clock menu more-arrow no yes mobile

Filed under:

Measuring the Diamondbacks defense: Catcher

After a longer than expected hiatus, due to a change of department at work [what are these things called "mornings" - whatever they are, I don't like them], here's the first installment in the D-backs defense series.

Mark J. Rebilas-USA TODAY Sports

You were, presumably, paying attention to the first article, Measuring the 2015 Diamondbacks defense: An introduction. Well, as far as catchers are concerned, you can pretty much forget all that, because measure catcher defense is as much an art as a science, even though there have been strides recently in terms of framing - or, as catchers seem to prefer it to be called, "presentation". Beyond that, you really need to take into account, for example, their talents at preventing stolen bases - but is it better to throw out a baserunner, or have an arm that stops the baserunner from trying to steal?

There's also pitch blocking - stopping wild pitches - game calling and even regular defense, when a catcher has to scurry out to snag a bunt and fire on to first. But the last of these don't happen very often. Sure, the Diamondbacks' catchers got credit for 1.211 putouts last year; but that includes one there for every K, unless they drop the ball. They only had 90 assists, about 70% of the total for the next lowest infield position, the 127 recorded at first base. And how do you even begin to rate game calling? Catcher's ERA is perhaps a start, but is heavily dependent on which pitchers you catch. All told, we don't know more than we do know, as this Fangraphs article makes clear.

Defensive Runs Saved

This measure combines stolen-base suppression and "good fielding plays". And it hates the D-backs, rating us overall as tied for worst in the majors at -7. The only one seen positively, was Tuffy Gosewisch at +3. Gerald Laird came in at zero, as you've expect from his whopping 4.1 innings of work, with Oscar Hernandez at -1, also in very limited playing time. Jordan Pacheco and Jarrod Saltalamacchia were at -2, with Welington Castillo at -5, though in his defense (hohoho), he worked close to twice as many innings as any of the others. This does, generally, seem to pass the eye test, though in terms of defensive butchery, I would likely rank Castillo better than Salty.

Stolen bases

Here are the stats for our five catchers, along with their amount of work.

Name                Inn SB CS CS%
Welington Castillo     626.1 29 11 28%
Jarrod Saltalamacchia  317.0 15  3 17%
Tuffy Gosewisch        311.1 19 11 37%
Jordan Pacheco         130.1  9  2 18%
Oscar Hernandez         77.1  1  1 50%
Gerald Laird             4.1  0      0
Team Total            1466.2 73 28 28%

There are some wide variations here. Strange how virtually nobody bothered to steal at all against our Invisible Catcher, Hernandez, with the Rule 5 pick conceding one stolen-base in 77.1 innings, when the team average was one every 20 innings. Take a look at Salty and Tuffy, who played almost the same amount. Does Jarrod deserve credit for 40% fewer runners trying to steal off him than against Tuffy? Or should he be penalized for throwing them out only half as often? Castillo shows up quite well here, with one stolen-base per 22 innings, and a throw-out rate exactly at the MLB average.

Pitch framing

Pitch f/X lets us know a specific pitch's location, but whether it's called a ball or a strike is not just due to how the catcher receives it. Home-plate umpires have different zones, obviously, but other factors that have been shown to have an impact include the batter, the count, the pitch type and even the ballpark. Filtering these out so that you only see the catcher's contribution is, clearly, no small feat. But unsung heroes like Mike Fast and Max Marchi have done it, and over at Beyond the Box Score, Dan Turkenkopf concluded that turning just a single ball into a strike was worth 0.133 runs.

That may seem a lot, but take a look at the situation where the first pitch in an at-bat is a ball. In 2015, National League hitters batted .265/.372/.426 after a 1-0 count. But if the catcher turns that ball into a strike, those figures change dramatically: after a 0-1 count, they hit .222/.263/.336, almost 200 OPS points less. Take the pitch f/X information, and you can figure out how many pitches a given catcher "should" have had called strikes; compare that to how many he actually did, and you can figure how many runs he saved or cost his team over the course of an entire season.  From StatCorner, here are the numbers for the 2015 catchers. It's not pretty

Name Sample zBall% oStr% +Calls PerGame RAA
Tuffy Gosewisch 2642 14.1 7.8 2 0.07 0.3
Gerald Laird 41 12.5 3.0 -1 -2.59 -0.2
Jordan Pacheco 1116 15.0 7.1 -8 -0.53 -1.0
Oscar Hernandez 635 13.0 5.7 -9 -1.09 -1.2
Jarrod Saltalamacchia 2614 16.6 6.3 -55 -1.63 -7.3
Welington Castillo 5111 16.3 6.8 -82 -1.25 -10.9

Sample is the number of pitchers. zBall% is the percentage of pitches in the nominal strike zone called a ball and oStr% the reverse; pitches outside which were called strikes. Compare those numbers to league average, and you can figure how many calls the catcher "got" over the season, per gave if you want. Finally, you can multip[y that figure by the .133 runs per call mentioned above and get the number of runs added above average. Again, Tuffy is the only one of our catchers to come out ahead. Both Salty and Beefy were...well, pretty sucky. Only four pitchers in the majors cost their teams more runs with their framing than Castillo.

But it's worth noting these numbers can be pretty volatile. Take a look at the numbers posted by Miguel Montero say. In 2012, he was decent, at +6.6. In 2013, he dropped to -2.3, but in 2014, he was the best in the entire major-leagues at +24.0. But this season, with the Cubs? He was still pretty good, but only +13.3, so ten runs fewer. It's clear there is a fair amount of noise in this particular stat, perhaps due to the various confounding factors still playing a part, which are hard to filter out.


It is, to some extent, a trade-off. Catchers who can hit and play great defense are very, very rare; more often, they tend to be like Peter O'Brien, and it's often a struggle for their catching to reach the level that allows their overall play to be a net positive. We see something similar with the numbers above, which tell us that Gosewisch was the best at fielding his position, and Castillo the worst, even as they put up OPS+'s of 47 and 116 respectively. Oh, to splice their genes together in a laboratory beneath Chase Field. In the absence of such a convenient facility [though maybe that is where the money from the TV deal is going?], we likely have to take the good with the bad.

This is why a platoon of sorts is not necessarily a bad idea. Perhaps we have our offensive catcher (Castillo) as the starter, but in the late innings of a close game, you can switch him out and bring a defensive one (Gosewisch), who can keep the running game in check and perhaps steal a few calls for your late-inning relievers. We see this often enough in the outfield, say; might it be worth doing behind the plate as well? The team seems to want to re-sign Saltalamacchia, but I'm not sure we wouldn't be better off with a better defender, to that end, or just keep our fingers crossed Tuffy will be ready for Opening Day.