Am I misusing WAR?

by erosen on May 6, 2010 5:23 PM EDT

I grabbed the team batting, defense and pitching WAR numbers from fangraphs and averaged the team totals by year:

2009 AL 36.4 vs NL 31.6

2008 AL 38.6 vs NL 32.4

2007 AL 37.0 vs NL 33.1

Is this a reasonable way to illustrate the ALs edge over the NL? If so, is it still the case?

2010 AL 5.9 vs NL 6.4 (which would be about 34.4 vs 37.6 at this pace)

Tweet Comment 6 comments | 0 recs |

Do you like this story?

Comments

As far as I can tell, yes, it's a misuse.

Fangraphs WAR does not, at least last I knew, make an explicit league adjustment. And win percentages across leagues are “fixed” (with the mostly trivial exception of interleague games).

What you are seeing is an artifact of how fWAR handles pitcher batting – it assigns them a batting value compared to the average hitter in the league, with no positional adjustment. This is slightly offset by the negative positional adjustment given to the DH (positional adjustment should “zero out” in the NL but be negative at about -1.5 WAR per AL team). But on the whole, the negative value for NL pitchers hitting more than offsets that.

by cwyers on May 6, 2010 7:24 PM EDT reply actions

The previous years' WAR values are probably somewhat descriptive, because they include the interleague games.

The AL has won more, and should rack of better stats during those games. (Your differences match the league differential pretty well, too.)

But so far in 2010, teams have only played intraleague games, so comparing the WAR for each league is no better than comparing AL WAR to Little League WAR. There are the same number of wins to go around in each league.

Beyond the Box Score Not a member of our saber-slanted community? Sign up here.

by Sky Kalkman on May 7, 2010 9:24 AM EDT reply actions