Maybe Justus Sheffield made a bad tweak

It definitely wasn’t a good tweak

When I arrived at Lookout Landing, the first thing that I did was write about Justus Sheffield. I wanted to write about Yusei Kikuchi, but that article topic had already been claimed. So, instead, I wrote about how Sheffield leaned into the sink of his four-seam fastball and opted to start throwing a sinker instead. I thought it might be something that complements his four-seam fastball, but that hasn’t been the case at all. His four-seamer has been nearly non-existent. And the results have been incredibly mixed.

Consider his numbers as a starter since switching to a sinker:

I realize that I just said that the results have been mixed. This table is hardly mixed! By these metrics, Sheffield has gotten worse across the board. That comes with the caveat that his 2019 is made up of just 36 innings pitched, but it doesn’t take long for metrics like K-BB% and CSW to tell us meaningful information. I think the argument can be made that moving to a sinker was neutral, but I think a more convincing argument is that it made him worse.

It’s instructive to think about why that is. If Sheffield was going to switch to a sinker, the end goal is that he gets better as a pitcher. I mean, obviously. What you probably want to see is at least one of two things. The first is that it’s a better pitch in isolation in some regard. That, or that it makes his other offerings stronger.

Despite the aforementioned sample size issue, the first question can be answered pretty simply. Sheffield has seen a marginal increase in CSW, going from 26.4% with his fastball in 2019 to 27.9% with his sinker from 2020 to 2021. That’s an improvement, but a marginal one at best.

Where he hasn’t improved is in contact quality. Again, a sample of 62 batted ball events isn’t the most meaningful sample, but Sheffield had a .344 xwOBAcon on his fastball in 2019. Since 2020, his sinker xwOBAcon is .401, which is a fair amount worse than the league average .371 xwOBAcon on sinkers since 2019. This isn’t to say that his fastball is superior to his sinker in contact quality — these numbers take a long time to become reliable. I do think it’s fair to say, though, that it’s clearly not worse. If it isn’t a better pitch in isolation, then it better help his other offerings play up.

Based on inferred spin axis data, his sinker is probably better than his fastball at mirroring the spin of his slider. But it’s a little more complicated than that. It might have had other reverberating effects.

Sheffield’s pitches over the past few years, by horizontal and vertical movement:

What differences are apparent are that Sheffield’s sinker gets much more arm-side movement than his fastball, with a little extra sink too. It’s so heavy that it overlaps in movement with his changeup. That wasn’t the only change though. He also shaved about three miles per hour off of his slider, and you may notice that he’s added several inches of glove-side movement.

That’s one potential explanation for the degradation of Sheffield’s slider; it’s that, because of the heavy sink of his sinker, his slider has had difficulty separating from it. In terms of vertical movement without gravity folded in, Sheffield’s fastball and slider had a differential of 7.06 inches, while his 2020 and 2021 figures are 5.38 and 6.48, respectively. So there’s less vertical separation, but in fairness, Sheffield widened his horizontal movement gap and velocity gap. Maybe it’s that they’re not separating enough vertically. Maybe his slider isn’t firm enough now. I think the stronger argument to be made is location-based.

Moving to a sinker has changed how Sheffield has located his pitches. Specifically, I want to focus on his fastball and slider locations.

First, his fastballs, between 2019 and 2020-2021:

And then Sheffield’s slider location, between 2019 and 2020-21:

As you might expect with a pitcher transitioning from a four-seam fastball to a sinker, Sheffield is pitching down in the zone more often. That’s meant that he’s had to expand the zone vertically with his slider too. He’s started throwing it farther out of the zone, and you may notice that he toyed with using it as a get-me-over pitch in the zone in 2020, which isn’t so relevant anymore.

If you’re the hitter, there’s not much to look for at the top of the zone. When Sheffield made the switch, I was optimistic that he would throw his sinkers at the bottom of the zone, and his four-seamers more elevated a la Luis Castillo. That’s pretty much what he’s done, but he’s done it so infrequently that hitters don’t have any issue making contact with whatever he’s throwing.

So there are several issues as is, both by process and outcome. You may have noticed that Sheffield has always thrown his fastballs — both his four-seamer and sinker alike — out over the plate. Part of the logic behind his switch was how natural the sinker felt for him. Scott Servais and Pete Woodworth both praised his command of his sinker when he debuted it, but that doesn’t appear to be the case. The percentage of fastballs he throws over the heart of the plate has leaped from 27.3% in 2019 to 31.0% since 2020. That doesn’t speak to strong command of his sinker.

The other thing is that slider has regressed significantly. Sheffield’s in-zone swinging-strike percentage has fallen from 19.4% in 2019 to 8.4% since moving to his sinker. Outside of the zone, it’s only fallen from 20.4% to 18.8%, but it’s important to note that swings and misses inside the zone accounted for 31.6% of his slider’s swings and misses in 2019. That number continues to fall drastically, and for a pitch with a 34.3% chase rate since 2019, he needs those in-zone whiffs back. He’s leaving them up in the zone this year, and he’s getting torched as a result.

There are several paths forward for Sheffield, most of which aren’t mutually exclusive. The simplest is that he takes his four-seam fastball and throws it at the top of the zone more. I’m not convinced this would be the thing to completely correct his struggles, but I think it’d provide a boost. And he can use all the help he can get.

The most compelling solution for me is that Justus moves back to his four-seam fastball as his primary fastball. Given my stance that his sinker isn’t superior to his four-seamer, the focus here shouldn’t be his fastballs at all, but how they affect his slider. If the past two years are any indication, his slider is better when he’s more heavily featuring a four-seam fastball. It wouldn’t be a bad idea to add a cutter to his repertoire over the offseason, either.

In his first two seasons, he’s pretty much looked the part of an average MLB starting pitcher — which is plenty valuable — but more recently, he’s bordered unwatchable. It’s important to remember that Sheffield is just 25 years old. Sometimes I have to remind myself of that. It’s just that, more than anything, maybe a starting pitcher leaning on an approach made up of below-average command, a below-average strikeout and walk percentage, and average contact suppression was never going to work. His future as a starter certainly isn’t unsalvageable, but it sure is looking bleaker by the day.