Devine at the 2012 MIT Sloan Sports Analytics Conference: Three takeaways from Day 1, three looks ahead at Day 2
BOSTON — We're at the halfway point of the 2012 MIT Sloan Sports Analytics Conference. After a jam-packed Day 1 schedule, most panelists and attendees scattered around the Hub in search of dinner, drinks and discussion of Descartes' Rule of Signs (probably).
Your Man, however, retreated to a neutral corner to think over what he'd just seen, try to put some rhyme to all that cold, hard reason, and nail down not only what piqued his interest from the first batch of sessions, but what he'd be looking forward to come the morning. (Besides a sweet continental breakfast, of course.)
Here's what I came up with. If you want to follow along with me for Day 2 of the conference, check out Ball Don't Lie's official Twitter account, @YahooBDL, and let me know what you're thinking.
THREE THOUGHTS FROM FRIDAY
1. Adam Silver has a personality, which is important. Like many NBA fans, my introduction to Adam Silver came in 2006, when it was reported that he would replace Russ Granik as the NBA's deputy commissioner and, in one of pro ball's greatest dumb little traditions, announcer of the selections in the second round of the annual NBA draft. And, like many NBA fans, my first extended exposure to Silver came during the league's fraught collective bargaining negotiations with the National Basketball Players Association and the resultant 161-day NBA lockout.
Day in and day out, Silver appeared alongside Commissioner David Stern to play ... well, I guess you can't really say "bad cop to Stern's good cop," so let's call it "worse cop to Stern's bad cop" in responding to union characterizations of the negotiations. When it came to the hard-line, fine-print details of what the owners would accept, Silver was often the one making the proclamations in a firm, resolved tone. (You may know it as the "I am the principal who is about to call your parents if you don't admit you threw the cherry bomb in the toilet, but actually I am going to call them anyway, I just want to hear you admit it" tone.)
Add all that up, and coming into Friday morning, my understanding of Adam Silver was that he was a smart, scary, cold and relatively nondescript force that seems poised to take over the NBA sooner rather than later. So it was a pleasant surprise that he was, as friend Beckley Mason of ESPN.com noted, "easily the most dynamic panelist" in the all-but-analytics-free (we'll get to that) conference-opening talk on the evolution of sports leagues.
Joined on the panel by NHL Commissioner Gary Bettman, über-agent Scott Boras, MLB Executive Vice President Rob Manfred and New York Giants Chairman Steve Tisch, Silver seemed at once authoritative and at ease in discussing the changing nature of leagues and league management. Touting the NBA's global focus, he casually mentioned that China is now the NBA's No. 2 market, behind only the United States, and that business there is going pretty well: "I think we just surpassed soccer as the No. 1 sport in the country."
Y'know, just the most populous nation in the world, with an estimated number of people playing basketball that's just a shade less than the entire U.S. population, and we're the top thing there. NBD, you guys.
Asked about who carries weight in contemporary labor negotiations, Silver got on the right side of the audience by playing to its interests. He referenced the critical role in the lockout's resolution played by analytical minds like Kevin Murphy, the players union's top economist, and noted the shift from lawyers to businesspeople at the negotiating table.
"The analytical people are more important than the lawyers themselves," Silver said.
In the case of the last round of NBA labor negotiations, their importance skyrocketed because of the widely publicized difference in opinion on the actual economic health of the league, which the NBA said was dire and the union said was probably quite a bit better than that. Bridging the knowledge gap wasn't easy, quick or painless, but to Silver's mind, having expert number-crunchers involved alongside the negotiators helped.
"I don't remember a time when analytics were so front and center in [a negotiation like this], with often competing models, frankly," Silver said.
In addition to playing to the crowd, Silver also cracked a couple of jokes. (OK, maybe not "jokes," but what passes for jokes at Sloan.)
Asked whether or not small-market teams like the Charlotte Bobcats can reasonably compete, both economically and on the court, Silver first promoted the deal the league just struck ("Part of it is the new CBA — we designed a system where we feel that well-managed teams have the opportunity to make at least some amount of money and at least be competitive") and then turned his attention to Bobcats owner (and now home-seller) Michael Jordan.
"I think there's no question that Michael is struggling a little bit in Charlotte," Silver said. "... He assures he is working as hard as he ever has in his life and is playing less golf."
Silver also took a gentle swipe at Bettman, who managed to veer briefly from his "the state of the NHL is strong" talking points to reminisce about working at the NBA in the early 1980s, when the NBA Finals were still being broadcast on tape delay.
"But Gary, you negotiated that deal," Silver interjected, eliciting laughs from the audience.
While the last few years haven't been the kindest to Stern's reputation as the greatest commissioner in American professional sport — and, in fact, the lockout may have done irreparable harm to that legacy — Stern has been widely beloved (or at least respected) by NBA fans due in large part to his persona. He is wry; he is withering; he is funny. Sure, he's fashioned himself into an emperor, but he's always been an entertaining one. Whenever he decides to abdicate his throne, the league will have lost a star. In a league of stars, where personality goes such a long way and the cults thereof matter so much, that absence will be felt.
At some point, Adam Silver is going to take the reins of the NBA from David Stern. He's not going to be that wry, that withering or that funny. He won't be the kind of star that Stern is. But on Friday, he showed that he can walk into a room to talk about something as dry as labor negotiations, set up on stage and show he's a person, actual and whole, rather than just a pointy-headed lawyer/business type. That matters.
(And if you doubted whether Silver's learned anything from working for Stern for the past six years, friend of BDL Zach Lowe notes at Sports Illustrated's The Point Forward blog that less than four months after striking a compromise on the collective bargaining agreement to end the 2011 lockout, Silver and the owners are already making noise about taking advantage of the opt-out clause available to them after the sixth year in that 10-year pact. The lesson: You're never done negotiating.)
2. Optical data tracking is still super sweet, and spatial and visual analytics could be the next big thing. Of all the analytics tools on display at last year's Sloan conference, perhaps the one with the most potential for massive dividends in the basketball world was STATS LLC's SportVU system. SportVU positions six special video cameras above the basketball court at different angles that capture, record and store all kinds of information on what's happening on the floor — the movements of all 10 players, the referee and the ball, who's running where, how much they're running, the height of the ball at different times and places, the flight path of passes, individual dribbles and about a million other things. It collects all this information, or optical tracking data, and then spits out reams and reams of stuff that look kind of like The Matrix. One of the big takeaways from last year's conference was that there's gold in them thar lines of code, if only you could figure out what kind of questions you should ask to extract it.
One of the two finalists for top honors in the research paper track of this year's Sloan conference, "Deconstructing the Rebound with Optical Tracking Data," uses SportVU data to track the height of the basketball after a missed shot. It traces the ball's descent it as it comes off the rim and goes down below 10, nine and eight feet off the deck to find out what happens to it, who's grabbing it, where they're grabbing it and what that can tell us about offensive rebounding.
University of Southern California researchers Rajiv Maheswaran (who presented the paper), Yu-Han Chang, Aaron Henehan and Samantha Danesis looked at about 11,000 field goal tries in an attempt to, as they wrote, "move beyond the outcome of who got the rebound." Instead, they wanted to learn deeper-dive stuff like how shot location impacted offensive rebounding rates, how likely a shot is to generate an offensive rebound and where players should position themselves to have the best chance of snagging offensive boards. Turns out that a shot from beyond the arc has the same likelihood of becoming an offensive rebound as one taken from just seven feet away, and both of them have a better chance of getting grabbed by a teammate than one taken from 11 feet, so if you fancy yourself a midrange shooter, you better be pretty damn good at it. (Or, y'know, take a few steps back and hoist a triple.)
One of the conclusions drawn by Maheswaran and his colleagues — that by emphasizing shot selection and firing away from areas where you stand the greatest chance of getting the carom, teams can grab a higher percentage of available offensive rebounds and potentially improve their effective field goal percentages, thereby getting a leg up in two of Dean Oliver's Four Factors — dovetails well with the project undertaken by the other finalist in the research paper track, Michigan State University associate geography professor Kirk Goldsberry.
The project, dubbed "CourtVision," aims to combine the practice of spatial analysis (look at how a given entity appears and acts in a given space, study clusters of activity, patterns, correlations, etc.) with the use of visual analytics (creating visual tools we can use to interact with a bunch of information, browse through complex sets of data, ask questions, etc.) to learn a whole bunch of stuff about what NBA players do — well or poorly, frequently or infrequently — from a ton of different spots on the basketball court.
As a case study and proof of concept, Goldsberry decided to try to use CourtVision to find an empirical answer to an often-debated question: Who is the best shooter in the NBA? He started by mapping out the "scoring area" — the 1,300-square-foot region on the court where 98 percent of NBA field-goal attempts occur. Then, he plotted out about 700,000 shots taken in every game from 2006 through 2011 — who took them, where they took them and what the outcomes of the shots were. Then he started making maps.
Goldsberry offers a new statistic for measuring a shooter's aptitude — "Shooting Range," a metric of how effective a player is at producing points from the greatest number of different spots on the court. A player's "range percentage" (Range%) measures the percentage of spots in that 1,300-square-foot scoring area from which a player will produce at least one point per shot attempted. The league leaders in Range% over the last five years mostly pass the laugh test — in order: Steve Nash, Ray Allen, Kobe Bryant, Dirk Nowitzki, Rashard Lewis, Joe Johnson, Vince Carter, Paul Pierce, Rudy Gay and Danny Granger.
When Goldsberry starts breaking down the maps, you start to see the value in understanding the unique spatial footprints and tendencies of every player, lineup and team in the league. You see just how fantastic Dirk is from the right baseline, an area in which none of the other top shooters even perform well. You see that Nash is a killer on wing three-pointers, but struggles in that right baseline spot that Dirk loves. You see that even deadeye Ray Allen has something akin to a three-point weakness — he's not so great from the left wing.
"Some players are good from some areas; some players are better from other areas," Goldsberry said. "We wanted to reveal those special spatial signatures."
The cool thing is, these specific deployments are just the tip of the iceberg. If the USC team can use SportVU for offensive rebounding data and Goldsberry can use CourtVision to identify the best and worst shooters, then why can't they use their systems to learn more about creating turnovers, or where most fouls happen, or, as Goldsberry noted, how to game plan?
"If I have Steve Nash's shot chart [and] if I have the Magic's defensive chart, then maybe I can figure out what kind of sets should be run to exploit their holes," he said.
The things Goldsberry, the USC team and others are looking at — finding ways to measure and visualize stuff that we might think or believe, but not know for sure — can not only help teams make better decisions; they can open the door to a whole new way of seeing the game. All we have to do is stop seeing The Matrix.
(For more on this emerging vein of analysis, head back over to The Point Forward for Lowe's brief but jam-packed breakdown of both papers.)
3. The conference just keeps getting bigger and bigger, which is both a good thing and a bad thing. This is less a basketball point than a more general observation, and it's a mouthful, so bear with me.
As I mentioned Friday, the 2012 edition of Sloan has more than 2,200 attendees, which conference organizers happily tout as a 50 percent increase in enrollment over last year's model and 13 times more people than the 175 who checked out the inaugural conference in 2007. The conference's "facts and figures" sheet offers a litany of exciting numbers (natch) detailing its growth — more than 700 students attending this year from more than 170 different academic institutions, 73 professional sports teams represented from six different pro sports leagues, eight discrete sports covered, nearly 200 combined submissions for the research paper and Evolution of Sport presentation tracks, and on and on.
Everything about Sloan seems to be gearing toward playing to larger audiences, from its ESPN-headlined roster of sponsors to its progressively nattier digs — last year's conference was held at the larger but more remote Boston Convention & Exhibition Center in South Boston; this year, we're at the much more centrally located Hynes Convention Center, right in the heart of Boston's Back Bay, simplifying travel and accommodations for heavy-hitting panelists and light-hitting bloggers alike. The bigger-better-faster-more thrust even extends to the quality of attendee swag bags; this year's is some handy ruggedized business that I'm pretty sure simultaneously keeps roasts piping hot and frozen margaritas nice and cool, even in the midst of a street fight.
Presentations in the giant main ballroom take place on and in front of a massive, glowing pink heart of a stage and dais. Klieg lights bathed the main drag of the convention floor as production workers prepared to film an episode of ESPN's stat-focused "Numbers Never* Lie" right in the middle of the hallway. For the second straight year, one of the most popular panelists at the show is Toronto Maple Leafs President and GM Brian Burke, who doesn't seem all that interested in analytics, but cuts an irascible figure on stage and curses on occasion, which is entertaining. (The same is true, albeit to a lesser extent, of ESPN color commentator and former NBA head coach Jeff Van Gundy, who eagerly stepped into a similar comic relief role during Friday's basketball analytics presentation.) There is a game room where you can see the Boston Bruins' mascot play air hockey.
Meanwhile, down a long hallway — in what's been branded the "NEXT Area," separated from the main ballroom and other primary panel spaces by a crush of humanity, an entire trade show, the TV setup and, as Jack Dickey wrote at Deadspin last year, throngs of "white guys in suits" itching for an elevator-pitch window of face time with one of the myriad front-office types in attendance — are the tiny-by-comparison rooms where the research paper presentations and Evolution of Sport talks take place. Where the lion's share of the actual, y'know, analytics discussion happens. Where the new theories and embryonic innovations are being shared, batted around and refined.
As writers like Jonathan Givony of DraftExpress suggested Friday, that's where the groundwork is being laid for the next crop of ideas that will help shape and enhance our understanding the way that, say, on-base percentage and Player Efficiency Rating have over the past two decades — where the stuff that could soon help us become way more knowledgeable fans than we are today is percolating. And it's shoved down the hall, out of the way, in the far corner of the convention center, behind the room with the air hockey table.
There are all sorts of devil's-advocate angles to this, of course. Logistics is a big one; more people want to see stuff like Bill Simmons going 12 rounds with Mark Cuban, so that gets the big room, and putting all the paper presentations together makes it easier for people who just want to see those to do so with ease. Also, I'd bet your paycheck that the folks giving the talks don't mind the lower relative profile, because the People Who Matter in front offices definitely know about the most interesting research being presented, and that's all that really counts.
Plus, it's not like the increased focus on the wider-net content has necessarily come at the expense of the nitty-gritty stuff; you may have to hunt a bit, but you can still find and attend those talks, and (in the ones I attended Friday) it looked like plenty of people were interested in staying on that track. If a conference this targeted in its focus can expand its scope a bit and still make good on its core mission, then why not steer more toward "something for everyone"?
To my mind, though, the chief devil's advocate argument is that without the infusion of the more broad-based/all-but-analytics-free discussions, higher-profile names, ramped-up production values and all the rest, there wouldn't be the level of interest, ticket sales, sponsorships and partnerships that make putting on this kind of show possible. Ditto for funding the $7,500 grand prize and $2,500 runner-up prize that go to the writers of the top two research papers.
Without this kind of show, where can that next wave of innovation actually break? If you want people to find religion, you could do worse than to start by getting them to church. But does it have to be, like, a megachurch?
THREE THINGS I'M EXCITED FOR SATURDAY
1. Adam Gold's "How to Cure Tanking" presentation in the Evolution of Sport track. Is unweighting the NBA draft lottery really the answer? Are other ideas worth considering? I will be there representing all fans of the Charlotte Bobcats.
2. The 1 p.m. to 2 p.m. EOS session. "Quantifying the Force of a Monster Dunk" and "Redefining the Positions in Basketball"? Yes, please.
3. Seeing how many members of Grantland I can get to autograph the free copy of "Grantland Quarterly" that came in my swag bag. I'm going to guess it will be "most, but not all." Wish me luck, gang!