I'll remember the 2010 Sloan Sports Analytics Conference, held Saturday at the Boston Convention and Events Center, as the point I realized that statistical analysis had become mainstream in the NBA. The role of analysis is a topic I write about as much as anyone, I suspect, including last Friday's column on "The State of APBRmetrics." Still, I apparently missed just how much things had changed.
What sold me? Not the 1,000 people in attendance, though it was a stunning turnout that Bill Simmons joked "broke the record for most dudes in one room" despite the 400 people on the waiting list who could not get in. It wasn't even the number of NBA team employees or consultants present in Boston, though between panelists and attendees half of the league was represented by my count. No, what surprised me was the respect accorded to the numbers by members of the two groups traditionally most resistant--coaches and former players.
The last panel of the day, on the subject of coaching analytics, featured former Dallas Mavericks coach Avery Johnson and long-time NBA guard Brent Barry, both of them now TV analysts. Admittedly, neither Johnson nor Barry is a stat guy by nature. Yet their experience with two of the leaders in the field--Johnson implementing statistical analysis at owner Mark Cuban's behest in Dallas, and Barry when he finished his career playing for GM Daryl Morey and the Houston Rockets--appeared to have largely won them over.
Johnson explained how the lineup data by which Cuban swears influenced his coaching, especially when it matched what he thought he was seeing on the floor, noted that the effectiveness of the corner three-pointer has influenced his philosophy of help defense and took a stat-friendly position on the benefit of fouling when up by three late in the game. (When Johnson asked for hands in the crowd, easily 80-90 percent said they would foul in a situation with five seconds left.)
I recall Barry being skeptical of analysis when I covered him as a player in Seattle, despite the power of self-interest: His True Shooting Percentages were typically off the charts and explained his value far better than conventional numbers ever could. Yet he praised the information the Rockets are able to offer their players should they want it, including a scouting report "as thick as a bible." Barry also argued that as younger owners buy into the league, support for analysis should grow.
Now, this was a conference of stat geeks, by stat geeks and for stat geeks, so to some extent hearing positive opinions is no surprise. However, Sloan's organizers wanted to foster some healthy debate on the subject, which explains why the day's marquee panel was titled "The Limits of Moneyball: What the Geeks Don't Get." At times, New England Patriots president Jonathan Kraft and Indianapolis Colts president Bill Polian offered that perspective, the later delivering a particularly withering critique of game analysis that undermined the good work that has been done by Football Outsiders and others.
The basketball guys on that panel--Cuban, Morey and Simmons--weren't having it. Cuban's main argument throughout the day (except when he delivered a read-between-the-lines pitch for a LeBron James sign-and-trade, as basketball analytics moderator Marc Stein called him out on doing) was for the NBA to get more heavily involved in the process of tracking stats, which he said "would save all of us a lot of time and money." Simmons, asked by moderator Michael Lewis to share what annoys him about the statistical revolution, noted only that at times the analytics community struggles to convey its message to the public in an understandable way.
There were certainly several different levels of understanding during the presentations that were part of the new research paper competition. I made it to the two finalists in the non-academic category, Joe Sill's explanation of regularized plus-minus (as seen on his site, hoopnumbers.com) and Brian Skinner (no, not the Clippers center) on how a traffic theory (Braess' Paradox) can help explain why the Ewing Theory might have a basis in math (sadly, Simmons did not make the presentation). Some of the math, especially in Sill's presentation, was clearly too technical for some of the attendees. At the same time, that kind of sophistication was demanded by those with backgrounds in formal statistics who posed rigorous questions at the end of each presentation.
The panels, especially The Limits of Moneyball and the basketball analytics session, were much more light-hearted. Members of competing front offices who are nominally competitors happily traded jokes with each other. Cuban repeatedly emphasized his belief that metrics are best used on the coaching side, while Portland general manager Kevin Pritchard said that statistics help the Blazers more in terms of player personnel. The basketball analytics panelists--in addition to Cuban and Pritchard, Boston's Mike Zarren, Denver's Dean Oliver and ESPN.com's John Hollinger--also expressed uniform excitement about the possibility of the video tracking being developed by the league that I mentioned in Friday's column.
One of the strangest articles I've read about the growth of statistical analysis in basketball was one in the Wall Street Journal last year that described the Nuggets' front office as "an anomaly" in the NBA in part because the team used gut instinct instead of the numbers. Leaving aside the fact that Oliver is as significant a member of the Denver staff as any statistical analyst outside of Morey, or that key adviser Bret Bearup (who was also at the Sloan Conference) pays attention to statistics, what struck me as odd was the notion that the use of analytics had somehow become the status quo for NBA teams.
I still don't think we're quite to that point, not when several teams, including some of the league's best, are uninterested at best and scornful at worst of statistical analysis. TrueHoop's Henry Abbott wrote Friday in his interview with Oliver that at least one GM could not comprehend why he would want to attend the Sloan Conference. But are we closer to that day even than we were a year ago? After Saturday, my answer is an unqualified yes.
A few other observations from the Sloan Conference
- Give credit to Oliver, one of the best at thinking of ways to get his point across, for coming up with a pithy quote that explains the importance of blending statistics and observation in the NBA. "Individuals see a basketball game better than the numbers," he said, "but the numbers see all the games."
Cuban was forthcoming as always, and a couple of players took a beating. He explained that Gerald Green "just doesn't understand the game of basketball," which was why he could put up numbers that were promising at an early age but never improved. Cuban also cited the signing of Evan Eschmeyer to a six-year, $20 million contract as an early mistake that happened when the Mavericks did not realize how inconsistent adjusted plus-minus results could be from year to year.
I thought one of Cuban's more interesting points was when he said that statistical analysis was more helpful at the bottom of the roster. This makes sense, since there tends to be agreement between the numbers and scouting on the truly good players but more discrepancies further down the talent distribution.
- For me, the biggest disappointment of the conference was not hearing more from Baseball Prospectus legend Nate Silver. Silver was placed on the coaching analytics panel to offer a contrast to the sideline veterans but ended up overshadowed by Johnson and Buck Showalter.
Follow Kevin on Twitter at @kpelton.
Kevin Pelton is an author of Basketball Prospectus.
You can contact Kevin by clicking here or click here to see Kevin's other articles.