@bradford_NBA what would you say to a Big Al hater who points at EFG% + TS% both being below average. Does that define his game?
— Timmy Hoskins (@TimmyHoskins) August 11, 2014
As I was attempting to answer this on twitter, I realized I needed a little more room. So I started a blog post. Then it grew, and I decided it belonged here. I feel it’s an interesting question and an example of my concerns about how readily available advanced statistics have become. Al Jefferson is a player that has interested me for some time. I have never been a fan of his game and was strongly opposed to the Bobcats signing him. In retrospect, I could not have been more wrong, which I’m more than happy to admit. My arguments against him revolved around his teams’ failures up until last season and individual statistics like those referenced in the above tweet. His defense, which was an obvious weakness (though much of that can be attributed to Ty Corbin and whatever it was he was trying to do), was an eyesore. At times he was a serious black hole on offense. However, having him on “my” team has made me re-assess his value and look a little closer.
The first thing to understand is exactly what EFG% and TS% mean. Both of these numbers are specifically shooting metrics. Many savvy fans could tell you that EFG% accounts for the added value of 3-pointers while TS% does that while adding value via free-throws. What I believe some fans fail to understand is how to interpret those numbers from a practical application standpoint. By definition, they are skewed towards players that shoot 3-pointers. In the right context, this makes a lot of sense. If you shoot 30% on only 3 pointers while another player shoots 40% on only 2 pointers, the 30% shooter will have scored 90 points on 100 attempts while the higher 40% shooter will have scored 80 points on the same number of attempts. Weighting shooting percentages based on shot selection levels the playing field, allowing us to more easily compare players. These metrics also give some insight into shot selection and a player’s understanding of the value of one shot vs. another, for example stepping in to shoot an 18 foot jumper rather than taking a step back to shoot a 3-pointer.
Now back to Jefferson and context. FG% generally favors players who take fewer shots and only score right at the basket. The top 13 FG% numbers in a minimum of 15 minutes per game were posted by Brandan Wright, DeAndre Jordan, Mason Plumlee, Chris Andersen, Andrew Bogut, Chris Wright, Andre Drummond, Ronny Turiaf, Tyson Chandler, Dwight Howard, Greg Stiemsma, Samuel Dalembert, and Al Horford. Of those, only Dwight Howard and Al Horford shot more than 10 times per game (Drummond was at 9.5). Horford lead that group in 3-point attempts per game at .4 per game.
Switching to TS% (because big guys draw a lot of fouls, we want to include free throws), the top 15 is as follows: Brandan Wright, Chris Andersen, Troy Daniels, Mason Plumlee, Kyle Korver, LeBron James, Pablo Prigioni, Kevin Durant, DeAndre Jordan, Brooke Lopez, Mike Miller, James Harden, Tyson Chandler, Andrew Bogut, Stephen Curry. These players break down into 4 categories: All-World players, elite shooters, low-usage post players (and by post I mean they catch lobs and that’s it), and Brandan Wright (so weird). Lopez had a 62.9 TS% but only played 17 games and has been closer to 55% over his career. You can see how TS% levels the playing field for shooters.
Jefferson is more than a low-usage dunker, but he’s not an elite shooter. By cherry picking a stat that, by design, rates him poorly, you would be misrepresenting his true value. Taking a broader approach paints a different picture. The statistic that really stands out to me as being elite is his ability to have the ball in his hands (29.7% usage rate) while not turning it over (7% TO rate). The only other players with a usage rate over 25% and a turnover rate 7% or lower are LaMarcus Aldridge (29.9/6.5) and Dirk Nowitzki (26.8/6.6). Additionally, he’s in the 94th percentile for rebound rate, 96.5th percentile in defensive rebound rate, and the 69th percentile in offensive rebound rate. Let’s keep going. Jefferson is in the 80th percentile for assist ratio among centers and 84th percentile in assist to turnover ratio among centers.
That’s a lot of stats. I didn’t put them in a table to prove a point, that you can go on and on about what makes Al Jefferson an elite player, especially at a position that is becoming less and less involved offensively. If you couldn’t get through all those numbers, consider the following. Analysts are constantly trying to find the basketball version of the Theory of Everything. The goal is to find a single metric that combines all other numbers in order to compare players and truly understand their value. I’ll put this one in a table for ease of reading.
PER and its derivatives love Jefferson, as does nba.com’s PIE. Kevin Pelton’s numbers less so. What’s the takeaway from all this? First, you can make a case for anything by selecting the right stat. At the root of the original question posed is confirmation bias. Someone’s not a fan of Al Jefferson for whatever reason, so when they find a stat that paints a negative picture of him, that stat becomes the basis for their argument. We all do this and, as fans, it’s a difficult tendency to overcome. But basketball isn’t black or white. Players aren’t just good or bad. Not everyone can be LeBron James or Kevin Durant, and everyone isn’t Austin Rivers.
Al Jefferson is an imperfect player. He can get tunnel vision at times. He doesn’t show any type of second effort on defense. Even in a conservative scheme, he struggles with pick and roll defense at times. But this is what I know. Charlotte jumped from 21 to 43 wins with only the significant additions being Al Jefferson and Coach Steve Clifford. Charlotte’s offense went from a 98.3 rating, 27th in the league, to a 101.2 rating, 24th overall. Not only did the offense improve, it improved while Kemba Walker and Gerald Henderson regressed from the previous season. Charlotte started the season first with Jefferson spraining his ankle, then Kemba hurting his arm. After getting healthy and comfortable with one another, Charlotte posted a 105.8 offensive rating, what would be a top 10 mark for the season. Again, the same roster from the previous year. In 2012-13, the Bobcats had the worst defense in the league. A part of that was the second worst defensive rebound rate in the league. I’ve already established Jefferson is an elite defensive rebounder. The next year, they had the 6th best defense in the league. A large part of that was being the best defensive rebounding team in the league. To play defense, you have to finish defensive possessions. Al Jefferson ends possessions.
Individual stats? Post all-star, a 105.1 individual offensive rating. With Jefferson on the court, the Bobcats had a 105.1 offensive rating. Again, top 10 levels of offense. Off the court, that number dropped to 102.2. Plus all the numbers outlined above.
Al Jefferson isn’t going to individually take a bad team and make them elite. Those types of talents are few and far between. He probably can’t be the best player on a championship team. But he can clearly be the best player on a good team (if you don’t think the Bobcats were a good team last year, you weren’t paying close enough attention). Jefferson was voted All-NBA 3rd team and it was no fluke. I fully expect, health permitting, Big Al to make his first all-star team and gain the recognition he deserves while continuing to lead an improved offense and team.