Trevor Rosenthal as Andrew Miller

When Trevor Rosenthal joined the St. Louis Cardinals in July 2012, he underwent what, at the time, was not necessarily a permanent role change. Since the 2011 season, when he emerged from Rookie leagues to pitch with the Quad Cities River Bandits of Class-A, Rosenthal had accumulated 42 starts and zero appearances in relief, but rather than put Rosenthal in the rotation with Jaime Garcia heading to the 60-day Disabled List, the Cardinals opted to implement Rosenthal, a more esteemed prospect than Matt Adams or Matt Carpenter in the 2012 Baseball Prospect Book as a starting pitcher, in the bullpen.

In the bullpen, Rosenthal matched his starting rotation earned run average in AA Springfield, contributing a 2.78 ERA in 22 ²⁄₃ innings. Seemingly following the path of Adam Wainwright in 2006 and Lance Lynn in 2011, Rosenthal made the 2013 Cardinals bullpen and was highly effective—in 75 ¹⁄₃ innings, he had a 2.63 ERA, an even more sterling 1.91 FIP, and went from medium-leverage reliever to Edward Mujica’s setup man to, late in the season and into the playoffs, unofficial closer.

For the next two seasons, Trevor Rosenthal served as Cardinals closer and was, if not in the absolute top echelon of MLB’s ninth-inning men, just a tick below that. After a somewhat diminished 2014, Rosenthal’s 2.1 FanGraphs Wins Above Replacement in 2015 trailed only Cody Allen, Aroldis Chapman, and Dellin Betances in fWAR among MLB relief pitchers.

Sure, he still had critics (what reliever doesn’t?), and his occasional bouts of wildness left him susceptible to more than his share of nail-biter outings, but Rosenthal was, on balance, excellent. The only reasonable argument for getting rid of Rosenthal was that he had tremendous trade value, certainly not that he was not a player worth having.

Rosenthal had become such a good reliever that discussion of moving him back into the rotation had been mostly tabled—it’s somewhat counter-intuitive that the proper course of action of a pitcher being exceptional is to assure that he not accumulate many more innings by starting, but there is a recent track record of high-end relievers who are former unsuccessful starters that suggest Rosenthal could be a case of a player who is much worse at starting than the overly simplistic “add one run to a reliever’s ERA to get his starter ERA” shorthand would imply.

But in 2016, the argument went from “Rosenthal shouldn’t start because he’s so good at closing” to “Rosenthal shouldn’t pitch because he’s bad now.” Rosenthal’s 4.46 ERA more than doubled his 2015 mark. His .425 opponent’s batting average on balls in play was probably a bit inflated by some bad luck, but he was also hit harder than ever before—his previous career-worst in Opponent Hard Hit Ball Percentage was 27.4%, and in 2016, this mark stood at 34.9%. Most startling, however, was the jump in Rosenthal’s walk rate; although he was never exactly a control artist in the vein of Koji Uehara or his Cardinals closing predecessor Edward Mujica, Rosenthal walked 14.7% of batters he faced in 2016.

Most of this ineffectiveness came before his late July stint on the Disabled List with shoulder discomfort. In his five appearances from September 16 onward, he was much better. He struck out two of three batters he faced against the San Francisco Giants in his first game back, and he avoided issuing any walks in three of the five games. But it was Rosenthal’s October 1 performance, his final outing of the 2016 season, that may have been a game-changer.

Following an ill-fated start from Michael Wacha, which was abandoned after just one inning, and an inning of relief from Miguel Socolovich, in a virtual must-win game against the Pittsburgh Pirates, the Cardinals turned to Trevor Rosenthal. For the first time since 2012, Trevor Rosenthal went three innings or more in a game, and he was fantastic. His 52 pitches were a bit high, but the results were all there—four strikeouts to just one walk, two hits allowed, and solid command that suggested that...maybe he could start? Maybe this is a Derek Lowe situation, a much rarer cousin to the bad starter-turned-good reliever phenomenon in which a struggling reliever becomes a “finishes 3rd in Cy Young voting” level starter?

Trevor Rosenthal as starter, while an undeniably interesting idea, is very unlikely. The Cardinals will enter the 2016-17 offseason with six established starting pitchers under contract (in alphabetical and therefore non-judgmental order: Jaime Garcia, Mike Leake, Lance Lynn, Carlos Martinez, Michael Wacha, Adam Wainwright) and several other options waiting in the wings (Alex Reyes, Luke Weaver, Tyler Lyons, Tim Cooney, Marco Gonzales). On a lesser team, Rosenthal might be worth the risk in the rotation, but the preponderance of good-if-not-great options makes this less likely.

However, what if Trevor Rosenthal is capable of being a middle reliever who can go a few innings at a time, not necessarily in the obligatory “save situation” but in high-leverage situations whenever such situations arise? From the modernization of the “closer” position until very recently (like, “within the last year or so” recently), this was the territory for bullpen depth, not pitchers such as Trevor Rosenthal with star potential. But we are living in a post-Andrew Miller world.

Well, not exactly, since Andrew Miller is still around and is still spectacular. But by virtue of circumstance, Miller has been able to redefine the 21st century relief superstar. Miller (like Rosenthal, a former starting pitcher) was an effective reliever for the Boston Red Sox and Baltimore Orioles but was relegated to non-closer duty due to the presence of Koji Uehara and Zach Britton, respectively.

In 2015, upon signing with the New York Yankees, Miller became the closer and, for a time, the prototype for what I’m now labeling the Andrew Miller style of reliever was his teammate, Dellin Betances. With Miller pigeonholed into the standard duties of a closer, Betances was more fluid—multiple inning outings, high-leverage appearances regardless of inning—and more valuable (2.4 fWAR to Miller’s 2.0 fWAR).

But in 2016, when the Yankees acquired fireballer/miscreant Aroldis Chapman, Miller was able to pitch in different situations. At the trade deadline, he was acquired by the Cleveland Indians, who already had a terrific closer in Cody Allen, thus allowing Miller to continue to reinvent modern perceptions of what an elite reliever is.

Indians manager Terry Francona took this to extreme levels in the postseason—Miller finished second behind only Corey Kluber among the team’s pitchers in innings during the playoffs. He won ALCS MVP. And he did all of this while earning as many saves in the postseason as Clayton Kershaw.

A critical variable here is Cody Allen. Terry Francona is a manager who has shown a willingness to take risks, but the presence of an established 9th inning guy meant that Francona could justify to traditionalists (in the sense of a tradition not yet thirty years old—reserving your best reliever for the 9th inning in order to accumulate one stat) that he was keeping a star in that role. Even if it were statistically the right thing to do, it would be much more difficult to pitch Andrew Miller in the 5th or 6th inning knowing the potential backlash of Mike Clevinger pitching the ninth.

Mike Matheny has a standard enough approach to bullpen management that it is unlikely he would be willing to implement Trevor Rosenthal outside of the ninth inning if he had no comparable options. But with the emergence of Seung Hwan Oh in 2016, the Cardinals seemingly have the flexibility that Rosenthal could pitch in a multitude of situations.

And with a growing sense that relievers are being paid not for saves but for general dominance, the Cardinals could manage to justify Rosenthal’s increasing arbitration costs by increasing his inning total, though not to nearly a starter’s level, and giving the Cardinals their own version of a dominant, World Series-worthy bullpen without having to invest in an expensive free agent or deadline acquisition.