Evaluating catcher defense is extremely difficult do to the nuances of the position. While statistics such as UZR and DRS are far from perfect, they offer significant insight into the quality of defenders at seven or eight positions. Catching is another animal entirely. While fielding batted balls and throwing to bases is part of what catchers do, most of their influence is more subtle. Rather than simply looking at their ability to prevent runs on batted balls, we have to divide up catcher defense into several parts to grasp it.
Stolen Base Prevention:
The idea of preventing stolen bases is the easiest to consider, but it remains difficult to do so in practice with the data available to us. What we want to be able to do is measure how a given catcher compares to an average catcher in terms of allowing stolen bases, while controlling for the parts of base stealing he cannot control. For example, studies have shown that the pitcher and runner play important roles in determining the probability a steal attempt occurs and how often that attempt will be successful. This means we cannot simply judge a catcher base on their number of steals allowed or their caught stealing rate.
There are two major public attempts to address this issues. First, available on FanGraphs and elsewhere, you can find rSB, or the DRS-based Stolen Base Runs Above Average. This stat measures, in runs, how a catcher compares to the average catcher in terms of preventing stolen bases, accounting for the fact that the best way to prevent a steal is to prevent the runner from attempting it.
Baseball Prospectus has recently debuted Swipe Rate Above Average (SRAA) and Takeoff Rate Above Average (TRAA) which communicate the same idea in percentage form. Their stats offer an indication, controlling for other factors, of how much more often a given catcher allows a stolen base or allows a runner to takeoff in the first place.
While this seems like it should be the easiest component to measure, we still remain uncertain about measuring the catcher’s role in SB prevention. rSB and SRAA/TRAA are good metrics, but they aren’t able to unpack everything. Statcast data which tracks pitcher time to the plate, catcher pop time, and base runner times will be immensely helpful in determining which catchers have saved runs and which catchers have just been around when their pitcher saved runs.
In addition to rSB, full catcher DRS tracks things like bunt runs saved which measure the role the catcher plays in getting outs in the traditional sense. This is a small percentage of what the catcher does, but future improvements such as their ability to execute tag plays at the plate would be useful. You can find this stat at FanGraphs or on Baseball-Reference.
Another obvious component of catching is preventing runners from advancing on wild pitches or passed balls. While WP and PB sound like different stats, they are essentially the same in their construction and effect. A pitch gets by the catcher and a run advances. Who is to blame matters, but the idea that it is binary and correctly evaluated by an official scorer is an outdated concept.
What you want to be able to measure is how well the catcher prevents WP/PB given the pitch they are thrown. The catcher can’t do anything about the fact that the pitcher threw a 57 footer, but they do control what happens next. The ability to block the ball and prevent runner advancement is in their purview and we want stats which measure exactly that. While they are not perfect, two stats provide that information.
At FanGraphs we host RPP, or Runs on Passed Pitches. This statistic makes use of PITCHf/x data and communicates how many runs a catcher has saved based on the difficulty of receiving the pitch. Baseball Prospectus has a similar PB-WP run saved statistic which attempts to answer the same question.
These stats are relatively strong compared to other catching statistics, but they still suffer from measurement issues. While you can estimate the way in which you think a ball is going to bounce, PITCHf/x doesn’t give you enough data to capture the true path of the ball. Statcast data may allow us to refine these stats.
While we’ve known about pitch framing as a concept for a very long time, we’ve only been able to measure framing in the PITCHf/x era. Gathering strikes and avoiding balls is the hallmark of a good pitcher and we know that umpires do not call uniform strike zones. The catcher, based on the way he receives the ball, can influence the call. Good catchers make sure strikes are called strikes and gets the umpire to call a few balls as strikes too.
There are many different ways to measure framing but the most popular numbers are posted at StatCorner and Baseball Prospectus. BP’s model is a bit more involved, controlling for more factors, but the general idea is that each pitch has some probability of being called a strike and catcher’s who get more strikes than expected are valuable.
Framing stats measure how many runs catchers save based on how many extra strikes they are able to get for their pitcher. The statistics do a nice job communicating how many extra strikes are called while a catcher is in the game, but they are potentially problematic when it comes to giving individual credit to the catcher. It is challenging to isolate the effect of the pitcher, umpire, batter, and catcher so that we can properly value the catcher. BP’s stats attempt to do this, but they use a modeling strategy that doesn’t totally capture the underlying process at work. For example, there is no way to distinguish whether the catcher received the ball well or if the umpire is calling an unusually large strike zone during that specific game.
Framing pitches is a very important part of catching, and we have an idea that the best catchers are worth 15 to 25 runs above average per season, but the numbers you see are definitely estimates.
Finally, game management (or game calling) remains the black box of catcher defense. No one has cracked this code. Catchers play a huge role in determining which pitches to throw and how a pitcher navigates a given lineup. Honestly, there is no public research that provides much insight into game calling. By all accounts, it should matter, we just don’t have any idea how to measure it.
There are certainly analysts working on this aspect of catcher defense, but right now we’re dealing with it theoretically rather than empirically.
Things to Remember:
● While FanGraphs does not account for pitch framing, Mike Fast has made some remarkable strides measuring a catcher’s pitch framing ability. If you’re interested, be sure to check out his research. Baseball Prospectus and Stat Corner have also published framing numbers. Our WAR does not include framing, but we intend to incorporate it in the future.
● There is no UZR for catchers.
● FanGraphs catcher WAR includes rSB and RPP.
Links for Further Reading: