Advanced stats you need to know
February 19, 2013 1 Comment
Keith Law, a baseball writer most known for writing ESPN pieces in which he evaluates up-and-coming prospects, wrote a piece today discussing some of the stats he can’t do without. ESPN The Magazine’s latest edition is called “the analytics issue,” and while its cover story is at least 10 years late and not necessarily accurate (WAR isn’t a be-all, end-all and the headline writer made a point to call some baseball lovers “RBI-addicted fans” :rolls eyes:), Law points out that there are plenty of other stats a player’s WAR is made up of that might be more indicative of a player’s future value than simply WAR alone. (All of Law’s pieces for ESPN are Insider-only and are behind a pay-wall, but here is the link in case you’re an Insider.) Law’s main point: Simple stats like slugging, BABIP and ground ball percentage are especially useful, though they aren’t super flashy sabermetrics.
In honor of Law, who’s found a way to make sabermetrics digestible and cool (while also having a sense of humor in the process), here’s a list of some super simple basketball stats that can reveal a lot about a player at any given position. In the coming weeks, I’ll take a closer look at each, giving examples of how it’s applicable. They’re all fairly elementary and I’ve brought up most of these numbers in posts past.
True Shooting Percentage – Often TS% is one of the numbers most revealing about a player’s offensive efficiency, shot selection and year-to-year consistency. It’s a metric that takes into account 2-point and 3-point field goals and free throws in order to provide an accurate measurement of his shooting ability. TS% separates streaky shooters from consistent, efficient ones. For example, J.R. Smith looks like he takes too many forced shots, but that’s hard to prove with the eye test. However, by looking at his TS% (when combined with the percentage of shots he makes off of assists) it’s clear that he’s a below-average shooter, and it’s likely a byproduct of his insistence to work in isolation. Kevin Durant, meanwhile, is a model of efficiency. Comparing Smith to Durant might not be fair, but Durant hasn’t had a TS% lower than 57.7 since his rookie year. (The league average this season is 53.6%. Smith has only topped that mark four times in his career, and in three of those seasons he played less than 25 minutes per game. The more minutes he plays, the more shots he takes, and the more shots he forces.)
FG% at the Rim – This stat in my opinion tends to work more against a player than in his favor, especially for point guards. For example, Brandon Jennings shoots almost 8 percent below the point guard league average percentage (60.9%) at the rim, where he takes 4 shots per game. It’s a stat that looks horrible on Jennings’ part (he’s only shot higher than 57% at the rim in one season). But Tony Parker and Jrue Holiday both shoot just under 70% at the rim, and both take more than 4 shots per game there. Only two point guards take more shots per game at the rim than Parker, but no one else in top 25 in attempts per game has a higher percentage, but that isn’t necessarily a stat that validates Parker as a top guard. Being below-average at the rim is much more devastating to a player’s value than being above-average is helpful. Point guards shoot a whole lot more now than they ever have, and they’re therefore expected to make shots at the rim, an area that shouldn’t be difficult to score from. Parker and Holiday live up to expectations, while Jennings only falls short.
Asst/TO Ratio – Now that point guards pass less and shoot more, the importance of limiting turnovers while racking up assists has diminished, at least to a certain degree. The league average among point guards is 2.07, meaning for every 2 assists a point guard dishes out, he turns it over once. Rajon Rondo, the league leader in assists, had a 2.83 ratio before tearing his ACL. Not surprisingly, as Rondo continues to increase his assist totals during his career, he has also turned the ball over more. Russell Westbrook, the top-scoring point guard, has a pedestrian 2.36 average, though it’s the second-highest rate of his career, and nearly a full point higher than last year’s poor mark. It’s worth noting that the two leaders year-in and year-out always tend to be Chris Paul and Jose Calderon, the two Gods of Assists in the NBA. Both always finish with a ratio at or above 4, a ridiculously high mark, indicating that they rarely turn the ball over and always find ways to put teammates in positions to score. Calderon’s rate is particularly interesting because he’s spent a majority of his career playing with bad players, though it’s worth noting that John Lucas, Toronto’s current backup point guard now that Calderon has been traded to Detroit, is third in the league in assists-to-turnover ratio, and starter Kyle Lowry has an above-average 2.72 rate. Those guys just enjoy passing, and Calderon has taken that infectious hobby along with him to Detroit.
Defensive Rebounding Rate – Really, this stat boils down to three things: size, leaping ability and effort. The DRR is used to determine the percentage of defensive rebounds grabbed by a player while he’s on the floor. The league leaders are almost always centers and power forwards, but this year Hornets small forward Al-Farouq Aminu is 9th in the league, showing that he tries really, really hard to rebound the ball. He eats glass, and that’s a good thing. Meanwhile, Dirk’s DRR this season is a ridiculously low 18.3, and according to basketball-reference.com it’s the lowest since his second season and more than three points lower than his career average. Dirk’s never been a leaper, but add a bum knee to his already limited athletic ability, and you can see why that’s the case. It’s always been difficult to measure a player’s “motor,” but taking a glimpse at his DRR at least offers some help.
PER – Short for Player Efficiency Rating, PER is basketball’s closest thing to baseball’s WAR. (We’ve come full circle.) It’s a very, very, very, very, very complex and imperfect number, though it typically grades players in a far, accurate way, as long as they put up decent measurable stats like blocks, steals, etc. (Here’s a link to its Wikipedia page. Scroll down to calculation to see the formula. You’ll need to scroll right to see the entire thing, a pretty good indicator that the formula is just too damn long.) Though the formula is admittedly not completely perfect, check out the list of the league leaders in PER year-by-year and it’s hard to argue that year’s leader wasn’t the best player in the league during that season. (Especially 2005-06.) The league leader in PER isn’t always the same as the league leader in Win Shares (for example, Kevin Durant leads the league in WS this season, though LeBron leads the league in PER), and often the top 20 lists look completely different. Tyson Chandler, for example, is 6th in WS but 20th in PER, which means he’s arguably more valuable to his team’s success than his stats indicate. There are fundamental differences in the way PER and Win Shares are calculated, but in terms of measuring an individual’s ability and not importance to his team (really, the difference between most outstanding and most valuable) PER is more accurate.