Player Impact Estimate vs. Wins Produced

The NBA has opened up a new stats platform.  Upon first glance, it seems pretty impressive.  I especially like the shot charts and the many filtering options.  However, one element in particular caught my eye: PIE.  Player Impact Estimate (PIE) looks like the NBA’s attempt at an overall statistic.  When looking at the formula, I noticed it looked a lot like Wages of Wins’ Win Score, which is closely related to it’s Wins Produced and WP48.  Some subtle differences, but it looks a bit like player Win Score divided by overall game Win Score.   People with more time than me will probably do a much more detailed comparison, but I thought I’d see how the Pistons ranked in each of the stats.

First the presented stats: PIE and WP.

Jose Calderon 7.6 Greg Monroe 14.8%
Andre Drummond 7.2 Jose Calderon 14.4%
Greg Monroe 4.2 Andre Drummond 14.3%
Tayshaun Prince 3.0 Will Bynum 10.8%
Kyle Singler 2.9 Charlie Villanueva 10.7%
Jason Maxiell 1.6 Tayshaun Prince 9.9%
Rodney Stuckey 1.4 Rodney Stuckey 8.9%
Austin Daye 1.3 Brandon Knight 8.8%
Charlie Villanueva 1.3 Austin Daye 8.6%
Brandon Knight 1.2 Khris Middleton 8.4%
Jonas Jerebko 0.4 Jason Maxiell 7.0%
Will Bynum 0.1 Kyle Singler 6.5%
Khris Middleton 0.1 Jonas Jerebko 5.8%
Viacheslav Kravtsov 0.1 Viacheslav Kravtsov 4.6%
Kim English 0.0 Kim English 3.5%
Corey Maggette -0.7 Corey Maggette 3.3%

PIE likes Monroe more as it puts him on the same level as Calderon and Drummond.  It’s after that the player ranks in the stats start to deviate.  PIE doesn’t have a position adjustment, so I wonder how much position adjustment can explain?

What if instead of using PIE, we use Total PIE (TPIE), since PIE is a per game stat.  TPIE = PIE * GP.

Jose Calderon 7.6 Greg Monroe 8.0
Andre Drummond 7.2 Jose Calderon 7.3
Greg Monroe 4.2 Andre Drummond 7.2
Tayshaun Prince 3.0 Tayshaun Prince 5.0
Kyle Singler 2.9 Brandon Knight 4.8
Jason Maxiell 1.6 Will Bynum 4.8
Rodney Stuckey 1.4 Charlie Villanueva 4.7
Austin Daye 1.3 Rodney Stuckey 4.3
Charlie Villanueva 1.3 Jason Maxiell 3.8
Brandon Knight 1.2 Kyle Singler 3.5
Jonas Jerebko 0.4 Austin Daye 2.6
Will Bynum 0.1 Jonas Jerebko 1.3
Khris Middleton 0.1 Kim English 0.8
Viacheslav Kravtsov 0.1 Corey Maggette 0.6
Kim English 0.0 Viacheslav Kravtsov 0.4
Corey Maggette -0.7 Khris Middleton 0.3

Closer still, though Monroe the point guards rank higher in TPIE than WP.

What if we adjusted per 48 minutes?  PIE48 = PIE / MP * 48.

NAME WP48 Name PIE48
Andre Drummond 0.349 Khris Middleton 1.344
Jose Calderon 0.249 Viacheslav Kravtsov 0.442
Khris Middleton 0.233 Andre Drummond 0.361
Austin Daye 0.150 Charlie Villanueva 0.321
Greg Monroe 0.113 Austin Daye 0.295
Kyle Singler 0.089 Will Bynum 0.288
Tayshaun Prince 0.087 Jose Calderon 0.247
Charlie Villanueva 0.085 Greg Monroe 0.215
Viacheslav Kravtsov 0.067 Kim English 0.210
Jonas Jerebko 0.060 Jonas Jerebko 0.199
Jason Maxiell 0.056 Rodney Stuckey 0.153
Rodney Stuckey 0.049 Tayshaun Prince 0.149
Brandon Knight 0.034 Brandon Knight 0.136
Kim English 0.008 Jason Maxiell 0.134
Will Bynum 0.007 Corey Maggette 0.113
Corey Maggette -0.139 Kyle Singler 0.111

Quite a mixed bag again.

Wages of Wins points out the strong correlation with WP and actual winning.  PIE apparently claims to do the same,  at a slightly lower, but still significant percentage.  For the time being, my trust is still in WP.  However, I look forward to seeing how much stock will be put into PIE in the future… especially after a round of scrutiny from the basketball stats community.


One thought on “Player Impact Estimate vs. Wins Produced

  1. I absolutely love your blog and find nearly all of your post’s to be exactly what I’m looking for.
    Does one offer guest writers to write content for yourself?
    I wouldn’t mind creating a post or elaborating on many of the subjects you write about here. Again, awesome blog!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: