Dynasty
Nastiest GIFs
The List
Dynasty
Nastiest GIFs
The List
+

Using xRV to Identify Potential 2022 Breakouts

The model thinks you should really pay attention to these seven arms.

Luke Smailes
Going Deep 1/22/2022

Back in October of 2020, I formally introduced Expected Run Value (xRV) before revamping it the following February in a version 2.0. The second edition added deception that was measured through attempts to quantify indistinguishability (through pitch tunneling), unpredictability (by analyzing how reactive each pitcher was to the count-state when choosing which pitches to throw), and unexpectedness (how well each pitcher was able to differentiate their movement profile of a given pitch from other pitchers that throw from a similar arm slot).

With MLB returning to a full 162-game schedule in 2021, there was a new robust set of pitch-level data to feed into the model. An interesting application method for me was to compare the 2020 numbers (albeit an abnormally small sample) of the model’s “underperformers” (i.e. those that had a substantially lower xRV than RV) to their 2021 results. Pitchers like John Means, Zack Wheeler, Tyler Anderson, Kyle Gibson, and Freddy Peralta all showed signals in their 2020 performances that either a bounce back or a breakout was a strong possibility in 2021, and that reality came to fruition.

The opposite was true for others like Zach Davies, Dallas Keuchel, Jesús Luzardo, Triston McKenzie, Taijuan Walker and Patrick Corbin who all outperformed their 2020 xRVs before regressing in 2021.

As we look forward to 2022, I’ve identified a crop of pitchers who could follow this same trend of seeing the difference between the expected and actual 2021 results manifest itself in this season’s results. But first, a couple of qualifiers. xRV, like most “x stats”, is designed to be descriptive rather than predictive, so I don’t want this to be misconstrued as model that is build to predict future success. Second, you may consider that some of my potential breakouts have already “broken out”. If that’s the case and my subjective definition doesn’t match yours, then my angle becomes that this pitcher’s 2021 success was legit and xRV likes him to build on it.

 

Dylan Cease – 2020 xRV Percentile: 22  • 2021 xRV Percentile: 97

Cease was one of my breakout picks heading into 2021 due to the offseason evidence that he was finally starting to spin his fastball more efficiently. The combination of working with Matrix Diamond Analytics to add data-driven development component to his offseason training while also having Ethan Katz as the main voice in his ear during the season seemed to do wonders for him. His lack of spin efficiency was essentially killing the potential of a fastball that was thrown, on average, as hard as any starting pitcher in baseball. In 2021, he ended up raising the vertical break by a massive three inches with a seven percent increase in efficiency. His command of his fastball still wasn’t good (38th percentile) but its stuff was good enough to compensate (90th percentile).

Cease also further separated the shapes of his slider (94th percentile xRV) and curveball (76th percentile xRV). He threw his slider slightly harder with a more traditional shape while his knuckle-curve was a little slower with more depth than in 2020. On average, Cease’s fastball and curveball had three feet of vertical break differential. That said, the model had his arsenal tunneling in the bottom half of qualified pitchers, so his pitches are just nasty on their own.

Even Cease’s changeup, a pitch that xRV doesn’t like holistically, got good results (.157 xBA, 47% whiff). He doesn’t command it well, and it was in the eight percentile stuff-wise. However, the pitch’s movement profile ranks high up in the ‘unexpectedness’ facet of deception (96th percentile). Of the 150 pitchers that threw 200 changeups in 2021, only John Means had more positive vertical break, and hitters innately don’t predict changeups to move that way. This unexpectedness attribute is built on a nearest neighbor algorithm that compares a pitch’s movement profile to those of other pitchers with similar release points. The almost 20 mph difference also helps this pitch succeed.

Can Cease continue the upward trajectory to a bona fide all-star starter in 2022, or did we already see his ceiling? My bet would be on the former. His pitch quality numbers give me confidence in that prediction, and as he gains more experience with his reshaped arsenal, I think his command takes another (much smaller) step forward.

 

Drew Rasmussen – 2020 xRV Percentile: 77 • 2021 xRV Percentile: 97

Even if you knew nothing about Rasmussen prior to Tampa Bay acquiring him from Milwaukee in May, you probably could have guessed he was filthy based on who was trading for him. He was effective last season mainly with just a fastball/slider two-pitch arsenal that xRV loves. In terms of stuff, the model has his fastball in the 66th percentile and his slider in the 84th percentile. He sat at 97 mph with his heater that also featured almost 18 inches of vertical break. His slurvy-slider was thrown hard at 86 mph, and opposing hitters walked away with a .159 batting average against it.

Add in the fact that he commands both pitch types better than 80% of the league, and Rasmussen becomes really exciting. The cherry on top is that Rasmussen also excelled at the most important factor of deception, pitch tunneling. His arsenal was tunneled in the 93rd percentile.

Rasmussen should be penciled in the Rays rotation no matter how they approach the post-lockout trade and free agent markets. From a fantasy prospective, the way the organization manages their pitchers can be less than ideal, but I think Rasmussen will force his way into a lot of consistent innings with good health.

 

Mitch Keller – 2020 xRV Percentile: 40 • 2021 xRV Percentile: 89

Keller is definitely the pitcher on this list that the model predicts the biggest turnaround for.  The almost two-run difference in his ERA and FIP is the first basic signal that better fortunes could be in store. xRV likes all three of his main pitches (his fastball, slider and curveball) in the 80th percentile or better. For a pitcher that struggled with walks as much as he did in his brief 2020 stint, the model also had his command of his arsenal in the 89th percentile in 2021 with his slider (95th percentile) as the headliner. His slider’s xRV was in the 80th percentile overall. If you can sense the theme here, xRV loves hard sliders that are commanded well. I think Keller also does a good job of diversifying the shapes of each pitch in arsenal as he, like Cease, tweaked his breaking balls over last offseason to better achieve this. He moved his slider from a hybrid shape to a more traditional one, and the result was about an 11 percent decrease in the pitch’s xRV.

Keller is going to get ample opportunity to prove himself given where Pittsburgh is in the rebuild, and there are some signs here that signal better times are ahead. Bumping up that slider usage, that’s currently below 25%, could be a pretty straightforward first step. A less-important but still relevant xRV deception attribute is usage unpredictability. For what that’s worth, Keller was in the 93rd percentile, meaning a did a good job of not letting the count state play a major role in determining his pitch selection. With him also flashing some upper-90s fastball in offseason bullpen sessions as well (probably means very little), Keller has some intriguing pieces that, if he can put them together, could lead him a 2022 breakout.

 

Carlos Hernández – 2020 xRV Percentile: 61 • 2021 xRV Percentile: 88

Hernández’s arsenal is headlined by his riding heater that sits at 97 mph and averages more than 17 inches of vertical break. He has above average command of the pitch and in fact, Hernández had above average command of his entire arsenal sans his changeup.  Stuff-wise, his best pitch is surprisingly his changeup (97th percentile) as he gets some of the same “unexpected high vertical break” effects that Cease does. His changeup, curveball, and fastball are all in the top third of the league in terms of stuff alone by my model.

All five of Hernández’s offerings are above average by xRV and only his changeup isn’t in the top third. He has a lot of options at his disposal to get hitters out. The interesting part about his arsenal is that he makes each offering work with different factors. His changeup has great stuff, but poor command. His sinker and slider both have poor stuff, but good command, unpredictable usages, and unexpected movement profiles (mainly the horizontal break from his over-the-top release). His fastball and curveball both just have upper third qualities across the board.

Hernández will be pitching 2022 at just 25 years old too in a home park that favors his style of pitching. Steamer doesn’t like his strikeout numbers to climb at all, but I think he out-performs that projection along with easily beating a near-5.00 ERA. He now has 100 major league innings under his belt and is coming off a season with consistently high-end pitch quality. I think he showcases the better control he had in the minors and increases his put away efficiency as he harnesses a robust arsenal.

 

Zac Gallen – 2020 xRV Percentile: 87 • 2021 xRV Percentile: 87

Gallen, like Hernández, had five above average offerings and four in the upper third. He also aligns with Cease in that xRV isn’t calling for a first time breakout, rather that his career success thus far is backed up strongly by his pitch quality and that model believes that Gallen can take another step forward. His 2021 ERA was a point and a half higher than his first 152 innings in the big leagues, yet his pitch quality remained the same.

He just about average in terms of pitch type usage unpredictability, but he excelled in the other two factors of deception that are baked into xRV. Gallen was in the top 25% of the league in terms of movement unexpectedness and ranked in the 96th percentile of pitch tunneling. Individually, his slider tunneled in the 96th percentile while his changeup was in the 94th and his slider in the 76th (in relation to Gallen’s main pitch, his four-seam fastball). Trevor Rogers and Andrew Heaney were the only two starting pitchers that were more deceptive than Gallen in 2021.

His best pitch by xRV is his knuckle curve (94th percentile) built more so on command and deception than it is stuff.

With Brett Strom now in Arizona, look for Gallen to take that next big step and wind up as a top-25 starter in 2022.

 

Luis Patiño – 2020 xRV Percentile: 61 • 2021 xRV Percentile: 85

In his first season with Tampa Bay, Patiño was effective at times for the Rays but he definitely didn’t reach his potential yet. Ay just 21 years old, Patiño flashed three offerings that graded out as above average (fastball, slider, curveball), while xRV didn’t care for his changeup and sinker.

His riding fastball is Patiño ’s bread and butter for good reason, as xRV holds it as his best pitch in the 90th percentile league-wide. At 96 mph with almost 19 inches of vertical break, Patiño could probably elevate his fastball more to take advantage of those qualities along with an upper third vertical approach angle. His slider and changeup tunnel well with his fastball, with the former generating impact horizontal break as well. The fastball is so good that anything Patiño can develop that plays off of it should become effective too. The heater’s stuff in the in the 82nd percentile, it was commanded in the 76th percentile, and it was tunneled (the most important factor of deception) in the 88th percentile with the rest of his arsenal. It’s filthy.

As Patiño continue to develop and mature, his stuff/command/deception potential is elite. The exploding fastball and with the hard, darting slider combination has already been effective. The main questions now surround the development of his supplementary offerings to better handle left-handed hitters, specifically the curveball and changeup.

Needless to say, Rasmussen, Patiño , McClannahan, and Baz is a very exciting crop of young pitchers for the Rays to build around.

 

Logan Gilbert – 2020 xRV Percentile: N/A • 2021 xRV Percentile: 84

When I noticed that Logan Gilbert posted an 84th percentile xRV rookie season, I assumed that the model loved  his high velo, high-ish vertical break fastball and its relationship with his slider, his other go-to pitch. He used these two pitches more than 85% of the time in 2021. While I was correct (his fastball was an 89th percentile pitch and his slider was in the 71st percentile), the model says that the success of both pitches was actually command-driven, but we aren’t even talking elite command either. His fastball was in the 64th percentile and his slider was in the 81st percentile of command respectively. Both pitches were right around average in terms of stuff and below average in terms of deception. It’s nothing that really jumps off the page at you, which is why the him being held in high regard by the model isn’t a clear concept.

In fact, Gilbert’s usage of his fastball was as predictable as almost every other pitcher in the league. The only pitchers that threw at least 500 fastballs in 2021 and were more predictable in their fastball usage were Clayton Kershaw (whose fastball had its highest Statcast-era xwOBA by far in 2021), Aroldis Chapman and Liam Hendriks. The latter two are clearly comfortable saying “Here’s my fastball. I’m going to throw it almost always and especially when I’m behind in the count. Good luck hitting it.” For Gilbert, his usage wasn’t necessarily surprising given his rookie status and lack of confidence in his third and fourth pitches, but this kind of usage of the pitch is something he’ll want to get away from as he matures.

His after-thought pitches at least show some potential. His changeup had good-looking peripherals last season and has solid velo differential with above average vertical break. His curveball is essentially a slow gravity-ball that Gilbert has very sparse command of. You’d think this pitch mix would result in Gilbert really struggling against left-handed hitters with 9th percentile changeup and a 29th percentile curveball, but he was actually better against them in 2022 than he was against righties.

At the very least, Gilbert’s sophomore campaign and the development of his arsenal will be interesting to follow. xRV liked him in 2021, but the dissection of why it liked him was a lot more confounding than the others mentioned in this piece. The fastball/slider combination has the chance to develop into a nasty combination, but as of now, xRV says they haven’t tunneled well yet (41st percentile). He’s my 2022 wild card.

 

Photos by Quinn Harris & Frank Jansky/Icon Sportswire | Adapted by Doug Carlin (@Bdougals on Twitter)

Leave a Reply

Your email address will not be published. Required fields are marked *

Account / Login