clock menu more-arrow no yes mobile

Filed under:

A Question for Someone Who Understands PER Better Than ClipperSteve

Why does Al Jefferson (PER 23.04) have a significantly higher PER than Chris Kaman (PER 18.31)?  I mean, I'm a pretty bright guy.  I understand things like pace and adjustments for minutes per game and stuff like that.  I mean, I wasn't a math major in College or anything, so formulas confuse me... oh wait a minute, I WAS a Math major in college!  

Look Al Jefferson's numbers are better than Chris Kaman's in exactly two areas - scoring and turnovers.  That's it.  OK, steals, but Hollinger himself says that steals are a meaningless stat.  Chris Kaman rebounds better, has more assists and blocks more shots.  And yes, Kaman rebounds better even when you consider rebound rate, and not just raw rebounding numbers (19.4 versus 19.0).  As for the scoring question, when you consider the fact that Kaman gets to the line more and shoots significantly better there, his Total Shooting Percentage is actually higher than Jefferson's (52.7% vs. 52.6%) despite Jefferson's better shooting percentage (49.3% vs. 47%).  So, OK, I know that if there's a knock on PER it's that it overvalues scoring some.  I'd even accept Jefferson rated slightly higher than Kaman.  I mean, Al's having one hell of a year.  But it just doesn't seem reasonable that Al Jefferson is rated  as the 14th best player in the NBA by PER while Kaman is 50th.  Does it?

Furthermore, if PER tends to overvalue scoring, then how is it possible that Marcus Camby's PER (18.46) is higher than Kaman's?  Yes, Camby is an astounding rebounder and shot-blocker, particularly on a per possession basis (his team plays at a high rate, but he does not play nearly as many minutes as Kaman).  But Camby scores significantly less than Kaman, and shoots a lower percentage also.

What's the deal?

Is turnover rate the be all end all of PER?  Can that be the one thing that is killing Kaman in this metric?  And if so, why does it not seem to hurt Dwight Howard and Dwyane Wade?

Here are the top ten true centers in the NBA by PER.  (Jefferson is not on the list because he's not a true center, but it was his numbers that really jumped out at me.)

  1. Amare Stoudemire -26.55 - ahead of Howard - I guess it does overvalue scoring.
  2. Dwight Howard - 25.04 - turnovers killing him too I guess.
  3. Yao Ming - 21.96 - turns it over almost as much as Kaman.
  4. Andrew Bynum - 21.35
  5. Brendan Haywood - 19.69
  6. Andris Biedrins - 19.60
  7. Shaquille O'Neal - 18.56
  8. Marcus Camby - 18.46
  9. Chris Kaman - 18.31
  10. Zydrunas Ilgauskas - 18.26

OK, so obviously Biedrins, Bynum and Haywood don't play a lot of minutes, and that always takes some getting used to.  The fact that none of them score as well, rebound as well or block shots as well as Kaman on a per minute, pace adjusted basis must only matter to me.  I mean, who wants scoring, rebounding or blocked shots from a center?

I've long been a fan of PER, but I'm now realizing that it was primarily because it served my purposes - it consistently valued Elton Brand as one of the best players in the NBA, an outcome I wholeheartedly supported.

But Kaman appears to have fallen into some PER black hole where the formula gacks.  Too many turnovers, not enough offense to offset those turnovers.  

Help me out here people.  What am I missing?  Kaman really is better than Brendan Haywood, right?