from Amazon
Tuesday, April 06, 2010
No matter how I slice or dice it, pitchers get worse every year…
Without giving you all the gory details, I have been working on aging curves for pitchers using the delta method corrected for survivor bias (my usual method). No matter what I do, I cannot get pitchers getting any better from year to year. I know this is pretty much what other researchers have gotten, but no one seems to believe it. Conventional wisdom that at least young pitchers get better is so ingrained in our heads. How often do you hear, “He is young. He will only get better?” And then of course you are inundated with examples of many a young pitcher who got better and better with age.
Yes, it is probably true that if you stay healthy, you will get better with age and experience, but I am taking about and looking at all pitchers whether they stay healthy or not. For every Felix Hernandez there is at least one Mark Prior. And whether you can tell whether a pitcher is going to stay reasonably healthy and thus likely to get better is another story altogether. (I think that in general you can’t.)
When I say, “No matter what I do,” I mean this:
All pitchers from around age 21 to 26 are completely flat in runs allowed (against linear weights against). After that, they simply skyrocket in a nice smooth pattern to the tune of around .2 runs per 9 innings per year.
If I break pitchers down into relievers and starters, I get the same thing. If I look at good pitchers (before I start to track them of course), same thing. If I look at pitchers who have already pitched for a while, I do get a longer plateau, but no one after age 28 or 29 gets better. In fact, even those who have pitched well for a while, get sharply worse after age 28 or 29. And those who do pitch well for a while, if they don’t get worse until age 28 or 29, they don’t get better before that - they basically stay flat. The best you can do is not get worse as a pitcher.
I also looked at “experience curves” independent of age, where I simply used years in the majors (any TBF per year) as a stand-in for age. Those curves were horrible. Pitchers (a mixture of ages of course) got decidedly worse every year in the majors.
Here are the basic shapes of the aging curves for all pitchers, using the delta method, corrected for survivor bias:
As I said, linear weights against is flat until age 26 (maybe a slight curve upward from 21-23 - it is hard to tell as the sample sizes are small and of course you have large selective sampling for young players that get a second year in MLB) and then a sharp, almost linear curve downward to the tune of around .2 runs a year.
K’s go down a little from 21-23 and then gradually go up until age 29 or so. After that, they drop precipitously - around .2 per game per year.
Walks go down from age 21 to 29, around .08 per game per year. After that, they go up .035 per game per year.
HR per BIP go up from the get go, .055 per game per year.
BABIP goes down slightly from 21-24 and then goes up after that, .005 per year.
I’ll put up a $100 bounty (for charity) for anyone who can find that pitchers get better with age or experience, using some reasonable subset of pitchers, not excluding health or role-change issues (e.g. it may be that pitchers who remain as starters age well), and of course it must be measured after the fact (i.e. no JC Bradbury-like) looking backward at pitchers who have already had long, illustrious careers). It is especially true for pitchers who have had long careers that they aged well.