Pitch Framing: MLB's Biggest Market Inefficiency, or Poorly Evaluated Skill?

Back in the heyday of the Moneyball Oakland Athletics, MLB front offices ignored a very important part of hitting: the ability to draw walks. Billy Beane and the A's took full advantage, acquiring players that were not very good on the surface, but were quite adept at taking the base on balls. In doing so, they managed to get productive players in their lineup for a very low cost. All of this is well known. This was a glaring market inefficiency and has since been covered up. MLB front offices are a lot sharper now, and it is very difficult to find significant market inefficiencies by examining past production. However, there might exist an exception to this rule: pitch framing.

I find it strange that MLB teams often ignore pitch framing, because it seems like a pretty common talking point among evaluators. Modern day catchers definitely put an emphasis on framing. Take Gary Sanchez for example. Gary is one of the least skilled defensive catchers we have ever seen, but he is smart. At the cost of his reputation, he sells out his blocking ability to frame pitches, and it works. He still isn't a productive framer, but he manages to not be a liability behind the plate because of his good approach. This anecdote supports the idea that teams put pressure on their catchers to be really good pitch framers. Teams care about framing, but to what extent?

The pitch framing skill gap is not what it used to be; just a decade ago, the difference between the worst and best pitch framers was around 50 runs. 50 runs is the difference between a league average player and a first ballot hall of famer in his prime. In 2021, the difference was just 31 runs, and that's despite Salvador Perez being a very weird outlier. Most catchers who suck at pitch framing, like Ryan Doumit and Carlos Santana (yes, he played catcher), generally suck at the catcher position in general. There is no chance that Doumit or Santana would have logged so many innings at catcher if they played today. Then you have Salvador Perez, who is by all accounts an incredible defensive catcher. Perez' issue is that umpires either have a personal vendetta against the Royals pitching staff, or they just want Salvy to look bad. Despite this, Perez gets playing time at catcher because of his impressive skills outside of framing. Without these skills, he would have certainly been moved to a DH/1B role.  Anyways, the point is that catchers place more of an emphasis on pitch framing than they used to. The Ryan Doumits of the world are gone, and even just a mediocre framer like Zack Collins is now one of the worst liabilities in the sport. Teams want their catchers to lie to umpires as much as possible, but are they willing to pay money for it?

The first graph shows all free agent signings among catchers since 2019 and their projected Steamer WAR not including their projected framing runs (Steamer didn't project pitch framing before 2019, so I can't move the sample back any further). Perez's massive extension with the Royals is not included, but it would only strengthen my theory.  The second graph shows the same thing except pitch framing is including in the projected WAR. It might be hard to tell from just eyeballing the graph, but there is a stronger relationship between salary and non-framing WAR (Pearson R value of 0.86 for non framing and 0.8 for framing).

To further examine this discrepancy, we should look at how pitch framing correlates with the residual for the second graph. The residual represents the difference between the actual salary and the expected salary given a projected WAR value that factors pitch framing.

As we can see, the more you were "underpaid" by teams, the higher the likelihood that you were a good pitch framer. This is a very good indication that teams are not even considering pitch framing as a genuine skill that can be predicted using statistics. The fact that you are more likely to predict how much a player's market value when you ignore their pitch framing acumen is enough to make this claim. However, there is more information to be considered.

The previous graph showed the residual for a WAR value that factored framing; the one you see on the right is the same thing except it uses a framing-agnostic WAR projection. As we can see, there is a *very* slight upper trend. All else equal, a catcher with good framing projections will tend to get slightly more than he *should. Maybe teams do want to pay for pitch framing, but their evaluation of pitch framing wildly differs from that of the public's.

It's hard to say which reality is true, but there are two possible scenarios here. The first is that teams do not use measurable framing statistics to project value for catchers. That much is pretty self explanatory. The second is that teams do very much care about pitch framing, but the measurement methods they use wildly differ from the public data that we have. This seems a little unlikely; it would make sense that evaluation methods differ, but framing is not particularly difficult to evaluate. We have pitch by pitch data with a strike zone overlay that can very quickly determine whether or not the umpire made the correct call, and with a large sample it should be pretty easy to see which catchers are stealing strikes. However, it is possible that teams control for a lot more than just the results of the pitches, and that creates a large gap between how players are publicly and privately evaluated. It's hard to say what is going on without actually asking the people making these decisions. All we know is that someone is misevaluating catchers.

*how much he should get if you assume that pitch framing is not a real skill with any value.