Return of the changeup

If you can't stick it, slip it

For those who were glued to the WBC, I’d like to bring you up to speed on a developing non-story from Spring Training camps. I’ll summarize it with a few headlines:

• Why Mets star Justin Verlander is emulating Nolan Ryan and dabbling with a changeup

• Paul Blackburn experiments with changeup in spring debut

• Expect to see more of Kopech's 'unreal' changeup

• Mariners’ Logan Gilbert adding a split-finger: ‘It’s going to be a phenomenal pitch’

• New changeup has Boyd boasting big results

Changeups. Pitchers who didn’t touch them last year are apparently breaking them out now. Usually, I’d say not to take the reports too seriously - every spring, pitchers claim they’re throwing more changeups and then don’t follow through. Case in point, Justin Verlander may indeed be emulating Nolan Ryan, but he has only thrown 8 changeups in 154 tagged pitches in Spring Training this year.

So is there any reason to believe the headlines? Is Verlander trolling us? Maybe. But either way, I’d like to look ahead and suggest that this year, a changeup revolution might really be coming.

Is this year different?

This year does feel a little different. While avoiding a slog through messy Spring Training data¹, here are two reasons why we should take pitchers seriously this year when they talk about throwing more changeups.

1. Hitters have been spun to death

Most of these pitchers claim to be working on changeups to give hitters a different look from the breaking ball-heavy mixes that are featured in today’s game. Given how many breaking balls we see nowadays, I do think there is something to that argument. Hitters gameplan just like pitchers do and can usually key in on a preferred pitch/location when they’re facing a mix that only breaks one direction (say, a 4-seam/slider combo). Chris Bassitt’s mix fits this mold and he thinks changeups are his solution:

[A]s he’s continued developing his cutter, Bassitt’s noticed left-handed hitters cheating to the pitch and trying to yank it into right-field on him. His answer to that adjustment will be throwing those left-handed hitters more changeups.

Bassitt throws a cutter, but the same logic could be applied to sliders. And it is hard to overstate how significant sliders have become. In a decade, slider usage has increased by 70%. This upward trajectory has pushed the slider into “primary pitch” territory. According to Pitch Info, more sliders were thrown last year than 2-seam fastballs were.

There’s a recent development that could kill the slider’s upward trajectory, though: a ban on a popular ingredient that makes a slider slide.

2. A sticky crackdown

Another headline from last week:

Can MLB stop ‘sticky stuff’ use? Players weigh in on new enforcement measures

Here’s the biggest reason the changeup might be about to make a comeback. “Sticky” makes a slider spin tighter and break more than it otherwise would, and MLB is doubling down on an effort to remove it from the game.

Where would a post-sticky world leave pitchers? Is there a pitch that thrives on a lack of spin? Ask Kevin Gausman:

“If I use ‘sticky’ on my split, it cuts,” he said. “I don’t know where it’s going. So I can’t use ‘sticky.’ Now, if I was a guy that primarily threw 40 to 50 percent sliders, I would probably be trying to search for some way to find that (extra grip)…

Gausman technically throws a split, but the changeup falls in the same low-spin, non-sticky, offspeed family. If a pitcher falls in the “40 to 50 percent slider” group that Gausman is referring to, it might be time to learn one. The work that pitchers are doing with changeups to “add another look” might also be described as “building an insurance plan.”

But changeups are hard to throw

Why do pitchers tend not to follow through on Spring Training changeup promises? Well, for one, the pitch is a really difficult one to learn (even over a several month span). The discomfort starts with the grip - most are uncomfortable - and extends to the arm/wrist action. Most traditional changeups that get a lot of sink and fade (toward a same-handed hitter) require heavy wrist/arm pronation.

While 4-seam fastballs require a pitcher to get “behind” the ball at release and breaking balls are produced by movement “through/around” the outside of the ball, a traditional changeup “turns over” the hand through the inside of the ball. For pitchers who don’t already pronate on another pitch (most any pitcher who doesn’t throw a heavy sinker), that action can be awkward.

Grayson Rodriguez pronating and flashing his palm toward the third-base dugout. Ouch. (per FanGraphs Prospects)

Getting the action right is hard enough. When pitchers come into Spring Training with a changeup goal, the next stage requires in-game execution. Because the changeup is such a “feel” pitch and so different from anything else, any newfound comfort can disappear the instant after a fastball or breaking ball is thrown.

That “feel” component is why you often see pitchers double-up on changeups; that is, throw a second right after executing a good one. Verlander’s 8 changeups that have been tagged in Spring Training starts include one back-to-back sequence and one back-to-back-to-back sequence.

What if there was an easier way?

In prior years, Matthew Boyd was one of the many pitchers who found it uncomfortable to throw a changeup because it was so different from the rest of his mix. Boyd had this to say in the “big results” interview I cited at the top of this post:

Not only is it physically different for my arm, but it’s hard to repeat something differently when you’re doing three [other] pitches with one way.

But Boyd may have found a loophole. What if there was a way to throw a changeup without heavy pronation?

So I throw a changeup just like a slider now, but using essentially the smooth part of the baseball to create no drag on one side, but seam is on the other side. And because of that, I get more movement than I did before, but the pattern of how my wrist is moving is like the other pitches. So it allows for the other pitches to be more consistent.

The interview goes on to suggest that Boyd’s new changeup benefits from the concept “seam-shifted wake.” I’m no expert on the physical dynamics behind this, so I won’t attempt to describe the mechanics in detail². I’ll do my best to simplify it here, though.

Those who are familiar with seam-shifted wake already know that the name provides a clue. The seam orientation of a baseball can affect the symmetry of its wake, and an asymmetrical wake can create additional movement. Seam-shifted wake explains why a knuckleball doesn’t spin but does move (unpredictably). On other pitches, seam-shifted wake can be honed to create predictable seam-induced movement.

In other words, seam-shifted wake can allow a pitcher to add sinking, fading movement to a changeup without pronating. The best visual comparison I can find between changeups that do and don’t benefit from seam-shifted wake was put together by Rob Friedman. Pay attention to the wrist movement and seams in each pitch:

Rob Friedman @PitchingNinja

Devin Williams Changeup release vs. Pablo López's Changeup release. Two different approaches. Both filthy.

Devin Williams is on the left; he is pronating so hard that it hurts just to watch. On the right is Pablo Lopez’s seam-shifted changeup, which isn’t heavily pronated but is pushed down and toward this right-handed hitter (partly because of the position of the seam on the top of the ball). The seams of Lopez’s changeup are doing the same type of work that Williams’ wrist is doing³.

This is how I read Boyd’s comments: he’s found a 2-seam changeup grip that sits opposite to his slider grip. I don’t know if it is thrown exactly like his slider (a pitch gripped like a slider and thrown like a slider should probably move like a slider), but I think he isn’t aggressively pronating on it.

Here’s one from last Saturday:

That had some good sink and fade on it. Check out Boyd’s hand just after release:

It doesn’t look like aggressive pronation to me. Boyd’s pointer/index fingers are extended and we can see his palm - it looks like the finish to a breaking ball. That’s pretty cool.

Who else is doing it?

Unfortunately, there is no surefire way to identify a pitch that benefits from seam-shifted wake with publicly available data. But we can get to a pretty good starting point by looking for gaps between a changeup’s actual spin axis (as captured by a camera) to an estimate of spin axis based on pitch movement.

The list below includes all 2022 changeups for which the observed spin axis was, by the hands of a clock, at least an hour different than the inferred spin axis. I wouldn’t get hung up on what an hour on the clock means; the idea is just to pick out pitchers like Lopez and Boyd - those who are likely leveraging the concept of seam-shifted wake:

*min. 100 changeups thrown

Logan Webb knows how well the pitch works when it works. He also has expressed how tricky it is to get the pitch right. A slight misalignment of the seams can turn the pitch into a beach ball, but the risk is worth the reward for him:

"When you get that seam-shifted wake the right way, it comes out and at the last second: bang," Webb said. "When it's not, it's a little slower and gives the hitter a better chance.”

Webb and Boyd overlapped briefly last year in San Francisco. Both presumably worked with VP of Pitching Development Brian Bannister (who is a big believer in seam-shifted wake). Boyd’s time with the Giants and former GM Scott Harris seems to have made an impression on the pitcher. Boyd followed Harris to Detroit this offseason.

Research on seam-shifted wake first hit the baseball world in 2020⁴, but only recently are we seeing signs that pitchers have begun embracing the concept as it applies to a changeup. The list above includes 12 pitchers whose changeups exceeded an observed-to-inferred spin axis gap of an hour (one that likely comes from seam-shifted wake)⁵. A dozen pitchers might not seem like many, but it represents a jump from prior seasons - there were only eight pitchers whose changeups met this criteria in 2021.

Time to head to the lab

If and when the stickiest substances that are turbocharging breaking balls find their way out of the game in 2023, the changeup might find its moment. For those pitchers who previously stayed away from the traditional changeup because its arm action feels too different from snapping off a breaking ball, seam orientation might provide a way forward. For some, it’s already doing so.

If you’re ever wondering how your favorite team could possibly use a multi-million dollar pitch design lab, this is one of the ways. Developing a seam-shifted wake changeup requires a high-speed camera system and multiple types of tracking (like a radar-based Trackman unit and an optical Rapsodo unit). Put all of those tools together, and with hours of fine-tuning, even a spin-heavy mix like Boyd’s can add a changeup with minimal disruption.

With Spring Training coming to a close and the time for development winding down, we’re about to find out who has actually built up a breaking ball insurance plan.

Have a great week - only one more until Opening Day.

---

Note: If you’re interested in identifying seam-shifted wake in real time in 2023, Baseball Savant manages this leaderboard (the source I used to pull the table above). Just remember that the complete picture of seam-shifted wake is a little more complicated than the leaderboard estimates.

1. We do have some Spring Training data, but it isn’t complete - Statcast data is available for some Spring games and not others. Even with complete game data, we shouldn’t make assumptions about 2023 regular season pitch usage based solely on these games. A lot of pitchers work fastball-heavy in early outings and gradually introduce the full mix.

2. Barton Smith and his research team began producing public research on seam-shifted wake in 2020. Smith credits student Andrew Smith for the seam-shifted wake term. This article and this presentation are great primers.

3. Changeups like Lopez’s are thrown with a “2-seam” grip. It is also possible to throw get seam-shifted wake to push a changeup away from a same-handed hitter by using a 4-seam grip, but that isn’t usually preferred (unless you’re Kyle Hendricks).

4. I think (that primer article above)

5. Because seam-shifted wake movement isn’t an all-or-nothing thing, I don’t mean to suggest that there is one hard observed-inferred spin axis gap that signifies seam-shifted wake. I just chose a cutoff that I thought would help pick out some good examples.