Evan Marshall and Jimmy Cordero: diamonds in the rough
Early in 2020, Ricky Renteria has signaled faith in Evan Marshall and Jimmy Cordero in high-leverage situations. Following last Thursday's off-day, Aaron Bummer and Steve Cishek, who are paid the big bucks to usually get these kind of outs, were presumably rested and available. However, it was Marshall and Cordero deployed protecting a one-run lead. They held down the fort, and Alex Colomé secured the ninth-inning save.
Protecting a one-run lead in Milwaukee on Monday night, it was Marshall who was summoned to handle the eighth inning. On Tuesday, Cordero was put into the same situation and was credited with a hold on the way to a six-game win streak.
So who are these guys? If we go back to their 2018 numbers, these two look like pitchers who might not even be able to crack a major league roster, much less as a playoff-hopeful team's seventh- and eighth-inning arms.
The Sox signed Marshall to a minor-league deal in late October 2018, with an invite to spring training. He pitched most of 2018 with Cleveland's Triple-A affiliate in Columbus and had a 1.13 ERA (2.50 FIP), but also pitched seven innings for the big league club with a 7.71 ERA (2.73 FIP).
Cordero was sought after by the White Sox organization early in 2019, but they had to wait their turn to get him. He was DFA'd by Washington on May 9 and was claimed by Toronto, the team that originally signed him back in 2012. The Jays DFA'd him two weeks later, before Seattle nabbed him. Finally, the Mariners cut him loose after pitching two-thirds of an inning for the Arkansas Travelers, the team's Double-A affiliate. That same day, June 7, he was claimed by the White Sox, where he has finally found success in the major leagues. Given their records, the Blue Jays and Mariners had waiver priority over the Sox in May 2019, so Cordero ended up throwing for four different organizations in less than a month.
As it turns out, this faith is warranted; Marshall and Cordero were both really good in 2019. And not just "ran-into-a-good-year," "BABIP-luck" good seasons. Their peripherals were very encouraging, and the White Sox have been able to get the most out of each pitcher's best offerings — Marshall's changeup and Cordero's sinker.
In limited opportunity with Cleveland, Marshall threw his sinker a majority of the time, a pitch that he had major issues commanding. It was either a hard-hit ball ...
... or significant arm-side misses. His sinker had respective xwOBAs of .469 (2015), .437 (2016), .359 (2017), and .408 (2018) while with Arizona, Seattle, and Cleveland. Those are less than ideal.
However, prior to 2019, the White Sox saw potential in Marshall's arsenal. His changeup produced a .130 xBA and a .122 xwOBA with a huge 55 Whiff% in a small 35-pitch sample.
The front office, in conjunction with pitching coach Don Cooper, began having Marshall throw his changeup as his primary pitch in 2019 while favoring his four-seamer over his sinker as its main supplement. He also lessened his curveball usage by about 5%, but it was actually his second-best pitch last season with an xBA of .235 and Whiff% of 37.8.
What Marshall ended up doing with his changeup was to replicate his 2018 success and by using it more, he had a very successful 2019 season. Marshall's changeup was quite literally one of the best individual pitches in baseball. Yes, you absolutely read that right. Let me explain.
Working with my classmate (and teammate) Jared White and my advisor/professor of economics at Coe College, Dr. Ryan Baranowski, we were able to create a random forrest model (I won't get into the specific details of how exactly it works) that tells us which pitcher's offerings had the lowest expected run values based on their "stuff" attributes, and their command attributes. This model was chosen due to the non-linearity of relationships between these attributes. Stuff attributes include horizontal and vertical movement, spin rate, and velocity. In general, we know the locations where it's optimal to pitch — away from the heart of the plate. So that's a little more straightforward. While we can confidently say from the model that stuff is more important in minimizing expected run value than command, command should certainly not be ignored.
This model tells us that Marshall's changeup was the 17th-best individual pitch last year, and the best overall changeup in MLB in 2019, slightly edging those of Luis Castillo, Aaron Nola, and Stephen Strasburg. I won't list all 16 better pitches, but the best was Blake Snell's four-seam fastball, followed by Jordan Hicks' sinker, and Walker Buehler's slider. Gerrit Cole's renowned fastball was fifth on the list. Out of 2,168 different pitches, Marshall's changeup is 17th, or in the 99th percentile.
So, how should this be interpreted?
Even though there were 20 pitchers with more vertical movement (when controlling for velocity and release point) and more than 100 pitchers with more horizontal movement (when controlling for velocity and release point) in 2019, Marshall's ability to command the pitch leapfrogs him over the others to the point where his is the best in terms of expected run value. For more context, Marshall's 47.7 Edge% was the third best in MLB in 2019 (min. 100 PAs), one spot above Kyle Hendricks.
This is borderline unfair for Shohei Ohtani.
When we consider Marshall's entire pitch arsenal, he ranks 30th out of 704 pitchers (96th percentile). His curveball was the 59th-best pitch (97th percentile). His sinker was 301st, which still puts it in the 86th percentile. Even in a season where his sinker was hit to a .299 xBA, 91.6 EV, and .378 xwOBA, the pitch's qualities tell us that it wasn't expected to get hit that hard.
That's one quality arsenal, especially for a relief pitcher and even more for a non-roster invitee in the previous spring.
This pitch looks like it could be a backdoor sinker or changeup for an awfully long time to the helpless Cheslor Cuthbert.
There was some fortunate timing this week, as Marshall was asked to speak with reporters before Wednesday's game, and he was able to speak to the prowess of this changeup early and often (courtesy of the White Sox):
So how about Jimmy Cordero? The White Sox made some significant adjustments to his pitch usage as well. In 2018, Cordero threw his four-seam fastball more than 40% of the time. While it was thrown at an average of 97.8 mph, it was straight as an arrow and was commanded poorly. Given its low spin rate (2160 RPMs), Cordero locating it primarily up in the zone was usually a recipe for disaster. His results confirm this, as opposing hitters had a .342 xBA and .442 xwOBA on it in 2018.
Much like Marshall, Cordero had a great pitch that he wasn't using enough. The White Sox identified it, and changed his usage when he joined the major league team in the middle of 2019. For Cordero, it was his sinker. He threw it a mile per hour slower, but his .271 xBA and .317 xwOBA were good enough for the Sox to take a shot on him on the waiver wire.
With the White Sox, Cordero increased his sinker usage by 29% while also adding 0.5 mph and about an inch of horizontal break. His four-seamer was kept around solely to attack left-handed hitters, not for righthanders as he used it in 2018. This is with the conventional idea that a sinker/two-seamer from a right-handed pitcher will more often run into the barrel of lefties.
I don't really know why his four-seamer/sinker usage was previously backwards. That initial fix was simple. He started consistently throwing ground balls with his sinker (it was hit at an average launch angle of -3 degrees), while also creating some firewood every once in awhile.
Cordero's 60.4% ground ball rate was the 11th best in baseball last season. His 45.8 Edge% was 26th best. These are elite-level results.
His changeup, a pitch he only threw 30 times in 2018, became his best secondary offering. He didn't give up a single hit on it, while also possessing a .096 xBA, a .115 xwOBA, and a massive 50 Whiff%. That Whiff% was the fifth-highest of any changeup in baseball (min. 50 thrown).
Cordero's changeup was at its best when it was thrown on roughly the same 235 degree axis as his sinker, but 10-12 mph slower. Roberto Perez is clearly disgusted with it as he walks back to the dugout wondering how that's allowed.
As for the aforementioned pitch stuff/command model, we have Cordero's sinker as the game's 215th-best pitch (90th percentile) and his changeup as the 406th-best pitch (81st percentile).
Overall, Cordero finished 2019 with much-improved peripheral numbers across the board, including a 50-point improvement in xwOBA and more than a two-run xFIP improvement.
Now in 2020, the two have continued their superb performances, albeit in a small sample size. This is particularly true for Marshall, who has thrown his changeup even more, up another 4% as of Tuesday. Its Whiff% currently sits at 43.5. His curveball, which has seen a 13% increase in usage, is at 60. Marshall is increasingly pitching to his strengths.
It's always good to see familiar White Sox foe, Alex Gordon, taken care of this easily by this changeup filth.
The organization has already gotten more than they could've asked for from these low-risk acquisitions. It is these kind of moves that provide a team with substantial surplus value that goes a long way when building a roster that is designed to contend. This organization hasn't had a lot of 'diamond in the rough' acquisitions through this rebuild, but now is the perfect time to discover them. If Marshall and Cordero can continue to build on their career resurgences, it could play a major role in the White Sox clinching a playoff berth and beyond.
Luke was born and raised in the south suburbs of Chicago and is a college student and collegiate baseball player at Coe College in Cedar Rapids, Iowa. While playing, Luke also leads the team's baseball analytics program that has successfully translated methods used in MLB to the Division-III game. As a result of this work, Luke's writing is analytically motivated, but will also focus on trades, roster construction, and offseason plans. For a White Sox organization that was late to the party in the analytical realm, their recent progression provides ample opportunity to critically analyze their decision making. He can be found on Twitter @CoeSoxMetrics.