A whole new world of NBA statistics
The opening, star-studded panel discussion, titled "Revenge of the Nerds" -- complete with music from the 1984 movie -- was over, and now much of the crowd had reassembled down the hall to prove the point.
There, before an SRO crowd, was Matthew Goldman, a PhD candidate in economics at UC-San Diego, holding court on the risk and efficiency among NBA teams as they attempt three-point shots in various late-game situations.
The audience at the recent MIT Sloan Sports Analytics Conference seemed engaged in the findings, which I would share here if not for the fact taht I would require a sophisticated understanding of math and diagrams and stuff, so . . . no.
Any questions? Well, actually, there were, so a line formed in front of Goldman after his presentation.
Now every sport is the subject of intense study that goes beyond intuition and deeply into data. But no sport is more of the analytics moment than basketball.
With more variables than baseball but fewer than football, hoops is hot among greater geekdom.
"Basketball is definitely the second wave," said Rockets general manager Daryl Morey, a founder of the conference, which since 2007 has grown from 100 true believers to 2,700 true movers and shakers earlier this month.
"Baseball will always be more able to be analyzed than our sport. We have more moving parts. But I do think we're sort of at the peak of analysis of basketball. I think there's still a lot to come as more data comes online."
No one knows more about the stats revolution than Cuban, who was on that "Nerds" panel, moderated by Lewis.
Cuban was on board in the early 2000s, and recalled other owners being baffled, even angered, by his theories. "It was crazy," he said. "You couldn't really have a conversation back then with anyone else in the NBA. Either I didn't know who to talk to, or they didn't know what I was talking about."
Now, "there's no friction whatsoever," Cuban said of a world in which NBA.com recently freed a trove of advanced data to anyone who wants to study it.
"Our view is the deeper a fan's understanding is of what is happening on the court, the more time they will spend following the sport, the more engaged they will be," he said.
Silver said analytics already are deeply embedded in NBA culture, among teams and fans.
"There's a huge human element to it as well, but there is no team at this point that doesn't see value to analytics," he said. "How heavily they rely on it differs from team to team, but certainly I think we're all seeing the confluence of this next generation, where it seems to fit our demo perfectly in terms of fan interest."
The new-wave statistics take getting used to for fans accustomed to points, rebounds and assists. Among the easier to grasp is net rating, which is similar to a plus-minus rating in hockey. Example: Before Amar'e Stoudemire's most recent injury, the Knicks averaged 3.0 more points per 100 possessions when he was on the court than not.
Another feature: Shot charts that break down teams' and players' percentages from 15 areas of the court.
That sort of thing is fodder for fans and journalists. But for general managers and coaches, it is part of an increasingly complicated, high-stakes game. (The Lakers were the only NBA team not represented at the Sloan conference.)
"I think we're nowhere in basketball -- literally nowhere," Morey said. "I think what teams are going to be running offensively and defensively [in the future] is going to be completely different."
The hot new tool is the ability to chronicle players' movements, at 25 frames a second, using technology developed for missile tracking. Half of NBA teams use "SportVU" to deepen understanding, especially in the difficult sphere of assessing defense.
So for example, "Tony Parker is really good at changing the shape of the defense," said Kirk Goldsberry, a visiting scholar at Harvard and an assistant professor of geography at Michigan State. "Now we can quantify that."
"I'm not making a judgment either way, just pointing to things that surprise you," said Brian Kopp, who oversees SportVU for STATS.
That is where the teams come in. Experts uniformly say their biggest challenge is not collecting data but translating it into digestible, acceptable form for players and coaches, who then must convert it onto the court.
It helps that GMs such as Morey increasingly have analytic backgrounds. But even sympathetic coaches argue that numbers go only so far.
On one hand, former Heat and Magic coach Stan Van Gundy said he accepted certain mathematical verities, such as the least efficient shot is a jumper from just inside the three-point arc.
"In Miami, they would be killing us because we didn't have guys with a midrange game," he said, "and I felt like saying, 'Thank God!' because of the efficiency of those things."
On the other hand, Van Gundy said, taking a rushed shot because the math supports it in a two-for-one possession situation late in a game can undermine a coach trying to establish a culture of smart shot-taking.
"A lot of analytics people think the game is a video game," he said. "What's going to separate teams, like in every other area of life, is how you use the information."
Few in the league would argue that.
"You have to have more of a people feel, an experiential feel on top of that," Celtics co-owner Steve Pagliuca said. "It's a great tool, but it's not everything."
TNT analyst Charles Barkley took an old-school stance when asked about analytics this week, saying: "I don't believe in that sort of crap . . . It comes down to what I see. When I watch basketball, I know what I'm looking for."
But analytics are here to stay, which is why ESPN was a presenting sponsor of the Sloan conference. When asked about basketball analytics, ESPN president John Skipper was able to spout defensive efficiency ratings of recent Final Four teams off the top of his head.
More and more, avid fans can do the same, which will fuel demand for more. The question is whether there will come a point at which everyone has the same data, making it difficult to find an edge.
The answer, for now, is that point is a long way off, especially for defense. But there has been progress. The Bucks' Larry Sanders emerged as an unlikely Sloan star thanks to his positive effect on interior defense.
"I think as we learn more about that, we'll get better, but are we at the Holy Grail yet? No, I don't think so," Pacers GM Kevin Pritchard said of analyzing defense.
"When progress happens in a certain area, it becomes harder and harder," Celtics assistant GM Mike Zarren said of the data arms race. "The question is, will we get other data, other sets of information that are useful? I don't know. There is no question we're not there now."
But we are getting there. John Hollinger, a journalist who dove into basketball analytics in the mid-1990s and created the player efficiency rating, worked at ESPN until late last year.
In December, the Grizzlies hired him as vice president of basketball operations.