Age and EWSL, 2004-06

This is Part III of my look back at at how Established Win Shares Levels fared in 2006. It’s time to look at the age adjustments. (I’ve looked at these previously here, here and here).
The great thing about doing something like EWSL as an ongoing project is that the data becomes progressively more stable over time: I now have three years of results to work from in evaluating how players tend to perform at each age relative to their adjusted Established Win Shares Levels, and thus can have progressively more confidence in the age adjustments I use going forward. For example, the more years of data I have, the less influenced it will be by a single generation of exceptional players born in a particular year.
Let’s start with the 3-year results for the non-pitchers:
Non-Pitchers 2004-06:

Age # WS EWSL %
21- 4 33 34 0.878
22 15 262 116.200 2.255
23 24 355 247.333 1.435
24 43 490 387.536 1.264
25 64 708 564.900 1.253
26 93 1065 934.770 1.139
27 89 1028 963.963 1.066
28 114 1416 1349.902 1.049
29 103 1221 1340.370 0.911
30 112 1329 1444.836 0.920
31 84 858 1014.036 0.846
32 89 902 1067.233 0.845
33 69 708 818.103 0.865
34 67 709 773.003 0.917
35 60 411 593.170 0.693
36 41 379 505.333 0.750
37 30 239 390.496 0.612
38 22 201 273.830 0.734
39 16 175 222.500 0.787
40+ 16 120 222.830 0.539

As you can see, the rapid rise of young players and their gradual fall from age 29 on is a powerful pattern, and one that grows smoother with each year’s additional data. 2006 was a good year for 27-year-olds and a bad year for 28-year-olds, so some equilibrium has been restored in that regard from the prior age adjustments showing 27-year-olds flatlining but then hopping up one last time at 28. After age 32, the number of players holding jobs really starts to drop off.
The train wreck at age 35 only grew more pronounced this season. On the other hand, additional data helped bouy up the 40+ year olds, whose numbers got devastated by Barry Bonds’ 2005. Here’s this year’s data on its own:
Non-Pitchers (2006):

Age # WS EWSL %
21- 0 0 0
22 4 58 26.20 2.214
23 11 198 134.90 1.468
24 15 155 125.97 1.230
25 18 160 127.10 1.259
26 30 401 358.67 1.118
27 21 282 219.13 1.287
28 28 289 318.67 0.907
29 29 364 396.57 0.918
30 45 550 636.47 0.864
31 25 270 266.67 1.013
32 32 348 439.40 0.792
33 16 140 163.67 0.855
34 26 323 357.17 0.904
35 19 133 209.17 0.636
36 15 102 152.00 0.671
37 6 36 89.33 0.403
38 8 69 110.83 0.623
39 6 48 65.50 0.733
40+ 7 60 84.83 0.707

As I’ve explained before, the nature of any established performance level will exaggerate the upward and downward trajectory of player aging, since a 25-year-old is still being partly compared to his 22-year-old self, while a 35-year-old is still being partly compared to his 32-year-old self.
Now, the pitchers:
Pitchers (2004-06):

Age # WS EWSL %
21- 7 42 30.000 1.400
22 17 128 105.000 1.219
23 23 177 172.230 1.028
24 40 266 256.870 1.036
25 57 442 376.800 1.173
26 76 474 463.800 1.022
27 87 592 617.560 0.959
28 92 617 596.800 1.034
29 79 499 576.536 0.866
30 69 423 498.400 0.849
31 63 416 507.170 0.820
32 55 292 425.630 0.686
33 49 275 383.300 0.717
34 39 205 272.966 0.751
35 29 118 194.336 0.607
36 23 123 147.670 0.833
37 20 147 153.663 0.957
38 22 161 204.996 0.785
39 19 141 162.333 0.869
40+ 26 232 269.336 0.950

2006 was a tough year for the established pitchers, at least the under-30 set. The one-year sample sizes get really small – for example, Jon Lieber was the 36-year-old starting pitcher, Steve Trachsel and Paul Byrd the only 35-year-old starters. In general, the rule still holds that the pitchers as a group start to fall off earlier than the hitters. The 2006 data:
Pitchers (2006):

Age # WS EWSL %
21- 2 19 13 1.462
22 6 36 41.60 0.865
23 6 36 33.30 1.081
24 15 96 107.77 0.891
25 23 169 158.17 1.068
26 17 98 93.30 1.050
27 29 230 228.90 1.005
28 33 202 218.40 0.925
29 25 149 174.77 0.853
30 26 144 188.27 0.765
31 20 155 174.17 0.890
32 12 49 71.83 0.682
33 18 95 143.50 0.662
34 14 92 115.80 0.794
35 8 40 50.67 0.789
36 9 46 54.67 0.841
37 3 24 20.83 1.152
38 5 35 41.83 0.837
39 6 61 55.50 1.099
40+ 13 86 115.67 0.744

Overall, as consistent with past data, the age/EWSL numbers are a powerful reminder of the tides of age pulling players down from 29 onward. Which is not surprising: in baseball, as in life, everyone comes up from nothing and goes back to nothing in the end.

2 thoughts on “Age and EWSL, 2004-06”

  1. I’ve always believed, but cannot figure out how to prove, just how much the decline at 29 or so is as much years of service as it is years of age. The study you show Crank is what is expected in a way. You start at 21, with few players, and of course, it grows. Unlike basketball, a 23 year old baseball player is considered very young. So setting a bottom limit of 21 is about where you can go. And by 32, the bottom drops out. So my question is:
    Is it just being 32, or is about 10-11 years what you can expect your body to produce? I think of Junior Griffey as someone who comes quickly to mind. I wonder about ARod, and haven’t taken it much further. But in the end, was Griffey just injury prone, or was he reaching a limit that he got to younger because he started younger. I guess you could look at how many 18 year olds started. They would have to be great or they wouldn’t have jobs (except maybe someone like Joe Nuxhall and Ed Kranepool), but guys like Feller, Gooden, Ott, Kaline. Not enough players probably for the data to have any meaning, but just wondering…

  2. re: age-related decline
    this data continues to confirm the pattern of age-related decline that Bill James first uncovered in his published work (or which he was the first to widely publish since others also were working on the topic) in professional baseball which shows that a baseball player reaches peak performance usually at age 28 and thereafter will suffer age-related decline, and sharp decline after age 35. The decline will be sharper for non-superstars than for superstars.
    The expression of the decline in terms of win shares and expected win share levels completes to a certain extent the theory as it was first formulated by James and others.
    These are important numbers. For example, what should be expected of a rookied like Ryan Howard, who came late to the bigs at age 25, after his second year of service? Obviously, that he will develop for three more years, and then start to decline. But since his arbitration and free agent years are three years away, it suggests that it would be imprudent to sign him to a long year deal if his skills are to decline. on the other hand, they are the declining skills of a great player.
    These are the kinds of questions that this type of analysis can help solve.
    Will Chase Utley decline sharply in the back end of the long deal the phils signed with him? Well, probably–though they are out of it before he hits 35.
    these and other burning questions for the hot stove league.
    –arthur j kyriazis, philly

Comments are closed.