Barmes 2011 WAR Causes a Question in the Accuracy of Defensive Metrics Used

Posted by on June 6, 2012 | 12 Comments

I don’t have an answer to the question. I’m asking for someone to help me answer the question. Here is the reason:

Clint Barmes posted a 3.1 WAR in 2011. Among position players, that put him at 62nd in all of the Majors in 2011. Exactly 201 position players had a WAR of at least one in 2011. Here is where Barmes ranked among those 201 players in offensive categories (these are where he ranked, not his actual numbers):

OPS+ 163
Runs 137
RBI 146
BA 172
OBP 170
SLG 159
OPS 167

It is weird me that a player who only started 2/3 of the games (114 starts total in 2011) at a tough defensive position could lift himself up and over 80 to 100 players (going from 163rd in OPS+ to 62nd in WAR) based solely on his defense. I have to be missing something. No offense to Seth Smith or anyone who comes up with these formulas because I’m quite certain they have a better feel for this than me and they have spent far more time on this than I have. But, this doesn’t even pass the straight face test. For example:

Paul Konerko (23 in OPS+) hit .300 had an OPS of over .900 and somehow achieved only 2.8 WAR.
Aramis Ramirez (33 in OPS+) playing a tougher position than Konerko had just 2.5 WAR
Yadier Molina (tied for 62 in OPS+) had just 2.6 WAR and won his fourth straight Gold Glove (so his defense was likely pretty good)
Brian McCann (right behind Molina tied for 66 in OPS+) was at 2.4

What I’m getting at is that one of three of the following would have to be true (maybe I’m missing one):
1. WAR overvalues defense in general or did so badly in the case of Barmes 2011 season
2. Clint Barmes had one of the greatest fielding seasons of all time in 2011
3. The defensive metrics used to build the dWAR component are not particularly good

Am I wrong? Can anyone explain this to me? According to WAR, Barmes defense allowed him to leap over clearly superior offensive players, including a Gold Glove winning catcher. That makes no sense to me at all.

Tags: ,
  • http://twitter.com/Schide Keenan Funk

    I had starting writing up a thing before I realized you were using Baseball Reference WAR and not Fangraphs WAR. Fangraphs is probably closer to your perceptions, as Konerko acquired 3.1 fWAR last year and Molina had 4.7. Although Barmes still had 3.1 which led to my confusion.

    You might want to read this WAR primer and then the links about the components and maybe switch over to fWAR because I believe that their numbers more accurately reflect what players are worth. http://www.fangraphs.com/library/index.php/misc/war/ Although it’s still not perfect.

    • RandyLinville

      Thanks Keenan. I have given that a glance before and will take a deeper read of it this weekend. I also look at the FanGraphs WAR data. I knew there were different formulas, but figured that directionally they were all pointing the same way.

  • Kevin_Creagh

    The positional value of a SS in much higher than a 1B and especially a DH, in the case of Konerko. The same holds true for Aramis Ramirez as a 3B.
    I would agree to some extent that the defensive metrics are given too much weight in the overall WAR calculation, but preventing runs is a key part of the game too. It really comes down to looking at the whole body of work for a player and identifying outliers that may have spiked a season’s WAR up from the rest.

    • RandyLinville

      I understand that defensively, SS is more valuable than 1B or 3B. But I would think you’d have to have Dick Stuart at first and a historically fantastic defensive season from SS to make an offensive zero playing SS (Barmes) pass a top 25 hitter playing 1B (Konerko) in terms of overall value. Barmes also passed a Gold Glove winning catcher who was a better offensive player. Molina, looking at OPS+, was the 62nd best hitter in baseball in 2011. He was a catcher. He won a Gold Glove. But in terms of WAR, he was tied for 80th.That’s the kind of thing that makes me think this stat is wacky.

      • john.alcorn

        The fangraphs experts will readily admit that catcher defense is undervalued in their current formula. BTW, I’d also heavily favor fangraphs WAR over BR. UZR for all its faults is superior to total zone IMO.
        At any rate, the key here is wins above replacement. A replacement 1B is much better offensively than a replacement SS. WAR is positionally adjusted, while OPS+ is not.

        • RandyLinville

          Thanks for the thoughts on undervalued catcher defense. I get that WAR is adjusted for positions. Just seems it overvalues defense.

      • http://www.facebook.com/people/Richard-YaZhynka/100000448392567 Richard Ya’Zhynka

        WAR is also positionally adjusted for offense. For instance, a shortstop with a .700 OPS has a much greater WAR value than a firstbaseman with the same OPS.

  • Lee Young

    I think we need a WAR war. lol

  • szielinski

    The fielding fWAR evaluations for qualified ML centerfielders is odd:

    http://www.fangraphs.com/leaders.aspx?pos=cf&stats=fld&lg=all&qual=y&type=1&season=2012&month=0&season1=2012&ind=0&team=0&rost=0&age=0&players=0

    I find it difficult to believe that burners like McCutchen, Maybin, Fowler and Stubbs earned their negative range rating. I also find it difficult to believe that Michael Bourn has superman-like range.

    Eventually and, perhaps, hopefully, MLB will provide Fieldingf/x data to the public. Till then, human observers are the best we have. But I’ll not wholly trust the fWAR fielding evaluations.

  • http://www.facebook.com/people/Richard-YaZhynka/100000448392567 Richard Ya’Zhynka

    On a WAR calculator at http://wahoosonfirst.com/war-calculator/ I gave a SS and a 1B the same numbers – a .315 OBP, a .385 SLG, average fielding, average defense, 600 PA, and 150 games. The WAR calculation for the SS was 2.4. The same numbers resulted in a WAR of 0.5 for the 1B.

    • http://www.facebook.com/people/Richard-YaZhynka/100000448392567 Richard Ya’Zhynka

      WAR is positionally adjusted for offense as well as fielding. Therefore, Barmes’ hitting at SS gives him a much higher offensive WAR than the same hitting numbers would give a 1B or a corner outfielder.

    • RandyLinville

      Thanks for the link – it is high on my list of things to look at this weekend. I’m I’ll spend more time there than I intend to!