NBA Stat Geek's Victory Plan: Sit Kendrick Perkins
While watching the Clippers/Thunder game Tuesday night, I heard something interesting. Clippers color commentator Michael Smith made the quip while watching Kendrick Perkins leave the game with his fourth foul, "Well, that's a bad thing... for the Clippers. The Thunder are better with Perkins off the court than with him on."
That's a crazy statement, even for someone who ended his career as BYU's leading career rebounder. Does Perkins truly mean that little to the Thunder? The big man who has started all of his 145 games in a Thunder uniform, including playoffs? Serge Ibaka's running partner down low should ride the bench?
But looking at the stats, maybe BYU didn't make Mr. Smith that crazy. In fact, after looking harder, I'm left to wonder why Kendrick Perkins is even on the court at all.
Kendrick Perkins by the Numbers
Numbers? There are numbers to analyze with Kendrick Perkins? Dude's only averaging 4.2 points, 5.9 rebounds, and 1.0 blocks per game. And that's all he's asked to do, because Durant and Westbrook and Ibaka and Martin are expected to take control, right?
Well, even when doing what little is asked of him, he's not doing it all that effectively. Take a look at the below chart, detailing the difference between two players in effective field goal percentage (eFG%), offensive rebound percentage (ORB%), total rebound percentage (TRB%), block percentage, offensive rating (ORtg, points per 100 possessions), defensive rating (DRtg, points allowed per 100 possessions), and John Hollinger's Player Efficiency Rating (PER).
You'd take the first guy in a heartbeat. It's not even close: Player A beats Player B in every major statistical category across the board except, well, assist percentage (and if you're choosing a big man based solely on how well he passes, then shame on you). So, that's what the man behind the scenes Sam Presti picked, right?
Yes, he did... back in 2010. Because, you see both Player A and Player B are Kendrick Perkins. Player A represents his numbers in his last full season in Boston, while Player B is his numbers this year. When analyzing Kendrick Perkins, you can't think of him as the same player that he was in Boston. Across the board, his percentages are down.
But Zach! you could easily say. It's possible that his percentages are down just because of his teammates! And to be fair, my exclamation point-happy friend, Serge Ibaka and Kevin Garnett are indeed different players; Ibaka is more of a shot blocker (taking the load off Perkins) while Garnett was more of a scorer in that '09-10 season.
Fine, have it your way. Then let's take these two players into account:
You're reading numberFire; you're a smart basketball consumer. You're probably one step ahead of me: that is indeed Hasheem Thabeet's numbers in 38 games played so far this year. Aka, the guy that All-Knowing Sam Presti has left on the bench in favor of Perkins.
Against opposing centers this season, Perkins has a net PER of -9.5, while shooting .055 eFG% worse from the floor and grabbing 1.5 less rebounds per 48 minutes. Thabeet, meanwhile, still has been poor but is only giving up -6.2 PER to opposing centers and is grabbing 1.0 less rebounds per 48 minutes. Considering Perkins has played 51 percent of the Thunder's minutes this season while Thabeet has played only 23 percent, it's safe to assume that Perkins is the main reason that the center position is the only spot on the floor where OKC has a worse PER than their opponents this season.
OKC's Best Option
OK, so maybe Kendrick Perkins isn't the best player night-in and night-out. In fact, he's actually kind of poor. But do the Thunder actually play better without him on the floor? Wouldn't they be too small without a big man?
Luckily for us, the Thunder have run those exact formations with Perkins on the bench. And the numbers with those rotations on the floor can provide the answers.
The Best Option: Play Small Ball - Westbrook-Martin-Sefolosha-Durant-Ibaka
In this lineup, sharp-shooter Kevin Martin graduates from sixth-man status and into the starting lineup. This pushes Thabo Sefolosha over the small forward position. Common sense would dictate that this small lineup would allow opponents to score easily, especially at the small forward position against the 6'7" Sefolosha.
There's a reason that we don't listen to conventional wisdom. In fact, of the Thunder's top five most played court configurations this season, this is the optimal one to their success. The Thunder have an ORtg of 123 points per 100 possessions with this five, 13 points greater than the starting five that contains Perkins. While this small-ball configuration does indeed give up a few more points, with a DRtg of 106 points allowed per 100 possessions, it's only five points more than the starting five allows.
Some other stats for your viewing pleasure, via 82games.com:
|eFG%||Opp. eFG%||TRB%||Net T/O%||ORtg||DRtg|
|Small Ball 5||.505||.413||50.0||-5.0||123||106|
Almost across the board, the Thunder are better with the small lineup. The only places where the Starting Five excels are reducing turnovers and allowing slightly less points. Strangely, the five without Perkins rebounds better, picking up 1.5 percent more of the available rebounds.
One could make the argument that the Martin-infused lineup is only solid due to the opponents they're facing - since Martin is not part of the starting five, he's often not facing the other team's best players. Plus, one could ask whether Perkins plays specifically because of matchups. Those are valid... to a point. Given a large sample size this season, these lineups have played against a variety of different opponents. In addition, playing 51 percent of minutes does not make Perkins a situational player - it makes him a go-to guy, inexplicably. Especially given the distance between the two sets of five, I believe it's statistically significant data that the Thunder would play better with Martin over Perkins no matter the opponent.
Why exactly is Perkins playing significant minutes, again? He's nowhere near the same player that he was with Boston, and he's not gelling with the rest of the team. In fact, without him on the court, the Thunder not only score more, but they also rebound more. Letting Durant and Ibaka play down low rather than keeping Durant outside and pairing Ibaka with Perkins seems like the way to go.
Unfortunately, salary concerns means that OKC is stuck with Perkins for a while. He is signed for two more years past this season, each at over $8 million per year. The old wisdom goes that if you're paying a guy, then why not play him?
But the conventional wisdom won't win your championships. If Sam Presti wants to have the statistically optimized roster play the majority of minutes, he'll bench Perkins. Now.