The Sabermetric Revolution, as Applied to Ryan Doumit

Allow me to try to simplify the sabermetric revolution as much as I can:

Late 20th Century: we are evaluating baseball players
Early 21st Century: we were so wrong about our baseball player evaluation!
Less Early 21st Century: we were so wrong about our baseball player evaluation, again!

First, there were players, then there were numbers. Then there were better numbers, then there were still better numbers. The numbers will only continue to improve with time, and a lot of the things we currently think we know about baseball will probably end up being at least partially untrue. Keep that in mind next time you express a particularly strong opinion. But anyway.

Several years ago, people started to care an awful lot about on-base percentage and offensive productivity. This was warranted, because it is important to get on base and be offensively productive. A little later on, people started to care an awful lot about defense. Turns out some of those OBP-happy sluggers were subtracting runs almost as fast as they were adding them. Whoopsadoodle.

I’d like to shift from the general to the specific, and I’d like to focus on Ryan Doumit as something of a case study. Doumit’s perceived value, I think, has moved around dramatically — not necessarily within the game, but among many of the observers. I know that, personally, the way I feel about Doumit now is quite a bit different from the way I felt about him in 2006 and 2007. And it’s not like Doumit as a player has really changed. What’s changed is the way we can look at him.

Doumit, in the minors, was a catcher. A switch-hitting catcher, with a career .984 OPS in triple-A. He could walk and hit for power and catch, and around the time he emerged, this was delightful. He came up with the Pirates and caught some and hit some. But it didn’t feel like he played enough; Doumit came off as being under-appreciated. Here was a guy who could hit while defending behind the plate, and that’s a rare breed. A lot of stat-savvy fans wanted their teams to trade for Doumit. A lot of stat-savvy Pirates fans wanted their team to play more Doumit. He did eventually get more playing time, catching 106 games in 2008, and his offense was good. There wasn’t a whole lot not to like about Ryan Doumit.

The Pirates, for their part, were never wild about Doumit’s defense. He played other positions, besides catcher, in order for the Pirates to keep his bat in the lineup. But we didn’t really know how to evaluate catcher defense, and Doumit didn’t seem that bad according to what was publicly available. It seemed like more evidence of under-appreciation.

Now we fast-forward, and here comes the pitch-framing part. Doumit, for his career, has been slightly below-average at throwing out base-stealers. He’s been slightly below-average at blocking pitches. In 2011, though, Mike Fast published some pitch-framing research that showed Doumit to be just dreadful when it comes to getting strikes for the pitchers. Doumit was shown to be the worst framer in the league, and the suggested effect was substantial.

This past year, Matthew Carruth conducted his own research that in large part agreed with Fast’s conclusions. Carruth also found that Doumit was the worst framer in the league. Between 2007-2012, Doumit received more than 28,000 called pitches. He was 1,104 called strikes below average, or -39 per 1,000 called pitches. This doesn’t look like something we can just sort of shrug off, like Doumit’s pitch-blocking. This looks like something that rather dramatically changes Ryan Doumit’s overall value.

Pitch-framing is still only somewhat understood, but while a lot of people have gotten familiar with Jose Molina as being the best, Doumit’s there at the other extreme. It in part explains why the Pirates were never thrilled with him, and though the Pirates did a lot of things wrong, it doesn’t look like they were wrong about Doumit as a catcher. If anything, one could argue they played him at catcher too often.

At one point, Doumit looked to be under-valued, because he could catch. Now he looks to be more appropriately-valued, or perhaps over-valued, because he can’t catch, really. He just does it when he’s asked to. We can take this beyond pitch-framing, if you allow me to get a little bit sloppy.

Every year since 2005, Doumit has caught, and in every year but 2006, Doumit has caught multiple hundreds of innings. What comes next is basically CERA, except without the “E” part. I know this is an unforgivable sin, but I can’t help myself. We have big samples of Doumit-catching time and not-Doumit-catching time. Let’s look at a table, leaving out 2006 (when Doumit’s catching numbers in a small sample were awful):

Year Doumit, CRA Non-Doumit, CRA
2005 5.08 4.71
2007 5.87 5.15
2008 5.41 5.57
2009 4.86 4.89
2010 5.39 5.69
2011 5.36 3.98
2012 5.64 4.98
Average 5.37 5.00

That’s a straight-up, un-weighted average. On average, when Doumit’s been behind the plate, his team’s allowed 5.37 runs per nine innings. On average, when Doumit hasn’t been behind the plate, his team’s allowed 5.00 runs per nine innings. This does not adjust for playing time or for pitchers caught, so of course it isn’t perfect. There’s a lot of potential error in here. But this could also be capturing everything — throwing arm, pitch-blocking, pitch-framing, game-calling. CRA is a hopelessly flawed statistic, but for whatever it’s worth, it isn’t kind to Ryan Doumit, which goes along with his defensive reputation.

Teams throw about 1,440 innings a season. Over 1,440 innings, a 0.37 RA difference is equal to about 59 runs. If we’re in any way capturing Ryan Doumit’s true defense as a catcher, he’s really not good at it, and his bat probably isn’t good enough to make things much better.

The evidence suggests that Ryan Doumit really just shouldn’t catch. Not for a contender, and probably not in anything but an emergency. Just to add on, let’s be kind of sloppy again. Over Ryan Doumit’s career, his team has won an average of 40.1% of the games he’s started (at any position, not just catcher). Over Ryan Doumit’s career, his team has won an average of 40.7% of the games he hasn’t started. He’s started 627 games, and he hasn’t started 668 games. He owns a career 106 wRC+, and 11.1 WAR.

Something tells me that WAR might be off. Something tells me Doumit hasn’t actually been much of a contributor. Something tells me that has a lot to do with his work behind the plate. Limited evidence suggests that Doumit can kind of handle himself in the outfield, and maybe at first base. Given his bat, he belongs in the league. But it seems like Doumit, for his career, has cost his teams a lot of runs, because they’ve put him behind the plate, where he hasn’t been any good. He’s been the sort of quiet-bad that adds up to some staggering figures, and while none of the evidence within this post is conclusive, it can’t be dismissed outright, either.

At one time, when we knew something, it seemed like Ryan Doumit was an undervalued catcher. At this time, when we know more, it seems like Ryan Doumit shouldn’t be a catcher at all, as he’s probably cost his teams way too many runs. Doumit’s actual career WAR might be zero, or it might even be worse than that. One thing, we weren’t wrong about: Ryan Doumit can hit. It’s a good thing for him that he can.

Print This Post

Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.

54 Responses to “The Sabermetric Revolution, as Applied to Ryan Doumit”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Jacob Smith says:

    He’s probably still worth a roster spot as a 5th OF / backup 1B who can play C in emergency situations. The fact that he can play C at the major league level, if very, very poorly, has some amount of utility value on its own, even without the bat.

    +10 Vote -1 Vote +1

    • Yeah, he makes sense as a part-timer and emergency third catcher. But, oh, the catching.

      Vote -1 Vote +1

    • AK7007 says:

      This. I always saw him as a Rays kind of player who could be inserted at any of several positions on each given day. He should catch less. But probably some amount of catching is correct for him, maybe even allowing a team to forgo a backup catcher altogether.

      It is also interesting that pitch framing does not appear to be a skill that players learn over time. It seems to be a skill, not a physical talent/attribute, and it would be surprising if it were not taught.

      Vote -1 Vote +1

      • Jacob Smith says:

        If I were a GM, I’d want an actual backup C on the roster. Doumit would be the “emergency 3rd C”, along with backing up 1B/DH and probably LF/RF (Related: can he play RF? I’ve never seen him in the OF).

        The situations where I would play him at C are mostly injury related, or 20+ inning game marathons where I’d pull the starting C for Doumit after some arbitrary number of innings.

        As an injury example, the starting C has a pulled hamstring, and is going to miss a week or so. The backup C starts 5-6 games, with Doumit getting a start somewhere in the middle as a day off for the backup.

        Vote -1 Vote +1

      • Tim says:

        He can stand in RF, at least, I’ve seen it. To my eye he didn’t look a whole lot worse than Jason Kubel, and he was certainly better than Delmon Young, not that that’s saying much.

        Vote -1 Vote +1

    • piratesbreak500 says:

      I remember when they showed the ERA of the catchers when Paulino was catching as a rookie (before he gained weight), and Doumit. It was a difference of about a run per nine.

      Vote -1 Vote +1

  2. Jeff says:

    Ryan Doumit then, John Jaso now.

    Vote -1 Vote +1

  3. jmart1951 says:

    Have you ever placed McCann under the same assumptions as Doumit?
    Is he substantially better than Doumit under the guidelines that you chose above?
    I believe that McCann’s arm is slightly below average and that his blocking skills are also slightly below average. He probably is a fairly good pitch framer and his game calling skills are probably a good bit better than Doumit’s, but game calling is probably generally included in your stats but maybe not to the level of importance it should have (not sure).
    Did you look at other catchers under the same criteria. I am just guessing here but this method of catcher evaluation might have a tendency to lower most catcher’s ratings (you would be more qualified than I to determine that).

    Vote -1 Vote +1

  4. I Agree Guy says:

    Anecdotally, from having watched him catch this last season, he is a pretty lousy receiver of the baseball and a poor all-around defender at the position.

    Ground-breaking, I know.

    Vote -1 Vote +1

  5. Bryan says:

    Interesting article. I take the core point of the article to be that because Saber is a constantly mutating field of knowledge, what was rational today might be irrational tomorrow, and vice versa. Take away message: be humble about analytics, don’t over play your hand, and triangulate with multiple sources of information (e.g. scouting report).

    Vote -1 Vote +1

  6. Tim says:

    Surely there’s a reasonably simple way to convert strikes/1000 to runs/game. Or maybe in the worst case we’d have to look at all the miscalls individually. But we shouldn’t have to handwave our way around the value there.

    Vote -1 Vote +1

    • Krog says:

      Absolutely. Until then, a simpler way to do it would be strikes/100. A starting pitcher throws about 100 pitches nowadays, so it is a better frame of reference. I have no frame of reference for what 1,000 pitches looks like! With Doumit, it means 3.9 missed strikes per 100 pitches, which would make any starter angry.

      Vote -1 Vote +1

  7. glib says:

    If MLB ever moves to electronic pitch tracking, all of a sudden Doumit’s value increases by +59 runs over a season. And I don’t know anyone who would be against.

    +8 Vote -1 Vote +1

    • matt w says:

      Jose Molina would be against.

      +6 Vote -1 Vote +1

    • Baltar says:

      Oddly, I was dead-set in favor of electronic calling of balls and strikes until this pitch-framing evidence came up. Now, I kind of like the idea of catchers having (or lacking) this skill. It has, in fact, greatly augmented my appreciation of the importance of catchers, and I now enjoy watching the catcher during games greatly.

      +8 Vote -1 Vote +1

  8. TT says:

    “First, there were players, then there were numbers. Then there were better numbers, then there were still better numbers. The numbers will only continue to improve with time, and a lot of the things we currently think we know about baseball will probably end up being at least partially untrue.”

    At some point you have to admit the numbers are often misleading when used to evaluate players. If they contradict the subjective judgments of professional baseball people, they are probably wrong. And compilations of numbers that have no discernible criteria for testing against the real world, like WAR, should probably be ignored.

    Unlike the evaluations based on solely spreadsheets, it does not appear that baseball’s evaluation of Doumit has changed much. The Twins used him as a DH and in the OF most of the time. They kept Drew Butera, a catch and throw guy, on the roster as a third catcher and he got almost as many starts as Doumit.

    That does not mean Doumit doesn’t have any value as a catcher. He does, even if his defense is marginal. But, like a lot of major league players, his value depends on the manager putting him in situations where his strengths will contribute and his flaws will be minimized. Thats the fatal flaw in looking at results and projecting value from them.

    Vote -1 Vote +1

    • adam w says:

      “At some point you have to admit the numbers are often misleading when used to evaluate players. If they contradict the subjective judgments of professional baseball people, they are probably wrong.”

      Looks like some things never change…

      Vote -1 Vote +1

      • KM says:

        TT goes too far, definitely. But there is a middle ground there… Dom Brown, for example, had pretty great numbers as a prospect… But it turns out when you see him that he doesn’t really have any sense of how to read a ball & take a good route to it in the field, and that he has a very funky swing that major league pitching will be able to exploit when they figure it out. He may still turn out to be an above average player, but certainly had his warts as well. Basically all I’m saying is that it’s worth weighing visual as well as statistical information in context…

        Vote -1 Vote +1

  9. jcxy says:

    Really interesting look at this topic through the lens of Doumit. Curious though…Jaso would have been more apropos though, right?

    Vote -1 Vote +1

  10. Ryan says:

    I don’t think that Doumit is as bad of a pitch framer as stats make him out to be. Now, I’m not saying he is good, but just not as bad. His most starts at C was 103 in 2008; in that year, the Pirates walked the 2nd most batters, and have always been an average to below average team in walking hitters.

    Vote -1 Vote +1

    • kiss my GO NATS says:

      Ummmmmm in case you missed it.. Bad pitch framing = more walks. Really bad pitch framing means many more walks. Perhaps the most walks in the league. You may be blaming pitchers for the catchers ineptitude.

      Vote -1 Vote +1

      • Ian says:

        Well, calls on the edge are on both the pitchers and catchers. Umps aren’t likely to give a call to someone that’s all over the place while they are more likely to give the benefit of the doubt to a guy who never walks someone, like Brad Radke. It is an interesting idea and would be kind of interesting to see how that might or might not play into it.

        Vote -1 Vote +1

      • Ryan says:

        In my comment I said that I still think Doumit is a bad pitch framer. However, even if you said that 50 walks were because of his pitch framing, they still were a bottom 5 in walks.

        Vote -1 Vote +1

  11. Danny says:

    this was really brilliant. also goes right to the core of why some teams are still old school, even decades after bill james.

    Vote -1 Vote +1

  12. DrEasy says:

    How would the walk rate of his pitchers compare with when he didn’t catch them? That would be one tangible (maybe better than CRA but still not great?) way to assess how he hurt his team due to bad framing.

    Vote -1 Vote +1

  13. Chad says:

    It might be interesting to look at how the other catchers on his teams did in regard to pitch-framing.

    Vote -1 Vote +1

  14. Stathead says:

    If you look at fangraphs war, it doesn’t actually do a great job of predicting team wins, which is kind of why I’m down on it (that and the error bars… oh god the error bars)

    Baseball-reference war does a great job of predicting team wins but it incorporates things like “clutch hitting” that don’t do a great job of predicting future performance.

    Vote -1 Vote +1

    • Baltar says:

      I’d love to see the stats on which you base this judgement.

      Vote -1 Vote +1

      • Stathead says:

        You can do it yourself. Replacement level is 47, you have the war for pitcher and for hitters. Add those numbers together and compare it to the team records. It’s not even close.

        Vote -1 Vote +1

  15. rotowizard says:

    As a Pirates fan that’s watch nearly every game he played in a black and white uniform, he is unequivocally the worst receiver I’ve ever seen. Pirates pitching was BAD when he was in town, but my god, he couldn’t get a strike on the corner called a strike to save his life. Or in this case, a base runner. That being said, he was a switch hitting catcher that could really HIT. In 2008 he had one of the best offensive seasons any Pirate catcher has had in the last 40 years. Only Manny Sanguillen (134) and Jason Kendall (136,140) have had a better wRC+ (127).

    Vote -1 Vote +1

    • rotowizard says:

      I also learned two things. 1) Jason Kendall was a lot better than I remembered and 2) Other than these few players, Pirates catchers have been really really bad.

      Vote -1 Vote +1

  16. Jason Kendall balled hard for a few years, until he Lattimore’d his ankle on first base.

    Vote -1 Vote +1

    • matt w says:

      It was really his thumb injury, and insisting on playing through his thumb injury, and having idiot Pirates management let him play through his thumb injury, that did Jason Kendall in. He was still great in 2000 after the ankle injury, but 2001-2 is when his value dropped.

      (Though then he was good again in 03-04. There’s a lot of reflexive bashing of the Pirates’ Kendall extension, but they pretty much got their money’s worth.)

      Vote -1 Vote +1

      • You’re right – I looked it up and he did turn in some good seasons after the injury. I misremembered after all these years. And I was like 15 then.

        I never really liked Kendall while he was here and it wasn’t until recently where I was like, hey, that guy was actually pretty good. Same thing to a lesser degree with Spanky Lavalliere.

        Vote -1 Vote +1

  17. odditie says:

    I love most everything Jeff writes, but I have an issue with the use of “sloppy” stuff.

    We know that Pujols is a great hitter, we know he has accumulated a lot of RBIs. There is no need to use a “sloppy” stat like RBIs to make your point because then you are applying something to explain Pujols that you wouldn’t use to explain R. Howard.

    I’d rather use objective views rather than trying to use bad information.

    Vote -1 Vote +1

  18. Pirates Hurdles says:

    So it took until 2013 for some to realize he was a terrible catcher hurting his team? Congratulations on getting up to speed with any Pirate fan who watched the games for years. This is a perfect case illustrating why scouting cannot be dismissed (not that I’m saying you are doing that Jeff, but some folks do, particularly the heavy fantasy folks).

    Vote -1 Vote +1

  19. matt w says:

    We do have to be cautious about the pitch-framing stuff, though; we don’t know what contributes to it. Could be the catcher, could be the staffs, could be other stuff.

    If the Pirates rank worst in pitch-framing for the third year in a row with their third different catcher, that might suggest it’s not the catcher.

    Vote -1 Vote +1

  20. Hurtlockertwo says:

    Is there a stat that shows how many games are actually lost because of bad fielding?? Is that what UZR shows?? Just asking. Bill James fielding bible lists runs saved/lost as an indicator of good/bad defense, but not the affect on wins/loses.

    Vote -1 Vote +1

  21. byron says:

    Didn’t Fangraphs already incorporate the pitch framing results into WAR? So isn’t Doumit’s 11.1 career WAR already reduced by his bad framing? And wasn’t this article completely pointless as a result?

    Vote -1 Vote +1

  22. Jeff Mathis says:

    This is the article I’ve been waiting for!!

    Vote -1 Vote +1

  23. Baltar says:

    That’s two great posts by Jeff that I’ve read just this afternoon.

    Vote -1 Vote +1

  24. phoenix2042 says:

    Is pitch framing included in WAR on fangraphs for catchers? The article said that Doumit has 11 career WAR, but his pitch framing cost 59 runs (about 6 wins). So does the 11 WAR listed on his player page include the 6 he cost his team on pitch framing, or is that 11 WAR before subtracting for framing?

    Vote -1 Vote +1

  25. Nadingo says:

    Nice take, Jeff!

    It reminds me of the great NL MVP debate of 2005. I remember wen statheads were united in protest when it looked like Andruw Jones might get the MVP on the strength of his 51 HRs and 128 RBIs when it was painfully obvious to anyone who really looked at the numbers that Pujols’ season, with its .430 OBP and 167 wRC+, was far more valuable. Of course, Pujols won the award, so a disaster was averted, right? You’d think so, until you check the 2005 leaderboards and find that, once Fangraphs incorporated UZR into the WAR calculation, Jones ended up edging out Pujols 8.3 to 8.2!

    Definitely my favorite example of bad numbers getting trumped by better numbers but then getting untrumped by even better numbers.

    Vote -1 Vote +1

  26. monkey business says:

    Here I thought the Twins 2012 staff wasn’t good and that’s why they have up lots of runs. Turns out, they just need a new catcher and everything will return to league average!

    Vote -1 Vote +1

  27. I think it should be mentioned here that the Fielding Bible looked into the catcher’s effects on pitching and their initial methodology (I think it was in Fielding Bible II), also found that Jose Molina has been a very good catcher for pitchers, corroborating the finding above about Jose being a good framer.

    That covered, I think, 2003-2008, so no Doumit on that list, unfortunately.

    This is the reason I hate all the comments taking WAR as the final word on this, end of story, when defense is clearly a huge unknown right now in the valuation of players, and particularly for catchers, where we are probably still in the early 1900’s relatively to offensive analysis.

    Vote -1 Vote +1

  28. MN_Oldgoat says:

    Fun to see these numbers change, then change again.
    Utilization of OBP, even if that’s not what he called it, was one of the things that made Weaver such an effective manager.
    I have referred to that “I was Framed” article several times since I first read it. Excellent article. Is it real? Seems to be. Does it make sense? As a former catcher, I think it does.
    It will be interesting to see if this sort of analysis continues, if it gathers momentum, and if it eventually standardizes how catchers are taught to catch in the future. I believe that some catchers have been taught this by other catchers for generations. Never thought to analyze it, but as a coach it is one of the first things I’ve tried to stress with young catchers.
    As a catcher, you try to influence the umpire. It’s just part of the game.

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

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