May 25, 2011; Dallas, TX, USA; Dallas Mavericks owner Mark Cuban celebrates with forward Dirk Nowitzki (center and Jason Kidd (right) after a victory in game five against the Oklahoma City Thunder for the Western Conference Finals of the 2011 NBA playoffs at American Airlines Center. Mandatory Credit: Matthew Emmons-US PRESSWIRE

Basketball Analytics from the MIT Sloan Sports Analytics Conference

Use your ← → (arrows) to browse

Matthew Emmons-US PRESSWIRE

This past weekend I was fortunate enough to attend the MIT Sloan Sports Analytics Conference in Boston. The conference was founded in 2006 and is co-chaired by Houston Rockets GM and MIT alum Daryl Morey and Jessica Gelman.

The conference is basically the most sports nerds ever gathered in one place during the year listening to panels and presentations on all of the cutting edge analytics in sports today.

This year the Godfather of sports analytics, Bill James, was in attendance. Baseball was really the first sport to delve into statistics and use analytics at the professional level and was most recently featured in the movie Moneyball.

Basketball is on the rise in this department. Basketball is different than baseball in that it is less measureable because there is less one-on-one competition in the games. Plus-minus is a useful stat for a player but there are 10 players on the court at once.

At the conference there were probably more presentations on basketball than any other sport. The big moment of basketball analytics that kept being brought up was the Dallas coaching staff’s decision to start J.J. Barea in the final two games of the NBA Finals versus the Heat last season.

The decision was largely influenced by analytics determining Barea’s effectiveness while on the floor with certain players and what Dallas’ offense was missing at the start of games (penetration mostly).

Right now in terms of basketball analytics were at a stage where there is a lot of data and much less information on how to best implement and interpret the data. We can see what percentage a player shoots at every exact location on the floor but how useful is that information and how do you use it to your advantage.

“A lot of analytics is just noise,” says Dean Oliver, ESPN’s Director of Production Analytics.

A big consensus from many of the panelists was just that. A lot of times analytics were just confirming what already was thought about something. And analytics aren’t going to turn the Charlotte Bobcats in an NBA Finals team either. Sometimes the effect they have can be very small.

“I asked Daryl [Morey] if we shot two-for-one every time how many wins would that translate into,” Jeff Van Gundy said. “He said about one every two years.”

Relaying the information and translating it into coaching is another thing. Van Gundy, who is very skeptical and not that into analytics, often was making remarks about how some things you just can’t tell players because you don’t want them over-thinking during games.

Gary A. Vasquez-US PRESSWIRE

Jackie MacMullan mentioned coming across a stat depicting Kevin Durant’s shooting percentage based on how many dribbles he took. He shot 55 percent when he took no dribbles, 58 percent with 1-to-2 dribbles, 39 percent with 3-to-4 dribbles and 40 percent with 5-plus dribbles.

“I wouldn’t tell a guy you’re 39 percent on 3-to-4 dribbles so dribble the fifth time to go up to 40,” Van Gundy joked.

It also depends on the player. Van Gundy said there are things you can tell Shane Battier and you know he will understand you but that’s not the case for every player in the league.

“Coaches might have 20 things they want to implement in practice that they believe will help win,” Mike Zarren, Assistant General Manager of the Boston Celtics said. “You can only focus on six in practice and the players might remember four of those things and execute one in the game.”

It almost sounds hard to believe something like that. But especially during this condensed season when the players have almost no time for practice, scouting in this way can be very tough.

I go back to Tyreke Evans and how successful he was in his rookie season in the NBA. One of the things about Evans when watching the games was that you knew he would always go to his right hand to finish, always. It was amazing to see players not adjust to this. Then I heard how sometimes it can take an entire year before scouting reports come out on a player and are implemented in games

Gary A. Vasquez-US PRESSWIRE

A popular area that basketball analytics cover is crunch time stats. We all know how Kobe Bryant has made the most game-winning shots of any current player in his career but he also takes the most.

“There’s no accepted definition of what crunch time is,” ESPN columnist John Hollinger said. “There’s so little data over one season that random noise is going to blow away anything you can determine as an actual skill. You really need to look at multi-year data to have anything even remotely convincing as far as making conclusions especially predictive conclusions.”

It’s also tough even to do that because players aren’t the same for their whole careers. They’re not robots and that seems to be sometimes overlooked in a way with these analytics.

“Guys aren’t static either,” Zarren said. “You become a better player over time.”

The game also changes during crunch time as there are less transition opportunities and defenses are usually set resulting in a lot of isolation offense.

“It’s not who elevates their game when the game is on the line it much more of who Wilt’s,” Zarren said. “The first thing you want to look at is the team. If the team is always good in those situations then there’s probably someone on the team that’s doing something worthwhile.”

One thing everyone can agree on is that the playoffs — when there are fewer games per week and you can prepare for one team at a time — analytics can become the most impactful.

We saw it last year with the Mavericks and we’ll probably see it again this year with someone. Every team has analytics. It really just comes down to who knows how to use it.

Analytics might make an impact on one or two plays throughout the course of a series and that doesn’t seem like much but remember that the Miami Heat were just two plays away from sweeping the Dallas Mavericks last year.

Use your ← → (arrows) to browse

Tags: Basketball Analytics Boston Celtics Dallas Mavericks Daryl Morey Dean Oliver ESPN Houston Rockets J.J. Barea Jackie MacMullan Jeff Van Gundy John Hollinger Kevin Durant Kobe Bryant Miami Heat Mike Zarren MIT Sloan Sports Analytics Conference Shane Battier Tyreke Evans

comments powered by Disqus