ROTOWIRE.COM NFL MLB / Minors NBA NHL GOLF RACING    COMMUNITY FORUMS PODCASTS MYROTOWIRE ASK AN EXPERT GAMES

What is all the fuss about PER?

For those that don’t know, I’ll be doing a series of blogs about advanced basketball stats.  Today, I’ll talk about John Hollinger’s Player Efficiency Rating (PER) stat because it is the most popular of the new advanced stats, the one that is becoming an accepted part of the basketball language.  I spend a lot of time on basketball message boards, and I have been seeing more and more people reference PER as a way of ranking players.  There are pros and cons to using PER, though, many of which are due to incorrect usage (many Know-it-Alls and Skeptics either misuse or malign the measure).  So let’s see if we can shed some light on things.


What is it: PER actually measures a combination of production and efficiency.  It looks at information found within the box scores (such as all of the fantasy stats), adjusts for team pace and individual minutes played, and normalizes based upon numbers across the league (league average PER is always set to 15).  The end result is a measurement that helps quantify many of the accomplishments that comprise a productive, efficient player (especially on offense).  For those that want more technical information on how PER is calculated, it can be found here.


Strengths: PER is a single number, which makes it easier to use to compare two players than looking at combinations of points/rebounds/assists/blocks/steals/percentages/etc.  Also, by normalizing for minutes played and pace, PER allows players in different situations to be compared more logically.  Finally, because PER is normalized to the league average it gives a way to compare players across time.  For instance, PER can’t tell us who was actually better between Shaq and Kareem Abdul-Jabbar…but it can tell us that over their careers, Shaq has been more statistically dominant with respect to his peers (career PER of 27.01) than Kareem was with respect to his peers (career PER of 24.58).


Weaknesses: PER does not account for defense outside of blocks/steals/rebounds, which would be a limitation for it being a “total player ranking” as it barely addresses half of the game.  It doesn’t address at all whether a player limits his opponents to worse statistics, or how the team’s defense performs with that player on the court.  Players like Bruce Bowen, for instance, that made their careers based upon strong defense are generally underrated by PER.  Likewise, PER does not take into account how an individual player’s numbers translate into team success, another weakness for using it as a stand-alone player rater.  Finally, because PER is normalized per minute played, looking only at PER can deceptively rank a bench player that plays very limited minutes above an impact starter.


Usage:  You can’t simply say “player X has a higher PER than player Y so he is better”.  Instead, you can use PER to say that player X has been more statistically productive than player Y, and from there you need to look to personal observations or other statistics to help determine how factors like defense, teammate quality, or team role would affect the measurement.  For instance, one can look at Michael Jordan’s PER of 31.1 in 1988-89 and Kobe Bryant’s PER of 28.0 in 2005-06 and surmise that Jordan was more statistically productive with respect to his peers in 1989 than Bryant in 2006.  But to truly try to gauge who was the better player those years, other stats that fill the holes of PER should be used.  PER is a great way to make a quick first comparison when trying to gauge players, but (like all stats) should always be used only as evidence within an argument and not as the argument itself.


Leaders: As of 1/6/09, the current leaders in the NBA in PER this season are:

1. LeBron James-CLE 32.2
2. Chris Paul-NOH 30.2
3. Dwyane Wade-MIA 29.4
4. Dwight Howard-ORL 26.1
5. Kobe Bryant-LAL 25.3


James' PER, if he can maintain it, would be the highest in NBA history.  The current high mark is 31.84 from Wilt Chamberlain in 1962 - 63 (Wilt has the two highest marks ever).  Jordan comes in with the third highest PER in history, 31.71 in 1987 - 88.  Here are the league leaders in PER for every season since 2000:


2007-08 NBA LeBron James 29.14
2006-07 NBA Dirk Nowitzki 27.59
2005-06 NBA Dirk Nowitzki 28.06
2004-05 NBA Kevin Garnett 28.20
2003-04 NBA Kevin Garnett 29.44
2002-03 NBA Tracy McGrady 30.27
2001-02 NBA Shaquille O'Neal 29.68
2000-01 NBA Shaquille O'Neal 30.23

Next time: the +/- stat

Comments

By: Dalton Del Don
On: 1/5/2009 6:16:00 PM
Good stuff Dre'.
 
By: Chris Liss
On: 1/5/2009 11:31:00 PM
Agreed - good stuff. Who's the king of PER this year? Garnett, LeBron, Kobe, Wade?
 
By: The Professor
On: 1/6/2009 6:48:00 AM
I went back and added the current top-5 in PER, and also the league leaders in PER for every season since 2000. I think I'll make that part of all of the future blogs, since it helps put the measure in perspective with people's perceptions.

LeBron, statistically, has been unreal this year. For him to have the highest PER ever...to get a feel for it, in Wilt's 62-63 campaign that currently is the best PER ever, he averaged 45 points and 24 boards per game. For awhile there were three players with a PER over 30 this year (LeBron, Paul, and Wade). Wade has recently slid a bit, and odds are that LeBron and/or Paul will slide under 30 eventually as well. But each of those three are playing unreal this season, when measured by PER.
 
By: nayfel
On: 1/6/2009 11:40:00 AM
I agree on your thoughts on some of the limitations. I think it can be very useful for people to say "X" player is performing at a higher PER than I anticipated, maybe he deserves more time but for people to assume that player is capable of performing up to thaat PER if givien double the minutes is just wrong. And taht is mostly how I hear PER being used.
 
By: Charlie Zegers
On: 1/8/2009 3:37:00 PM
It's sort of like OPS.

Baseball writers and commentators who don't understand sabermetrics in the slightest will throw around a term like OPS because it's a single number they can sort of understand.

Yes, John Sterling, I'm looking at you.
 

Leave a comment

Commenting is restricted to registered users only. Please register or login now to submit a comment.

Tell Someone

  • Digg it
  • submit to reddit reddit
  • Add to Mixx!

Recent Favorites

How BABIP Affects the Value of Strikeouts and Walks
The league-average batting-average-on-balls-in-play (BABIP) has been on a downward trajectory for the last half decade, something that has implications for the values of different at-bat outcomes. BABIP is down precipitously this season (.289), though it's hard to tell to what extent that's due to all the shifting defenses (better data on where balls are hit have enabled this), or is simply just an April cold-weather/small sample anomaly.
NHL Playoff Observations -- Monday
Just a few things that caught my eye ...
  • Alexander Semin was on the wrong side of the grocery stick most of the night. You know, that spot between the back-up goalie and the rest of the team ... the one where the dead wood goes so it doesn't get in the way of on-the-fly changes.
  • Honey badger don't care. Honey badger doesn't give a $%!&.  T.J. Oshie gave up four inches and 50 pounds to Dustin Penner, but he didn't care. Then again, honey badger didn't care when he hit Rick Nash a few years ago, either. I wonder if the NHL would let me get a Blues #74 jersey with Honey Badger on the back.
Buy Lowest
Every year, we hear advice about how it's time to "buy low" on struggling stars like Albert Pujols, Giancarlo Stanton and Robinson Cano. "Get them now for cheap before they heat up!" Except that the market for superstars - at least in competitive leagues - usually doesn't move much after one bad month. (Whether it should in these cases is another question). It's more likely that the Pujols and Stanton owners are doubling down, figuring that only a hot streak from a player of that caliber can undo the damage that's already been done.

That doesn't mean you can't take advantage of the market's early overreaction, but you'll have to aim a little lower and take on quite a bit more risk. The players listed below are those who should be available at an actual discount.

Fantasy Rookies
Now that the 2012 NFL Draft is complete and rosters look about the way they will when training camps open in late July, it’s time to start the speculation train full steam ahead to fantasy draft days. Many players taken in this class will have a fantasy impact, some immediate, some in small spurts as rookies and some not for years to come.
Cowboys Draft: First Blush Thoughts
My initial reactions to the 2012 Dallas draft crop.

RSS Feeds