Paulino vs. Doumit

I'm annoyed by all this "calling a good game" hooey we keep hearing as justification for playing Paulino, a man who can't hit or play defense and has exhibited no interest in trying to do either, over Doumit, a man who can definitely hit and while just as bad defensively at least puts forth some effort. So, I thought I'd do what I could to see if I can measure any such thing.

I'm not a big believer in ERA, or even RA, as a good measure of a pitcher's ability. Especially the ability of a pitcher who relies strongly on his defense, given how bad the Pirates' defenses have been for as long as I can remember (and how many of our recent pitchers rely on said crappy defense). So I measured effectiveness by the percentage of plate appearances against a pitcher that resulted in a strikeout, a walk, or a home run. I also looked at percentage of runners caught stealing. This helps strip out the effect that big fluctuations in BABIP can have. Also, this is how baseballreference.com presents pitcher splits when broken down by catcher, so it's easier for me to play with the data in this way.

I looked at Pirate starters who have started at least 7 games for the Pirates over 2006-2007. This gave me a list of Duke, Snell, Gorzelanny, Maholm, Chacon, Armas, Santos, Perez, and Wells. Using baseballreference.com's splits, I looked at each pitcher's line with Paulino, Doumit, or Cota behind the plate (hey, we've heard the same damned things about Cota, too).

Now, the simple results (with total plate appearances):

Now, there are a number of problems with this data. The most obvious, to me, is that Paulino's is less prone to being victimized by the fluctuation that a smaller sample size will endure. Also skewing the data is that Paulino does tend to catch the better pitchers more often than Doumit does. Let's compare Snell and Gorzelanny.

It appears that Paulino also has the edge here, but remember again, Doumit's figures are more likely to be skewed (especially the Snell line) because there's much less data to work with.

Let's translate those numbers into a per-nine-innings basis, which are a lot easier to understand (for me, anyways). There are an average of 38.86 plate appearances per nine innings this season. ("Trans. innings" is translated innings.)

At this point, we can attempt to correct for the biases in the pitchers each catcher caught by weighing each pitcher equally. After that, we get:

So what do these statistics tell us (or me, anyway)?

1. There is no clear, meaningful statistical proof here that Paulino is any better a game caller than Doumit is. The only thing I can say with any degree of confidence is that Doumit's starters walk an extra batter about every 18 innings. Using linear weights, a walk is worth about .4 of a run these days. That's 1 run every 45 innings. Using baseballreference.com's RC/27 formula, Doumit at the plate, for his career, would be worth .9 of a run more than Paulino over those 45 innings. That other .1 of a run is more than made up for by a) Paulino's 2006 being a flukey offensive year, skewing his offensive value upwards, b) Doumit at least makes a friggin effort out there. And this doesn't take into account the fact that Doumit's pitchers, in the evenly-weighted sample, strike out more batters than Paulino's.

   EDIT: I had to rethink this after realizing that I had forgotten that baseballreference's RC/27 is how many runs a team full of that player would score. Doumit creates 1 run every 6.00 plate appearances. Paulino creates 1 run every 10.38 plate appearances. Figure 4 plate appearances every 9 innings for each player. Over 9 innings, Doumit creates .67 run versus Paulino's .39. While Doumit's "poor game-calling" is creating 1 more run every 45 innings for Pirate opponents, Doumit's bat is creating 1.4 more runs than Paulino's is.

2. There is a TON of noise in trying to figure out the impact a catcher has on his staff. I haven't looked at relief pitchers, I haven't controlled for day vs. night games (Doumit does a lot of catching Sunday day games). I haven't controlled for home vs. away (if there even is a different, I have no idea). I don't have passed balls or wild pitches included. I haven't controlled for quality of opponent. It may be that Paulino's pitchers, tending to be the better pitchers, face tougher lineups more often. I don't know.
3. EDITED: I am more confident in saying that this so-called "gamecalling" skill (if it exists) is so insignificant that it is utterly buried by the noise from other, far more meaningful influences on the batter-pitcher contest than I am in saying that Doumit's bat is .4 of a run more valuable than Paulino's finger-wiggling over 5 games (or almost 13 runs in a season). Given this much noise, .4 (or 13) are just too small in their respective contexts to really be meaningful.
4. Even with the weighting, Ollie Perez really screws with Cota's numbers.