Article on Aging at Baseball Prospectus

I have an article up on aging at Baseball Prospectus today.

How Do Baseball Players Age? Investigating the Age-27 Theory

Recently, there’s been a decent amount of chatter regarding how baseball players age, and I have to admit that it’s mostly my fault. In a study that was recently published in Journal of Sports Sciences, I find that players tend to peak around the age of 29; this finding has been met with resistance from some individuals in the sabermetric community, where 27 has long been considered the age when players peak. Will Carroll and Christina Kahrl graciously asked if I would be willing to defend my findings on Baseball Prospectus. I agreed, and I thank Will and Christina for the opportunity to do so.

7 Responses “Article on Aging at Baseball Prospectus”

  1. Tom T says:

    I promise one day I’ll get around to reading this study instead of just asking ignorant questions, but did you separate the effect of experience from aging? In other words, maybe the upward slope of the curve comes from experience/learning (not age). The downward slope of the curve could come from age (declining skills). If this were true, predicting a peak might depend less on age than on at-bats. Teams might care more about getting young players some experience. Just wonderin’.

  2. JC says:

    Good question, Tom. I investigated this in the study, sometimes including age and experience in the same equation and looking at them separately, and there seemed to be no measurable experience effect. I suspect there is a small one, but it appeared to be overwhelmed by aging as best I could tell.

  3. Nick Steiner says:

    JC – I have a few questions about the study, and I don’t have an active BPro subscription so I thought I would post them here. Based on reading the BPro article (which a friend sent me), it seems to be that yo have a different definition and use for your study than MGL and others do.

    It appears that you want to try and separate aging from injuries, and only including players who fit a certain PA requirement and age restriction allows you to only observe age (and possibly other factors) related curves rather than age + injury.

    Can you please verify if that is correct before I proceed?

  4. JC says:

    I’m not exactly clear on what you are asking. I’ve e-mailed you the study. The distinction between aging and non-related injuries is relevant to explaining why the mode peak understates the true aging peak. In my estimates, I don’t differentiate as to why players decline.

  5. Nick Steiner says:

    Right, so you are trying to isolate the effects of aging from injuries and other factors.

    Here is my, (and MGL, Tango, et al) problem with your study. By using players who were in baseball until they were at least 35 years old, you are, by definition, not including players who peaked early and were out of baseball before age 35. That will naturally inflate the observed peak age. How can you not see this?

  6. Nick Steiner says:

    Also, thank you for emailing me the study.

  7. JC says:

    1. The aging-attrition distinction explains why using the mode method will bias the peak age downward. That’s it. I don’t use the mode method for this reason. This has nothing to do with my aging estimates as I just explained to you.

    2. I do NOT require players to play until 35. I can’t see how you could make such an erroneous statement given all that I have written on the subject. I only look at player performances between the ages of 24 and 35 so as not to bias the estimates from exceptionally good players who will be getting more playing time at younger and older ages.

    3. How can I not see? Good grief! How many times have I explained why the sample cutoffs are not biasing the peak age? Look at the player skills: strikeouts peak in the early-20s, walks in the early-30s—nine years apart for pitchers. We don’t see any bias there. I also drop the requirement for inclusion into the study to 1000 career PA and the age estimates don’t differ much from the larger sample. I have written two posts devoted to this issue on the issue on this blog.

    You didn’t even bother to read my study or the numerous posts I have written explaining the study; yet, here you are lecturing me. What’s the point of responding to you if you’re not going to even bother to read it? At least Icarus had wings of wax; you’re just flapping your arms. Time is scarce, and I am going to waste no more of it engaging you.