I stumbled across Thomas Boswell’s most recent Washington Post chat in which he wrote [all errors sic]:
The Fangraphs valuations just don’t come close. Something’s wrong with their method and I think its probably UZR. Folr example, they value the whole 84-win Rays team in ’09 at a salry of $229M, but they think the 80-win Brewers, just four less wins, are “worth” only $116M using their stat methods. Huh???? Even worse, they say the A’s and Blue Jays, both with 75 wins, are worth $162M and $176M, but the Astros and Pads, who won 74 and 75 games, are only worth $98M and $96M. Obviously, to me, there’s a big problem here. …
… Go on, explain those salary valuation. It can’t be done. Any method or theory that proposes to assign “win value” to every player has to be consistent with past win totals. Duh.
First, it’s worth getting the numbers correct. In 2009 we don’t have the Brewers at $116M. It’s actually $128M, but in any event, there are always going to be some outliers.
It’s true that the Blue Jays and Athletics were two of the major outliers last year, but the problem is not with UZR, at least in the Blue Jays case. UZR considered the Blue Jays as one of the six worst defensive teams in baseball. The A’s aren’t being overvalued by UZR either at just about average.
If you look at the Pythag Records for 2009, the Blue Jays and Athletics are at -8 and -6, suggesting they should have won 8 and 6 games more respectively based on their runs scored and runs against. Team WAR is also going to reflect that.
Since 2002, our WAR data has a .86 correlation with actual wins. So while we’ve said it before: WAR is consistent with past win totals. Just because you’ve cherry picked a few outliers doesn’t change the fact that the numbers do add up.