Wins Above Replacement (WAR) is Fine: Jim Caple’s Master Class in Strawman Argumentation

Over at, the excellent Jim Caple published an article entitled “Let’s Corral the WAR Horse” in which he argues against the use of WAR as “THE definitive evaluation of a player’s worth.”

It’s a good article, but I’m not sure who it is that he is arguing against.

I have been a proponent of advanced statistics since I was in high school, discovering sites like the baseball cube and Baseball Think Factory.  The Cube provided unbridled access to statistics at players across all levels of the minors and even in college, and Baseball Think Factory provided an incredible forum for people interested in baseball to discuss and analyze most nuanced parts of the game.  The work at BBTF has helped lead to advanced things like WAR, ERA+ and VORP, but also simpler things such as studies about run expectancy (whether it’s good to bunt when down by one with a man on first in the ninth inning).

Unless you were under a rock last Fall, you know that the AL MVP debate raged between supporters of middle-of-the-order stalwart Miguel Cabrera (Triple Crown winner, 44 HR, 139 RBI) and center-field rookie wunderkind Mike Trout (125 runs, 50 SB, gold glove).  Those who supported Cabrera tended to look to traditional stats and to narrative, while those who supported to Trout tended to point at the most advanced statistics, like WAR, to explain their case.

As Mr. Caple explains, WAR embraces defense (as well as baserunning) to come up with one final number which attempts to approximate the value each player provides to his team.  Last year, Trout posted a WAR of 10.4 while Cabrera, with his poor defense and baserunning, posted a WAR of only 6.8.

Caple dismisses the use of WAR because 1) it is complicated, 2) the two main sources of WAR (Baseball-Reference and Fangraphs) provide different calculations of it, and 3) it necessarily encompasses assumptions and theory about player value (“But if we say a shortstop has an Ultimate Zone Rating of 12.4, well, it means he’s likely a very good fielder but we also have to assume that the theory behind UZR is indeed an accurate measure of fielding ability.”).

Caple makes good arguments against using WAR as your only tool to evaluate players.  The problem is, I don’t know a single intelligent person anywhere who would actually do that.

In support of his argument that people look at WAR alone, he provides no evidence.  His only support, aside from anecdotal support, is to link to an off-the-cuff blog post called “Mike Trout is your AL MVP (WAR says so)“.  However the article isn’t even an advocacy article — it is merely a note, and it merely looks to explain the disparity between the two players by looking at their contributions in hitting, defense, and baserunning.

You will always be able to find dummies who will misuse a wonderful tool.  You will always be able to point out the flaws in statistics we use to evaluate things (i.e. batting average is only hits, OPS weights on-base equally to slugging, RBI depend on the runners on base).  But it would be a mistake to turn the discussion about the usefulness of WAR into some anti-intellectual crusade.

WAR is one of many tools that an intelligent baseball fan can use when establishing their opinion on a player’s performance.  There will be outliers, such as the one between Derek Jeter and Brendan Ryan which Caple points out in his article (one site shows Jeter more valuable, the other shows Ryan more valuable).  But there are outliers when it comes to ANY statistic.  Who would you rather have had last year, Ben Revere with his .294 average (and only 675 OPS) or Edwin Encarnacion, with his .280 average (and 941 OPS)?  A lumbering designated hitter with a 900 OPS would likely be less valuable to a team than a gold glove shortstop with an 800 OPS.

In addition, there is a point where the disparity between two players simply reaches beyond actuarial error.  Although a 1 WAR gap between players could perhaps be attributed to the volatility and uncertainty of fielding and baserunning statistics, a 4 WAR gap can certainly not be.  This is where the debate between Trout and Cabrera becomes the most pointed.

You can disagree as to whether Mike Trout deserved the MVP last year, but no matter which way you slice it, he had a season last season for the ages, combining power (30 HR), speed (50 SB), on base skills (.394 obp) and defense (gold glove) to comprise the ultimate player.  Valuing those attributes is not new.

When a player with speed at a premium position (CF/SS) posts an incredible year, the odds are that he will win the MVP or finish close.  Rickey Henderson twice posted a WAR of 9.8.  He won the MVP once, and came in third another time.  Ken Griffey twice posted 8.9 WAR or greater.  In those seasons he won the MVP and came in fourth. Alex Rodriguez posted a WAR of 9 or higher four times, finishing first, first, second, and third in the MVP vote.  And this is not a new phenomenon — in many seasons, older generations of fans would take Mickey Mantle over Ted Kluszewski or Andre Dawson over Jack Clark, or a young Barry Bonds over Eddie Murray in landslides despite statistical parity.  Similarly, I would have voted for Albert Pujols (8.2 WAR) or Carlos Beltran (8.0 WAR) over Ryan Howard (5.0) WAR in the 2006 NL MVP vote.  Pujols was just as good offensively as Howard, plus did it with the glove.  Beltran was a tick behind offensively, but was a gold glove center fielder.

Even for people like me — who view WAR as a useful tool rather than a be-all-end-all statistic — can see a gap between a 10+ WAR season and a 6.8 WAR season as beyond the margin of error presented by the thoughtful critiques that Mr. Caple points out in his article.  I’m willing to say confidently that Trout was better than Cabrera last year, not because of WAR, although it’s nice that WAR agrees with what I feel like I know, intuitively, about the game.

WAR, rather than simplifying things, is an incredible tool which can be used (loosely) to compare players on different teams, across leagues, and across eras.  But please, put away the pitchforks, because I don’ t believe that any thinking fan wants to replace the MVP vote with a spreadsheet.  (See 2009 AL MVP for example).


Brian Mangan is an attorney who lives in New York.  He loves statistics, but also really loves baseball (and both plays and watches games!).  He does not live in his mother’s basement.