Fellow FanGraphs writer Jeff Sullivan wrote about Matt Shoemaker last week, being one of the first (to my knowledge) to note that Shoemaker had recently ramped up the usage of his splitter.
Sullivan noted Shoemaker's increased use of the splitter in all counts. Which is great, because it's arguably his best pitch. His slider is good, too -- both induce an almost-equal percentage of whiffs per swing -- but it's the splitter that has coerced a meager .117 isolated power (ISO) in his Major League career. That's a big part of it. More splitters means fewer other things, and those other things, as Sullivan noted, have generally been bad.

There are other nuances in usage that piqued my interest that I'd like to disentangle, though. Let's borrow a couple of tables from Brooks Baseball:


Between his first seven starts and his last three starts (and excluding last night's start), Shoemaker doubled his splitter usage versus left-handed hitters and tripled it versus right-handed hitters. He never threw it for first-pitch strikes against righties but now does so 20 percent of the time. Against righties, that rate almost quadrupled. A good way to succeed as a pitcher is to work into advantageous counts. Throwing first-pitch strikes 73.6 percent of the time will do the trick. That Shoemaker uses his best pitch to do so further serves as a boon to his effectiveness.
Perhaps most notable, then, is Shoemaker's insistence on attacking hitters with his splitter when behind in the count. Earlier in the season, he blindly relied on his below-average four-seamer, a pitch that allowed a .302 batting average and .206 ISO last year. Now, he's trusting a pitch that has allowed an almost-spotless .027 ISO this year to turn counts in his favor.
As aforementioned, with more splitters comes fewer other pitches. Shoemaker has completely abandoned his curve, and he has eliminated his slider versus lefties. In Shoemaker's career, those two pitches have allowed a whopping .302 ISO versus lefties. Now throwing more splitters, Shoemaker no longer has to rely on them. He hasn't even thought about them. He traded his two worst pitches against lefties for his best pitch against them. It's such a simple fix. Meanwhile, the slider is less vulnerable against righties. Allegedly, it has struck out more than 25 percent of hitters during his career.
In addition to the increased strikeouts, all of Shoemaker's peripherals have improved. The ground ball rate (GB%) on his splitter increased from less than 40 percent prior to his May 21 start to more than 60 percent afterward. That's a huge deal for a player with serious fly ball issues. And it might be sustainable, too, given he's spotting the pitch a bit better now than he had been in April and early May. He's also throwing the pitch for strikes almost three-fourths of the time -- much better than the near-coin flip from his first seven starts. It's so simple, but more strikes and fewer balls correlates nicely with more strikeouts and fewer walks.
Something I can't explain, though, is that the whiff rates on Shoemaker's four-seam and sinker have doubled and almost tripled, respectively, since ramping up his splitter usage. That's something I can't explain, and I think we'll see the strikeouts wane a bit if those numbers start to normalize. Shoemaker only struck out six hitters last night, so maybe we're already on our way toward something more normal.
But maybe these outcomes have resulted from the overwhelming changes to Shoemaker's arsenal. Maybe the change in sequencing and offerings have played up the effectiveness of his otherwise lackluster "primary" pitches that really should have never been his cornerstones. I can't completely buy into that narrative just yet. Because when the effectiveness of those pitches regress, we'll understand why he's not truly, consistently, a 12 strikeouts-per-nine (K/9) kind of pitcher.

Until that narrative fleshes itself out, there's no reason to think Shoemaker can't ride his splitter to success. It's his best pitch, and he's throwing it a lot. He's trusting it in good counts, and he's trusting it in bad counts -- but, because of it, he's not finding himself in many bad counts, either, helping him work really deep into his starts. He's not relying on conventional pitching wisdom -- throw fastballs behind in the count, throw junk ahead in the count. His junk was never especially good against lefties, and his heat was never especially good against anyone.
Shoemaker just looks confident. Whether or not he says so, it doesn't really matter. The pitch usage charts show it. He's confident in an arsenal that relies on an off-speed pitch, and he doesn't feel compelled to mix in a fourth pitch against lefties. That same conventional wisdom probably tells us Shoemaker can't succeed on this kind of sustenance, but nothing else has worked all too well for him so far, so why not? A pitcher should use his best pitch most often. The argument is a little more nuanced than that -- sequencing, frequency, and depth of arsenal play a dynamic role in the effectiveness of every pitch -- but it can be reasonably simplified as such.
Shoemaker has completed the seventh inning in four straight starts -- something he had never done before. Shoemaker had some strikeout success like this in late July last year, but it came on the tails of a 12.6-percent swinging strike rate (SwStr%). Shoemaker generated a 21.1-percent rate in his three starts from May 21 to June 1 -- a much more believable rate relative to his strikeout success.
With all that said, Shoemaker has struck out 37 and walked zero - ZERO! - in his last 30 1/3 innings spanning four starts. He allowed two home runs last night, as he is wont to do, but he cruised through the first six innings -- something that hadn't come easily for him since his unexpectedly good 2014 campaign. He has induced swings on pitches outside the zone at a roughly 40-percent clip, and hitters are absolutely flailing. Encouraging hitters to turn tons of balls into tons of strikes is a good way to maximize outs and minimize baserunners. All of it's good for a 1.92 xFIP, but if you're not a believer, last night's 3.26 xFIP will do just fine. Or even his 3.55 xFIP on the season -- which includes all of those rough starts to begin the year. Strand rate (LOB%) and batting average on balls in play (BABIP) can really mess with a guy's outcomes.
Given Shoemaker still costs next to nothing at this point, it'd be wise to invest in him and wait to see if the magic wears out. I don't think you can guarantee me it will. Regression will hit, sure. It always does. But this is not the Shoemaker of old, and we shouldn't treat him as such.
A version of this article also appeared at FanGraphs.com.