Skip to content

Being anti-analytics in 2015 is plain lazy

Action Images

With Charles Barkley's anti-analytics rant only weeks old, ESPN's Analytics Rankings debuting, and the annual MIT Sloan Sports Analytics Conference set for this weekend, advanced statistical analysis is once again at the forefront of NBA debates.

The thing is, the majority of metrics being referenced when discussing advanced stats aren't that advanced at all. They're just smarter, newer and more readily available than they were years ago.

For example, because NBA broadcasts still cite team offensive and defensive rankings based on raw points per game, the newer, more sensible approach of measuring per possession - or per 100 possessions - is seen as advanced.

Of course, there's nothing advanced about it. Teams play at different paces, which affects point totals. Measuring how effective a team is at scoring on each possession they use is much more indicative than how many points they score in general, just as measuring how effective a team is at preventing scoring on opposing possessions is more indicative than points allowed.

Offensive and defensive ratings (Points for/allowed per 100 possessions) also take more factors into account. For example, while field goal percentage allowed is a solid measure of a team's defensive ability, it doesn't consider defensive rebounding.

If a team forces two misses, but the attacking team grabs two offensive rebounds and scores on their third try, the result is the same as if they had scored on the first. The defensive team has allowed a basket before recovering possession. The fact that the attacking team shot only 33.3 percent on that possession isn't relevant. That they scored at least two points on one possession is.

If a player scored 20 points of 10-of-15 shooting, he'd probably be praised for an efficient night. But what if that player turned the ball over five times and got to the free-throw line on three possessions? Then it turns out they actually scored 20 points on 23 possessions, or less than a point per. That's not efficient at all, and the player's individual offensive rating of 0.87 (or 87 per 100) would reflect that.

As for rebounding, Rebound Rate measures the percentage of available rebounds that a team or player grabs while on the court. Rebounding opportunities vary from team to team and game to game. A player who grabbed 10-of-20 available rebounds did not have a better rebounding game than a player who grabbed six of 10 in similar minutes elsewhere, even though their per-game or even per-minute numbers would imply he did.

Truly advanced stats are metrics like Player Efficiency Rating, Win Shares and even ESPN's Real Plus-Minus. They're arrived at using complex formulas and aren't as easily explainable. Though they hold value in their own way, an old-school coach or player not embracing them would be understandable.

But if you don't see the value in basic efficiency, if you don't see the sensibility in modern per-possession rates over per-game measures, or in player tracking data and lineup combination data - all of which is publicly available - that doesn't make you old-school.

It makes you unwilling to accept common sense on account of having to learn something new.

In other words, it makes you lazy.

Daily Newsletter

Get the latest trending sports news daily in your inbox