Currently, FanGraphs’s version of catcher WAR uses a sum of RPP (runs saved or added by saving/allowing passed balls) and rSB (runs saved or added by catching/allowing basestealers) plus a positional adjustment to calculate the defensive part. Baseball-Reference uses DRS (which uses RPP and rSB), as they do with all other position players, to calculate the defensive part in their WAR. But neither of these include a key part of catcher defense: pitch framing.
I am of the opinion that pitch framing should absolutely be included in catcher WAR. It has shown to be a repeatable skill, and we have a fairly good way to calculate it. Lots of stuff has been done with it – lots and lots and lots of stuff – and it’s now all the rage in the sabermetric world. (By the way, if you’re going to read any of those articles, I highly recommend the one linked from the word “with” – it’s a great explanation, albeit a little complicated, to how pitch-framing calculations work).
For now, though, neither FanGraphs nor Baseball-Reference uses pitch framing in WAR. Baseball Prospectus doesn’t use it in their WARP, either. There is good reason for this – it’s pretty new, and it could turn out that we’re doing something wrong calculating it. But I tend to think that all this extensive research that is very consistent with itself means that it’s probably pretty accurate. So I’m going to add pitch framing numbers to WAR (fWAR, to be exact), just to see what it’s like. I’ll add the runs each catcher has saved (or cost) his team by framing to their defense and re-do the WAR calculations. I took the catcher framing numbers from Statcorner, since theirs are more easily exportable than Baseball Prospectus’s.
This should be easy. After all, FanGraphs WAR is just batting runs plus baserunning runs plus defensive runs plus positional runs plus league runs plus replacement runs, all divided by the runs per win value, which was about 9.25 for 2013. So I just tacked on the framing runs for each catcher to their total runs and redivided by 9.25 to get their new WAR.
|Name||Runs saved with Framing||fWAR||WAR w/ Framing||Difference|
Right away we see that this sends the WAR of some of the elite framers skyrocketing. Yadier Molina becomes more valuable than Miguel Cabrera due to the 2.1 wins that framing boosts him by, and Jonathan Lucroy becomes more valuable than Clayton Kershaw due to the ridiculous 3.4 wins that framing boosts him by (at least if you’re working with fWAR). Chris Stewart goes from barely above replacement level to a three-win player. Others, of course, see their value plummet. Wilin Rosario goes from average to replacement level by virtue of the 2.1 wins that framing costs him. Ryan Doumit goes from barely above replacement level to… well, really, really terrible.
A few things to bear in mind as you look at these numbers: the Statcorner numbers, which I used, are a little different from the Baseball Prospectus numbers. Obviously, I can’t know which are more accurate, and again, I only used the Statcorner ones because they were easier to export. Generally, though, the two data mostly agree with each other, so it’s not that big of an issue.
We also see the distribution of talent spread out a little bit, as it appears that the better catchers (generally speaking) see their WAR go up more often that it goes down, and the worse catchers usually see their WAR go down instead of up – thus implying that better catchers are also better framers. To verify the fact that the numbers are more spread out now, we can check the standard deviation of the two samples of WAR for these 70 catchers: the standard deviation of the regular fWAR is 1.6, while the standard deviation of the “enhanced” WAR is 2.0. As a parting thought, here is a graph showing fWAR against framing ability (the correlation is slight, with an r^2 of only 0.02 or so, but it is still there):