This is one of those posts where a question got into my head, and I’m going to write through the process of answering it, whatever the outcome. I figure if I was curious enough to dig into it, maybe a few of you would be curious, too. And it wound up being a nerdy post about two pitchers who might not be offended if they were, themselves, given that label. So it fit.
Basically, when I saw that Kyle Hendricks and Alec Mills were two of the three finalists for the NL Pitcher Gold Glove, I then observed that their 3 Defensive Runs Saved each were the highest in baseball. That, in turn got me wondering something about pitchers who frequently “beat” their FIP: to what extent does a pitcher’s own defense help contribute to the lowering of his ERA, relative to his FIP?
In other words, when we look at a guy who has a better ERA than his FIP (Fielding Independent Pitching), might we consider “crediting” that pitcher with his own defense, rather than totally stripping it out in FIP? I get that FIP is attempting to describe for us how the pitcher pitched regardless of anything that happened on defense, but it kinda feels like that pitcher, as a defender, is more like an exception. Yeah, we would want to know if Joe Pitcherface had a superstar stud shortstop and center fielder behind him, which helped lower his ERA relative to his FIP, in part because we want to know how Joe Pitcherface might perform in front of a different set of defenders … but what if Joe Pitcherface is himself a superstar stud defender? That’s one defender that Joe Pitcherface will always take with him, no matter what team he joins.
All right. So I have no idea how this actually goes, but I did know that each of Hendricks (2.88/3.55) and Mills (4.48/5.44) had ERAs far better than their FIPs this year. Most of that we would attribute to (1) good quality of contact, (2) good defense, and (3) luck. But maybe some chunk of it is actually just those guys being studs on defense this year?
To get at that question, I’m just playing things like this: if each guy “saved” three runs with his defense this year, what happens to the FIP calculations if we chop of three runs? Welllllll that’s kinda tricky, since FIP doesn’t actually use runs in its formula: (13*HR + 3*BB – 2*K)/IP + C, where C is a constant to adjust the FIP number to make it look like, and track to, ERA by adjusting the league-average FIP to match the league-average ERA.
Ah ha! Since FIP tracks to ERA, we can just do it this way: if Hendricks’ FIP implied 3.55 “runs” per 9 innings over 81.1 innings, then he “allowed” about 32.1 “FIP runs” this year. And if Mills’ FIP implied 5.44 “runs” per 9 innings over 62.1 innings, then he “allowed” about 37.7 “FIP runs” this year.
Now, we subtract three runs for each guy, and go backwards to calculate new FIPs, this time accounting for the defense each pitcher provided himself.
That leaves Kyle Hendricks with a new “FIP” of 3.23, and Alec Mills with a new “FIP” of 5.01. Hey, that made a pretty big difference!
Obviously you still have a pretty good chunk of the distance between ERA and FIP that must be explained by the trio of contact quality, other player defense, and luck, but now we can say with some confidence that each of Hendricks and Mills earned at least some part of their ERA by virtue of their own excellent defense! Heck, for Mills, it was nearly half the distance!
Another way to look at this. If you were observing a pitcher’s season and you feared that his FIP was going to tell you more about his next season’s results than his ERA, you would look at these shifts in Hendricks and Mills and say, yeah, those FIPs are a lot higher than their ERAs, so maybe I wouldn’t expect them to duplicate the ERAs … but I also probably shouldn’t expect those FIPs, since they bring their own defense with them.
Of course, small sample advanced defensive metrics are rough, so you’re gonna have to caveat for that. But the point here wasn’t really to be all that precise about 2021 predictions. Instead, it was just to see if these Gold Glove finalists’ defense probably helps bridge some of that gap between ERA and FIP. And I think it pretty clearly does! Baseball and math and stats are fun!