Written by: Justin Choi (@justinochoi)
Follow us on Twitter! @Prospects365
When I first got into sabermetrics, I was fascinated by the concept of defense independent pitching. Developed by Voros McCracken nearly two decades ago, it posited that pitchers have little control over their BABIP and should therefore be evaluated by their ability to control home runs, walks, and strikeouts. Today, we use FIP, a modern iteration of his idea.
Later on, I learned that FIP isn’t a universal rule, as some pitchers manage to outperform it throughout their career. Still, the metric remains a good proxy for gauging a pitcher’s luck, or lack thereof. For the former, we can download the data from the Fangraphs leaderboards to calculate the FIP-minus-ERA differentials of pitchers. As of writing, here are the top 5 in that category amongst starters (min. 30 IP):
As planned, on the very top is our titular Dylan Cease, starting pitcher for the Chicago White Sox. A high FIP-minus-ERA differential doesn’t always signify regression, but when it’s as drastic as the one he carries, something is wrong – or at least a bit fishy.
So why does FIP hate Dylan Cease? Despite a blazing fastball that averages 98 mph, Cease doesn’t strike out hitters – just 5.87 per 9 innings, far below the league average. Furthermore, his lack of command leads to walks (3.91 per 9) and home runs (1.76 per 9). In short, he’s terrible at controlling the inputs used to calculate FIP.
And yet – we can’t just ignore his 3.33 ERA. Earlier, I said that certain pitchers can outperform their FIP, and pitch-to-contact wizards like Kyle Hendricks comes to mind. Despite his low career strikeout totals, Hendricks has managed to tame hitters by inducing soft contact. Doing so doesn’t lower his FIP by much, but in this case can we claim that Hendricks is getting ‘lucky’ on his batted balls? More often than not, no.
However, Dylan Cease is no Kyle Hendricks. Amongst the 48 pitchers who qualify for an ERA title, Cease’s groundball rate of 37.6% ranks 38th. That puts him in the company of Lance Lynn and Matthew Boyd. As Cease continues to challenge hitters with his fastball, it’s unlikely that his batted ball profile changes in the near future.
For a while, this is where my analysis ended. Dylan Cease was another pitcher, one of many, who walked a tightrope to either fall off or make it to the end. But what I knew didn’t add up. How could he still have a 3.33 ERA? During the second attempt of a deep dive, I focused on the fact that Cease gave up relatively few runs despite his walk and home-run rate. Did he, for example, allow home runs when men weren’t on base?
The answer is yes! Cease has allowed 9 home runs so far, 6 of which were solo shots. That might be part of the answer. At the same time, though, it makes sense that a pitcher will allow the most contact without runner(s) on base – because that’s when he throws the most pitches. After all, Dylan Cease’s wOBA on contact with no runners on base (.376) isn’t notably higher than the league average (.366). So we’re not done here.
What about his walks? This is where we take an interesting turn. It turns out that after a runner is aboard on first (i.e. after a walk), Dylan Cease turns into one of the league’s best pitchers! Don’t believe me? Below is a long table that shows his wOBA by base state, with league averages included as a comparison point. Also included are the pitches he’s thrown in each base state to indicate the sample size. Pay attention to the highlighted numbers:
What this table illustrates is that in specific situations, Dylan Cease is adept at escaping trouble, at least compared to a fictitious ‘average’ pitcher. That certainly would help lower his ERA. It’s also quite strange. Different situations don’t increase or decrease the wOBA of possible outcomes – the value of a home run is still around one-and-a-half runs regardless of a particular base state. Yet, the distribution of Cease’s wOBA against is uneven.
It’s possible that Cease’s approach against hitters changes from the stretch. This time, I checked if the percent usage of his four pitches – fastball, slider, curveball, and changeup – differed by each base state. Due to concerns with sample size and for simplicity’s sake, I’ve only included the five most common base states. Here are the results:
The results are a little ambiguous, but overall it seems like Cease is more reluctant to rely on his fastball with men aboard, opting to use his slider instead. That’s a great decision – the slider is objectively his best pitch, sporting 42.1 inches of vertical drop to induce whiffs. Here’s the thing, though. When Cease drops his fastball usage below 40% in certain situations to throw more sliders, that’s when he gets hit the hardest:
The combined sample size is less than 100 pitches, but it nonetheless exacerbates the problem: there seems to be no consistency between Cease’s pitch usage and the wOBA he allows. We’ve looked into the numbers in search for answers and ended up with more questions instead. For example, if his pitch usage with a runner on first isn’t too off from when there are no runners, why is his wOBA against so much lower?
In addition to the 6 solo shots, I discovered that 14 of his 20 walks were issued with the bases empty, providing a slightly more complete answer. Still, I’m grasping at straws. There is no rhyme or reason to Dylan Cease’s run prevention – instead, a perfect combination of good and bad luck exists, which somehow allows him to outperform his peripherals.
Lastly, I considered pitch location, which is also possible for Cease to change depending on the situation. But research revealed that any differences aren’t significant enough to explain his odd splits. For example, here is where his fastball and sliders are headed when the bases are empty:
And here’s where they’re headed with a runner on first:
If anything, the increased concentration of fastballs down the middle should make him more susceptible to hard contact. As the wOBA tables above illustrate, however, that hasn’t been this case so far.
Ultimately, Cease’s results seem to be based more on unique circumstances as opposed to skill. No surprise ending here. For reasons beyond my comprehension, he has defied FIP again and again rather than succumb to its pull of regression.
His lowly .233 BABIP is an easy answer, but not a complete one. He allows a bunch of solo home runs and bases-empty walks, yet when trouble arises, line-outs and groundouts suddenly materialize. He does get hit hard – especially with runners on 2nd and 3rd – but he’s also thrown 20 pitches with the bases loaded without allowing a single base-runner afterwards. BABIP we can explain; the rest we can not. That’s the wonder of Dylan Cease.
Many White Sox fans and analysts have raised concerns about Cease. They consider him a ticking time-bomb, and that’s correct. In the movies, a digital readout would be attached to the time-bomb, spelling out the exact hour, minute, and second at which it would explode. The heroes would then try their best to dismantle the bomb before it destroys a thing or person of significance.
In baseball, we don’t have that luxury. We don’t know when, if ever, Cease will start allowing run after run. We are unable to predict or prevent it. You can’t just walk up to a pitcher and ask him to suppress his home run rate and improve his command. It’s impossible. But in a sick, twisted way, that’s what also makes baseball entertaining to watch. Dylan Cease might continue to be good into the next few weeks, into the playoffs, and even into the next season. Alternatively, everything might come crashing down in his next start.
This isn’t to say Cease is a bad pitcher – his curveball and changeups are all over the place right now, but there’s some upside to him, especially within that potentially lethal fastball-slider combo. Rather, my point is that whatever happens to his ERA in the near future… don’t be surprised.
Follow P365 MLB and KBO Analyst Justin Choi on Twitter! @justinochoi
Follow us on Twitter! @Prospects365
Featured image courtesy of photographer John Antonoff and the Chicago Sun Times
All statistics from Baseball Savant and Fangraphs