Luis Garcia Is Two Pitchers in One

After losing Gerrit Cole in free agency after the 2019 World Series and Justin Verlander to a torn UCL just six innings into the season, the Astros found themselves in desperate need of pitching help last year. Zack Greinke helped fill the void left by Charlie Morton’s departure, and Lance McCullers Jr. returned after missing all of 2019, but besides those two pitchers, there were more questions than answers in Houston’s rotation.

Insert a quintet of young pitchers with varying degrees of experience: Framber Valdez (107.2 MLB innings to his name), José Urquidy (41 MLB innings and a stellar upper-minors track record), and three pitchers with a lot of blank space on their résumés in Cristian Javier, Enoli Paredes, and Luis Garcia. With the help of these five and some clever piggybacking, the Astros overcame a mediocre regular-season record to oust Minnesota in the wild-card round and Oakland in the ALDS before falling to Tampa Bay in the ALCS.

Of that group, I want to focus on Garcia, for two reasons. First, he is the team’s current leader in pitching WAR, narrowly edging out Greinke, with a 2.82 ERA supported by a 28.1% strikeout rate and 7.7% walk rate. Second, his rise to the big-league club was the most surprising of the five guys I mentioned above. He signed with Houston out of Venezuela at age 20 (which is old for an international amateur) for a mere $20,000. He had not pitched above high-A before his MLB call-up, though he did dominate the opposition at every level. In 2019, he saw a velocity bump and struck out almost 36% of the hitters he faced in 43 innings pitched at low-A, then whiffed 39.4% of batters he faced in 65.2 innings one level up.

Garcia’s performance put the baseball industry on notice. He went from not appearing on our 2019 Astros prospect list to the No. 19 spot in the 2020 edition and a 40 Future Value grade. After losing Verlander, Urquidy, Chris Devenski, Roberto Osuna, and Brad Peacock to injury in the first month of 2020, a desperate Houston plucked Garcia from its alternate site.

Garcia tossed just 12 innings for the Astros and posted a 2.92 ERA, but he struck out only 18.4% of opposing hitters and handed out free passes at a 10.2% clip. That’s changed this year, as he has added almost 10 percentage points to the strikeout rate and trimmed a few from the walk rate, creating a more sustainable profile that supports a park and league-adjusted ERA 31% better than average and an adjusted FIP 11% better than league average. He got there with a deep mix of five pitches, all of which he throws with some frequency and all of which are at least average overall:

(For context, a league average CSW rate is about 28%.)

The four-seamer is Garcia’s most-used pitch and gets a below-average number of swings and misses (the average for fastballs is about 10%), which can be attributed to below-average velocity and, correspondingly, spin (though he does get a couple more inches of ride than average). He also throws his four-seamer in the zone more than 60% of the time, almost 25% more than league average, which dampens its whiff rate.

I do think this aggression with the fastball, though, allows the curveball to play up. The two pitches have spin axes that are just about 180 degrees apart; he mirrors them well out of his hand. As for the rest of his arsenal, the cutter has been able to generate a ton of whiffs despite rarely being in the strike zone, and the slider is his best pitch by CSW% at a slightly below-average rate as a percentage of overall pitches but at an enormous rate as a percentage of all swings.

Looking at how often he uses each pitch, you would think that Garcia is among the most unpredictable pitchers in MLB. But check out his pitch usage by batter-handedness:

It is almost as if he is two different pitchers depending on where the opposing batter is standing. Against left-handed hitters, he basically scraps his slider and rarely throws his cutter, primarily going with the four-seamer, curveball and change. Leaning on the latter two makes sense, given that curveballs and changeups work well against opposite-handed hitters. Turfing the cutter, on the other hand, is mildly surprising, since that is a pitch type that generally displays neutral platoon splits. When a righty steps into the box, meanwhile, Garcia rarely throws the changeup or curveball and relies on the four-seamer, cutter and slider. When you peek at his overall arsenal, you would think he has an egalitarian pitch mix of five offerings. Instead, he is two three-pitch pitchers put together.

The divergence in Garcia’s approach against righties and lefties raises the question: What are his platoon splits? With the caveat that he has thrown 70.1 innings this year and only faced 285 batters, he has struck out 24.8% of the left-handers he’s faced and walked 4.8%, and against right-handers, it’s 30.6% and 10%, respectively. The K-BB% figures are similar (within 0.6 percentage points), though the shapes are different. The overall results, however, tell a different story:

Compared to the rest of the league, Garcia has a massive split in his wOBA results by handedness. That comes down to getting swings and misses 54% more often against righties than lefties, which he does by rarely throwing the ball in the strike zone against righties aside from his four-seamer. Against lefties, though, he attacks the zone with that pitch, and it simply hasn’t worked:

Left-handed hitters have a .434 wOBA against the four-seamer and have whiffed on only 6.2% of those pitches. That’s part of why Garcia’s walk rate is muted when facing lefties, but his approach has resulted in fewer strikeouts and worse overall results. Worse yet, in two-strike counts the fastball is still overwhelmingly his most-used pitch despite how well his other pitches have performed:

If Garcia were more aggressive trying to induce chases or reduced the usage of his fastball, the split would probably begin to narrow, as the curveball and changeup have performed extremely well against lefties. On the other side of home plate, his cutter has been unstoppable. The book should start to come out on that pitch, though, as he only throws it in the zone 26.4% of the time against right-handed hitters. A correction there would lower his chase rate and, in turn, the swinging-strike rate, hurting his success against righties and narrowing his platoon split. The same holds for his slider, his other potent offering against same-side hitters, which has generated a ton of called strikes even though he only throws it in the zone 43.2% of the time. Opposing right-handers are swinging at only 33.5% of his sliders, but that should creep up as they realize how effectively Garcia can steal strikes with that pitch.

Another source of his platoon splits narrowing? The expected results for each handedness have been relatively similar, no matter how you slice it:

The differences here are stark. Whether talking about overall xwOBA, xwOBAcon, or xwOBA on pitches in and out of the strike zone, Garcia has vastly underperformed against lefties and overperformed against righties. And while xwOBA is not a predictive stat, it can help us see and describe the fact that his platoon split has not been as drastic as you would believe looking at the surface-level numbers.

Garcia is a fascinating pitcher. He was signed at an unusual age for very little money, came out of nowhere to appear on the prospect radar, made it to MLB without ever throwing a pitch at or above Double A, and has been a stalwart for a club with the best run differential in the league thus far. And his approach is as atypical as his path. He is comfortable throwing five different pitches, but depending on where the opposing batter stands, he’s two different versions of the same pitcher. The story of his rise to prominence is a good one in and of itself. The method by which he has gotten here, however, is just as fun.