# The Underpaid and Overpaid in 2008-09

Steve Walters’ brilliant discussion of general managers in baseball provides a nice jumping off point for the annual review of “overpaid” and “underpaid” NBA players.   Such a review begins with observations that are quite similar to what was said last summer.

According to the USA Today, the NBA spent \$2.2 billion on player salaries in 2008-09.  With 445 players taking the court, the average salary in the Association was \$4.85 million.

Although this is a large sum of money, one suspects it could be even higher.  The NBA has instituted a variety of salary restrictions.  There is the famous salary cap, which limits how much each team can spend on its payroll.  There is also a maximum salary, which is a cap on individual player salaries (and yes, this should be called a salary cap).  Finally there is a cap on rookie pay, which limits how much a first round draft choice can earn on his first contract. All of these restrictions are designed to limit a player’s pay.  And whenever we observe an employer trying to limit an employee’s pay, we suspect that those employees might be exploited.

What does it mean to be exploited?  In 1933, Joan Robinson argued a worker is exploited if he or she is paid a wage less than the revenue being generated.  Given this definitions, if a worker is paid \$10 an hour, but only generates \$8 of value, then that worker is not exploited.  But if a worker is paid \$10 million, but generates \$12 million in value, then by definition he or she is exploited.

Given this definition, which players in the NBA are the most exploited? And on the flip side, which players are exploiting the NBA?

To answer these questions I am going to follow the same methodology I have employed the past two years.  Here is the description of this methodology:

The standard approach is to simply regress team revenue on wins (and other stuff). But I think there’s a problem with this approach for the NBA (a problem I wish to avoid getting in to for a blog entry), so I am going to estimate the value of a win differently.  We know from our research on revenue and attendance [mentioned in The Wages of Wins, which is soon to be published as a paperback :) ] that players primarily generate value in the NBA via their ability to generate wins.  And we also know that a player’s wage is only for the regular season.  Consequently, we could say that the value of one win in the NBA is simply the amount of money the league paid its players divided by how many wins these players produced in the regular season.

Such an approach makes three assumptions.  I am assuming that all players in the NBA are collectively paid what they are worth (which may be true if the union bargained effectively), players are only paid to produce wins (which is a reasonable assumption given the research cited above), and the value of a win is the same for all teams (okay, not true, but two out of three ain’t too bad).

I have noted in the past that I am not thrilled with this approach.  In fact, last year I stated: “After thinking about it a bit more, I think I can do better.  But I don’t feel like working through my idea for a blog post.”  When I read those sentences this year I wondered why I never bothered to write this “better idea” down someplace.  I am sure it will come back to me someday.  In the meantime, though, we are left with the same old methodology.

We have a measure of how many wins each player created (i.e. Wins Produced).  And if teams spent \$2.2 billion on players, and these players produced 1,230 regular season wins, we can argue that each win is worth \$1.755 million.   With a value of win in hand, all we need to do is to multiply the value of a win by each player’s Wins Produced.   This calculation gives us a measure of what a team could have paid for a player’s productivity; a measure which can be easily compared to a player’s salary.  Once again, players who produced more than they were paid are, by definition, exploited or underpaid.

So who were the most exploited in 2008-09?  The answer is in Table One.

Topping the list of underpaid players is Chris Paul [who also led this list in 2007-08].  According to the USA Today, Paul was paid \$4.6 million in 2008-09.  For this money, the New Orleans Hornets received 28.2 wins.  At \$1.7 million per win, the Hornets should have paid \$49.5 million to Paul; so Paul was underpaid by nearly \$45 million (at least, that’s the answer if we accept this particular methodology).  HoopsHype says that Paul is scheduled to make \$13.5 million in 2009-10. If this was Paul’s salary in 2008-09 he would have only been underpaid by about \$36 million. Such a mark would still lead the list of underpaid players, but the gap between Paul and LeBron James — the number two player on the list — would be much smaller.

When we look over the list of underpaid we see the work of two institutions in the NBA; the rookie salary scale and the limit on how much can be paid to any individual player.  The limit on rookie pay allows teams to dramatically underpay players like Paul, Rajon Rondo, Brandon Roy, David Lee, Kevin Durant, and Al Horford. LeBron, Dwight Howard, and Dwyane Wade are each paid quite a bit, but nowhere near what these players would command if each win cost their employer \$1.7 million. And then there are players like Dominic McGuire and Jamario Moon.  These players are underpaid because what they do is underappreciated (or underrated).  If a general manager employs one of the players listed in Table One he is either smart or lucky.

In contrast, if a general manager employs a player listed in Table Two he is not so lucky.

Topping the list is Jermaine O’Neal, who is a repeat winner of the Most Overpaid Player (MOP).  To be fair, this year I set the cut-off at 2,000 minutes played.  O’Neal didn’t play this many minutes in 2007-08, so he would not have been the MOP last year by the 2008-09 standard (that award would have gone to Boris Diaw).  Nevertheless, O’Neal didn’t contribute much in 2007-08.  Despite his high salary and lack of production, both the Toronto Raptors and Miami Heat thought it was a good idea to trade for his services.

Once upon a time, Jermaine O’Neal was a productive player.  In 2002-03 he posted a 0.194WP48 [Wins Produced per 48 minutes], a mark that is nearly double what we see from an average player.  As recently as 2006-07, he was still above average. Across the last two years, though, O’Neal has struggled; and before the next season he will turn 31.  So it seems likely that the days of a productive O’Neal are over.  His days of getting paid like a star, though, are not.  HoopsHype indicates that O’Neal will get paid nearly \$23 million in 2009-10.  This means it’s possible O’Neal will repeat his MOP title in 2009-10.

In fact, this is likely.  HoopsHype reports that Jermaine O’Neal will rank third in the NBA in salary in 2009-10.   Allen Iverson – who was the runner-up for MOP this past year – will take a substantial pay cut if he joins another NBA team.   Al Harrington may be able to repeat his third place MOP ranking, but it’s unlikely that his productivity will fall enough for him to catch O’Neal.  No, if O’Neal manages to log 2,000 minutes again in 2009-10, it looks like this title stays with him.

Beyond 2010, though, a new name will emerge.  O’Neal will probably take a pay cut after this season.  This means a player like Andrea Bargnani, Al Thornton, or Spencer Hawes could rise up and take a future MOP title.  Once again, though, that probably won’t happen until 2011.

Let me close by noting that this is the end of the underrated, overrated, underpaid, and overpaid columns for this year.  So look for something different in the next column.  Not sure what that topic will be, although I am sure it will have to be something different.

- DJ

Our research on the NBA was summarized HERE.

The Technical Notes at wagesofwins.com provides substantially more information on the published research behind Wins Produced and Win Score

Wins Produced, Win Score, and PAWSmin are also discussed in the following posts:

Simple Models of Player Performance

Wins Produced vs. Win Score

What Wins Produced Says and What It Does Not Say

Introducing PAWSmin — and a Defense of Box Score Statistics

Finally, A Guide to Evaluating Models contains useful hints on how to interpret and evaluate statistical models.