What the Data Says: White Sox Rotation

Unlikely or not, the White Sox starting rotation has been an early strength of the team in 2021.

Lucas Giolito has been all of spectacular, dreadful, and also has a couple starts in-between. Lance Lynn has proved to be a worthwhile acquisition through a mere three starts. Dallas Keuchel’s peripherals look pretty typical of his successful career, even with some early command issues. Dylan Cease struggled command-wise before a compete game shutout on Thursday night where he looked special. If we were playing another 60-game season, Carlos Rodón would be one of the midseason favorites for the AL Cy Young. He’s been flat-out fun to watch. And then there’s Michael Kopech, the phenomenon who has created controversy as to what exactly his role should be for the rest of the campaign with his electric start.

All told, the White Sox starting staff currently leads the American League in starting pitcher ERA, FIP, xFIP, and fWAR.

This group is shaping out to be one that gives the White Sox a strong chance to win on a nightly basis. It specifically has that one quality that was clearly missing last year — depth — and high-upside depth at that.

---

Lucas Giolito

A couple of rough April starts should not deter people from believing that Giolito is a bonafide ace. His stuff has actually gotten better compared to the previous two seasons, but it’s been his command that has driven his recent struggles.

His fastball vertical break is up an inch and a half, and it now sits at an elite 20.1 inches. That represents the highest 4-seam fastball vertical break of any starting pitcher in baseball. What should you do with that kind of movement? Keep it near the hitter’s letters. He’s certainly made that a point of emphasis, but he has also too often found the heart of the plate. This is seen when looking at his whiffs vs. base hit allowed on his fastball.

Giolito’s fastball Whiff% is actually down 9%, this year even though it’s moving more. He needs to get it elevated to take advantage of it.

To pair with a more oft-elevated heater, Giolito needs to get his changeup down. He does have the unique ability to have success with high changeups that are located by design, but even so, he generally gets his best results at knee-level, like virtually everyone else. Giolito delivered them lower last season and enjoyed a higher Whiff% and an xwOBA that was 37 points lower. Giolito’s changeup really doesn’t move relatively well at all. It never has — which makes its success even more impressive. This year, he’s added 2.5 inches of horizontal run to the pitch, but even then, that still only puts it in the 20th percentile of movement vs. the average 82 mph changeup.

Giolito has the unique ability to throw his changeup with the exact same conviction as his fastball. The spin axis differential of the two pitches is also only 15 degrees. So if the hitter can’t tell that a changeup is coming from Giolito’s delivery, and they can hardly differentiate the spin, the 11 mph velocity differential is tough to handle. The way it moves, given Giolito’s release point (another factor of deception included in my latest xRV model), is also rather unexpected. It sits in the 75th percentile of movement unexpectedness, as hitters inherently predict a pitch’s movement based on where exactly it is released from.

To round out his repertoire, Giolito’s slider looks a little different this year, and it hasn’t been working well so far. The pitch’s velocity is up 1 mph, but its Whiff% is down 20% and is xwOBA has ballooned from .227 last year to .386 currently. It already possessed a cutter’s shape in 2020, but now its spin axis is all the way up to 180 degrees on average, creating less drop and less horizontal movement. It would most likely benefit Giolito to get the pitch’s shape closer to the 2020 version (roughly 160 degrees — 11:15), or even closer to a traditional right-handed slider’s shape near 90 degrees (9:00).

Gio’s 5.68 ERA likely sticks out to most as a representation of failure through five starts. Those command/execution issues have produced a 7% increase in barrels, 70-point xwOBAcon spike and an 8% decrease in whiffs. Underneath those result in a small sample, however, are that aforementioned fastball movement increase and an xERA that sits much lower at 3.76.

He’ll be fine.

---

Lance Lynn

Through his first three starts, Lynn has been just what the doctor ordered in terms of a true No. 2 behind Giolito. With one of the most unique arsenals — one that features four-seam, sinker, and cut fastball variations almost exclusively — Lynn is extremely unpredictable in their usages. Specifically, Lynn’s pitch usage is extremely non-count dependent. That means ahead or behind, Lynn usually has his entire arsenal at his disposal. Though, a 68% first pitch strike percentage (85th percentile) has meant that he is usually pitching from ahead. All three of his fastball options were in the 60th percentile or better in terms of count unpredictability. Currently, every pitch in that arsenal carries a sub-.300 xwOBA.

Lynn is posting a career-high whiff rate on his four-seamer and at 36.8%, which ranks in the 95th percentile in MLB. In 2020, it was an 82nd percentile pitch according to Expected Run Value (xRV). His cutter also has a career-high whiff rate this year (26.3%), and is another pitch that xRV liked a year ago (93rd percentile). His sinker has 83rd percentile drop. They’re all good pitches, and watching his starts consistently helps one realize how surgical he is with each of them.

Lynn mostly has ditched his curveball, but we may see it a little more as the season goes on. He’s only thrown five of them this year after 8% usage a year ago. His unsustainable 32.5 K%-BB% has been fun to watch, but even as that normalizes, Lynn’s start-to-start dependability (along with Keuchel) in-between Giolito’s brilliance and the exciting, but sometimes volatile efforts of Cease, Rodón and Kopech is comforting.

---

Dallas Keuchel

Keuchel has gotten hit pretty hard through his six starts this year. His Barrel% has gone from the 85th percentile in 2020 to the 36th percentile this season and at 4.38, Keuchel has his highest xFIP since his rookie year in 2012. Not great at all, but there are still some telltale signs that make me confident in saying that we will likely be dealing with the “same old Keuchel” through 2021.

Keuchel’s success has always largely been built on control and command of the strike zone. That’s mostly still here, and it looks like that may be more important than ever, as he’s currently running his lowest K% since also 2012.

Keuchel’s Edge% has dropped from 46.1% to 44.6%, but even the latter is in the 74th percentile. His secondaries have generally been located well, and those high sliders you may have noticed are by design. Keuchel has subtly dropped those into the zone to get ahead of hitters early in the count.

There’s also been some early-season growth in his “stuff” that Keuchel hopes will stick. He’s getting almost an inch more of drop on his sinker, sitting now firmly in the 93rd percentile among all sinkers vs. average. That’s resulted in his launch angle against going from 0 degrees last year to -6, and a 3% increase in his already-elite ground ball rate. At 55.9%, it’s currently the fifth highest in baseball among qualified pitchers.

Every one of Keuchel’s options in his five-pitch arsenal has a higher xwOBA than last season. His cutter is at .460, and he’s dropped its usage by 12%, in favor of more sinkers and changeups. Getting it up higher in the zone has been its death sentence. Keuchel will need either his cutter or slider, given their movement profiles, to get better results to support his sinker and changeup. His tough third inning on Friday night was seemingly a result of Keuchel trying to be too fine, rather than trusting his stuff to generate weak contact. Just get that sinker at the bottom of the zone and let hitters do what they’ve been doing over the past nine seasons — beat it into the ground.

Uncharacteristically, Keuchel has also struggled with lefties early, as they’re slashing .400/.438/.600 with a .445 wOBA compared to righties hitting .205/.295/.318. with a .272 wOBA. What’s noticeable is the amount of center-cut sinkers they’ve seen that have been hit hard — they’re slugging .857 against it. It looks like small sample size randomness to me and not a major cause for concern, although mixing up the pitch usage strategy against lefites probably wouldn’t hurt. With Lynn and a transformed Rodón now in the fold, Keuchel settling in to a consistent mid- to back-end rotation role seems probable.

Depth through the rotation, you say?

---

Dylan Cease

Prior to the season, it was my belief that Cease reaching his own All-Star potential would be one of the keys to the White Sox reaching their 95 win-plus potential; I even made the bold claim in the South Side Sox Slack channel that Cease would in fact be a 2021 American League All-Star. Given Ethan Katz as the new voice in his ear this offseason, paired with his work with Matrix Diamond Analytics, I was privy to some data signaling that his stuff would be improved in 2021. After an 11-strikeout performance against Colorado that rounded out Cease’s spring training, he showed a little of that his ceiling could be before backing that up Thursday night.

Even with better “stuff,” the key for Cease is and will continue to be his command. A lot of his problems over the past two seasons stemmed from command issues, but also a heater with inefficient spin that significantly limited the pitch’s rise. The basic question that I found myself asking after last season was how an upper-90s fastball could be so hittable, especially given not one, but two nasty breaking pitches in his back pocket? My Expected Run Value metric (xRV) also only had the pitch in the 42nd percentile of all pitch types.

When examining his first five starts of 2021, Cease’s increased spin efficiency from 82% to 90% jumps off the page (20th to 47th percentile), and what that does is take his vertical break from 15.7 inches up to 19.3 inches. Add in the fact that he’s also added an extra inch of run and 65 rpm of raw spin on his heater, and you realize the pitch is transformed.

I can keep walking through the chain of effects ...

That increase in vertical movement shapes a flatter vertical approach angle that is now in the 76th percentile. A flatter approach angle makes it more challenging for hitters to get on-plane to square up a pitch, especially when a pitch is executed at the top of the zone. Given these qualities, working up in the zone is imperative to maximize this flat, 96 mph, 19-inch pitch. Cease has done just that and has seen a 5% jump in whiffs on that fastball.

Aside from the fastball, Cease has tweaked the shapes of both his slider and curveball in an effort to further differentiate the two. He’s taken the slider from a 52-degree spin axis and turned into a more traditional 69-degree pitch, where vertical break is traded for increased horizontal break and velocity.

Cease’s curveball has moved in the opposite direction, as he’s added more true top-spin to give the pitch a more 12-6 shape, which has increased its vertical break. Between his fastball and curveball, Cease now has 38 inches of vertical movement differential with almost perfectly-minored spin. With mirrored spin directions, it’s more challenging for hitters to anticipate a pitch’s movement as it’s in flight.

In smaller samples, Cease’s command of his slider and curveball have both been good enough to allow them to play off of his improved fastball. His slider is getting 85th percentile whiffs at 49%.

Overall, Cease’s contact quality has improved, but it’s still not great. Coming into Thursday night’s start against Detroit, Cease had a 15% walk rate and a disastrous 39.2% first pitch strike percentage (the latter being the second-worst among pitchers with at least 10 innings). In his complete game shutout, Cease threw 71% first pitch strikes. It was a fantastic development, and Dylan mentioned that he and Lance Lynn discussed some of the specifics that were holding him back through his first four starts.

We’ve seen encouraging signs and the progress is evident (including a massive three-run FIP improvement!). The question is, can Cease string successful starts together against better opponents?

---

Carlos Rodón

My biggest takeaway from what we’ve seen from Rodón is that the guy is finally healthy. Yes, Katz has likely had a positive effect on him too, but after both a shoulder and Tommy John procedure changed the trajectory of the No. 3 overall pick’s development, Rodón looks to be completely back.

The best evidence for this is Rodón’s lively fastball. It’s really been a completely different pitch, in a fantastic way. The velocity is up two mph to 95, vertical break is up two inches, and horizontal break is up one. This has driven Rodón’s fastball’s whiff rate up 21%.

While Rodón’s slider isn’t “transformed” stuff-wise, its relationship with the new heater combined with much better command have made it a deadly pitch. It’s always had good movement. This year, the slider getting 84th percentile horizontal movement above average. His slider’s whiff rate is now at 49%, up 23% from last year, and it now sits in the 88th percentile. He has yet to give up a hit on it through 105 pitches, and it has a .048 xBA and .173 xwOBA.

Sheesh. Fastball velocity does wonders for the rest of your arsenal.

Rodón’s third option, his changeup, has a .236 xwOBA. It gets decent horizontal break, but really is nothing special in terms of movement. However, righthanders now have to deal with 95-98 mph heaters and wicked back-foot sliders, so the changeup’s success makes sense. He’s kept it low in the zone, as it has paired well with Rodón’s other two offerings.

Overall, a .130 BABIP won’t hold, especially with a below-average hard-hit rate. But nonetheless, Rodón has been a complete revelation this year, and is tracking to be this rotation’s third starter.

---

Michael Kopech

This is the Michael Kopech we’ve been waiting almost four years for. His fastball/slider combo has been devastating.

Kopech’s fastball is absolutely filthy. We can say that technically his fastball velocity is up from 2018 (95.4 to 96.3), but we know that he was pitching at least some of that ’18 season with a damaged elbow. The velocity sits comfortably above average, but it’s the movement that separates his fastball: It gets more than 18 inches of vertical break and almost 12 inches of horizontal break. To put that into perspective, the only pitchers this season with at least 18.5 inches of vertical break and 10 inches of horizontal break on their fastballs (minimum 100 pitches) are: Trevor Bauer, Gerrit Cole, Matt Barnes, Justin Dunn, Jeff Hoffman and Ian Kennedy. Cole is the only one who also boasts 96+ mph velocity.

Kopech’s fastball has a .172 xwOBA and a 36.7% whiff rate (95th percentile), up 20% from 2018. He’s also getting more extension off the mound, which should be helping his vertical break and vertical approach angle. Kopech’s fastball/slider combo looks to be benefitting from some spin-mirroring effects. The slider hasn’t shown crazy movement, but it’s getting 71st percentile whiffs and has a .191 xwOBA.

That’s all well and great, but Kopech’s stuff has never really been in question. His success thus far has largely been a biproduct of his vastly improved command. He’s using that vertical movement at the top of the zone efficiently, while his slider and changeup command have also been good. He’s generating 69% first-pitch strikes, which ranks in the 89th percentile league-wide. For the question marks that have come with Kopech’s ability to control any of his pitches, this kind of execution is very good to see.

How about some more perspective on Michael Kopech? He has a 39.7% K-BB% which ranks third among pitchers who have made at least two starts this year. The Top 10 of this category looks like this:

1. Corbin Burnes - 45.4%
2. Jacob deGrom - 44.7
3. Michael Kopech - 39.7
4. Gerrit Cole - 39.5
5. Joe Musgrove - 32.7
6. Lance Lynn - 32.5
7. Tyler Glasnow - 31.5
8. Dustin May - 31.4
9. Shane Bieber - 30.3
10. Trevor Bauer - 29.1

*Carlos Rodón ranks 11th at 28.4%

That’s phenomenal company.

I like the role that Kopech will likely be in for the rest of the season — multi-inning reliever who has the ability to make spot starts as needed. The reality is that he’ll need to be on an innings limit, so the club will be unable to fully unleash him until 2022. With both Lynn and Rodón as impending free agents, Kopech likely will get that shot. What will be interesting is if any of the current five go down with an injury for an extended period of time and the White Sox are in the thick of a pennant chase. Would Tony La Russa plug Kopech in for 3-5 innings per start and attempt to keep him within the innings limitation bounds, or is he forced to turn to the likes of Reynaldo López, Jonathan Stiever and/or Jimmy Lambert?

I’m certain that decision wouldn’t stir up White Sox Twitter too much ...