The Talent Pool and Marginal Value in the NBA

Posted on 07/14/2010 by


One of the most commonly held truths in the NBA is the following: “You need the big man to win” . This  illustrates to me that the perception of value  by position in the NBA is fixed. The fixed nature of this notion to me is disconsonant with the way markets work in the real world.

Markets in the real world are fickle,variable things. Values and Prices change and move based on an almost limitless and not easily quantifiable set of variables. Predicting them as investors, the Fed & other world-wide institutions have sadly found out is a nontrivial challenge. However in sports, we have certain advantages .

Markets in sports have the advantage of being confined to a discrete set of parameters and results. Points, Wins, Losses and other easily measurable statistics are available for analysis and proper valuation of assets (independent of whether or not this analysis is done, the data exists). This is what allows us to make a reasonably accurate  quantitative assessment of an NBA players’ value to his team.

One of the innovative features of Wins Produced is that it focuses on the Marginal Value of an NBA player vs others playing the same position for the particular set. According to Wikipedia, marginal value is:

1. a value  that holds true given particular constraints,

2. the change in a value associated with a specific change in some independent variable, whether it be of that variable or of a dependent variable, or

3. [when underlying values are quantified] the ratio of the change of a dependent variable to that of the independent variable.

What this means for Wins Produced is that a player is always compared to his peers and his value is measured based on the competitive advantage in wins he provides to his team. So knowing a players individual statistics in a vacuum is not enough to evaluate him, you have to know who he will be playing against,the talent pool, to make a truly informed decision.

So now we can finally get to some pretty charts. When we look at the Average player value measured in ADJP48 (the oft mentioned position adjustment) over time we get the following:

PrettyWhich looks like this in a Chart:

What this data shows is that the level of talent at a given position is not fixed over time but rather an ever evolving variable. The marginal Value of a player to a team is a function of the talent pool of peers competing against him. The value of an employee (the player) to the employer (NBA Teams) is heavily influenced by other available employees.

Let’s pick some examples to illustrate what this means. I decided to use the best seasons available for players who played solely in one position (except for SG where I adjusted up from 98% beacuse of scarcity).

Name                     Pos         Year       Team     Sum WP    Total WP48

Chris Paul             PG          2009      NOH      28.81         0.461

David Robinson C             1991        SAS        29.25        0.454

Larry Bird             SF          1987       BOS        25.79        0.412

Charles Barkley  PF          1993       PHO       24.59        0.413

Dwyane Wade     SG          2009      MIA       21.61        0.340

We will call these five seasons our Standards Seasons by Position. These were all great seasons when they happened  but what would their value be if we moved them over time:

That Would be a hell of a Team

It’s clearly evident from these charts the talent floor in the NBA is not fixed. As impressive as Chris Paul season in 2009 was it would have been 4.4 wins better in 1978 when the quality of point guards was much less.  There’s a long period (about 85-05 ) where big men were scarce but now a dominant big man is not as much of a competitive advantage (sorry Magic fans). Right now the biggest marginal values are to be had at Shooting Guard, Point Guard & Small forward (sorry everyone who isn’t a Heat fan).

Big men (C & PF) have won championships and probably will again but right now the little men look to be Kings.

Posted in: Uncategorized