The Twins went 3-5 last week on their eight-game road trip through Detroit and Toronto. Despite that tough stretch, they only dropped a half-game off their division lead and saw no meaningful change in their playoff odds (still ~58%, down slightly from ~60% last weekend. My major takeaways from the standings this week are that the Twins are fortunate to play in the American League Central division and that Rocco Baldelli’s club is resilient in a way that seems to make them greater than the sum of their parts.
That the skeleton crew roster caused by COVID (in more than one way) went into Toronto and won a series by knocking around Kevin Gausman in the rubber match is nothing short of impressive. Early in the week when the injuries and illnesses kept piling up that kind of series win seemed improbable. Most of us probably just expected the Twins to go into survival mode until the healthy reinforcements can come back. Now, I think they did slightly better than that.
Gilberto Celestino was activated from the COVID IL Saturday and played yesterday. Joe Ryan was considered “a tick behind” Celestino and is expected to ramp up his workload early this week. Carlos Correa is on a similar timeline. The team will also get another four regulars back when they return home from Canada. It was a tough week, but in the big picture, the Twins are mostly in the same (good) position they were when the week began.
This week: vs. New York Yankees (3 games), vs. Tampa Bay Rays (3 games)
2. Celestino’s Development
Speaking of Celestino, what a difference a year makes. The young outfielder experienced significant growing pains in his brief rookie campaign with the Twins last year, but he’s grown and developed from those experiences to be an important contributor so far this season.
Celestino was pressed into emergency duty much earlier than anyone would have wanted last season, jumping to the Twins from AA-Wichita. Over 23 games and 62 plate appearances, Celestino was clearly overmatched at the plate, hitting an almost 80% below league average .136/.177/.288, (22 wRC+). This season, Celestino’s .350/.416/.400 line over 91 PAs (through June 5) is well above average (145 wRC+).
His line is probably a little bit misleading because he’s been quite fortunate in the batted ball luck department (.459 BABIP) and his expected slash stats per Statcast are a fair bit below his actual numbers. His contact quality numbers (i.e., exit velocities, launch angles, hard-hit rates, etc.) also do not show any demonstrable changes from last season that would help to explain his strong start.
That said, from the eye test, Celestino looks much more comfortable and competitive at the plate. He has struck out a little bit less often and walked a little bit more thanks to some improvements in controlling the strike zone. His overall contact rate is up a few ticks to 84% of his swings, meaning his whiffs are down. Celestino did an admirable job for a rookie jumping from AA to avoid chasing pitches out of the zone last season (26.3% vs. league average of 28.3%) and he’s improved significantly in that category this year (16.2% chase rate).
Perhaps even more than his solid performance at the dish is that Celestino is proving to be a well-above-average defender. Statcast has him in the 89th percentile in its Outs Above Average metric and he’s graded above average in both center and left fields. The plus defense in multiple spots, combined with average-ish or better offense, is a luxury to have from a 4th outfielder and the overall future outlook for this 23-year-old outfielder is pointing up.
3. Larnach’s Year Two Rebound
The same kind of story applies to Trevor Larnach. He started out very strongly for the Twins last season — despite having played just 43 games above A-ball — but struggled as the season went on and the league adjusted its approach to pitching him. The book on the former Oregon State Beaver was that he would mash fastballs, so pitchers fed him an unhealthy diet of breaking balls and changeups. Larnach swung and missed on about 55% of his swings against those offspeed pitches and eventually stopped hitting fastballs, too. For the season, Larnach ended with a .233/.322/.350 line (89 wRC+) with 19 extra-base hits over 301 PAs.
This season, Larnach currently is tied for 2nd with Max Kepler (and trailing Byron Buxton) on the Twins with 1.3 fWAR. Through 37 games and 131 PAs, Larnach has slashed .261/.336/.470 (130 wRC+) and knocked 16 extra-base hits, including this rocket off a frisbee breaking ball yesterday:
Swinging and missing is still a major part of his game (and always will be). With that heavy three true outcomes profile, as I explained in the link above last season, Larnach has to do a lot of damage when he does make contact and his batted ball data this season suggests he’s doing better in that regard. His barrel rate (6.9% per PA), sweet spot rate (37.8%), and 92.0 mph average exit velocity (up from 90.3 last year) have all increased from last year and indicate he’s making higher-quality contact. The average exit velocity and Larnach’s expected wOBA on batted balls (.475 as measured by the mouthful stat known as xwOBAcon) rank within the top 8% across the game.
Maybe more impactful is that Larnach has not been a complete black hole against spin. He’s still much weaker against breaking balls and changeups than fastballs, but he’s raised his numbers across the board against breaking pitches (.256 BA, .538 SLG) so far this season, including hitting seven of those extra-base hits.
Larnach has also added value in the field, drawing average or better marks across all three of the major defensive metrics (+6.2 UZR, +11 DRS, 0 OAA) and already racking up three outfield assists.
In total, he’s not been a dramatically different player than he was last season, but he’s made incremental improvements in all the areas he needed to. And, like Celestino, he looks much more comfortable and in control compared to last season.
4. Kirilloff Time?
It was next to impossible for much of last season to talk about Larnach without also talking about Alex Kirilloff. Like Larnach, Kirilloff had an outstanding but short-lived debut before injury struck. Since re-injuring his wrist last May, Kirilloff has not hit with any authority. You can divide his major league career into three stretches and clearly see the impacts of the injury:
Pre-wrist injury (44 PAs through May 3, 2021):
- 96.5 avg. exit velocity (EV), 11.2° launch angle (LA), 171 feet avg. distance, .357 ISO
Post-wrist injury playing through (189 PAs after May 3, 2021):
- 89.9 Avg. EV, 6.0° LA, 150 feet avg. distance, .189 ISO
Post-wrist surgery (32 PAs in 2022 Season):
- 86.7 Avg. EV, -2.1° LA, 121 feet avg. distance, .000 ISO
You can probably infer from those numbers that the wrist injury completely wrecked Kirilloff’s ability to lift the ball with any power. As his percentage of ground balls has gone up, his hard contact has gone down:
Now, though, there might be some light at the end of the tunnel. Since being optioned to AAA St. Paul for a little rest and rehab after a new cortisone injection in the ailing wrist, Kirilloff’s bat has come to life:
Since May 26, Alex Kirilloff is hitting .438/.538/1031 at Triple-A St. Paul with four doubles, five homers, 13 RBIs, 14 runs and seven walks in 39 plate appearances. He was in Detroit on the Taxi Squad earlier this week for Tuesday only, was sent back. #MNTwins— DanHayesMLB (@DanHayesMLB) June 5, 2022
Kirilloff hit home runs on Friday, Saturday, and Sunday (two) for the Saints. It seems he’s again hitting the ball in the air and might be forcing the issue for the Twins to bring him back to the majors. Perhaps we’ll see him on Tuesday for the start of the series with the Yankees.
5. Twins Links
- Devin Smeltzer has been just what the Twins needed in the starting rotation the past few weeks. Smeltzer’s 1.93 ERA and five useful starts also mean his story has drawn good coverage. Dan Hayes had a great profile piece on Smeltzer’s recovery last season from a hard-to-find herniated disk that caused him to collapse during a bullpen for The Athletic.
- Randy Dobnak has suffered another setback in his rehab from surgery to fix a torn pulley in his right middle finger. Meg Ryan has great details about the original injury and a subsequent tear in the same finger. I hadn’t realized that Dobnak had suffered two different tears in that finger.
- Right-hander Jorge Alcala started a rehabilitation assignment with Class A Fort Myers earlier this week. Here’s hoping that goes well and he can soon add some much-needed depth and competence to the Twins’ battered relief corps.
6. Around Baseball Links
- Miami’s Jesús Sánchez launched a 496-foot home run at Coors Field last week, which prompted MLB.com to recap the ten longest home runs tracked in the Statcast era. The list includes two Miguel Sanó blasts
- If you have had a chance to check out the new Friday night MLB broadcast on Apple TV+, you no doubt have taken notice of the real-time data and probabilities that they’ve integrated into the broadcast. If you paid close attention, you might have noticed some situations where those probabilities did not make sense — such as a player’s chances of getting on base increasing while he fell behind in the ball-strike count. FanGraphs’ Ben Clemens, with the help of the crowd, analyzed those probabilities. It turns out that they might not be very good.
- The CEO of the company that produces those numbers for MLB Network, Kelly Pracht of nVenue, joined the Effectively Wild podcast at FanGraphs to discuss the company’s work and Clemens’ findings. It was an interesting listen that did little to convince me that the models being used right now to produce the probabilities for the broadcast are not just the result of overfitting and interpreting signals from noise. That said, I love the idea of real-time probabilities as part of a baseball production and have often wondered whether articulating the batter’s percent chances of doing different things during a plate appearance would be more accessible to casual fans than the number-salad of unintuitive stats we use today. But, it has to be done well to work. This is a step in the right direction and hopefully can be tweaked and tuned for better accuracy going forward.