FanPost

Using win shares to predict the value of free agents

I spent a little time last week trying to come up with a simple formula to estimate the value of free agents. By no means am I trying to crack the code, I just want something that is easy to understand and gives a ball park range of how a player should be paid.

By taking a closer look of win shares across the NBA, it turns out that the distribution is exponential. Similarly, looking at the plot of the salaries of all players across the NBA also reveals an exponential distribution. Since both metrics have a similar shape, we can pretty easily convert their win share metric into an expected salary metric and compare with the players actual salary.

CDF salary CDF winshare

A little more in depth explanation can be found here as well as the visualization for all players in the NBA. Fun tool to play around with.

Here are the expected salaries for the Wizards free agents based on this calculation:

Marcin Gortat: $12.5mm (Signed for 12 million)

Trevor Ariza: $12.3mm (Ain't gonna happen for that much, although people tend to underestimate his contribution)

Trevor Booker: $6.6mm (Since we didn't extend a qualifying offer for 4.6, there is no way he'll receive 6.6mm)

Kevin Seraphin: $1.2mm (Signed for the QO of $3.9mm, seems high)

Other interesting highlights:

Jodie Meeks: $7mm (Signed for 6 million)

Lance Stephenson: $11.4mm (Not happy with what Indiana is offering)

Greg Monroe: $9.1mm (Looking to receive the max, interesting to see how this plays out)

thumbnail nba

This represents the view of the user who wrote the FanPost, and not the entire Bullets Forever community. We're a place of many opinions, not just one.