Friday, August 14, 2009
WAR and WSAB
Posted by Dave StudemanLong-time readers of the Hardball Times may know that I've done a lot of work with Win Shares, creating Win Shares Above Bench and ranking the all-time greats.
Wins Above Replacement, or WAR, is newest win-based stat of choice (I'm still a fan of WPA, too). I wrote about the difference WAR can make for particular players earlier this week, and I got to wondering how it did overall. Me being me, I graphed it.
Here's a graph of WAR (on the x axis) and WSAB (on the y) for all position players from 1900 to 2008:
As you can see, the two stats not only track very closely, but they almost intersect at zero (I divided WSAB by three to make them equivalent to the WAR scale). If you're a math nut, the R squared between the two is .96. That's the good news.
However, as I showed the other day, the difference between WAR and WSAB can matter a lot for individual players. Also, the variance between the two is greater for individual seasons, as you would expect. That's the more discerning news.
I've got no other point. Just thought you'd like to know.
Dave was called a "national treasure" by Rob Neyer. Seriously. Comments about this article can be sent to him through the miracle of e-mail.
It would be very interesting to see, once the measures are put on the same scale, the players they disagree on the most.