The fall of the swingman

We usually think of starters and relievers as different entities. The starter is the pitcher who starts the game, and the reliever is the pitcher who comes in to relieve the starter. It’s wonderfully simple. But starters don’t always start, and defining what is and isn’t a starter can sometimes be difficult. Pitchers are often used in different roles, even during the same season.

Take Clay Buchholz’s 2016 season as an example. He threw 139¹⁄₃ innings over the past year, but not all of them came as a starter. In fact, only 21 out of his 37 appearances came out of the rotation — meaning that in 2016, Buchholz made 16 appearances out of the pen.

Due to bad performance, and the Red Sox high expectations, on May, 29th, Buchholz made his first appearance as a reliever. It was the first time he had made an appearance out of the pen since 2008, and it was only his third appearance out of the pen in his entire career.

Buchholz, though, isn’t alone in this regard. In fact, as I was able to find out, pitchers whom one might label as starters, pitch innings out of the bullpen more than one might think, but that trend is drastically decreasing. ^[1]

The trend has somewhat stabilized over the past few years, but it’s undeniable that in general, fewer and fewer starters are pitching innings out of the bullpen.

This makes a lot of sense, because pitchers nowadays are very different than they were just 10 years ago. More and more pitchers are becoming specialized, meaning they’re pitching fewer and fewer innings. It also doesn’t make sense to have your starter constantly roaming from being a starter to being a relief pitcher. Both tasks require different skills and different levels of exertion, and the more you play around with a pitcher, the bigger the chance he’ll getting hurt.

That’s why, when I found that in previous eras of baseball, more than 50 percent of starters made at least one appearance out of the pen, I was a little shocked.

But maybe I shouldn’t have been. A number of studies and articles have been written on how the game is changing. And pitchers may be at the apex of that change. Roles between starters and relievers are being more defined and divided. You rarely see a starter make a bullpen appearance nowadays, unless they started underperforming or are maybe rehabbing from an injury.

That said, just because this phenomenon isn’t as common as it used to be, doesn’t mean that it doesn’t happen — and in fact, some pitchers do walk that fine line, between being a starter and being a relief pitcher.

In 2016, Luis Perdomo, Jhoulys Chacin, Ross Stripling, Patrick Corbin, Albert Suarez and Clay Buchholz led all starters in relief innings pitched. The commonality between all of these pitchers is their performance: Every single one of them had a DRA above 4 (except for Jhoulys Chacin, who had a DRA of 3.94, which isn’t a whole lot better).

While these pitchers are interesting test cases on their own, the hurler who really stood out to me was Tanner Roark.

It makes sense for Buchholz to be moved to the bullpen. He had a 6.08 DRA in 2016, and the Red Sox decided early on to put him in the pen. Roark, on the other hand, pitched more than 200 innings in 2016. Granted, he wasn’t blowing people away, but if you are committed to having him as your starter then why would you make him pitch out of the bullpen, even once?

Well, it turns out Roark was going to get some extra rest thanks to the Nationals’ schedule, and therefore he made himself available to pitch out of the pen. The game was on July 10th and the Nationals were facing the Mets. Dusty Baker called for Roark with two outs in the sixth, and he came in and pitched 2¹⁄₃ innings.

This was an isolated incident, but it got me wondering how common it was. How often do starters who pitch a lot of innings end up throwing innings in relief?

The relationship here makes a lot of sense. The fewer innings a starter pitched, the more likely he is to compile some innings as a reliever. The reason is quite simple. These are usually pitchers who either aren’t performing very well or don’t have a lot of confidence from their manager. That’s why managers probably don’t mind, playing with their roles every now and then.

Ken Dixon, Don Aase, and Juan Eichelberger all drew a fine line between starter and reliever in those respective seasons. Each one of them pitched more than 50 innings in relief, which in today's game is around the normal workload of a relief pitcher. Dixon and Aase pitched more than 100 innings out of the rotation, while Eichelberger pitched 80; those, again, are not unheard-of workloads for starters in today’s game.

But these pitchers pitched in a different era. Each one of those seasons occurred in the early to mid-80s. And Wilbur Wood had his insane season in 1973, where he threw five innings in relief and 354¹⁄₃ as a starter — numbers we’ll never see again.

And, if we examine the pitchers who threw at least 200 innings and who threw at least one inning in relief, we can see that the occurrence is almost nonexistent nowadays.

This is no surprise. It’s cool when Clayton Kershaw makes a bullpen appearance in the playoffs. But if he does that during the regular season, then it’s kind of less cool. At that point, it’s a little weird…unless, say, it’s to clinch a playoff spot in the final days of the regular season, and then it becomes cool again.

At any rate, you don’t want a “workhorse” — a pitcher that you depend on every week to make starts — to also make a bullpen appearance. There are many reasons why, but mainly, you don’t want to risk an injury, especially when your team is so dependent on that pitcher.

The fact that more than 50 percent of starters threw 200 innings and made at least one relief appearance in the ‘60s and early ‘70s is madness, and once again speaks to the ever-changing and evolving game of baseball.

An earlier version of this article incorrectly attributed the data to Baseball-Reference instead of Baseball Prospectus. Please don't cancel our BP subscription, Aaron.

***

^[1] A “starter” was defined as a pitcher with more than six appearances who made more than 50 percent of his appearances as a starter (GS/G < 0.5); the rest were defined as relievers. All data came from Baseball Prospectus.

Julien Assouline is a contributor at Beyond the Box Score. He’s written for Baseball Prospectus, The Hardball Times, and BP Milwaukee. You can follow him on twitter @JulienAssouline.