In the three seasons that followed a disastrous rookie season, Twins starter José Berríos had solidified his place as a solidly above average major league starter pitcher. From 2017 to 2019, Berríos posted a 40-27 record, ranked 13th in innings pitched (538.1 IP), 42nd in ERA (3.80), 36th in FIP (3.87), and 19th in total fWAR (10.3) among 115 qualified pitchers. More granular batted ball data told an even more positive story. Berríos’ average exit velocity allowed (86.8 mph) tied for 12th best, and his hard-hit percentage (31.6%) tied for 19th. All in all, Berríos had become a modern-day workhorse, averaging 30 starts and nearly 180 innings pitched per year, with very solid (if not “ace” level) run prevention stats.
But still, we wanted and sometimes expected more. He had shown inconsistent flashes of being more. Because of his first-round pedigree and frequent public displays of putting in the work to get better, I think we have all been slightly holding our breath for Berríos to make the expected ascension to full-fledged playoff “Ace”.
That’s why the beginning to his 2020 campaign was so disappointing. On Opening Day, in a sign of what was to come, Berríos was spotted two different four run leads and gave them back. Ultimately, he lasted only four innings and allowed five runs with only 1 strikeout, settling for a no decision in a game an “Ace” would have won. His next four starts were generally similar, with walks and home runs allowed up, and strikeouts down. Through five starts Berríos’ uncharacteristic (and uninspiring) line looked like this:
Most of those figures would easily be career-worsts. Even casual observers could see that Berríos was struggling with his command. The walks made that obvious but process oriented stats like first pitch strike percentage and zone percentage supported that conclusion as well. His percentage of first pitch strikes was down to 60.4% from 65.2% in 2019, and his overall percentage of pitches in the strike zone had decreased to 39.0% from 43.4%. No doubt due in part to the lagging command, Berríos was also struggling mightily against right-handed hitters, allowing them a .400 / .446 / .580 triple slash line that worked out to a .436 wOBA.
Despite the struggles, there were a few signs that Berríos was still improving. After those first five starts, his swinging strike rate was 11.1%, which was in line with his career best. He was also displaying increased velocity on all his pitches, averaging around 95 mph on his four-seam fastball, 94 mph on his sinker, and pushing his breaking ball up to 84 mph. Those numbers would all be career-highs. After his fifth start, a 4 inning, 4 walk, 4 run dud on August 15 against Kansas City, Twins pitching coach Wes Johnson suggested to John Shipley of the St. Paul Pioneer-Press that Berríos’ stuff was better than the past and he was having a hard time harnessing it. “You look at his actual velocity and his stuff; his movement on his pitches is better than it ever has been, and I think that’s an adjustment,” Johnson said. “We’re doing some drills. We talk about it all the time. He’s throwing the ball fine. He’s really close, I think.”
Three weeks and three Berríos starts have passed since Johnson made those comments, and we can now look back and assess that he was right. In his last three outings, Berríos has looked much more like an improved version of himself:
Those results have been supported by an increased first pitch strike percentage (64.8%), an even further increased swinging strike rate (15.4%), sustaining of the velocity increases mentioned above, and a correction of his right-handed hitter struggles (.152 / .222 / .152). Johnson suggested Berríos was only an adjustment or two away and the numbers since those comments suggest he has found it.
What did he change?
Mound Position / Release Point
I hadn’t noticed anything different with Berríos’ delivery and arm slot while watching the recent games on TV. I had noticed his better command (and resulting better outcomes) and thought his delivery seemed more comfortable. Last Wednesday against Chicago, Twins’ broadcaster Bert Blyleven commented about Berríos’ starting location on the pitching rubber being to the extreme third base side. It was a passing comment, the kind meant to add detail to a broadcast, and Blyleven didn’t expand on it any further. But it was enough to pique my curiosity and make me want to see if something was different. Thanks to Fangraphs and Brooks Baseball, there is plenty of good data available to analyze for this question.
See the significant change in the plot? His horizontal release point reveals a clear change that occurred with his fifth start of the season against Kansas City – one start prior to his results starting to turn around. That is clearly more than just usual variance. Beginning with that fifth start Berríos has been releasing his pitches nearly a full foot farther to the third base side than he was earlier this year.
Here’s another view, from Fangraphs. See how there are two distinct groupings of data points, about a foot apart?
From this data we can also quickly see his vertical release point has not changed. Those plots are showing normal variation around 5.4 vertical feet or so. Given that his vertical release has not changed measurably and we haven’t seen any visual evidence on the broadcasts of a dramatic arm slot shift, I think we can confidently conclude that this release point change is not due to his mechanics. This looks like a deliberate shift of his position on the rubber.
To support that claim, I went looking for some visual evidence. Here’s a picture taken before he made this adjustment, during his third start against Pittsburgh, on August 4. Look at his right foot and its location on the rubber.
His whole foot is on the rubber and you can even see a little bit of the rubber peaking beyond the toe of his right cleat. From the cleat marks left in the dirt, you can also see he was landing with his front foot almost directly down the mound in line with the third base edge of the rubber.
Next, here’s a photo taken after his adjustment, during his 7th start, at Cleveland on August 26.
While the angle this photo was taken from is a little more to the third base side than the one above, Berríos is at a similar time in his delivery and you can see he has shifted his starting position to the third base side of the rubber. Here, only the heel of his right foot remains in contact with the rubber.
Finally, this one was taken directly behind home plate when he faced Chicago on September 2.
This one does the best job of showing the full degree of the change. The cleat marks up near the rubber clearly show that he was starting with most of his right foot off the rubber to the third base side. In comparison to the Pittsburgh start, you can also clearly see how far his landing spot had shifted toward third.
This all makes it clear that Berríos has made an adjustment to his mound position. The change did not yield immediate dividends – after all, his first start utilizing this new mound position was that frustrating outing against Kansas City. But there were some signs in that start that the change might work out well. For one, Berríos threw the highest percentage of pitches in the strike zone that he had all season to that point (42.9%). He struck out 7 of the 20 batters he faced that day, also his highest rate of 2020. Of the 9 pitches that he allowed the Royals to put in play the average exit velocity was just 87 mph, and the maximum exit velocity allowed was 98 mph, which was easily the lowest amount and speed of hard contact he had allowed all season. All the results in his three starts since suggest this change has been a good one.
Pitch Mix
In addition to the position adjustment on the mound, Berríos also tinkered with his pitch mix around the same time. Beginning with his fifth start against Kansas City, Berríos started to swap out four-seam fastballs for sinkers and curveballs. Focus on the relationship of the dark blue (four-seam fastball) and red (sinker) lines in the plot below.
By Fangraphs Pitch Info data, in his first four starts Berríos threw his four-seam fastball more than his sinker in each start, averaging 32% overall. In the four starts since he’s used his four-seamer just 18.3%. Correspondingly, his sinker usage increased. In his first four starts he used his sinker 20.7%. Since, that figure has increased to 29.4% and has been utilized more often than his four-seam fastball in each game. His outstanding curveball (yellow) has picked up the remainder, increasing from 27.7% to 33.7%.
Why it works
Rather than credit one adjustment over the other, I think it’s the interplay of the mound positioning and pitch mix changes that has led to José Berríos’ recent improvement. The logic behind the pitch mix adjustment seems to be pretty clear – emphasizing his best pitches. Fangraphs’ calculated pitch type linear weights make plain that his four-seam fastball has been far and away his worst offering so far this year.
By minimizing that offering, he’s found slightly better results with it. With the new mound position, he’s also gotten improved results with sinkers and changeups, while mostly maintaining his already outstanding breaking ball.
Changing his starting place on the mound has changed the hitter’s visual of his curveball. As a result, he’s now getting more swings and misses on his best strikeout weapon.
Against right-handed batters, it looks like it’s going to be in the strike zone before it sweeps away and out of the zone late in its trajectory. Against lefties, it looks like it’s going to be a ball outside the zone before it breaks over the plate. In either case, it’s been getting an awful lot of awkward swings, as our friends from Chicago showed us last week:
José Berríos, Disgusting Curveballs (2nd Inning). pic.twitter.com/4o0ZrfL0ew
— Rob Friedman (@PitchingNinja) September 3, 2020
In some ways, these adjustments are tried and true. As the video above shows, Berríos’ curveball is a big sweeper, featuring a significant amount of horizontal movement. He’s maintained its shape while increasing its velocity the past few years. It’s to the point where you could now reasonably think of his curveball more like a traditional slider. If you accept that and start to think about what pitches pair best with big breaking sliders, you quickly would come to the idea that you should want something that breaks hard the opposite direction, like a sinker. Sinker-slider is one of the most common pitch pairings in baseball for a reason – because they work well together. The combination gives weapons to work both sides of the plate and keep hitters honest. If they start to cheat by leaning out over the plate to get to the breaking ball away, you can do this:
José Berríos, Nasty 95mph Two Seamer. pic.twitter.com/htMyRC3lqK
— Rob Friedman (@PitchingNinja) September 3, 2020
For Berríos, the sinker and curveball play well together because of his unique breaking ball. Shifted further on the rubber has just maximized the effect of their opposite movement profiles, making them more effective. He also has the above average changeup and retains the now mid-90s four-seam fastball to work up in and above the zone when hitters are cheating down. So far, the mid-season changes he’s made seem to have made his whole arsenal better and more optimized for the pitches relative to one another. Strikeouts, groundballs, and limiting walks are a recipe for pitcher success. As a result of his adjustments, Berríos has improved in all three areas. He’s striking out a career best 10.07 batters per nine innings for the season (and 11.72 K/9 in his last three starts) and getting ground balls at an increased rate. His 2020 season total is up to 44.1%, which is a career best, and it’s been an even higher 51.3% in his last three starts. He’s cut his walks from 4.81 per nine during the five bad starts to a more palatable 3.57 per nine in his last three outings. That’s still too high, but perhaps it will come back down towards last year’s 2.29 as he gets more comfortable with his new mound position.
The combination of changing his starting position to get his pitches on a more difficult line for the hitter to discern ball or strike and working his pitches off one another more effectively has been the key that’s unlocked José Berríos in 2020. It remains to be seen if these are the tweaks that will help him reach the next level we’ve so long expected, but it’s another piece of evidence that he’s capable of and committed to making needed improvements.
John is a contributor to Twinkie Town with an emphasis on analytics. He is a lifelong Twins fan and former college pitcher.