Where is the line that tells you whether a pitcher is a starter or reliever? It’s subjective, but there are a few benchmarks that baseball folks have used over time to determine a pitcher’s role. Do they have more than two pitches? Do they have extreme splits? Does their velocity hold as their pitch count increases? Depending on the answer to any of these questions (and more), you’ll have a decent idea of a pitcher’s fate as a starter or reliever.

Sometimes, though, pitchers get moved to the bullpen simply because the other five options at any given time seem more viable, and not necessarily because they didn’t possess the skills to dish out six quality innings. The latter firmly describes Michael King.

King was a starter throughout his entire minor league career and even started a few games during his debut season in 2020 before settling into a long relief role in 2021 thanks to a crowded Yankees rotation. But the version of King we saw in the minors and those two seasons is not what he is now. For that, we have to reference his 2022 season, when he fully embraced his ability to supinate and started throwing his sweeper as his only breaking ball, upping the usage to 30%; the year prior, he only threw it 10% of the time. Before suffering a season-ending elbow fracture, he posted a 2.29 ERA and 2.23 FIP in 51 innings, delivering several multi-inning outings out of the bullpen.

On top of his sinker and sweeper, King displayed his ability to mix in his four-seamer and changeup at the right times, both of which have been crucial to his success as a starter; more on that later. That gave him a four-pitch mix where each pitch was at least average and more than one that was plus. That’s the exact kind of arsenal you’d expect a rotation arm to have, but coming off elbow surgery and bullpen success and facing another full starting five, the expectation wasn’t that he’d get the chance to pitch in that role.

Luckily for him and the Yankees, King started the season looking close to his elite form in ‘22 despite coming off elbow surgery. The one regression he did see was a lost tick of velocity across the board, but perhaps that was best for his health and longevity. A cross-body delivery like his can be volatile if the body isn’t optimized for it; maybe dialing it down a tick allowed him to manage the wear and tear of a full season better. Either way, his repertoire was still effective, and he proved it in April, delivering several multi-inning relief outings in which he threw as many as 40 pitches and often between 25–35.

At the time, the Yankees were being careful with King, giving him at least three days (and usually more) in between appearances; rarely did he throw with only two days of rest until June. Then, as the Yankees’ starting rotation was put to the test at multiple points in the summer, King’s presence became more important. He began to have longer outings and planned long relief days in late July and early August. By the end of August, he was being stretched out into deeper pitch counts. It’s possible that decision came more out of need than want, but it was the perfect time to experiment, as the Yankees had fallen out of the wild card race.

Before King’s August 24 start, his season-high pitch count was 44. From every start after that one, his cap increased:

On the surface, this looks very positive. It’s only been a month, but King has held his velocity through each of the starts, even as he’s hit the century mark in pitch count. Aside from the performance, looking for signs of fatigue is crucial to assessing his viability as a starter long term, and as he continues to work at this level of volume, it’s good to see him keep up the velocity, especially since this is the most he has thrown in one season since 2018.

To zoom in a little bit more on his velocity trends, let’s look at his averages by inning as a starter:

Keep in mind that we’re working with small samples here, and that King has only worked past five innings once. Even with that, the dropoff wasn’t too bad in his dominant 13-strikeout performance against Toronto. He mitigated it by varying the velocity between his two fastballs and properly working in his sweeper and changeup. As you work deeper into games and lose a little bit of juice, it makes sense to switch up locations, vary pitch speeds even within a given pitch, and mix in different pitches. That’s the beauty of being a four-pitch guy with a knack for dotting corners with your heater: You provide hitters with different looks and can stay unpredictable even as they’ve already seen you multiple times.

That brings me to the next key factor in analyzing King’s transition to starting: the pitch mix change.

Out of the bullpen, King often used his sweeper as his primary weapon, which made sense; he was only in for one trip through the order, and the pitch was often a good tool for getting ahead in the count or getting a whiff on the outer third, especially against right-handed batters. But that’s exactly it: the pitch is most effective in one specific part of the zone. He can’t manipulate the location as much as he can with his sinker. By relying more heavily on that sinker, he’s taken on the old school mantra of establishing his best fastball, then working off it. Although the pitch has negative run value as a starter, he needs it to keep hitters guessing locations and speeds. And while he does use his four-seamer quite frequently, don’t get it twisted: King is who he is because of his ability to command his sinker in every part of the zone.

That’s another reason why he was such a good candidate to start. Not many pitchers in baseball can command a Corey Kluber-esque frisbee sinker, but King is one of them, and that gives him a unique look relative to the rest of the league. Because of this, he can run sinkers over the back door to righties (and front door to lefties) better than any right-handed pitcher in baseball. Among all righties with 100 pitches on the glove-side third of the plate since the start of 2022, nobody has a higher called strike rate (74.8%) than King; the next highest is Brusdar Graterol at 67.9%. The cross-body movement paired with the pitch’s run makes it very difficult for hitters to identify it as a ball or strike. This unique trait is always at the forefront of hitter’s mind and forces them to wait a long time to make their swing decision.

You can see a perfect example of this in action in this at-bat against Bo Bichette from Wednesday night:

Pitch 1 (0-0, sinker)

Pitch 2 (0-1, sweeper)

Pitch 3 (1-1, four-seamer)

Pitch 4 (1-2, four-seamer)

Pitch 5 (1-2, sinker)

Bichette has elite feel for his barrel, and as a hitter who peppers the opposite field gap better than anybody in baseball, this should have been a good matchup for him. In theory, King’s arm-side run on the outer half might work perfectly for Bichette’s bat path, but execution like this makes that obsolete. King’s strength may be that pitch on the outer third, but he opted to run a sinker inside to start the at-bat. This is a prime example of knowing your hitter: bang him inside and force him to try to open to pull the ball. By following up with the sweeper, he tested whether Bichette would attempt to cheat for the inside heater and chase, and though Bichette held off, King got a good read on his approach.

The next swing is telling of how King’s execution on the previous two pitches kept Bichette guessing, as he got a center-cut four-seamer but was too late to get his barrel to it. The logical next step from King was to dial it up more to see if he could get a whiff, and that is exactly what he did.

Note that this pitch got up to 95.6 mph — three ticks harder than the sinker inside on the first pitch. This is something King has done very well in his move to the rotation. He wants the four-seamer to function as a firmer heater that can work up in the zone to make sure hitters don’t sit on the sink. Based on Bichette’s swing, that plan worked very well. After this, Bichette was forced to speed up; he hadn’t been on a single fastball all at-bat and needed to do something to make that happen, but he cheated too much on the inside sinker that was running inside and couldn’t hold up in time once he recognized it as a ball.

The thing is, these types of uncomfortable swings are the norm for King. He has the stuff and command to expose weaknesses in a variety of hitter types, even those that should be a bad theoretical matchup for him. To me, that’s what it means to be a starter.

King isn’t the type of guy who lights up a radar gun, but his stuff is wicked. If it helps put it into perspective, he is in the same territory of overall Stuff+ as Grayson Rodriguez and Dylan Cease. That’s good for 13th among all pitchers with at least 90 innings. It’s been clear for almost a year now that he has the stuff and command to warrant a starting rotation opportunity. He got his shot and didn’t look back; in 28.1 innings as a starter, he has a 0.72 FIP and 1.27 ERA. It’s safe to say his spot isn’t going anywhere.