What's a fair market price for Agent Zero?

Having read about the Washington Bullets imminent offer to pay +$100 million over six seasons to free agent point guard Gilbert Arenas, I decided to try to determine what his actual fair market value is by using offensive and defensive win score calculations to try to estimate how many wins he's likely to produce and how much each team was willing to pay each player per win. Now remember, this is purely an academic exercise... completely detached from the reality of the wildly distorted NBA labor market.

To establish Gilbert's fair market value, first I established the likely market rate teams will be willing to pay per win next season. I came up with the number $1.8 million. This number comes from the actual market value teams paid per win last season, $1.67 million (total NBA salary divided by total number of wins), and then I raised that by the likely increase next season in total NBA salaries, which should be about 8 percent (based on what the increase has been the last couple of years).

Next, I tried to estimate how many wins the Bullets could expect from Gilbert next season. This is a dicey proposition because of Gilbert's recent injury problems, but I took a stab at it. Here's how.

To determine Gilbert's likely offensive contribution, I averaged Gilbert's Win Score per 48 over the past 4 seasons, and that came out to 7.6. To determine his likely defensive contribution, I averaged his counterpart opponent's Win Score per 48 over the same time period, and that came out to 6.9.

Now, I needed to estimate the number of minutes he could be expected to play. I took the average of the Bullets total player minutes over the last 4 seasons and the average percentage of those minutes consumed by Gilbert. Putting those two averages together, I estimated that Gilbert would play about 2499 minutes next season.

Applying his offensive and defensive Win Score averages to that amount of playing time, and assuming the average Win Score per 48 for NBA point guards would remain at this season's 6.9, I calculated that Gilbert Arenas was likely to add 6.4 offensive half wins to the Bullets next season and 5.2 defensive half wins, for a grand total of 5.8 wins overall. At the anticipated market rate per win of $1.8 million, that means a fair contract offer to Gilbert Arenas for next season would be about $10.4 million. Projecting into the future, I guess you'd have to anticipate some deterioration in Gilbert's Win Contributions, but, giving him the benefit of the doubt, I would offer him a 6 year, $62.4 million contract. In other words, more than $40 million less than the Bullets are prepared to offer.

And given that Arenas made more than that per annum on his last deal, I estimate it would take him less than 2.2 seconds to turn my offer down. As I said, this was a purely academic exercise.