’t marked by his athletic prowess a la Jason Williams, but rather a blend of ability and brains. He consistently made the right defensive rotations and rarely attempted low-percentage shots. Indeed, the Battier-style of play is pretty much the accepted standard for your typical Duke teams. Capable and smart.
Ironically, the Duke basketball brass might not actually be that smart by today’s sports standards. The team currently employs zero official basketball quants (think Dean Oliver of the Denver Nuggets) and does not delve into unorthodox statistical analysis (efficiency ratings, adjusted plus minus ratio, etc.) as a means of strategy evaluation. At least, that’s the team’s public relations stance. When The Chronicle’s Ben Cohen did a column on tracking offensive and defensive efficiency, he got this quote from assistant coach Chris Collins: “The numbers we use a lot are turnovers and offensive rebounds. The other key is we try to get ourselves to the free-throw line. Those are probably the main ones that we look at—and obviously, well, shooting the ball.”
If we are to take Collin’s words at face value, then the team is doing only the most rudimentary statistical basketball analysis possible. Turnovers, offensive rebounds and fouls essentially mean nothing in a vacuum: they’re variables dependent on a range of factors, from pace to score of the game (What happens to John Scheyer’s assist/turnover ratio when Duke is up 10 with 5:00 left and K elects to run the shot clock down?) to lineups (What is Duke’s offensive rebounding like when the Plumlee brothers are at the 4 and 5? How about Thomas and Zoubek?). All of these factors, according to the basketball’s best and brightest—namely, Oliver, Morey, Cuban, Kevin Pritchard, John Hollinger and Mike Zarren (Celtics’ Basketball Operations Analyst)—can be relatively accounted for and weighted appropriately with enough patience and tinkering. Sure, these new statistical tools might not be perfect, but they’re certainly better at describing what actually happened in a game than the box score as constituted. The traditional box score might as well be thrown out the window. Points, free throw discrepancy, assists (which, by the way, are wholly determined by subjective scorekeepers), blocks and rebounds are worthless without context. Statistical analysis is the only empirical means of providing a way to corroborate what our eyes see with numbers on paper—the basketball guys at Sloan made this idea abundantly clear.
Yet, as I scrolled down the program and looked for Sloan’s Duke representatives, only one name appeared: mine. To be fair, I’m not sure any ACC schools sent their basketball people to Sloan. The analytics movement is mostly a pro-phenomenon, which makes sense. In the NBA, sample sizes can be large (five and six year analyses are possible), and data is less diluted. What I mean by the latter term is that everyone’s strength of schedule in the NBA is more or less the same, whereas in the NCAA comparing ’09 Memphis with ’09 UNC is virtually impossible: Memphis played in a watered down, woefully bad C-USA, while UNC competed in a conference chock-full of tournament teams and lottery picks. Perhaps Duke Basketball has no desire to invest in analytics because it has no legitimate application.
I for one am not sold on that train of thought. At the end of the day, the hard numbers, if interpreted and mathematically manipulated correctly, can tell a compelling sports story. A story, I would posit, that is more logical and accurate than any type of qualitative analysis. The statistical movement is now a staple of MLB and dominating the general management of the best NBA organizations. Not coincidentally, the eight NBA teams heavy on quantitative methodology—the Celtics, Lakers, Rockets, Thunder, Nuggets, Mavericks, Magic and Cavs—all are in prime position to contend for championships in the next several years. These teams unite around the idea that the same prized concepts we use in science and math should necessarily be a part of sports. These organizations’ decision-making processes are no doubt more rigorous and objectivity-based than their peers (just take a look at some of the research papers presented at Sloan—might I recommend Brian Skinner’s paper, The Price of Anarchy, which links hoops offensive schemes with Braess’s Paradox?). A strong analytic-minded managerial core is not sufficient to win a championship (cue discussion of talent and execution), but I’d argue it’s becoming progressively necessary.
College sports has yet to embrace Michael Lewis and the “Moneyball” line of thinking. Duke is in prime position to lead the way, boasting some of the best number-crunching undergrad and grad students in the world who will work for virtually nothing in order to get closer to the ground K and his cronies walk on. I doubt that K, Collins or any of the coaching staff knows much about Markovian chains, linear regressions or noise in data sets. However, there’s a plethora of students here who do, and they should be utilized effectively, even if it’s just in an experimental fashion. At the very least, Duke’s varsity teams should take a look at the value of analytics in their respective sports and see what the movement has to offer.
I’ll even save a few seats at Sloan next year.