keyboard_arrow_uptop

You’ll notice that Tim Lincecum isn’t very good these days. He’s actually quite bad. It’s hard to watch him sometimes, especially when he’s matched up against a good opponent, knowing he’s become so likely to lose the battle. That’s what happens, I suppose: pitchers get older, they get worse.

Not his pitching, though. I’m talking about his hitting. Lincecum has one measly hit this season, a little groundball single through the hole between shortstop and third base. He has struck out 12 times in 16 official at-bats. Just six players—five pitchers, and Khris Davis, who R.J. brilliantly describes as “Chris Davis with more K”—have a lower contact rate on pitches in the strike zone.

This isn’t really about Tim Lincecum, who has just 16 official at-bats. It’s about the aging curve for pitcher’s hitting. We basically know the aging curves in baseball, and most either go up-up-up-peak-down-down-down (like hitters hitting) or they go down-down-down-down-down-down (like defense) or they’re a bit more complicated but generally resemble one of those (like pitcher’s pitching).

But there’s another possible aging curve that you could imagine but don’t see much, and that one goes up-up-up-up-up-up. For instance, maybe team leadership, if we could ever measure it, would be an up-up-up.

Pitchers hitting used to be an up-up-up. Pitchers used to come into the league terrible at hitting, then get progressively better as they aged, at least as a cohort. In the 10 years from 1954 to 1963, for instance—a period chosen basically at random, except that I hoped to avoid any extreme swings on the league-wide offensive environment —this is how each age tier (tiers also chosen arbitrarily) performed:

Age PA OBP SLG K rate
17-23 10176 0.187 0.187 34%
24-26 18337 0.203 0.203 31%
27-29 18177 0.204 0.205 29%
30-33 15015 0.214 0.215 28%
34+ 7546 0.214 0.218 27%

So you see a trend and your brain fills in reasons for why this would be*, whether the mechanics of it make any sense. In this case, sure, it makes sense. Pitchers aren’t hitters, and they don’t have nearly the exposure to elite pitching that hitters do. They’re overmatched when they arrive, and probably terrified out of their minds, as we all would be and as we all would be. Through years of exposure they get a bit more comfortable. It might help that they also get stronger, just as regular hitters do, though we don’t see the same decline in the 30s, so probably more likely it’s about getting accustomed, about adjusting to the first few hundred major-league fastballs and curveballs.

This makes sense.

There’s another aging curve shape that we haven’t seen, and that one goes like this: straight-straight-straight-straigh-straight-straight. Pitchers hitting used to be a straight-stright-straight skill. In the 1970s and 1980s, the game had changed for pitchers hitting: there was a DH, for instance. Pitchers overall were getting worse with the bat. And the aging curve had changed to no curve at all, just a line. Here’s 1974 to 1983:

Age PA OBP SLG K rate
17-23 6036 0.184 0.181 35%
24-26 13431 0.191 0.187 34%
27-29 10601 0.191 0.195 31%
30-33 7701 0.196 0.194 31%
34+ 5534 0.188 0.199 27%

If I’d started with this, instead of going back to the 1950s, our brains would fill in explanations for why this is the most logical way for pitchers’ hitting to change. After all, pitchers aren’t hitters; they’re already performing at something close to the bare minimum of what an extremely athletic, baseball-experienced person can do. Few individual pitchers do significantly worse than these averages because it’s hard to do significantly worse; inevitably, there will be walks, there will be bleeders and bloops, there will be the occasionally squared-up gapper, there will be sacrifice bunts, and the hundreds of outs that come in between are just the norm.

This makes sense, and yet the previous explanation for how pitchers age as hitters also made sense. So which one is the right one? Or, a better question might be, is either one the right one?

Because, back to Lincecum, who was never a good hitter but who has gotten worse as he has aged. He doesn’t actually signify anything on his own, but he is a convenient stand-in for the larger pool of pitchers hitting. Since 2010, this is the aging curve for pitchers swinging:

Age PA OBP SLG K-Rate
17-23 2666 0.174 0.186 39%
24-26 7670 0.174 0.176 39%
27-29 6018 0.173 0.176 42%
30-33 5018 0.169 0.175 35%
34+ 3514 0.167 0.164 34%

The one thing we see is that pitchers make more contact as they get older, no matter the era. But overall production now goes down-down-down-down-down-down, just like fastball velocity and running speed and metabolism and familiarity with pop culture references. Pitchers’ hitting has become, like most things, a sad slide from birth to death. (This is what Nate Silver found when he looked into the matter in 2008.) How would you explain this, assuming this was something to be explained? Probably that, since pitchers have no actual skill, they’re able to get some production out of their natural athleticism, which goes down as they age. And there’s little motivation for them to work to reverse the decline. There’s no motivation, really, especially since most of them will spend at least some of their careers in the American League. You might also consider them increasingly savvy, realizing that baseballs are exceptionally scary and dangerous, and the older they get the more they realize hitting just isn’t worth it.

Of course, each of these explanations would make a lot more sense if not for the exactly opposite explanations we used to explain the previous eras.

So there are two takeaways: It’s easy to fill out a table with a narrative, whether it’s correct or not; and, if you find one that is correct, it might by then be expired, anyway.

*Survivorship bias might come into play here, though I’m not concerned about it because pitchers are not selected (or rejected) for their offensive skills. Unless there is a correlation between hitting and pitching skill in the majors, or pitcher health and hitting skill in the majors, I wouldn’t expect the older cohorts to be skewed because of hitting success. Notably, if this were the case, it isn’t anymore.