What happens when you replace one problem with another and no one notices? In the world of pitcher injuries, we may have done exactly this. In the late 1990s, there was an emerging problem in Major League Baseball. Starting pitchers were often being asked to throw 130 pitches in one night, and this seemed to be associated with those same pitchers later suffering catastrophic injuries. At least, that seemed to be the way it worked in the 1990s and 2000s. Prior to that, we don’t have the data to really run detailed analyses (we lack pitch count and injury data), though from what we can fathom, practices that were later called “abusive” for starters weren’t necessarily associated with more injuries prior to the 1980s.
But in the 2000s, there was a reckoning around pitch counts. Teams became more conscious of the research surrounding pitch counts and started toying with the number 100 as a good upper limit for a good day of work. Even then, there were still some mega pitch-count games. But then, suddenly, the problem solved itself.
There’s a certain logic, and maybe even a correct one, to pitch count limits. Pitching is a violent motion for the arm and the strain that comes from it builds up over time. Eventually, for some pitchers, something snaps. The human body needs rest. But suddenly, teams seemed all too willing to give their pitchers that rest. The number of starts taking place on “long” rest increased and as the graph shows above, the average number of pitches thrown by a starting pitcher dropped.
The problem was “solved” by the realization that bringing in a reliever was a better option and there was no need to have pitchers extend much past 100 pitches per game. But rates of pitcher injury didn’t go down. For that, we may be able to thank Sam Peltzman. If you’re searching your memory banks for some random 1980s pitcher, that’s not who Peltzman is. Peltzman was a professor of economics (and is now a professor emeritus) at the University of Chicago who noted that when seat belt laws were introduced in the United States, it didn’t cut the highway death rate as much as expected. Perhaps the problem was that feeling better protected by their new seat belts, drivers felt more empowered to take more risks when they drove. The idea is that life is full of risks and so all of us have some risk tolerance that governs what we are willing and not willing to try. Introducing a new safety measure makes everyone more safe, but it doesn’t affect the initial risk tolerance. People therefore take risks that they previously wouldn’t have, optimizing their level of risk tolerance rather than their level of safety.
In baseball, we were left with a vocabulary of pitch counts, but didn’t anticipate that pitchers might say “Well, if you’re going to limit me to 100 pitches, I’m going to throw the pitches that I do get to throw harder.” There’s a word that I will never forget from my days of clinical practice (I hold a degree in clinical psychology but have long since abandoned actively seeing patients). That word is “iatrogenic.” It means that the treatment for a condition causes either the problem to get worse or causes a different problem.
We turn our eyes now to the children. Last month, MLB wrote a research report that didn’t contain a lot of research and then didn’t really release it to the public, though it was widely reported on in the baseball press. Whether MLB’s concern for the kids is driven by public relations or a genuine concern for pediatric health (probably a bit from each column), the report does call out the continuing youth epidemic of major elbow surgeries, particularly Tommy John surgery.
In 2014, MLB partnered with USA Baseball to produce a series of guidelines termed “Pitch Smart” which provided age-based guidance for how many pitches a young person should max out at and how many days of rest a young person needed afterward. The recommendations—and let’s note the rather tentative nature of that word “recommendations”—were based on information provided by pediatric orthopedists and other experts, and for the most part have become the de facto pitch limit framework in youth baseball. The big-name organizations (Little League, Pony League, American Legion) have all instituted the guidelines.
In 2016, the National Federation of State High School Associations adopted a similar policy based on pitch counts. Before that, most state athletic associations based their rules on innings limits (if they had any regulations at all), rather than pitch limits. In 2014, I did some research on whether Major League pitchers from states whose high schools had no restrictions were more likely to sustain injuries. They were. It’s hard to draw a causal relationship from that sort of study, but it’s hard to ignore the possibilities there too.
It’s important to note the timeframe there. The Pitch Smart guidelines were released in 2014, right before MLB saw a decrease in its own average pitch count for its (adult) starters. Around that time, the archetype of the one-inning reliever had been laid down, but the emphasis that MLB teams placed on shuttling players back to and forth from the minors to accommodate the heavy bullpen workload required to maintain the “max effort all the time” movement that was emerging. A decade later, the art of pitching is different than it once was.
It’s left us with a vocabulary composed entirely of pitch counts and rest days when that’s not really the problem anymore. Research—and here I’m talking about verifiable academic research published in peer-reviewed science journals—has shown that even Little League pitchers who had MRIs taken before and after their seasons and were fully compliant with the guidelines showed damage to their elbows. The researchers in that paper noted that the biggest risk factor for damage was year-round pitching, rather than any pitch count shenanigans. One other major problem is that pitching isn’t the only sort of throwing that Little Leaguers do. Little League pitchers will often play field positions when not pitching, and those extra throws don’t count as pitches, but your elbow doesn’t know that.
Pitch counts are a blunt instrument. There probably are some kids (and adults) whose arms can handle a higher-than-average workload, though figuring out who’s who is probably well beyond any current technology. (Wouldn’t you know it, the kids who are most likely to be over-used, according to the guidelines, are the kids who throw the fastest. Funny that.) Even if we had that technology, the only way it could really be used is to personalize a pitch limit, and I don’t think people would welcome that. Why does one pitcher get to throw 50 and another 20?
But pitches are, erm, count-able and quantifiable; so are rest days, and that might be the biggest problem of all. When the issue was “pitchers are hurling too many times” it was just a matter of finding a good number for them to stop at and then having everyone in the ballpark sing a verse of “99 Bottles of Beer on the Wall” as they threw each pitch. When the beer runs out, the reliever runs in. We can make the rules the same for everyone and play to the averages. It’s all very linear.
In my book The New Ballgame: The Not-So-Hidden Forces Shaping Modern Baseball, I wrote a chapter called “Is Analytics Ruining Baseball?” I answered the question with “yes, but not in the way you think.” One of the flaws of the analytic movement has been conceptualizing everything in baseball in terms of questions that have an answer. If I asked you how many pitches were “too many” for a young person, we might disagree about method, but we are competing to see which of us can come up with the best methodology to produce the answer. Maybe it’s 50 or maybe it’s 60, but we will both come to the end of that exercise with a number.
Not everything in life is linear like that. Putting a number on the problem didn’t solve it. It’s possible that it made things better, but it also may have just taken time for kids (and let’s be honest, their parents) to figure out ways to re-establish the levels of risk that they were previously comfortable with. If we’re going to be serious about reducing pitcher injuries, we need to think of these problems dynamically. You can institute pitch counts in games, but you certainly can’t police what young pitchers do in their free time. You can tell them that they’re limited to 80 pitches today, but can you mandate that they limit themselves to pitching only eight months of the year?
One of the functions of culture is to produce ways in which individuals can express their risk tolerance. We have bran flakes and raw milk, certificates of deposit and penny stocks, smart cars and motorcycles. We can add to that pitch counts and pitching labs. The problem of pitcher injuries isn’t going to be solved only with putting the right rules in place. It’s going to have to be an expression of culture. It’s going to run headlong into questions of how much risk is worth it when the possible payoff is individual glory and team achievement. The data tell us that it will have to start at the youth level, which means that the MRI machines are going up against cultural norms of parental responsibility.
It’s going to be a bigger fight, and a whole lot messier, than we might imagine.
Thank you for reading
This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.
Subscribe now