June 17, 2015
Reflections on a Golden Changeup
I wasn’t a Lance McCullers fan. When he was called up, I wasn’t expecting great things. He had middle-of-the-rotation potential, but I thought his more realistic role was back-end starter or late-inning reliever. I wasn’t alone.
In April of last year Chris Rodriguez saw McCullers twice, giving him a 60 OFP ceiling and a 55 realistic grade as an eighth-inning reliever. Ron Shah saw him in July and had the same outlook. McCullers was the seventh-best prospect on the Astros’ Top 10 according to the BP Prospect Staff, though you could argue that says more about the Houston farm than McCullers.
McCullers was called up in mid-May, and Chris Crawford had this to say about him:
The short version, in other words, is that the underlying story has always been the same: McCullers had two 70-grade pitches, but struggled with control and getting left-handed hitters out. His dynamic fastball-curveball combination could play perfectly well in relief, but as a starter that limited two-pitch repertoire would make life difficult.
So we’ve established that I’m not crazy. The industry consensus (including Kiley McDaniel’s grades in addition to the reports listed above) on the tools was:
That looks like a reliever to me.
I was wrong, though. We all were. While it’s only been 30-plus innings so far in MLB, McCullers has been dominant. He has a DRA of 2.62 and a cFIP of 81. Both are in the top 25 among starters with at least 30 IP this season.
That's from a tweet by Cespedes Family BBQ in which they wondered if that was a sinker, as MLB Advanced Media had tagged it in Gameday. McCullers cleared that up: “Believe it or not... That's my CH lol”
That’s not a 40-grade changeup. Not even close.
Realistically, the pitch is probably more of a throttled sinker, according to BP’s Harry Pavlidis, but if McCullers uses the pitch as a changeup, then it's a changeup. And if McCullers throws that changeup, then his fastball-curve repertoire just got a huge shot in the arm.
He’s not throwing the pitch very often, just about 11 percent of the time, but even having it in his back pocket is valuable. As you'd expect, he’s used it much more heavily against left-handed hitters (they’ve seen 84 percent of them). The new changeup is a fitting addition to his filthy repertoire: His whiff rate is over 40 percent on the pitch, matching the dominance of his curveball. The changeup is making McCullers look like an ace, or at least a potential future one.
A GIF is a GIF, though, and baseball is a game of repetition. How does the pitch shown above compare to McCullers’ usual changeup? Below is a scatter plot of all of McCullers’ changes with the one from the Melky Cabrera at-bat highlighted and annotated:
(The vertical movement in the chart above includes the effect of gravity on the pitch.)
The pitch at which Melky Cabrera hopelessly flailed was the fastest changeup McCullers has thrown this season, but his average change is still quite hard: The 90.5 mph mean puts McCullers in the 99th percentile for changeup velocity among starters. Only Aaron Sanchez throws his harder. McCullers does not throw his fastball 100 mph, so that changeup velocity means that he has less differential between his fastball and change than typical, or than is typically considered ideal, but note that his approximately 5 mph differential is about the same as we've seen from the current version of Zack Greinke, who from 2012 to present has a differential a hair under 5 mph. Greinke, you'll recall, has been one of the 10 best starters in baseball over that span.
Moving on: McCullers’ typical change has about eight inches of arm-side run, just a bit above the median starting pitcher (55th percentile), but exceptional drop (a “rise” of only 1.8 inches), clocking in with roughly 3.2 inches more vertical movement than the average starter's changeup, placing him in the 89th percentile. These figures are closer to those of Charlie Morton's sinker (2013–15: 92.5 mph, 9.7 inches of run, and 1.8 inches of “rise”) than Morton would probably like to admit. Greinke's changeup + Morton's sinker = pretty good.
Does McCullers need to throw every change like the one he used to victimize Cabrera, or is his average changeup good enough for him to exceed the 55 realistic potential we slapped on him prior to this year? The first step in answering that question is to look at how the Cabrera change compares to McCullers' average offering: It was roughly 3 mph faster than his average and had about four more inches of drop, though it was around his mean horizontal movement. That said, you could argue that it wasn’t even the best changeup he has thrown this season. McCullers has thrown a few 90-plus mph changeups with more drop and many 90-plus changeups with more horizontal movement than the one Cabrera whiffed on.
If McCullers continues to hone his command, his average changeup should be a good enough offering to allow him to maintain his breakout. The raw talent is there. The only question is whether McCullers will be able to continually tweak and improve the pitch as a main component of his arsenal.
About 10 days ago, Evan Drellich quoted McCullers describing a conversation with his High-A pitching coach, Donnie Alexander:
Hey, we’re going to commit to this changeup thing. It’s going to cost me runs, it’s going to cost me walks. It’s going to cost me some balls flying out of the park, but it’s going to be better for me down the road.
That piece also includes a number of other interesting tidbits:
The promising note here is that McCullers is dedicated to improving as a professional by exploring and committing to his changeup. He’s also shown that he’s not willing to give up easily, trying out several grips before settling on one, a modified claw grip, that seems to work for him.
That’s not to say that everything is smooth sailing from here on out. The control of his change is an issue: Right now it’s a ball more than 55 percent of the time. More broadly, he’s currently walking fewer batters than he did at Double-A, and nearly half as many as he did in High-A, so there’s potential for some regression there. He’s also currently sporting a .241 BABIP that would rank among the 20 lowest for a qualified starter this century were it carried over a full season.
It’s also important to note that at the end of the day, the sample sizes we’re dealing with are incredibly small. When you’re looking for a breakout, though, you need to be willing to make a judgment on this level of information. Once McCullers has thrown 200 changeups, dominating (or not) through the rest of this season, it’ll be too late to peg him as a breakout pitcher. The tools and the soft factors support the conclusion: McCullers has the raw stuff to get the job done and he has clearly shown an attitude and approach that suggests he will continually improve and adjust as necessary.
I misspoke earlier. Nobody really got McCullers wrong. His old show-me changeup wasn’t going to make him a successful starter. Nobody saw the new pitch. McCullers had a hole in his performance and he experimented until he found the right changeup grip to fill it. It’s not a magic bullet, but even a league-average change would make an immense difference for him. Based on what we’ve seen, this new changeup has the potential to be much better than a league-average pitch.
This is precisely why it’s incredibly difficult to predict breakouts. When you look at the most recent reports on McCullers, review his PECOTA projections, and consider the realities of a two-pitch pitcher in today’s game, all signs point to a late-inning reliever. The reality, though, is that McCullers and those close to him have been working on fixing those flaws. Sometimes players surprise you, but when you look back, it’s only surprising to us on the outside looking in, who don’t get the whole story until well after it’s been written.