When They Were Young

Generally speaking, aging curves of players have shown that up until 28 years old someone can be expected to improve.  Once past that benchmark a skillset will begin to decline.  For some it will be more gradual than others but this is what those in the analytical community have come to expect.  On Thursday we looked at the only ten pitchers from 1956 until now that have been primarily starters from the age of 38-44, finding that Nolan Ryan has the highest average game score while Roger Clemens’ 134 ERA+ slightly outdoes the 131 of Randy Johnson for the lead.  Also of note is how all ten pitchers posted an above average ERA+, with David Wells and Tommy John’s 101 as the minimum.
I mentioned how there would be an inherent bias in that only the pitchers effective enough on the mound would be able to hang around as a starter this late into their years so it should almost be expected to find quality numbers.  We then looked at all pitchers age 45+ and saw the group whittle down to five, with Jamie Moyer emerging as the best of the bunch.  This catalyzed a new train of thought: how do the age 38+ seasons for these ten pitchers compare with their first seven or so seasons?  Were they better via these metrics as old-timers?
The goal involved comparing the first seven years to the age 38-44 years, however some players started their careers as relievers so, as long as they became a starter within a reasonable amount of time (2-3 years) I went ahead and recorded their statistics during their first seven years as a starter.  The only pitcher this could not be done for was Charlie Hough, who was a reliever for a significant portion of his career’s beginning before becoming a starter a bit later on.  I also could not use the Average Game Score as a measure because Warren Spahn’s career began prior to 1956; before that point the Retrosheet game logs are fewer and further between in terms of existence or proofing (some may exist but are not 100% sure to be correct).
Here are the comparisons, with A representing their early years and B representing elderly prformance, all in terms of their average yearly performance:

  • Nolan Ryan A: 207 IP, 1.82 K/BB, 113 ERA+
  • Nolan Ryan B: 208 IP, 2.75 K/BB, 114 ERA+
  • Randy Johnson A: 205 IP, 2.14 K/BB, 121 ERA+
  • Randy Johnson B: 158 IP, 4.59 K/BB, 131 ERA+
  • Roger Clemens A: 216 IP, 3.10 K/BB, 125 ERA+
  • Roger Clemens B: 179 IP, 2.96 K/BB, 134 ERA+
  • Warren Spahn A: 267 IP, 1.46 K/BB, 107 ERA+
  • Warren Spahn B: 246 IP, 1.90 K/BB, 105 ERA+
  • Phil Niekro A: 267 IP, 2.48 K/BB, 125 ERA+
  • Phil Niekro B: 265 IP, 1.76 K/BB, 114 ERA+
  • Charlie Hough A: 106 IP, 1.30 K/BB, 106 ERA+
  • Charlie Hough B: 221 IP, 1.31 K/BB, 107 ERA+
  • Gaylord Perry A: 281 IP, 2.79 K/BB, 122 ERA+
  • Gaylord Perry B: 213 IP, 2.27 K/BB, 106 ERA+
  • Jamie Moyer A: 122 IP, 1.64 K/BB, 93 ERA+
  • Jamie Moyer B: 210 IP, 2.20 K/BB, 107 ERA+
  • David Wells A: 176 IP, 2.52 K/BB, 109 ERA+
  • David Wells B: 162 IP, 3.45 K/BB, 101 ERA+
  • Tommy John A: 194 IP, 1.88 K/BB, 118 ERA+
  • Tommy John B: 160 IP, 1.27 K/BB, 101 ERA+

Looking at the results for each of these ten pitchers we see that nine of them tended to post better numbers in their younger years.  Sure, there are instances of these nine posting better K/BB ratios as elder statesmen, but their durability and ERA+ would decrease.  Nolan Ryan’s ERA+ and IP stayed the same but his K/BB shot up when older.  Randy Johnson actually had a better K/BB and ERA+ while older despite many more IP as a youngster.  Roger Clemens had better IP when young, a better ERA+ when old, and the same K/BB essentially.
Warren Spahn had better IP when young, a better K/BB when old, and essentially the same ERA+.  Phil Niekro was better at everything when young.  Charlie Hough is tough because his early years were as a reliever, but, still, he posted virtually idential K/BB and ERA+ numbers.  Gaylord Perry joined Niekro as posting better counts in all three categories when young.  David Wells had better IP and ERA+ numbers when young but a much better K/BB later on.  And Tommy John posted better numbers of these three metrics when young.
That leaves just Jamie Moyer, who was the only one here to post better numbers in all three categories when older.  He was below average as a youngster and increased that to 7% above average as an old-timer.  Not only is Jamie Moyer potentially the best 45+ pitcher in the Retrosheet era (1956+) but he is the only one of these ten to legitimately improve in all three of these areas (durability, strikeouts to walks, ERA+) from his first seven seasons to his age 38-44 seasons.


2 Responses to When They Were Young

  1. Dave Evans says:

    I just did a paired t test on the data and although the data slightly suggests that on average this population gets a slightly lower ERA+, the p-value of .596 gives us no room to make a conclusion. So so far there is no evidence that this populations skills increase or decrease between the two stages of their careers. Like you said, the population is tricky because it has some inherent bias, but it may be interesting to see if there are any common factors from these ptichers careers, like pitch types, GB/FB, etc.
    I have a feeling its pretty hard with the small sample size, but it would be interesting to see if their is a particular skillset in pitchers that ages better, kind of like how people believe that three outcome guys age worse than contact hitters

  2. Ten players is always too small a sample to determine anything tangible, even without running tests of significance. That wasn’t the point here. This was just a look at these old-timers and their rates in first seven years vs. last seven years. With such a small sample we of course could not learn anything tangible about the population as a whole, but I’m not comparing each guy to the population, but rather to himself. It’s similar to the clutch stat at Fangraphs, in the sense that it measures performance of a player against himself.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: