Yesterday, Tom Verducci wrote about the danger of pitchers going over their accustomed workloads:
Get ready for the down side to all that young pitching success. It’s called the 2007 season. More specifically, it’s the Year-After Effect, the price teams almost always pay for pushing their young pitchers too far. And we could be due for a huge crash next season. I’ve been tracking the YAE for about a decade now. It’s based on a general rule of thumb among executives and pitching coaches: young pitchers should not have their innings workload increased by more than 25 or 30 innings per year …
When I’ve looked at major league pitchers 25-and-younger who were pushed 30 or more innings beyond their previous season (or, in cases such as injury-shortened years, their previous pro high), I’ve been amazed how often those pitchers broke down with a serious injury the next season or took a major step backward in their development.
It’s an interesting theory in a field sabermetricians have just begun to scratch. Can we predict injuries? Do some teams push their pitchers too hard? Well, those are questions we can at least begin to answer.
Verducci says that he’s been tracking these numbers for a decade, so let’s look at pitchers between 1996 and 2005. That should limit us to studying the modern pitcher, anyway. Here’s what I did: I looked at all pitcher seasons under age 25, and tracked their performance in the next two seasons.
I then split those pitchers into two groups: Those who threw 30 or more innings in their second season than they did in their first, and those who did not. There were 270 pitchers in the first group, 709 in the second. I then looked at how each group performed in its third season; according to Verducci’s theory, pitchers from the first group should see a larger decline in their performance as compared to the second. In essence, the second group acts as a control.
Now there are some obvious flaws in my study. I don’t include minor league innings or postseason data, which Verducci does. I only look at whether a pitcher throws 30 or more innings than he did in the previous season, rather than in any season in his career. However, even with those issues, if the Year-After Effect exists, it should show up at a group level, and given that, it will show up in this study. So on to the results:
IP H/G ERA HR/G BB/G SO/G IP H/G ERA HR/G BB/G SO/G Year One 51.51 9.08 4.76 1.05 3.81 6.62 65.83 9.15 4.80 1.15 3.67 6.23 Year Two 129.71 8.93 4.35 1.05 3.48 6.49 42.83 9.16 4.77 1.12 3.57 6.50 Year Three 111.50 8.98 4.39 1.06 3.36 6.57 47.68 9.10 4.47 1.10 3.35 6.54 Change 2.16 0.99 0.92 1.01 0.88 0.99 0.72 0.99 0.93 0.96 0.91 1.05
The first half of that chart tracks pitchers who threw at least 30 more innings in year two than they did in year one. As you can, see the average is quite a bit more, as they went from about 50 innings to almost 130. Meanwhile, the second group actually threw less innings, on average, than the first one did, losing an average of 23 innings per player.
Verducci is right in a way, as the first group’s innings do decline somewhat, while the second group actually sees a bit of an increase in innings from their second year to their third year. But, in terms of overall improvement, the two groups are virtually indistinguishable.
We can’t really look at numbers from year two because they biased by selective sampling, but we can fairly compare numbers from year one to year three, which is what I’ve done in the last row, which shows the percentage change in each category from year one to year three.
As you can see, both groups start out with almost identical performance in year one (4.76 ERA versus 4.80 ERA) and end up with very similar performance in year three (a 4.39 ERA for group one versus a 4.47 ERA for group two). The only very big difference in the change category is innings pitched: The “overworked” guys throw 216% more innings in the third year than they do in the second, while the control group throws 72% as many.
Seems like Verducci is wrong. Except, you might argue, that the first group is likely composed mostly of starting pitchers while the second group is largely relievers. The relievers would be expected to pitch less innings the third year even if the Year-After Effect existed, even though the second group averaged more innings pitched in year one than did the first group because we’re probably missing a lot of minor league innings for the first group. As well, the relievers would be expected to drop out of the major leagues more easily because an average ERA for a reliever is not nearly as good as an average ERA for a starting pitcher.
OK, so what happens if we limit ourselves to pitchers who threw at least 100 innings in year two? Actually, a funny thing. The pitchers who best their career high by at least 30 innings go on to throw 90% more innings in year three than they do in year one, and those who didn’t only throw 78% as many innings. What’s more, while the YAE candidates have an ERA 9% lower in year three than it was in year one, the guys who were accustomed to the big workload do not improve their performance at all.
The bottom line: a dramatic increase in innings on a young pitcher elevates the risk of injury or a setback to their development.
But the evidence points to the opposite. Pitchers who see a large increase in workload are more likely to continue to be successful than those who don’t. It’s important to remember that correlation does not mean causation—just because throwing a lot more innings than a pitcher ever has before is correlated with future success does not mean that managers should be riding their young pitchers hard—but it does imply that Verducci’s argument is incorrect, and there is absolutely no reason that we should expect these YAE candidates to do worse because they’ve overworked.
Verducci cautions readers to watch out for big declines from Matt Cain, Gustavo Chacin, Zach Duke, Scott Kazmir, and Pat Maholm. I say draft them in your fantasy draft and get your hopes up! Except when it comes to Duke and Maholm—Pirates players are hopeless, no matter what the numbers say.