Michael Soroka wasn’t sinking

Michael Soroka returned to the majors after nearly three years away on May 29, 2023. After two outings, and despite the Braves more-than-intimating that they wanted him to be up for good once he got the call back to the bigs, he was optioned back to Triple-A Gwinnett exactly a week later. There’s a lot to be said about Soroka’s brief two-start ride through Oakland and Phoenix, much of which rightly needs to focus on the Braves’ stubborn insistence that he persevere beyond the eighteenth batter of the game. But while the Braves definitely shouldn’t have forced Soroka in that situation, there’s a more-pressing problem for the kid: his sinker, plainly put, isn’t sinking. In fact, nothing is sinking, and that’s apparently a problem.

Before we talk about Soroka’s small-sample, two-start woes, it’s helpful to get a refresher on what made him so good that he was able to post a 4 fWAR season in 2019, which spanned the sixth through 34th starts of his career. One of those things was being able to outpitch his xFIP- (87) through some positive HR/FB stuff (78 FIP-), but you don’t run an 87 xFIP- for a whole season without executing a gameplan, and Soroka definitely did that.

This chart is Soroka in bubble (and spiky bubble) form.

The idea is clear: throw stuff that sinks. We’ve got a sinker that fades and sinks more than usual, a changeup that basically moves like the sinker but warbles around ten miles per hour slower, a four-seamer for some reason (much like Bruno, we continue not to talk about four-seamers from sinker guys), and a slider with really sharp drop, almost a one-plane curve. With this profile, you pitch down, and then you go even further down for whiffs. Did Soroka pitch that way? Yeah, pretty much.

Against righties (ignore the percentages in the image, which are for all batters), Soroka was 50/24/17/10 sinker/slider/four-seamer/changeup, in the following locations.

The sinker was your usual jam/rollover pitch, on the inner edge. The slider was your usual chase finagler, though Soroka sometimes used it as a get-me-over. The four-seamer arguably didn’t serve much of a purpose, but was thrown in the same as another jam job that could get pop-ups if hitters expected it to be a sinker. (The spin on Soroka’s two fastballs was similar, and both had slight spin deflection in opposite directions.) The changeup wasn’t used much against righties and was largely just a surprise/panic pitch.

Against lefties, Soroka leveraged more of his arsenal: 38/24/22/16 sinker/slider/four-seamer/changeup, in the following locations:

The sinker was still thrown as a roll-over pitch, but to the same gloveside edge. The slider profile changed, working in different respects as either a “too far to pull” pitch, a pitch for whiffs below the zone, or even a Charlie Morton-esque backfoot (or back quad, at times) special. The four-seamer was a get-me-over that sometimes fished for up-and-in whiffs. The changeup was essentially the sinker but less frequent (and slower).

How well did this plan work? Pretty well. At worst, if you really care about contact management, it did better-than-average on an xwOBACON basis. Batters slammed pitch after pitch (namely, the sinker, but also the slider to some extent) into the ground, generally on the pull side where the defensive positioning was easy to align. Soroka didn’t strike many guys out, but he minimized to the walks to the point where it didn’t matter too much, and pitching to contact rarely hurt him. Righties didn’t have an xwOBA greater than .302 against any of his four pitches; lefties teed off on his two fastballs but struggled mightily to make contact against his slider and changeup. Soroka was actually, in total, kind of naff against lefties in his awesome full year, but given that he had an easy time with the other 60 percent of the batters he faced, it wasn’t a huge deal.

So, “deep” into a two-start sample in 2023, with a completely reworked delivery and everything else, what has Soroka’s essence been? Well, it’s not great, for starters (or relievers, or anyone).

The sinker... isn’t sinking. The changeup also isn’t sinking. In fact, the changeup doesn’t really make much sense — it’s way slower than the average changeup, which means gravity alone should force it to drop way more (these arsenal charts are not adjusted for velocity/gravity), but it’s just. not. dropping. The slider seems same-ish. The four-seamer being more generic, and less of a weirdo “hey I throw a four-seamer despite the massive natural sink my arm action and grip impart onto the ball” offering underscores the fact that for all of his changes, Soroka doesn’t actually make the ball go down anymore.

On its own, this isn’t a giant problem. Soroka’s slider has fine shape, and he’s thrown a ton of them. While the xwOBA looks nasty (.454 on the year), that’s really just walks, as it has an xwOBACON of .270 on a whopping three balls in play (two weak flares for hits, a weak groundout, none hit above 84 mph off the bat). You could throw a ton of well-located sliders with some random stuff for a change of pace/spin, and probably survive.

Unfortunately, Soroka hasn’t located, and he doesn’t have random stuff. He’s had actively detrimental, problematic stuff. Let’s go to the pitch location chart, this time for all batters, because of the sample size.

The slider is not generally in the zone. The four-seamer has the same idea as it did before, but lacks “rise,” and is thrown in the zone. Corbin Carroll homered on a four-seamer that just wasn’t up enough. But the real calamity is the sinker. Look at that location. This is a pitch that’s supposed to sink, and is being thrown essentially down the middle. If it sinks, it actually ends up easier to hit. The changeup has the same problem: it’s definitely not sinking, and it’s being hung by default.

The net result is a .394 xwOBACON on the sinker, a .350 xwOBACON on the four-seamer, and a .332 xwOBACON on the changeup. None of those present an escape option: their shapes, for lack of a better word, currently suuuuuuuck, and their locations are even worse than the pitch shape. Even if Soroka could locate the slider, he probably couldn’t survive throwing any of these other pitches, and he’s not locating the slider. The pitches still tunnel well, and there’s some seam-shifted wake on everything to play with, but those are all fringy toppings to an unappetizing, spoiled butter, moldy bread sandwich.

Now that Soroka’s been optioned, I guess he’ll try to figure out some way to exist in the majors with better location, or better stuff, or, well, anything that seems workable. Theoretically, if this is discoverable, there’s no reason why he couldn’t have discovered it in the majors, but I guess the Braves were unwilling to let him get hammered in the middle innings of games while he did so. The good news is that there is a way out. Better slider command. Making things sink. Hell, even making the changeup go back to a sensible, “natural” rate of sink could do the trick. Not throwing the four-seamer if it’s going to pepper the zone. There are options. It’s just that Soroka, for all of his Herculean effort in making it this far, has even more work to do.