Sports Reference Blog

A Discussion of WAR Wherein I Ardently Attempt to Avoid any WAR-Related Puns

Posted by sean on November 21, 2017

This article assumes a lot of prior knowledge about the discussion of Wins Above Replacement, you can catch up here

First off, none of us are here without Bill James. We are all at our very best merely Chaucer or Joyce to his Shakespeare. All sabermetrics predating him flowed into his work and all sabermetrics after him carries echoes of his work.

To the discussion at hand.

At its nub, I see this debate/argument/kerfuffle falling along the lines of an argument pitting Results vs Process. This debate goes back at least as long as statheads have gathered online to discuss MVP votes and what the award means. In five minutes of digging, I found this 1997 discussion from rec.sport.baseball. Every year, this exact same debate happens on websites and online forums. EVERY SINGLE YEAR. Every single discussion of this piece (that I've seen), whether on Twitter, BaseballThinkFactory.org or in the comments on Joe Posnanski's piece eventually distills down to this discussion.

To highlight that, I was struck by how Bill James ended his piece.

"What creates value for a baseball player is winning games."

This is a formulation that never, ever would occur to me and very likely would not to the other publishers of WAR who have weighed in. Dave Cameron, for instance, stated:

“What did each player do, as an individual, to help his team try to win games?”

Bill is saying (I'm paraphrasing): "Winning games [or Wins] creates value for a baseball player." Note that it's undetermined as stated if value ONLY comes from wins, but that's a small point.
Dave is saying, and I would also say: "Player value creates wins." I think when we talk about "luck" or "variance" we are saying that some portion of wins also comes from something outside of the players' control, so it's not a necessary and sufficient (in a mathematical sense) relationship. Wins aren't only created by direct player value.

From those very different paradigms, the entire argument flows and each side makes decisions based on those differing views.

Now through years of arguing and debate, I have come to the conclusion that these two differing approaches, when considering how to value a season already in the books, are largely a matter of taste and worldview. Looking forward, I think the latter is better, but in looking back I'm fine with a person taking either approach to evaluating the value a player added to during a season. In interacting with Bill on Twitter, I believe that he's steadfast that his view is correct and ours is wrong (or "nonsense", "misleading", "in error" to use his words). I disagree and believe each viewpoint has merit and is, on this issue, largely one of personal preference.

In order to place my marker on what I think our WAR metric's usefulness is, I'm going to elaborate on my views here.

I believe:

  • Baseball-Reference.com's WAR tells us how many wins a typical team would lose if the player was replaced with a replacement-level player.
  • I believe that value assessment is relevant to the MVP vote (though not necessarily definitive).
  • Were I to vote, my criteria would be: "If I was a GM on April 1 and knew all of the players' performances for the season ahead, which player would I choose?"
  • Again, if you don't like that, I harbor you no grudge, and you can choose differently, but
  • You also are never going to convince me that your way is the One True Path™ that we all must follow.
  • As Jonathan Judge mentioned in his piece linked above, the WAR values should be viewed as containing error bars. I've never invested the time necessary to make this improvement (truth be told: all of my statistics is pre-bayesian, so I'm a dinosaur in that regard).
  • So when WAR is as close as it was at the top of the AL and NL this year, I'd probably look at things like team success, WPA (clutch play), and more in making my ballot.

I think that covers it.

Two last points--one that amused me and one that didn't. First, Altuve and Judge were very close in WAR and Judge had the edge if you combine B-R, BP, and FanGraph's numbers. Reading Bill's article you would think that Judge won in near-unanimous fashion, but Altuve won in a romp. Judge led Altuve on two of 30 ballots. The voters knew what was what. They knew Altuve and Judge were the top two and then considered them relative to each other and came down very heavily on the side of Altuve. So any view that the MVP has or will become a rehash of the WAR leaderboard seems overwrought.

Lastly, there was one thing that made me a bit melancholy in reading Bill's initial piece and Joe Posnanski's follow up. Both Bill and Joe relate that Bill was reticent to comment on flaws he saw in others' work due to not wanting to "punch down" on young analysts and made a conscious decision not to engage: "At that time it was my policy not to argue with younger analysts. I was much more well-known, at that time, than they were, and it’s a one-way street." Initially, I found this rather patronizing, though I understood what Bill was saying. After thinking about it more, it made me a bit sad. We can discuss this stuff without arguing. We all share a mutual purpose in learning what is true and what is not. We all love baseball. I can imagine that going to war for years with Elias and the baseball establishment could lead to cynicism and a winner-take-all viewpoint, but I am disappointed that my peers and I missed out on Bill's mentorship and expertise.

postscript: I have a lot of stuff to do and very little bandwidth to keep discussing this, so I apologize if I don't engage with your (I'm certain) very well-reasoned arguments.

Editor's Note: You have my permission to republish any or all of this (please do!) provided you link to this original post and attribute it to me. Sean Forman

7 Responses to “A Discussion of WAR Wherein I Ardently Attempt to Avoid any WAR-Related Puns”

  1. Jason Winter Says:

    I've written my share of scathing blog posts and takedowns and the like, but I'm just a (relatively) no-name schmuck. If I were a bigger name, and I had an issue with some of the policies or whatever of the people/companies I'm writing about, and thought that their opinions carried *any* merit -- as opposed to some Joe Morgan-esque, old man, "this is how things were in my day, dammit" -- I might make some attempt to contact them to try and hash things out.

    That's what kind of saddens me about all this. Bill and Joe are certainly big enough names, and I'm sure they could both have easily gotten in touch with you and talked about this and then present their take on the matters. Instead, they both presented their arguments like "angry bloggers" with no access whatsoever and a grudge.

    Also, this, from Bill's first piece:

    "But. . . I have held my peace on this for 20-some years. . .that argument is just dead wrong. There are five reasons why it is wrong."

    " And fourth and finally, the connection between wins and other statistical accomplishments is the basis of statistical analysis."

    5=4? Now that is some *really* advanced statistical analysis!

  2. StevenKeys Says:

    "♪ WAR, huh, good god, what is it good for, absolutely nothin' ♫ (War, Starr, '70, Gordy)!"
    Unless you're in the club, i.e., junior sports journalist & fantasy fanatic. Then a must-use.

  3. section 34 Says:

    I am a Baltimore Orioles fan. We have hated analysts flaunting WAR in our faces for several years now.

    It's not the statistic itself, which the Orioles have intermittently proven is flawed along the lines James suggests. I actually like WAR as a shortcut for comparing players from different eras. What I hate is the smug authority with which analysts use it. WAR is not a real statistic that measures what happened; it's a useful shortcut and approximation. But in the hands of certain writers it is the end of the discussion, not the beginning.

  4. Bill James vs. The Noise Says:

    […] https://www.sports-reference.com/blog/2017/11/a-discussion-of-war-wherein-i-ardently-attempt-to-avoi… […]

  5. An (Re)Introduction to JAWS: The Formula to Help Determine Who Should Make the Hall of Fame | SoCal Sports Net Says:

    […] have seen an interesting debate between leading lights of the sabermetric community—Bill James, Forman, FanGraphs’ Dave Cameron, Baseball Prospectus’ Jonathan Judge—over the importance of […]

  6. Tim Says:

    I don't like the way  most of the Cy Young/MVP voters are just basing everything on WAR. Like Scherzer over Krenshaw, for instance. Why have voters, if that's the case? Also, the fact that there are two different WAR stats, Baseball Reference and Fangraphs, and they had some pretty wide differences, like difference winner in AL and huge difference in how valuable Andrelton Simmons's defense is. I think that traditional stats still are valuable, and aren't based on discretionary formulas.
    If a guy had 40 homers and 100 RBI , for instance, there is not gonna be another formula that says he had 30 homes and 120 RBI 

  7. Gordon Arsenoff Says:

    WAR values should be viewed as containing error bars. I've never invested the time necessary to make this improvement [...].

    I've gotten pretty good at computing appropriate error bars for stuff, including WAR components. I have at least an initial sense of how we'd do this for B-R WAR. E-mail me if you'd like to discuss setting up a project to get this done.

Leave a Reply

HTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>