WAR is a heavily debated statistic. There’s controversy over which version of WAR is better — BP, FanGraphs, or BR — and there’s debate about whether or not ERA, FIP, or xFIP should be used to calculate WAR for pitchers. The following study uses fWAR.
As much debate as there is, WAR is a fairly accurate predictor of a team’s record even with the inconsistencies and flaws in some of the stats. Before we get into the data, here’s how to project a team’s wins with WAR:
In any given year there would are 2430 wins total between all 30 teams (81*30 = 2430). This year FanGraphs gave out a total of 1150.7 WAR, leaving the league replacement level at 2430 minus 1150.7, or 1279.3.
To find the replacement level per team, divide 1279.8 by 30, which equates to 42.6. So, to project a team’s record based on WAR, add 42.6 to their total WAR. Now, let’s explore the 2012 WAR data.
Here is a list of every team’s batting WAR, pitching WAR, and total WAR:
|Team||Batting WAR||Pitching WAR||Total WAR|
Last year I determined that the correlation between WAR and actual wins for the 2011 season was 0.88. This year the correlation dropped to .83. A lot of that drop had to do with the Orioles, who outperformed their WAR wins by 18.5 games.
One standard deviation between WAR and actual wins for this year is 5.40 wins, which is less accurate compared to the 4.61 I found last year. Here are some more facts about this year’s WAR numbers compared to actual wins:
- 18 of the 30 teams were within one standard deviation.
- 28 of the 30 teams were within two standard deviations (10.80). Only two teams — Baltimore (18.5) and Milwaukee (11.3) — were more than two standard deviations away.
- WAR standings would give us the same order of standings for one division: the NL East.
- Four of the ten playoff teams were outside the top 10 in total WAR.
Here are the 2012 regular season standings side-by-side with the 2012 WAR standings:
The purpose of WAR is to provide a stat that quantifies a player’s total value in one easy to use metric. WAR does a good job with that. Using WAR to project a team’s record for the next season won’t give you an exact number, but it can give you a general idea of what kind of numbers a team needs to compete. Unless, of course, a team can outplay their WAR by 18.5 wins like the Orioles.