Friday, March 2, 2012
Fact or fiction: Basketball analytics
ESPN.com and the TrueHoop Network
The TrueHoop Network has stormed Boston for the MIT Sloan Sports Analytics Conference. Five bloggers give us their thoughts on the Basketball Analytics panel that featured ESPN.com's John Hollinger, ESPN Boston's Jackie MacMullan, ESPN.com's Dean Oliver, NBA analyst and former coach Jeff Van Gundy and Boston Celtics executive Mike Zarren.
1. Fact or Fiction: If your team isn't investing heavily in analytics, you should be unhappy.
Jim Cavan, KnickerBlogger: Fact. If you're not looking to establish analytics as a significant cornerstone for your front-office agenda, you're doing it wrong. That's not to say it has to be the entire focus. Stats are undeniably getting better, more interesting, and -- it is hoped -- more reliable all the time.
Danny Nowell, Magic Basketball: Fact. Look, not everybody is going to use them well or extensively, but the fact is analytics are tools most teams have and which, in some cases, are credited with major successes. Disagree with analytics all you want, but if your opponents are committing to them, and they say it helps them, you'd be foolish not to follow suit.
Michael Pina, Red94: Fact. With almost everything in life you'd like to gather as much information as possible before making an important decision. Running a multimillion-dollar NBA organization is no different.
Jared Wade, Eight Points, Nine Seconds: Fact. Compared to the investment teams make each year in salary for players and coaches, hiring a few smart guys to analyze data is a relatively cheap way to gain a competitive advantage. This doesn't mean you base all your personnel moves or on-court strategy on stats. It just means you can make more informed decisions.
Charlie Widdoes, ClipperBlog: Fact. Even the heaviest investment in analytics costs almost nothing to NBA clubs. To not take advantage of the information that is available is just a symbol that the organization isn't doing all it can to win.
2. Fact or Fiction: JVG's claim that "the game is as much of a first-quarter game as a fourth-quarter game."
Cavan: Fact. And semantics. Van Gundy certainly has a point in claiming that what a team does or does not do in the early quarters sets a tone that has a trajectory inextricably linked with what happens late in the game. Or something. If LeBron James hits a game-winning shot (this is earth, by the way) over Kyle Korver because Luol Deng's two first-quarter fouls caused him to foul out down the stretch, did the Heat win because of LeBron's shot, or because Deng wasn't in to smother his grill? To me, it's a question with thousands of answers and no answers at the same time.
Nowell: Fact. I'm inclined to agree with Van Gundy, but there's an interesting debate going on here in Boston, as some research has shown that late-game situations matter a good deal more. While I think points scored early matter just as much as those scored late, I'm less sure of my opinion than I was before today.
Pina: Fact. If you believe in an immeasurable factor like "pressure" factoring in to a game's outcome, this may lean toward fiction, but the fact remains that a 12-footer from the first quarter weighs the same as one in the fourth.
Wade: Fiction. But just slightly. Execution and playmaking do become more important when you have no time left to make up for mistakes. It's not that first-quarter mistakes are OK. It's just that they are reparable. But if you screw up with three seconds left, you're probably going to lose.
Widdoes: Fact. While the circumstances may differ, particularly down the stretch of close games, the points count the same in the first as they do in the fourth. As Van Gundy said, he'd rather establish a lead early and maintain it than attempt to mount a comeback late.
3. Fact or Fiction: Communicating basic analytics to coaches is more important than having the most cutting-edge data.
Cavan: Fact-ish. It's one thing to develop cutting-edge data, it's an entirely different thing to know what to do with it -- which trades to pursue, draft picks to make, rotations to implement. Think splitting the atom. But convincing stats-neutral coaches to absorb some of the more established and peer-reviewed metrics while leaving room for the "eyes" and "guts" to work their more ethereal magic seems like a pretty good middle-ground approach.
Nowell: Fact. Analytics mean precisely nothing unless they can be used to make important decisions. That means big decisions, yes -- things like trades and drafting -- but it also means small decisions -- like in-game adjustments. If these things are made useful to the people who stand to benefit from them in the moment, we will get better basketball.
Pina: Fact. While having exclusive data can give you a competitive advantage, what good is it if the players can't benefit? It's like holding a winning lottery ticket but being unable to cash it in.
Wade: Fiction. Being on the cutting edge is paramount because the next generation of analytics will make the current crop look downright primitive. Comparing the video and spatial analysis that may exist in a few years to adjusted plus-minus, for example, may be like comparing a Rolex to a sundial. And the teams that already have an analytical culture will benefit the most.
Widdoes: Fact. Unfortunately, good information is useless if it can't be applied to the product on the court. Dean Oliver alluded to the importance of communication when he said that his reports for coaches would rarely include numbers, because they weren't as easy to digest.
4. Fact or Fiction: Advanced stats will be commonplace among most fans in 10 years.
Cavan: Fact and Fiction. Ten years from now, it may very well be that things like PER, WS48 and the like will have become more trusted and accepted by the mainstream public. But it's also true that today's groundbreaking metrics will become tomorrow's points per game. As such, it's hard to believe methods quantifying the "hot hand theory" by way of body-scanning lasers (copyright 2012 by Jim Cavan Industries) or whatever will be easily graspable. In short, the public will always dismiss the most advanced stats initially, even if they come to accept them down the road.
Nowell: Fact. The backlash now is strong and vocal, but I think fans want to see their teams win. Winning teams use advanced stats. Ergo, fans will root for advanced stats.
Pina: Fact. In the past five years, analytics have forced their way into the mainstream (as proven true by the popularity of this very conference). One would think 10 years from now it'll be further engrained in the fabric of fandom.
Wade: Fact. It's hard to believe that people will still consider stats like team points per game or regular field goal percentage for high-volume 3-point shooters to be worthwhile. I doubt rebound rate will replace rebounds per game like it should, for instance, but there will eventually be a general acceptance and usage of many "advanced" stats.
Widdoes: Fact, at least for the numbers we consider to be advanced today. We've seen it in baseball: Television coverage has begun to incorporate newer numbers and gradually they become accepted in the conversation. Basketball statistics are inherently less definitive, but we are getting there.
5. Fact or Fiction: The analytics movement is getting out of control. Just watch the games!
Cavan: Fiction. Analytics can actually help us find the beauty, the poetry in the games, because they help us see things we might not have otherwise known were there. Sure, front offices employing more and more in the way of advanced stats might change the look, feel and makeup of your team. But will it do so any more than a horrible trade or dumb draft pick? Probably not.
Nowell: Fiction. I'll say what every analytics proponent says: Some of this stuff is pretty arcane, and not all of it is relevant to specific in-game situations -- every analyst should also, of course, watch the games. However, with so many people invested in the league and the advantages analytics offer, there is no reason to keep your head in the sand.
Pina: Fiction. As long as people still watch basketball players play basketball each night, there's no harm in having advanced statistics there to give them a better understanding of what's developing before their eyes.
Wade: Fiction. Sure, some NBA writers misuse analytics. And giving more data to people who barely grasp, say, pace-adjusted team ratings is like throwing gasoline on a fire of ignorance. But those on the forefront of this movement don't believe they have found the Holy Grail. They understand the limitations of numbers and use them as a means to enlighten and add nuance.
Widdoes: Fiction. Even for the most educated fan, it's difficult to see this transformation, if it exists, taking place on the court by just watching the games. We may find out after the fact that J.J. Barea is starting because "the numbers" say he should, but that may not be common and certainly isn't obvious.