Al Jefferson and Diminishing Returns

Posted on October 15, 2010 by


Sekou Smith at has snapped.

While debating the merits of Al Jefferson over Carlos Boozer as the low post catalyst for the Utah Jazz, someone informed me that Jefferson’s numbers on losing teams don’t compare the stats Boozer put up in a winning situation in Utah the last six years.

I snapped. Seriously, I lost it.

When told that solid numbers on a bad team mean nothing, I couldn’t hold my tongue. It’s the most ridiculous thing I’ve heard.

Smith went on to make two arguments.

1. Numbers are numbers, and whether you are on a good or bad team these numbers have the same meaning.  

2. The Jazz will at least be as good with Al Jefferson as they were with Carlos Boozer.

Let me address the second argument first. 

Here are the career numbers of Al Jefferson and Carlos Boozer.

In terms of Wins Produced, Boozer has consistently offered a higher level of production.  In fact, in Jefferson’s best season – with Boston in 2006-07 – he offered a lower level of production than we see from Boozer’s career averages. 

What explain the difference? Three months ago Robbie O’Malley actually discussed this issue.  According to Robbie, the key difference is shooting efficiency.  Yes, both players can score. But Boozer gets more of his shots to actually go in the basket.  Consequently, Boozer has a bigger impact on team wins.

So moving from Boozer to Jefferson is not a step in the right direction for fans of the Jazz.   Yes, Jefferson is cheaper. And he is probably less of a headache (fans in Utah grew very tired of Boozer’s attitude).  But Boozer does produce more wins than Jefferson (and fans everywhere really like wins).

But is it the case that the production we see from Jefferson will actually decline because he is moving to a better team?  On this point, Smith is sort of correct (and sort of not correct). 

In both the Wages of Wins and Stumbling on Wins we discuss the subject of diminishing returns.   Economic theory – and the empirical evidence – tell us that as the productivity of a player’s teammates increases the production we see from the player will fall.  But the effect is small.

To see how small, let’s estimate how the move from Minnesota to Utah will impact Jefferson’s numbers.  Our story begins with what every player did last year with the Jazz and Timberwolves.

Carlos Boozer produced 16.2 wins last season in 2,673 minutes.  The remaining Jazz produced 38.9 wins in 17,083 minutes.  And this means that Boozer’s teammated posted a WP48 [Wins Produced per 48 minutes] of 0.109. 

In contrast, Jefferson only produced 6.4 wins in 2,463 minutes.  Jefferson’s teammates played 17,316 minutes but only produced 8.9 wins. This means that Jefferson’s teammates posted a WP48 of 0.025.  So it is the case that Jefferson’s teammates were much worse than the players who played with Boozer.

Now what does it mean to move from a team where the players are quite bad to a team with much better teammates?

For an answer we turn to our study of NBA coaches.  This study considered the impact a variety of factors (beyond coaching) had on player performance.  The list of various factors we considered included the productivity of a player’s teammates, or more precisely, teammate WP48.  This study – across 30 years of data – indicated that teammate WP48 had a statistically significant and negative impact on player performance. 

The coefficient on this factor was -0.300.  And this tells us that the Jefferson’s WP48 should decline by 0.025 as he moves from Minnesota to Utah [-0.300 * (0.109 – 0.025)].

One should note that the change in the quality of Jefferson’s teamates is rather large.  And this impact appears somewhat small.  To see how small, consider how much Jefferson’s and Boozer’s numbers jump from season to season.  Boozer has been consistently above average – and except for seasons where he missed significant time due to injury – Boozer has been consistently above 0.200.  But movements of 0.025 in the WP48 numbers are not uncommon.

A similar story can be seen for Jefferson. With one exception, Jefferson has been an above average player who fails to clear the 0.200 mark.  Again, though, changes of 0.025 are not uncommon (three times we have seen a larger year-to-year change for Jefferson).

In sum, Smith is sort of right.  What we saw from Jefferson in Minnesota is probably close to what we will see from Jefferson in Utah.  In other words, Jefferson will probably be above average but will probably not be able to clear the 0.200 mark (and if he does, probably not by much). 

Unfortunately for fans of Utah, Boozer consistently did more.  Again, this is because Jefferson’s shooting efficiency has consistently lagged behind the mark we see for Boozer.  And that means that Jefferson will probably produce fewer wins than Boozer (and therefore, Utah will probably not be as good in 2010-11 as they were in 2009-10).

– DJ