One hundred twenty-four starting pitchers have thrown at least 100 innings in Major League Baseball this season. Of those, just 16 are 24 years old or younger, in terms of official baseball age. Matt Harvey is among the 16, albeit toward the older end of the spectrum.
The table below shows all 16 of those young hurlers, along with the number of pitches they have thrown per start this season, and finally, where that figure ranks among the 124 total pitchers who have topped 100 frames:
Young Starters, 2013
Pitchers Per Game Started
Rank (of 124)
Only Sale has thrown more pitches per outing than Harvey. Sale’s day will come, but on Monday, it was Harvey who took to the shelf with a partially torn ulnar collateral ligament (UCL), which might or might not require surgery, but absolutely represents malpractice on the part of the Mets.
It’s true that Harvey is almost 25, the age at which, by general medical and research consensus, the restrictions can come off a bit. It’s also true, though, that he’s only almost 25, not 25, He was and is part of the under-25 injury nexus.
My thesis here is that the Mets asked far more of Harvey than they ought to have, and that the fact that he was within the top 15 percent of all pitchers in pitches per game started before turning 25 has relevance to the fact that he got hurt. It’s not a terribly popular one: Jonah Keri shot me down on Twitter over it. Still, I think I can defend it.
The question at hand is whether the risk of injury increases for young pitchers who pitch more than their peers no matter what, or whether some level of moderation in their workload allows teams to minimize the injury risk. In essence, the tension between my position and Keri’s is that he (and most of the people I respect in baseball’s analytical circles) perceives the risk level associated with any given workload to be absolute, while I perceive it to be relative.
In other words, take a pitcher and clone him. Put both pitchers through identical childhoods. There is to be no physiological difference between them when we begin this thought experiment. Send one through time to 1993, and have each begin their career today.
Have each throw exactly the number of pitches Harvey has thrown in his 26 starts this season:
Matt Harvey, Game Log, 2013
Now, is one guy at greater risk than the other? I would argue yes. I would argue that a pitcher in 2013 bearing this workload is working harder, relative to his peers, and that he’s therefore working within a system of instruction, preparation, treatment and management that can’t support him as easily. I would argue he’s at greater absolute risk.
I don’t think, though, that that matters too much. To me, elbow injuries are like lightning strikes. They’re sudden, violent, acute and unpredictable, but they’re not random. Of the people who get caught in any given storm, there is always a most likely person to get struck.
Teams used to send young pitchers out onto rooftops to adjust TV antennae. Now they have them wearing soft-rubber shoes and staying low. They’re getting better at sheltering them, but they can’t protect them completely without keeping them indoors (off the mound) in the first place.
The Mets didn’t do a good enough job of keeping Harvey low. In fact, he was the young arm (other than Sale and arguably Bumgarner) at the highest elevation in the storm. He was at much less risk to get hurt than Kerry Wood ever was, but still, he was at more risk than almost anyone else in his peer group, and Terry Collins put him there in the service of nothing.
Nothing would have been lost if Harvey had vacated the mound 10 pitches sooner in two or three of those games. Nothing would have been lost, even if the Mets lost one or two of those games precisely because Harvey got the hook. The Mets had no chance at anything this season. They needed to recognize that better.
Harvey should have been on the Fernandez diet, throwing 95 pitches per game. Unlike Fernandez, Harvey is old enough that they could really cut him loose even next year, have him occasionally reaching 120 pitches in the service of more important wins. This year, though, this lost year in Queens when there was no hope of a playoff berth and Harvey still had some growing and developing to do, they should have been miles more careful with him.
The counterargument will be that the Mets sought to protect Harvey in a different way. They gave him extra rest days, fairly regularly. Officially, he went 13 days between starts on July 8 and July 21. (Officially, that’s true, but he did pitch in the All-Star Game right in the middle of that stretch.) They also gave him two extra days after his start April 29, pitching him next on May 7.
I fail to see that this is effective protection from elbow injury. It’s a fine way to manage his seasonal workload and total usage, which are crucial objectives in the prevention of shoulder injuries. Elbow injuries, though, are not about how much one pitches. It’s about how much one pitches once fatigue begins to set in and the maintenance of sufficient stuff gets more difficult. Elbow drag is the term, and almost universally, elbow drag increases as a pitcher crosses the 110-pitch plateau.
Harvey threw 121 pitches in that April 29 start, and 105 on May 7. He threw 121 on July 8, and 112 on July 21. The recovery time the Mets gave him probably allowed him to take the mound after the rests without pain, without swelling, without any sense of trepidation. I doubt, though, that it actually and actively reduced the chances that he would hurt his elbow. At the very least, I firmly believe that asking him to throw fewer pitches in each start in the first place would have reduced those chances more.
If Harvey were older, his ligaments might be stronger and he might be more physically developed. That’s not even the primary reason, in his particular case, that I advocated waiting before loading him up this way. Rather, the primary reason is that Harvey had some mental development yet to do, some things to learn about the modulation and regulation of the effort he puts into his pitches.
Harvey’s velocity jump upon arrival in the big leagues is well-documented. You can find an intelligent conversation on the topic in Tuesday’s Effectively Wild, the daily podcast from Baseball Prospectus. Harvey the prospect threw in the 92-95 range, occasionally touching higher. Harvey the Met has averaged nearly 96 miles per hour on his fastball this season, touching 100 more than a few times. He also throws a slider that can touch 92 miles per hour and sits at 90.
First of all, a significant velocity gain suggests something problematic to me: that a guy is willing, even eager, to test and touch the top of his register. I don’t apply it universally or with a heavy hand, but whenever I see that sort of change, my first thought is that this guy is much too interested in throwing hard.
Secondly, that slider just scares me. Any non-fastball a pitcher can throw that hard sends up a warning sign in my mind. Throwing 92 with the torque and torsion required to produce a slider can’t not put one at some elevated risk, in my opinion.
My main point with this is that you have to protect against all injuries, not just certain kinds. It sometimes means, at least when one is not contending, both stretching the time between starts and ensuring no single start rises to the level of a threat to the pitcher’s health. It means not waiting to see 121 in the pitch-count column before pushing a guy back. It means keeping 121 out of that column altogether.
Heck, it’s not like the Mets even protected him all that well from an absolute workload perspective. Here, at the close of his season, he ranks seventh in baseball in innings pitched. Given a full season, and assuming that no shutdown was forthcoming absent the injury, he was on pace for roughly 220 innings. That’s more than I would ask any young arm to throw under the competitive circumstances. Even so, it’s not the main issue. The main issue is how deep the team let some of his starts run.
Long relievers could have significantly reduced the chances that we would end up here. If Collins felt like he could go to an arm who could bridge some gaps for him, the game on April 29—when it took Harvey 121 pitches to get 16 outs while allowing four runs—would not have become what it became. On July 21, when the Mets had a 5-0 lead by the end of the fourth inning, Harvey’s day could have ended after six innings and 97 pitches, instead of after seven and 112. Collins should have stopped those things from happening anyway. He should have gotten Harvey out of each start sooner. It didn’t help, though, that the modern bullpen is built for power, not endurance, and not, honestly, for relief, either.
A lot of this might sound like hindsight bias. It isn’t, or at least, not much of it is. I decried pushing Harvey past 110 pitches each time it happened. I balked at the Mets’ determination to squeeze innings and strikeouts out of him at a time when those innings and strikeouts mean nothing to the franchise in terms of reaching the playoffs. This is an unpopular stance, but I’m standing it. Matt Harvey’s injury was not random, and the fact that it happened is at least partially the Mets’ fault, and his own.Next post: Bryce Harper: Still a Better Bet Than Any Other Player in Baseball
Previous post: Left Outs: How the Death of Platoons and a Right-Handed Game Are Tamping Down Offense League-Wide