clock menu more-arrow no yes mobile

Filed under:

A Thought: Game Calling

As a pretty good rule of thumb, if a player wants to make it all the way to the Major Leagues, he needs to be in possession of at least one skill. There's no other way around it. Major League roster spots aren't like A's at Harvard; due to their limited nature, they don't just get handed out like candy. A player has to earn his opportunity, and he has to do it by demonstrating that his ability makes him capable of making a contribution.

It is because of Rob Johnson's seeming attempt to flout this rule that we've spent a lot of time talking about catcher defense and, more specifically, a catcher's ability to call a game. Johnson, you'll recall, posted a phenomenally low CERA and allowed hitters to post a .660 aggregate OPS, vs. Kenji Johjima's .777. Being that Johnson can't really hit or catch and doesn't have a super arm, it's been argued that game-calling is his primary skill.

Which, hey, that could be true, in that his other tools are kind of anti-skills. We've gone to some length, though, in examining the history of CERA research, and nobody's ever been able to find a real effect. The best work on the subject has failed to uncover any compelling reason to make use of the statistic, as study after study has shown that, if a game-calling effect exists, its impact is so small that it can't be adequately measured.

And this, I think, is where a lot of us stand when it comes to how we feel about CERA. It's neat, but it's not predictive and it's not useful, and everyone's better off not paying it a lot of attention. It will mislead far more often than it will reveal.

But here's where I think people could stand to be a little more careful. The leap from "game calling can't really be measured" to "game calling isn't a skill" is not a large one, but it's one with a significant implication. Believe in the former and you ignore CERA and go about your business. Believe in the latter and you believe that, when it comes to calling pitches and setting up behind the plate, all catchers are identical.

Which is silly. While no catcher who calls an unthinkably bad game will ever make it to the bigs, the very fact that you can call a bad game - that you and I would call a worse game than Rob Johnson - implies a spread in ability. And though the spread may not be that broad in the Major Leagues, we have to believe that it exists at least to some degree. Because catcher brains aren't identical, because there's thought that goes into every pitch, there will be, among big league catchers, a best game-caller and a worst game-caller.

We would expect better game-callers to post lower CERAs, and we would expect worse game-callers to post higher CERAs. Adjusted for pitchers and opponents, of course. Obviously, if you're calling a better game, that means your pitchers are allowing fewer runs. And here's where it gets interesting. Catchers catch a lot of innings. There's nobody out there who's, say, a full CERA point better than average, but do you think a study could really pick up on a difference of 0.2? What about 0.1? With all the variables and all the adjustments, do you think that, if there were a spread from -0.1 to +0.1, any study would be able to catch it?

Over 900 innings - the ~average of the top 30 catchers in innings caught - a 0.1 CERA effect would be equal to ten runs. So if such a true-talent spread from -0.1 to +0.1 did exist, that would come out to a spread of 20 runs over a full season, or roughly two wins. 

The implication being that, even given a spread that small, we could be talking about the best game-caller being two wins better than the worst game-caller over a full season by true talent, on game-calling alone.

Please don't read this and think the take-home message is something weird in support of Rob Johnson. That's not what I'm saying. We can't believe in CERA, and as such we can't believe in Johnson's game-calling being a major strength. But just because we can't use individual CERA figures doesn't mean we can't entertain the notion that a game-calling skill does exist, and in the event that it does, even a seemingly small spread would mean a significant difference in value between the best game-caller and the worst game-caller in baseball. Whoever those guys may be.

Hopefully PITCHfx is eventually able to shed light on this subject and give us a good answer, once and for all.