Pitchers’ workloads are certainly a topic we’re used to hearing about as baseball fans. We live in the pitch count era after all, and every game has a pitch count indicator on the screen showing how many the starter has thrown. We’ve gotten used to starters getting the hook right around 100, even if they’re pitching well. We also know that it is to avoid injury to this most injury-prone of positions. It’s never been shown very clearly that higher pitch counts lead to injury, but there’s enough worry that teams want to play it safe with these prized assets. This is even more true with young pitchers: they often aren’t allowed past 85 or 90 pitches if the team is especially worried about their arm.
We also know the other reason why: pitchers just aren’t going to keep doing as well if you leave them in for that long. Past 100 pitches, pitchers are usually well into their third time through the opposing team’s batting order, if not their fourth. We know that each additional time hitters get to see the same pitcher in the same game, the better the hitters do against him. And we know that, of course, pitchers get tired as they throw more pitches, and their velocity drops, and with it, their effectiveness.
But should there be another consideration here? We know the long-term reasons for limiting pitch counts, as well as the short-term ones. But what about the medium term: how does a starter’s pitch count affect how he’ll do his next time out on the mound?
Over at Baseball Prospectus, Russell Carleton (a.k.a. @pizzacutter4) looked at this question back in 2013. He found that past 100 pitches, every further pitch thrown leads to more home runs and more singles being given up next time out, as well as fewer balls in play meekly falling for outs. But his study was only focused on the extreme upper end of pitch counts, inspired as it was by Tim Lincecum’s brilliant 148-pitch no-hitter. That matters, but I also want to know what happens before a starter gets to 100 pitches. There’s no reason to think the effect of workload only kicks in after 100 pitches have been thrown. Will a pitcher do better next time out if his pitch count is kept significantly below 100? I decided to find out.
First, I took a list of every pitcher season in the past three years in which the pitcher made at least 30 starts. I chose the past three years because that’s roughly the length of the current run-scoring environment — ever-growing strikeouts but also high home runs — and there was also a precipitous league-wide fall in complete games from 2015 to 2016, after relative stability in the years prior. I set 30 starts as a minimum because I didn’t want the dataset contaminated by pitchers who swapped between starting and relieving, and therefore threw fewer pitches per start than true starters, or by pitchers who had injuries that they were returning from, which would possibly have left them on pitch counts. For that latter reason, I also removed from the dataset any pitcher seasons in which the player had spent any time on the disabled list, or any pitcher seasons that came directly after one in which the pitcher suffered a major season-ending injury that he was still working his way back from come spring. Finally, I removed knuckleballers, since their wear and recovery, of course, works very differently from that of normal pitchers. This left me with a total of 133 pitcher seasons from 2016 to 2018.
For each pitcher season, I looked only at the starts the pitcher made on four days rest. First, I calculated the pitcher’s overall FIP on four days rest that season. Next, I divided all of his starts on four days rest into buckets by how many pitches he’d thrown in his previous start: 0-9, 10-19, 20-19, and so on upwards. I then calculated his total FIP that season in each of those buckets, and I expressed it as a percentage of his overall four-days-rest FIP. Finally, I totaled the results in each bucket for all 133 player-seasons in the dataset. Here are the results:
The first column is the range of pitches thrown by the pitcher in his last start. The second column is the number of innings in that sample that were thrown on four days’ rest from throwing the number of pitches in the first column. The third column is just the percentage of the total innings represented by the second column: that is, 1.7% of innings in the dataset were thrown five days after throwing somewhere from 60 to 69 pitches. The fourth column is the finding: the ratio, for all of the innings thrown following that level of workload, of FIP to the pitchers’ overall FIP that season. That is, five days after throwing 60 to 69 pitches, pitchers FIPs tended to be 91% as high as their overall seasonal FIP. Finally, the fifth column is simply the fourth column multiplied by the league-average FIP over the past three years, 4.23, to put the results in more familiar terms.
Let’s put the results in even more understandable terms. Let’s make a graph!
Clearly, things are pretty muddled by the small sample sizes at the extremes. Let’s make another graph, but this time, only of the buckets from 80 pitches to 119, the four buckets that have at least 10% of the data each and that together contain 91.2% of the data.
So what do we make of this data? I see two main takeaways, one direct and one less so.
First, it’s pretty clear that throwing more pitches does, indeed, tend to limit a starter’s effectiveness his next time out. And not only at very high pitch counts (where this study, curiously, contradicts Russell Carleton’s, though I can’t imagine throwing more than 110 pitches actually helps pitchers in their next start, so there may be a hidden bias here), but also at moderate, common levels, where the data shows a huge loss in effectiveness from throwing 80-something pitches to 90-something, and another, smaller loss from 90-something to 100-and-something. Managers shouldn’t just have a pitcher’s next start on their minds when they’re deciding whether to let him go out for the next inning after already having thrown 110 pitches; they should be considering their pitcher’s next start at all points during his current start. That is to say, managers shouldn’t only try to keep their starters from racking up very high pitch counts; they should try to limit their pitch counts long before 100 even comes into view. This has limits, of course — there wasn’t any meaningful amount of data on the very low end of the pitch range, and it’s very plausible that if a starter throws 20 pitches, he’ll be more rusty than rested his next time out — but once a good starter gets to 60 or 70 pitches, his manager should be looking to get him out as soon as it makes sense in the game situation.
What I mean here, specifically, is that it makes no sense to ride your best starters past 60 or 70 pitches in a blowout. To illustrate a striking example of this, let’s take a look at Chris Sale last summer. From June 24th to July 27th, Sale went on a well-publicized tear. In six starts, he threw 39.0 innings, struck out 67 batters to 6 walks, and he allowed 1 (one) run. During several of these starts, Boston’s offense was just as hot as Sale, scoring 5, 9, 8, and 6 runs while he was still in the game. Yet even while Sale was allowing 0 or 1 runs in each of these starts, manager Alex Cora left him in to throw 93, 101, 99, and 99 pitches, respectively, with these enormous leads.
What’s the point in this? When you have a 7-0 lead going into the other team’s half of the 6th inning, there’s no reason to be using your best pitching resources. Just as you would never bring in your ace reliever — Craig Kimbrel, in Boston’s case — to such a game, you should also never keep riding your best starter. The upside is a tiny increase in win probability — which, for Boston, had already reached 98.2% entering the bottom of the 6th in this June 29th game in New York! The downside, as we’ve now seen, is a fairly significant deterioration in Sale’s performance in his next start, which could be much closer than this one.
To be clear, what I’m proposing isn’t only that managers decrease how long they leave their top starters in in close games. Rather, it’s that they start treating their best starters more like their best relievers: ration their usage based on leverage, getting a lot of innings out of them in close, important games and keeping them out of blowouts. Cora should’ve removed Sale a few innings earlier during each of those June/July blowouts. But sometimes, he should leave him in longer. Consider another of Sale’s starts last season, in Toronto on May 11th. Sale threw 9 innings and dominated, striking out 15 Blue Jays with no walks while allowing three runs. However, because Sale had already thrown 116 pitches, Cora took him out when the 9th inning ended in a 3-3 tie and went to the bullpen. By the 12th inning, he was on his third reliever of the night, and Brian Johnson gave up the game-winner to Luke Maile. Cora could’ve avoided putting the game in Johnson’s hands that early if he’d kept Sale in, at least for the 10th. And perhaps he could’ve left Sale in for the 10th if he wore him out less during blowouts, leaving him more rested and more effective for starts in which his ability was more needed. Managers shouldn’t just use their best starters less: they should use them more efficiently, shifting pitches away from blowouts and towards close games, so that they’re more rested, more effective, and available for longer in the latter.
But Sale’s case also hints at the second, and less direct, of the two conclusions we can draw from this study. Sale, quite notably, has tended to falter late in seasons compared to his performance in the earlier months. He also made two trips to the disabled list in 2018 with shoulder problems. This study didn’t say anything about the long-term effects of high pitch counts, it’s true. Its analysis of their impact stopped five days after they occurred. But impacts can be inferred, as well as directly shown. And if pitching longer wears pitchers out in a way that causes them to be less effective in their following start, then, when accumulated, it quite likely also wears them out in a way that causes them to be less effective over the following months. And, quite possibly, it increases their susceptibility to injury. Perhaps this isn’t too relevant for ultra-durable starters without these seeming wear-and-tear issues: using Max Scherzer, for example, more efficiently in the ways I lined out before is probably the only change Washington needs to make with him. But for someone like Sale, who seems to tire out over the course of every season, having him throw fewer total pitches (particularly in blowouts) may help keep him off the disabled list and fresh for September — and October.