In Pursuit of True Justus: The Case of Sheffield’s Encouraging Inning

Searching for truth and Justus in a Post-Paxton world

Let’s get down to business, shall we?

The prosecution has accused Justus Sheffield of being “destined for the bullpen.” The case is precedented by a rich history of similar fastball/slider pitchers with inconsistent command. The defendant claims that the “command will come” and the third pitch “will develop” enough to be a major league starter. He is younger than AA prospect Justin Dunn, they insist. (I’m already tired of writing like this, but the law is the law).

I present Justus Sheffield’s 4/26 start for consideration. People of the jury, we all agree that Justus Sheffield being a good starting pitcher is incredibly important to this franchise for maximizing its competitive window (and for shutting up Yankees fans, the highest priority).

In the Spring, Sheffield was everything we could want: aggressive, strike-throwing, changeup developing. Like some of you, I believed he was fixed.

However, in Tacoma Sheffield struggled with the same problem he had always struggled with: command and consistency (6.86 BB/9, 7.63 FIP).

Despite the struggles, Justus was called upon to make his Mariner debut. He looked much like the Sheffield who’d been pitching in the minors, throwing only 39 of 74 pitches for strikes and walking four in three innings. His pitch breakdown was 65% fastball, 30% slider, and 5% changeup. This alone does not bode well for the Sheffield as Starter case.

However, if we take a more granular look at the tiniest of sample sizes, we see something intriguing:

The outlier here is the third inning, concluded with a brisk 16 pitches, 62.5% of them for strikes. That’s great, as league average strike rate is ~60%. In that inning alone he threw two of his four changeups, mixed his pitches, and threw strikes. In that inning, he performed like Justus will need to perform to hack it as a starter.

Let’s take a look at how it happened.

I trace the third inning success to the last batter of the second. After struggling with command in the second inning and walking Logan Forsythe on this not-great slider:

Notably Progressive pitching coach Paul Davis made a visit to the mound.

As a self-proclaimed body language expert, I have transcribed the gist:

Davis: you look nervous, are you nervous?

Sheffield: no

Davis: OK, it’s just that you seem nervous

Sheffield: I’m good

Davis: It’s ok to admit that you’re nervous.

Sheffield: I…

Davis: Remember I am not some stuffy coach. I am your spiritual guide. Let my guide your spirit, Justus.

Sheffield: kicks dirt I might be a little nervous.

Davis: KNEW IT! Stop. Stop being afraid. I command you as spirit guide.

Sheffield:

Give or take.

Before Paul whispered in Sheffield’s ear, through his first 26 pitches he’d thrown 5 sliders for 0 strikes. After? 6 consecutive sliders and 5 strikes.

The next inning, Sheffield was efficient and retired Jeff Mathis, Delino DeShields, and Rougned Odor. Not exactly a murderers row, but they’re still major league hitters. Here’s how he attacked them.

Sheffield v. Mathis

Starts with a slider to get ahead:

And then a fastball that misses its spot but mimics the location of the slider enough to get a swinging strike.

He misses low with the next fastball before coming back with one of four changeups on the day.

It hangs up in the zone and looks crushable, but it’s a third-look for Mathis in this at bat and it kept the ball off the barrel.

Sheffield v. DeShields

In the next at bat, Sheffield struggles to get ahead of Delino DeShields, missing up and away with a fastball. His 1-0 pitch is another changeup.

Not a great result, but not a terrible pitch—with Statcast spotting it just below the zone (Omar Narvaez does not try particularly hard to frame it) and good, late sink. Justus lobs fastballs at him until a low 3-2 fastball in the same location as the change is topped to the left side.

Sheffield v. Odor

Our last batter was the only non-Danny-Santana strikeout. Here’s the plot.

We can see the first slider (yellow) started off the AB in the zone for a swinging strike. Similar to the Mathis at bat, and Santana before him, Justus gets away with an elevated slider because it catches the hitter off-guard. He comes back with the fastball (red) up in the zone for a foul. After trying to get him to chase the next two sliders, Sheffield strikes out Odor with a breaking pitch up in the zone that Statcast has listed as a curveball. Let’s see.

There’s not a lot of movement on this failed slider, and the location is quite literally the opposite of the intended target, but when you put a hitter on the defensive by working ahead and showing them different looks around the strike zone, it helps cover up mistakes.

So that was it. The good inning. It required Sheffield to pitch ahead (two of his six first pitch strikes on the day) mix all his pitches, and keep the ball in the strike zone. This is a fairly obvious path to success and I’m sure it was always the intention, but it’s something Sheffield wasn’t able to do consistently in the other innings.

Not only that, but the consistency of the inconsistency was worrisome. Here’s his strike plot, which you’ve seen before:

Justus has a pattern. He misses with fastball up arm side, and misses with the slider down glove side. The inability to move the ball around the zone has also led to predictable pitch sequencing.

The inability to throw a slider behind in the count or a fastball when ahead will severely limit Sheffield’s ability to get through a line-up multiple times. Again this is a tiny sample, but he only has 2.5 pitches to work with right now, so his options are already limited.

Let’s look at the consistency of his pitches through a different lens: Spin Axis and Velocity

The fastballs and changeups are bunched reasonably tightly, meaning their spin axis was relatively consistent, even if the velocity changed. I wasn’t sure what an average plot looked like (I’m new to this!), so I went through Brooks and looked at other pitchers with slider-heavy repertoires. No one’s looked as wild as this. This is wild.

Those sliders resemble a connect-the-dots of a shark.

Those sliders vary in axis from 150-350 degrees. But the movement of the sliders are not radically different as you (I) might expect.

Here’s the thing, this might be intentional. Sheffield’s breaking ball tripped up Statcast and it might be a way to give hitters different breaks in order to artificially expand his arsenal (as Elias used to do with his arm angles). Here’s the other thing, though: maybe he should stop. Without consistent spin and break, locating becomes that much more difficult. If Sheffield is going to live with two pitches the third time through the lineup, he’s got to be able to throw them both for strikes when behind in the count.

OK, so in the case of the Sheffield v. Scouting Community, we still don’t have enough evidence to make a determination one way or the other. It was three innings. But with all the pressure on Sheffield to perform as a starter in the Post-Paxton era, it is concerning that the pitch mix and wildness showed up exactly in the way we’d hoped he’d fixed. He’ll have to trust that (very hard, 88mph) changeup and get ahead of hitters to realize his potential.

If he doesn’t develop the changeup and keeps throwing two pitches 95% of the time, there’s still a chance he could cut it as a starter. I look at Brad Keller of the Royals who is not great, but throws 65% fastballs and 30% sliders with shaky command like Sheffield (though he also throws a sinker). Or maybe there’s a Patrick Corbin lurking in there, who has made his money throwing two pitches outside of the strike zone that hitters can’t seem to lay off. But the list of successful two-pitch starters is very short.

If Sheffield is going to avoid the bullpen, he needs to pitch like he did in the third inning more consistently. While we aren’t close to a final judgement yet, the case for Justus Sheffield as mid-to-top of the rotation starter just got harder, not easier, to defend.