Aging Strikeouts: You’ll Never Be This Good Again

When Johnny says his last lines in “The Outsiders” — “Stay gold, Ponyboy. Stay gold.” — there’s more than a slight touch of mortality in the moment. There might even be outright pessimism about the directive. After all, the Robert Frost poem he’s referencing finishes: “Nothing gold can stay.”

Turns out Johnny and Frost know a little something about pitchers and strikeout rates. Thanks to the inestimable Jeff Zimmerman, we have strikeout aging curves for both starters and relievers. As dawn turns to day, it seems, pitchers also lose their gold.

Here is a more detailed explanation of how Zimmerman did his hitter aging curves.
He repeated a similar process for this graph, and the sample goes back to 2002.

What a stark graph. Zero is the peak, so your average pitcher’s strikeout rate pretty much peaks when he debuts, or shortly afterward. Starters might be able to maintain that peak until 25 years old, or so, but they never really improve the rate. Relievers? Let’s just say you better smoke ‘em while you got ‘em (and then let ‘em walk).

That’s probably the biggest take-away from this graph: If a pitcher’s strikeout rate naturally declines from day one, you probably don’t want to build your team on free-agent pitchers. Despite the high attrition rates for pitching prospects, it seems to make sense that you’d develop your own pitching, rather than aquire — by definition — post-peak free-agents. This also might put a little tiny check mark in Yu Darvish‘s corner, considering the fact that he’s 25 — but that’s a story for another time.

This graph is yet another reason to avoid signing free-agent relievers to multi-year deals. You’re not going to like the third year of that Brandon Lyon deal, Houston. And we told you so. It’s not like he was falling off a nice peak.

Remember this effect when evaluating similar young pitchers, too. Do you like Brandon Beachy and Vance Worley? They pitched a similar number of innings last year, are a year apart and had decent control. Neither was a top prospect, so both were pleasant finds for their teams and are under team control for a good chunk of time. But the Braves should be less willing to trade their young starter. Beachy is coming off a 28.6% strikeout rate, which might represent his peak. Worley, on the other hand, debuted with a 21.5% rate. Even if Worley has a marginally better ground-ball rate, his strikeout rate is statistically likely to get worse — and it’s coming off a lower peak.

But, of course, there’s true talent and then there are results. Someone like Mike Minor, in particular, could see a true-talent decline and still have better results in the future. No one is is suggesting that every pitcher in the league last year will show worse overall results going forward. The bouncing ball knows no master.

Still, when it comes to those fascist strikeouts, we know that you’ll probably never be as good as you just were.

Print This Post

Graphs: Baseball, Roto, Beer, brats (OK, no graphs for that...yet), repeat. Follow him on Twitter @enosarris.

32 Responses to “Aging Strikeouts: You’ll Never Be This Good Again”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Cliff says:

    Okay, but there’s no breakout for different types of pitchers. I imagine different styles age differently.

    Vote -1 Vote +1

    • JDanger says:

      I always thought that when Ralph says “stay gold, Ponyboy” that he was talking about his heart. Like, he should “stay gold” in his heart. But I realized the other day that he’s not. He’s talking about his hair. If you go back and watch the movie again you’ll see that Ponyboy has gold hair in that scene. He dyed it to get away from the cops. And ralph was like, thats cool hair.

      Vote -1 Vote +1

      • baycommuter says:

        uh, I assume that’s a joke, it’s clear what’s meant in the book, though the movie is kind of a mess…. When Ponyboy grew up, did he become Horseman?

        Vote -1 Vote +1

  2. CJ says:

    This begs the question “what is the aging curve for BB%?” A declining K rate isn’t so bad if it is accompanied by a declining BB rate. My impression is that young pitchers often arrive in the big leagues with their peak BB rate. My guess is that the pitchers who carve out long career generally improve their control to offset a declining K rate. But I don’t have any hard evidence to support this view. So, I hope that this study is accompanied by a study of the aging rate for BB%.

    +6 Vote -1 Vote +1

  3. Matt says:

    Just throwing it out there, but for young starters like Beachy and Worley, is there a tendency for decreased walk rate as a starter ages, and could that counter a declining K rate, enabling sustained results overall? You did mention Worley/Beachy had decent control, so for them decreased BB rate might be negligible.

    Vote -1 Vote +1

  4. delv says:

    Are these “K-rates for all pitchers, averaged” regardless of length of career? I figure that pitchers are more likely to get a chance to be in the majors at age 20-24 if they have good strikeout rates, and less likely if they’re contact pitchers. On the flipside of things, there’s a traditionalist tendency to accept “pitching to contact” from older pitchers, so maybe non-K guys are more likely to be given ML innings when they’re older, regardless of performance. Basically, I wonder about how the sample here is composed and what it really tells us. The way you’ve presented the data, we are made to believe that these numbers are true OF EACH PITCHER, but what it seems that you’ve in fact shown us are numbers true OF ALL PITCHERS (TOGETHER).

    Vote -1 Vote +1

    • Eno Sarris says:

      I don’t see where I suggested it was true of all pitchers — just that it was statistically likely. An aging curve is always an aggregate, meaning there are always outliers.

      Vote -1 Vote +1

      • delv says:

        Noting your use of the term “outlier,” I think you’re underestimating the portion of players that pitch for 4 years or less. I would venture to guess that they’re a rather significant portion of the sample and that their piece-meal contributions to the aggregate can’t be taken as part of a temporal curve, at all, but rather as a collection of points dependent in their existence more upon the circumstances under which they got to participate than upon their age at the point in their careers when they got to.

        Vote -1 Vote +1

      • Barkey Walker says:

        It really matters how you do weighting in situations like this. If I pitch 9 innings do I get the same weight as someone who turned in 180 innings or 1/20th the weight? It is more complicated for these, “aging” charts since each increase in years is across two years.

        Vote -1 Vote +1

    • Jeff Zimmerman says:

      Some answers to the questions here:

      Delv: This is the tendencies of all pitchers that pitched in the two age brackets. So with the old pitchers, it is the “pitch to contact” pitchers from 32 to 33 that saw a decline. If they are the only pitchers to survive that long, then they are the only ones to sample.

      Walker: The %’s are weighted by the harmonic mean of the total TBF. The more batters a pitcher faces, the more their stats will be weight. There is a link to the method used under the image.

      Vote -1 Vote +1

  5. Roycliff Hallalee says:

    “That’s probably the biggest take-away from this graph: If a pitcher’s strikeout rate naturally declines from day one, you probably don’t want to build your team on free-agent pitchers.”

    Wouldn’t an xFIP/age or WAR/age plot be a better measure of pitcher aging?

    Also, the main benefit of home grown pitchers are dollars saved.

    Vote -1 Vote +1

    • CircleChange11 says:

      Why wouldn’t the conclusion be to have an all rookie rotation?

      Pitchers have the highest potential for value when they are young and strikeout a lot of players. Their arms are live and healthy.

      I see this thread and my mind jumps to Gooden’s 11 WAR campaign in his second season.

      Pitchers may get “smarter” and try to get batters out with fewer pitches. As they age, the velocity may dim and pitcher shift to trying to get batters out with more movement (weaker contact). Scouting may catch up to pitchers.

      Furthermore, young pitchers tend to make it o the big leagues on velocity or outstanding stuff. How many rookie and/or young pitchers make it to the majors because they’re “crafty” or “offspeed specialists”? Not many. Not only that but up to that point the young flamethrowers probably haven;t had to really use much else to be successful. As they get more experience, they develop other pitches, and it’s probably not a coincidence that it coincides with a decrease in velocity and strikeouts.

      Tim Lincecum may be the poster boy for the information contained within the graph.

      Vote -1 Vote +1

  6. test says:

    Included in there, and I don’t see any indication that it was controlled for, is that very good young pitchers often get to pitch out of the bullpen for a few years. This could be because teams have innings limits on them, or their secondary pitches aren’t ready for starting, etc, but how much of that age 21-22 peak is due to future starters airing it out in the bullpen for a year or two? Bard and Feliz come to mind this year as possible moves to the rotation.

    This has two effects – the obvious one is the bullpens have a lower percentage of the “good arms” as age gets older. Starting rotations, however, also gain from this, since that’s where these guys end up. That could explain the maintainence of SP K% over those age ranges, at least in some small part.

    Vote -1 Vote +1

    • Baltar says:

      Relievers moving to the starting rotation after they’re already in the majors? Maybe a long time ago this happened. Today, it’s so rare that it’s statistically insignificant.

      Vote -1 Vote +1

  7. Hurtlockertwo says:

    Throwers turn into pitchers, what a concept.

    Vote -1 Vote +1

  8. supgreg says:

    Too many of you are confusing a declining K rate with declining efficiency of a pitcher.

    The point is to show that K rates decrease as you age, and that a pitchers best K potential lies before he is 25. It is not saying young pitchers post the best ERA, FIP or xFIP.


    Higher K rate doesn’t mean better pitcher!

    Vote -1 Vote +1

    • FrankTheFunkasaurusRex says:

      usually does

      Vote -1 Vote +1

      • Antonio Bananas says:

        That’s only 1 component. Would you want to not trade for Gio Gonzalez at this point? The evidence from this article would suggest he’s going to get substantially worse and very quickly because of his already really high walk rate.

        Vote -1 Vote +1

      • Tom says:

        I think people are hung up visually on the graph…

        A 2% drop from age 25 to age 30 is not exactly a rapid decline. For example that’s a K rate plummeting from 9/9IP to ~8.8/9IP.

        Vote -1 Vote +1

  9. Tom says:

    How large is the age21-22 reliever sample?

    It looks like there is a substantial decrease (2%) after that first year and that might be making the starter and reliever curves look so different…. If age22-23 was the starting point it appears the curves would look a lot more similar (especially age 22-29)

    Is this real or a more of an artifact? Like say maybe future quality starters pitching out of the pen when they are first called up? Or maybe the quality of guys that make the bigs in the pen at age 21-22 being better? (A reliever with a lower K rate is probably more likely not going to make the bigs at that young an age and thus will depress the curve later on)

    Anyway – good stuff…. if you do this for BB, HR’s and maybe even groundball rates you can look at how FIP and XFIP ages.

    Vote -1 Vote +1

    • Eno Sarris says:

      the smallest sample size I know: 15000 BF, and yes I do believe it’s on the young end of the curve. There are definitely some interesting things going on on the peripheries — if you wanted to say that K% peaked at 25 that might be as true as anything I said. Still, I think that’s earlier than people think.

      Vote -1 Vote +1

  10. Luke says:

    I think Tom makes the point best. This graph makes it look a lot worse than it is. A starter who records 200 k’s at 27 years old will see a decrease to 196 k’s by age 31. That’s one strikeout per year – hardly worth writing someone off for. And as others have mentioned, BB% also decreases with age, so the effects could be negated.

    Vote -1 Vote +1

  11. Antonio bananas says:

    Most 21-23 year olds are super prospects right? They have insane stuff and that’s why they’re called up so early. So, my theory is that it’s not so much that guys “peak” early, it’s that once a guy has the physical tools, the ‘mental’ aspect is overrated. Or maybe it’s the other way around, maybe once a guy has Major League stuff and knows how to pitch, his productivity is pretty flat. Like the guy above me said, 2% isn’t really that much. If pitchers actually got worse overall on average from age 21, wouldn’t WAR and just about every other statistic also show it?

    Vote -1 Vote +1

  12. justanotherfantoo says:

    For some reason, these principles don’t seem to apply in Philadelphia. There, pitchers just get better on arrival — whether it’s from the minor leagues or another team — and continue to get better with age.

    At the current rate, when Halladay and Lee are in their mid-40s, they’ll be pitching over 300 innings a year to a 1-something ERA, an OPS against below .500, and a WHIP below .900. Hopefully, for the sake of all of the other teams, they never figure out how to make it last through the post season.

    Vote -1 Vote +1

  13. Siddfynch says:

    Students in statistics classes are taught to be on the lookout for disingenuous graphs like this….the axis is limited to only those numbers that contain the full range of the data, not a range based on anything else. No reason not to use 100% as the full range; restricting it to 12% is a great way to ensure the author’s point is made, but isn’t otherwise very defensible. Whoever authored it, consider your pinky slapped.

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>