On the morning of the trade deadline last summer, the Twins were in first place in the AL Central, a game up on the surprisingly pesky Guardians and two games in front of the heavily favored White Sox. The front office pored over the trade market, assessed their own needs, and looked for matches with other clubs looking to trade current talent for future prospects.
Then, the Twins did what first-place teams do at trading deadlines. They traded a package of mid-level prospects, half of whom were still several years away from the majors, for an ace reliever, who, at the time, had a 1.68 ERA, a very good 2.99 FIP, more strikeouts than innings pitched, a ground ball rate of 60%, and averaged about 98 miles-per-hour with his fastball. And, that ace reliever, Jorge López, was under team control for two additional, cost-controlled seasons, to boot.
Absent any other context, a deal like that is one from the general managing for dummies textbook. You make it almost every time.
But context matters. Especially when that deadline acquisition posted a 4.37 ERA / 4.35 FIP with the peripheral stats to match and played a big role in that first-place team fading in the standings down the stretch.
The context around López is even more important because he had not always been a shutdown reliever. In fact, most of his career prior to 2022 had been a major struggle.
But he had matriculated slowly through the minor leagues and then taken several years and three different clubs to finally stick in the majors. All while going through the unimaginable experience of caring for and finding treatment for his infant son, Mikael, who was diagnosed with two chronic autoimmune diseases that have required several years of intense medical care, including an intestinal transplant, and additional treatments similar to those applied to leukemia, including chemotherapy and a bone marrow transplant.
Really great story by @alec_lewis from June on the other player the #MNTwins acquired in the Jorge López deal, his 9-year-old son, Mikael, who was born with two chronic autoimmune diseases.https://t.co/h3l5sKjgWa— DanHayesMLB (@DanHayesMLB) August 2, 2022
Entering 2022 López was seen as far from guaranteed to make the Baltimore roster because he had pitched 350 major league innings and supplied a cumulative 6.04 ERA and 5.15 FIP to that point.
That context makes the questions about López this offseason louder. Was the dominant version of him that the Twins thought they were acquiring a small sample fluke? Were the Twins hoodwinked into buying high on a pitcher that regressed back to the pitcher he always was? And, most importantly, what changed, and is any of it fixable for 2023?
Baltimore Bullpen Stalwart
Shortly before the 2022 season, Baltimore traded two of their expected high-leverage bullpen members, Cole Sulser and Tanner Scott, to Miami. That move prompted Manager Brandon Hyde to proclaim to the media, “You’re gonna see Jorge López pitch in big spots.”
That statement was newsworthy because López was thought to be a big question mark. The rebuilding Orioles had claimed him off waivers from Kansas City in 2020 and had tried him as both a starter and a reliever for parts of two seasons with minimal success while working to implement some mechanical adjustments to shorten his arm swing and more efficiently leverage his hip rotation in his delivery.
In a bullpen stint at the end of the 2021 season, the Orioles had seen those adjustments click when López’s 93-94 mph sinker started coming in at 97+ mph. They also knew the off-season work he’d done to try to cement those gains.
Hyde called on López against the middle of the Rays’ order in the 8th inning on Opening Day and quickly settled on López as his primary fireman.
López wielded his newfound high-90s velocity to be dominant in a way that he’d never been close to before. Through the end of June, he had a 0.73 ERA, 2.30 FIP, 64.0% ground ball rate, 27.1% strikeout rate, and 9.7% walk rate while holding opponents to a .125 average. Among 178 qualified relievers at that point in the season, he was tied for 7th in fWAR and 6th in win probability added.
Early July Hiccup
When the calendar flipped to July, things started to change for López. Ironically, the Twins played a role. López was tasked with protecting Baltimore’s 2-1, ninth-inning lead against the Twins on July 1st. Eventual AL batting champion Luis Arráez led off with a single, then Byron Buxton did this to a López slider:
The next night, López again came on with a one-run lead in the 9th. Jorge Polanco immediately tied the game with a homer and the Twins rallied to a walk-off win behind three more hits from Alex Kirilloff, Gary Sánchez, and José Miranda.
Two days later, in Texas, López entered a tied game in the 9th inning and promptly gave up a leadoff home run to Marcus Semien.
While López seemingly righted the ship the rest of the month, his July ERA was 4.76, with a 5.23 FIP, a .304 opponent batting average, and a 48.4% ground ball rate.
López in Minnesota
López’s first appearance for the Twins came the day after the trade deadline and he locked down a tidy, three-up, three-down save against Detroit. It was an outing that our Jon Gamble called “the fastest, most stress-free ninth inning of the Twins season” and prompted Sandwiches to satirize López learning the Twins’ heart-stopping ways in the bullpen.
But, López had a messy second appearance in a wacky Twins win over Toronto and then blew a late-night save (with some defensive “help” from Nick Gordon) in an extra-inning loss in Anaheim a week later in a game that was the beginning of the end of the Twins’ playoff aspirations. The rest of his season was more like July than the three months prior:
There was always going to be some regression from his excellent early season start because no one can sustain a .180 BABIP, but the disappearance of swinging strikes and strikeouts stands out in the data above. In general, López became more hittable later in the season.
Not The Usual Suspects
Statcast data shows that, by and large, there were not any obvious changes in López’s stuff. His velocity held consistently around 98 mph all season. He mostly maintained the same pitch mix: sinkers about half the time, his knuckle-curveball and changeup as the primary secondaries 15-20% of the time each, and 10% sliders for the rest. There were no significant changes in his spin rates on any of those pitches.
That said, it’s clear that something changed with the key part of his breakout — his sinker:
His knuckle-curveball and changeup essentially produced similar results. But his sinker missed fewer bats, was elevated more easily and allowed more extra-base hits.
More Vertical & More Down
In terms of horizontal movement, not much changed. But vertically, there is a clear trend that López’s sinker got more movement as the year progressed:
By season’s end, his sinker was dropping three more inches than it had at the start of the year. You might think, for a pitcher that relies on a sinker as his primary pitch, that more sink would be a good thing. But López doesn’t only use his sinker down in the zone in the traditional manner. One of his adjustments with Baltimore was taking that high-velo sinker up in the zone to his arm side (up and in to right-handed batters).
That extra downward movement coincided with more of his sinkers being located belt-high or below in the second half, as this comparison shows:
Before July 1st, 44.7% of his sinkers were located in or above the top third of the strike zone. After, 35% of his sinkers were up.
It’s not clear if the extra movement caused the pitches to be located lower or if it was an intentional approach change to locate them lower. What is clear though, is that the movement profile and location changes coincided with a subtle lowering of his vertical release point as the season went on:
The Y axis on the chart makes this look like a bigger change than it is, but it’s clear that López dropped his arm slot a few inches as the year went on. That likely played an important part in creating the movement changes and could have affected his command.
Purposeful or not, those changes proved meaningful because of how batters produced against López’s sinkers down in the zone compared to sinkers located up:
When López got his sinker up in one of the highest five sections of the zone graphic above, hitters batted .113 / .254 /.132 (.202 wOBA). When it was located in the lower eight sections, their line was .301 /.370 / .458 (.361).
Spin Direction, Vertical Approach Angle, and Deception
The installation of the upgraded Hawk-Eye cameras in the Statcast system in 2020 enabled measurement of the direction pitches were spinning as they were released from a pitcher’s hand. That data could be taken with the existing measurements of pitch movement to better understand what creates pitch movement and gave us a better ability to explain pitches with “deceptive” and “late” movements that seemingly deviated from what would have been expected. One of these effects has been popularized as “seam-shifted wake” from research by Dr. Barton Smith and Dr. Alan Nathan, Driveline, and others.
Jorge López’s sinker is one of those pitches, especially when it’s up.
In the first half, at release (spin-based movement in the chart below) López’s average sinker last season had an expected spin direction at 1:30 on a clock face. But, when it was measured at home plate (observed movement), it was spinning at 2:07 per Alex Chamberlain’s Pitch Leaderboard.
Generally speaking, the greater the deviation between the two measurements, the more deception the pitch has. When López threw his sinker lower in the second half, the deviation narrowed to 1:45 to 2:07.
Along those same deception lines, the vertical approach angle — something that makes Joe Ryan’s fastball unique — of López’s sinker shifted slightly from -5.4° early in the season to -5.6° later. That’s noteworthy because the league average vertical angle for sinkers last season was -5.8° and fastballs that are “flatter” than average tend to miss more bats, especially up in the zone.
Taking that together with the sub-optimal location changes helps to explain much of López’s loss of swings and misses. His lower sinkers were less deceptive and more hittable, which is clear in the chart of contact rates below:
Adjusting for 2023
Going into next season, the Twins are going to be relying on López to play a high-leverage role in their bullpen once again. On one hand, we can take away that López’s second-half struggles were not the result of any major reduction in stuff. The tweaks and improvements he made to break out with Baltimore were still present in Minnesota and it’s pretty clear that he’s not the same pitcher that was once waived by the Royals.
On the other hand, López does have some work to do to again thrive in a late-inning, high-leverage role. He’ll need to make some adjustments to either correct the things that changed with his sinker – which should be a reasonably doable thing – or lean into them and adjust his approach.
More than anything else, he needs to get back to pounding the top of the zone with his sinker. That will require mechanical and release point tweaks to flatten his sinker back out or find a new target to get it to land where he needs it to. If he’s not able to adjust that pitch to elevate it again, he’ll need to change, perhaps by embracing the extra sink and working further down in the zone in a more traditional sinker manner or by reducing his sinker reliance in favor of more breaking balls and changeups. Whatever the course, he has to consistently locate that sinker at the top or bottom of the zone, away from the hitters’ belt level. Doing that should help him to again miss more bats and be more like the Baltimore version of himself the Twins acquired.