Sloan Sports Analytics Conference, Day 1: Ambition And Curmudgeons

Here are some of the things I learned during the first day of the MIT Sloan Sports Analytics Conference in Boston:

Bill James is a golden god. There is almost no question that a sports nerd convention drawing 2,200 people (and more come each year) would have been impossible without James' contributions to the baseball world, and all 2,200 people knew it. After he took part in the "Baseball Analytics" panel, the crush swarming the stage was incredible. By estimate, he fielded approximately 14,376 "What's your advice on…" questions from ambitious MIT students after the panel, and I assume he answered just as many after his filming of a live B.S. Report with Bill Simmons to close Day 1.

Brian Burke is a good quote … and not a fan of Moneyball. The general manager of the Toronto Maple Leafs, Burke took part in two seminars yesterday: "Hockey Analytics" and "Art & Analytics of Negotiation." A quote from the former: "Goddammit, there's more to this than looking at statistics and picking players out of a goddamn hat." A quote from the latter panel (in describing his negotiation style): "I try to be a prick all the time."

Players play. A visibly uncomfortable Rocco Baldelli took the stage with superagent Scott Boras, golden god Bill James, Cleveland Indians president Mark Shapiro, Houston Astros GM Jeff Luhnow and moderator Rob Neyer. When asked about the role of analytics in the way players approach the game, Baldelli simply said "Most players, they go out and they play." Fair enough.

Jackie MacMullan is not Mark Cuban. In kicking off the "Basketball Analytics" panel, the always entertaining MacMullan apologized for Cuban's absence (he was unable to make it to town for Day 1, and considering the disgusting weather outside right now, we'll see if he makes it for Day 2). MacMullan simply said, "I'm sorry I'm not Mark Cuban. We're all sorry we're not Mark Cuban." Still, Cuban's absence paved the way for a new star of the conference: Jeff Van Gundy.

Jeff Van Gundy is a funny, funny man. The "Basketball Analytics" panel included MacMullan, Van Gundy, Celtics assistant general manager Mike Zarren, ESPN's stats guru Dean Oliver and ESPN's other stats guru John Hollinger. Van Gundy dominated. He talked about the role stats could play in getting a message across to a player -- stats can be used as evidence that both the coach is right and that he's not just picking on a player for no reason -- and explained that, when he didn't have a statistic readily available for assisting his point, he would just make one up. Oh yeah, and …

Jeff Van Gundy is not a fan of Linsanity. After acknowledging that four of the five people on the panel were employed by ESPN, he said the following: "ESPN bludgeoned you with stories of Lin's 'struggle.' 'Struggle' compared to whom?" He went on to discuss how Lin was still in the NBA for a good portion of last season, how minor league baseball players encounter a struggle far worse than anything Lin has dealt with to date, and how other NBA players have taken longer, harder paths to the big time than he has. "Let the story breathe a little bit before you just make it up." (Meanwhile, others struggled to make cogent remarks about Linsanity at all. Scott Boras compared Linsanity to the Ichiro Suzuki story, while Michael Wilbon said that Lin is "making people look at basketball in a different way," meaning Lin's story has illustrated how international the game has become. Um, Lin was raised in Palo Alto, guys. Not China. Not even Taiwan.)

Gary Bettman hogs the mic. Taking part in the opening panel, "In the Best Interest of the Game: The Evolution of Sports Leagues," with Boras, MLB Executive Vice President Rob Manfred, NBA Deputy Commissioner Adam Silver, New York Giants co-owner Steve Tisch and moderator Michael Wilbon, Bettman had the habit of … talking for a while. He insisted that the NHL was "recovered the day we reopened," talked about how well the NHL is doing today (and with evidently $3.2 billion in revenue over the last year, it is difficult to argue with that), explained the dilemma of Olympic Hockey (when the Olympics are in North America and/or feature a U.S./Canada final and all the viewer interest that accompanies that, stopping the NHL season for weeks because of the Olympics is a very good thing. When it's in Europe, and neither Canada nor the U.S. make the finals, it is a major hindrance), and talked about how far sports leagues have come since he started in the NBA 31 years ago. Silver was able to get in some jabs and quick wit, however. When Bettman was discussing how the NBA finals were shown on tape delay 30 years ago, Silver jumped in by giving him hell for negotiating that deal. Bettman wagged his finger and, without smiling, insisted that he inherited the deal. Uncomfortable moments can be fun!

Nerds poke holes in things. Throughout the conference, contest winners have been presenting sports-related research papers to a pretty large audience. The first two had the following titles: "Experience and Winning in the NBA" (PDF) and "Using Cumulative Win Probabilities to Predict NCAA Basketball Performance." (PDF) Both presentations were limited in scope, due simply to how large the scope could become (the first included a huge sample of NBA player data, the second included years of college basketball play-by-play data) and the fact that the writers are smart, busy college students. Still, the urge to take them down was evident in the Q&A sessions afterward. "Don't you think you missed some causation by…" "Don't you think your sample size is too small?" "(Paraphrased) Actually, I have six more follow-up questions." Et cetera. Never change, nerds.

There are no absolutes in analytics. So stop asking. Enjoy the beautiful, beautiful gray area.

Chemistry matters. Perhaps the most interesting conclusion of the first research presentation was the effect that roster continuity can have on NBA playoff success. While there is no statistically significant tie to continuity and regular season wins, keeping the same team intact (for the most part) can add an extra few wins to your postseason ledger. Obviously this doesn't lead to any sort of black-and-white "DON'T MAKE ANY MOVES!" conclusion, but it does suggest that there is value in not making moves you don't have to make.

Scott Boras is a member of #teamcake. The superagent, who looks like a direct cross between Gene Chizik and Tom Delay in person, made multiple "making the cake" recipes in discussing player valuation. Take that, pie lovers.

Coaching analytics do not exist. The seminar entitled "Coaching Analytics" featured Van Gundy, former NFL coach Eric Mangini, Bill Simmons, Detroit Lions defensive end Lawrence Jackson and was moderated by Sloan Conference organizer and General Manager of the Houston Rockets Daryl Morey. In no way were "analytics" even slightly discussed, but it was certainly a fun conversation, mostly because of Van Gundy and his personal instigator, Simmons. When Daryl Morey asked what mistakes are typically made in hiring coaches, Simmons predictably (and justifiably) went after the trend of "recycling coaches" instead of giving a new guy a shot at the top job. This irked former coach Van Gundy, who asked for examples. The first was the Charlotte Bobcats' hire of Paul Silas, and Van Gundy quickly interjected that "the coach is not the problem in Charlotte." (That the topic ended with Morey agreeing that "GMs should get fired way more often" showed both that the conversation was entertaining and that Morey is unafraid of losing his job anytime soon.)

Basketball is catching up. Thanks mostly to James, advanced stats in baseball got an enormous head start. But the amount of talent and brain power attacking the game of basketball right now is both staggering and exciting. The research presentations relating to basketball appeared to get by far the highest attendance, and Twitter blew up during the "Deconstructing the Rebound with Optical Tracking Data" presentation (PDF) at the end of the day. As James said during the B.S. Report, "In a lot of ways, [basketball nerds] are better at getting their stuff on the floor." As he put it, "It is easier to do entry-level analysis on baseball than basketball," meaning those working on basketball are pretty damn smart. So they've got that going for them.

The Internet has been huge. As James put it, "I was basically blogging before people were blogging." In his first four years writing his abstract, selling it through an ad in the back of the Sporting News, he never once sold even 1,000 copies. His trajectory picked up after a Sports Illustrated profile, but he was still having only so much of an impact after a decade. But in the last decade, the Internet has been very, very good for both James and the proliferation of sports statistics. The reasons are obvious: 1) It is so incredibly easy to find like-minded individuals in the Internet age, and 2) wow, is there so much data readily available today. Hell, Jeff Luhnow referenced Fangraphs.com during the "Baseball Analytics" discussion. That alone says something. It's a good time to be a nerd, and that was constantly reaffirmed on the conference's first day.