Miguel Cabrera and Maybe Deserving RBIs

Jeff Sullivan of Fangraphs just wrote a very interesting post about the way Miguel Cabrera changes his approach with runners on base, and in particular the way he changed the way he changes his approach with runners on base in 2013. The basic observations are that Cabrera in general swings more with runners on; that historically this increased aggressiveness has applied both to pitches in the strike zone and to pitches out of the zone; and that in 2013, this stopped being the case, that is, he was more aggressive in swinging at strikes with runners on, but not in swinging at balls. That is to say, in 2013 he managed to combine his more aggressive approach with runners on base with an increased feel for the strike zone and plate discipline. He also seemed to increase his swing rate on pitches on the interior part of the plate in particular, which he is legendary for punishing like a fiend out of hell.

Not said in the article is one of the things I think is most interesting about this observation. Because what we have here is an observation that in what you might call RBI opportunities, Cabrera actually does change his approach, and in some ways that might make him somewhat more likely to drive runners in. Moreover, in 2013, he changed that change in his approach for the better, and indeed absolutely crushed the ball with runners on. He also drove in 137 runs, and until he got injured late in the season he was contending to drive in a run per game. Now, the official position of the sabermetric community is that RBIs are an essentially worthless metric of individual value, that they vary based almost entirely on the strength of the hitters in front of you, i.e. not based on how good you are. Of course a better player will have more RBIs, but among two players who have similar overall batting lines a difference in RBI totals will be a reflection of team strength, not player quality. "Clutch" hitting is more or less a statistical illusion. That's the thinking.

But is this true in a case like Cabrera? Swing rates are among the least noisy data in baseball; the sample sizes are huge compared to things like at bats or whatever, and there are just a lot less confounding factors like defense or whatever. And we see very clearly that Cabrera has a very different approach with runners on base. He is also very successful with runners on base, and was particularly so with runners on base. Result, he had 137 RBIs in 652 plate appearances. Baseball-References says that an average MLB player would've had 67 RBIs over a similar span. Now, certainly, Cabrera had more runners on than an average player would've, but not that much more. He had a little under 25% more runners on third, and a smaller figure for runners on first and second. Being a bit overly generous, then, we might attribute a 25% increase over an average figure to his team. That would get him to 84 RBIs. He got to 84 RBIs in his 82nd game of the season, after his 378rd plate appearance. Yeah.

Now, y'know, the remaining 53 RBIs above average expectation are probably mostly just attributable to the fact that he hit .348/.442/.636 last year, which is a lot better than average. His wRC+ was 192. (Note that that doesn't actually mean he was nearly twice as good as an average player, since a slash line of .000/.000/.000 isn't a wRC+ of 0, it's -100 (just ask Justin Verlander); a 192 wRC+ means he generated 46% more value than an average player.) But he hit .366/.483/.733 with runners on base, for a wRC+ of 222, and .397/.529/.782 with runners in scoring position. wRC+: 243. His BAbip didn't even increase that much in those situations, going from .358 with the bases empty to .352 with runners on and .386 with runners in scoring position. Everything we can observe about his performance last year, very much including the data on his approach at the plate, suggests that Miguel Cabrera significantly elevated his game in RBI situations last year. He was great with the bases empty; he was Bonds with runners on. (Okay, Bonds in a down year, but within Barry Bonds range, which is plenty.)

And maybe that means he should get some credit for all those runs he drove in. It's really easy to say that, almost always, the RBI number doesn't really give us any useful information. That's one of the first things sabermetrics ever said, and it remains one of the most important. But the field has gotten to the point where it does and should say much more subtle things, and one of those things might be that, for this one unbelievably-good-at-hitting-pitched-baseballs player, the general rule might not hold. And maybe, therefore, we're okay with his MVP award.*


*Although I would've been more okay with it if he hadn't essentially missed the last month or so of the season, at least in terms of production. Before he got hurt he was at least keeping up with Trout by context-neutral stuff like WAR.