Exploring Optimal Fastball Traits Through the Approach of Luis Castillo

How Luis Castillo mastered the art of pitching.

“See the ball; hit the ball,” said the great Pete Rose. If only it were that easy––at least against Luis Castillo.

You may know the Reds ace for his highlight-reel changeup, missing bats with extreme levels of arm-side run and vertical drop that few pitchers rival. Perhaps you are fond of his slider, located optimally for both a high ground-ball rate and a high called-strikes rate. While the value of good breaking and off-speed pitches cannot be understated, Castillo’s fastballs are not only a primary source of success but also a blueprint for deception that pitchers of all ages can study to understand how to maximize the difficulty of both seeing the ball and hitting the ball.

Fastball velocity can carry pitchers far in their journey to the pinnacle of baseball, but optimizing fastballs only begins with velocity. The reason Castillo’s fastballs serve as an ideal model for developing optimal fastball characteristics is that he couples his velocity with a pitching approach that reflects a nuanced understanding of not only how to retire hitters but also how to control both opposing quantity of contact and opposing quality of contact. A pitcher with Castillo’s fastball velocity is certainly capable of reaching the big leagues, but it is Castillo’s pitching approach — centered around tunneling his fastballs and optimizing his fastball shapes and locations — that has facilitated his mastery of the art of pitching.




Castillo throws a four-seam fastball at a 26.5 percent usage rate and a sinker at a 24.8 percent usage rate, but to most hitters, detecting which fastball variant is coming their way becomes plausible only at a relatively late stage in the pitch’s flight. To that end, Castillo has implemented the recently popularized idea of pitch tunneling — throwing two pitches along nearly identical trajectories before they diverge — into his fastball strategy.

Luis Castillo’s Pitch Characteristics at Release, Courtesy of Alex Chamberlain

Out of Castillo’s hand, both his four-seam fastball and his sinker feature nearly identical velocities and extensions at practically the same release point. Additionally, the initial spin axes of the two pitches are extremely similar, indicating that hitters should struggle to identify the pitch not only at the point of release but also through the early stages of the ball’s journey home.

Luis Castillo’s Pitch Shape Metrics, Courtesy of Alex Chamberlain

To Castillo’s benefit, his four-seam fastball features far more induced vertical break and far less arm-side run than his sinker. Whereas his sinker is an especially effective deception tool horizontally for Castillo, his four-seam is his vertical weapon. In other words, Castillo’s four-seam fastballs that appear to be sinkers are extremely difficult to connect with because they do not sink, and Castillo’s sinkers that look like four-seam fastballs are very difficult to barrel because of their additional run into right-handed hitters and away from left-handed hitters. Therefore, although hitters may be able to identify a fastball out of Castillo’s hand, their inability to pick up on the specific variant limits both their ability to make contact and their ability to produce on contact.


High and Flat Four-Seam Fastballs


The list of reasons Castillo’s tunneling is such an effective asset extends far beyond the break profiles of his fastballs. While a four-seam fastball with much more induced vertical break than a sinker is difficult to catch up to, the flat shape of Castillo’s four-seam fastball — juxtaposed with the steeper shape of his sinker — adds a new dimension to the effectiveness of Castillo’s deception; it controls opposing launch angles.

Controlling opposing launch angles is an extremely difficult maneuver for pitchers because the optimal launch angle for hitters depends heavily on the vertical approach angle of a pitch. In its truest form, hitting is a test of one’s ability to match the plane of the bat to the plane of the ball. A bat with an attack angle equal in magnitude to the vertical approach angle of the pitch is considered perfectly squared up, whereas a bat path that is much flatter or steeper than the pitch path will likely result in either weak contact or a whiff.

Therefore, to maximize their effect on opposing launch angles, pitchers must throw their pitches at vertical approach angles that are too difficult to match — specifically, vertical approach angles that fall outside of or near the boundary of the 2-to-16-degree range deemed standard for MLB hitters by Blast Motion.

Nearly every MLB pitcher who throws a four-seam fastball deploys it at a vertical approach angle between -3 and -7 degrees. That said, while the variance of attack angles is likely far greater than that of four-seam vertical approach angles, the variance in the effectiveness of fastball vertical approach angles is stark. Since MLB attack angles are centered around 8 degrees, fastballs that approach the zone closer to the -7 degree boundary are far easier to square up than fastballs that approach it at the -3 degree boundary. Thus, while steeper fastballs have a suboptimal shape for generating weak contact, flatter fastballs will almost always be met with attack angles of greater magnitudes than their vertical approach angles, inducing swings under the ball that generate weak fly balls and whiffs at high rates.

Furthermore, Castillo’s four-seam fastball shape is optimal for inducing whiffs and fly balls because his four-seam is extremely flat, a product of his high velocity and low release point. Not only is the vertical approach angle on his four-seam near the -3 degree boundary at -3.9 degrees, but also, his height-adjusted vertical approach angle (vertical approach angle above average) is 0.64 degrees, indicating that his fastball is 0.64 degrees flatter than fastballs located similarly.

To optimize Castillo’s four-seam location, one must conceptualize the effectiveness of a flat four-seam at different heights. Although Castillo’s four-seam is consistently flat relative to its location, low four-seam fastballs will naturally be steeper than high four-seam fastballs. Thus, to maximize flatness, Castillo would leave his fastball in the upper third of the zone or above it.

Additionally, let’s revisit the aforementioned definition of hitting: matching the plane of the bat to the plane of the ball. Any four-seam fastball left over the middle of the plate, where swings are most natural, would be far easier to connect with than a four-seam left high or low. Because high four-seams are flatter than low four-seams and are as difficult to catch up to as four-seams in any other location, high four-seam fastballs currently reign supreme in the art of pitching.


The numbers confirm the proposed effect of Castillo’s high and flat four-seam fastball. Thrown at a vertical approach angle above average of 0.64 degrees and a high-location rate of 60.6 percent, Castillo’s four-seam generates fly balls at a 52.4 percent clip, well above the league-average 41 percent four-seam fly-ball rate. Additionally, despite leaving the pitch in the zone at a respectable 47.2 percent rate, Castillo leads all of the 277 pitchers who have thrown at least 150 four-seam fastballs in four-seam whiff rate, with a tremendous 45.2 percent mark. In fact, no other pitcher in this bunch has posted a four-seam whiff rate over 35 percent this season, a testament to the utter dominance of Castillo’s high and flat four-seam fastball.


Steeper Sinkers


Castillo’s sinker shape tells a slightly different story. Whereas Chamberlain’s research shows that steep sinkers at the bottom of the zone are optimal for controlling launch angles — inducing a high proportion of ground balls — Castillo has recorded a strong 59.3 percent ground-ball rate on his sinker despite its flat 0.43 degree vertical approach angle above average and its relatively average 41.8 percent low-location rate.

Sinker Ground-Ball Rate Leaders Among Starting Pitchers (min. 150 sinkers)

Comparing Castillo to the 15 starting pitchers who rank atop the league in sinker ground-ball rate reveals two telling truths. First, Castillo’s sinker is significantly flatter than each of the sinkers above. Second, of these 15 sinkers, only Robbie Ray’s is flatter than average, and even Ray’s sinker features a nearly average degree of flatness.

Although Castillo’s 59.3 sinker ground-ball rate does not rank among the very top of the league, his high ground-ball rate on a very flat sinker promotes the notion Castillo does not align with previously developed theories about ground-ball-inducing steep sinkers; rather, he is an exception to the rule.

To understand the primary reason Castillo breaks a rule applicable to nearly every aforementioned pitcher, consider the arsenals of the five leaders in sinker ground-ball rate among pitchers with flatter-than-average sinkers. Robbie Ray, Kyle Wright, Aaron Sanchez, Michael Lorenzen, and Frankie Montas all throw four-seam fastballs at release points, extensions, velocities, and spin axes nearly identical to their corresponding sinker characteristics.

Conveniently, their four-seam and sinker movement profiles all differ by at least four inches of induced vertical break and at least five inches of horizontal break and feature steeper sinkers than four-seam fastballs, reintroducing the concept of tunneling and consequently bringing the discussion of Castillo’s success full circle. Because these pitchers tunnel their flat sinkers with a flatter four-seams, they generate swings over their sinkers, consequently inducing ground balls at rates competitive with those of some of the steepest sinkerballers in baseball; in this department, Castillo is the cream of the crop.

The progress Castillo has made since his first full season in 2018 has been nothing short of admirable. He added almost a full tick to his four-seam fastball and lowered his four-seam release point by 0.4, which flattened his fastball from 0.31 degrees above average to 0.64 degrees above average. Additionally, he balanced his sinker and four-seam fastball usage to maximize the effectiveness of his tunneling. Perhaps the most notable adjustment Castillo made is nearly doubling his high-location rate on his four-seam fastball from 31.5 percent to 60.6 percent, wholly buying into one of the rapidly developing theories in modern pitching.

Castillo has always had the talent to be successful, but the maxim “hard work beats talent if talent doesn’t work hard” will always prevail. Thankfully, Castillo’s hard work has skyrocketed him to the summit of the baseball world.

Aidan Resnick

An aspiring sports analyst, Aidan is a sophomore at the University of Chicago, studying statistics, computer science, and economics. In 2019, he attended the Wharton Moneyball Academy, the Carnegie Mellon Sports Analytics Conference, and the MIT Sloan Sports Analytics Conference, which inspired him to pursue sports analytics. Since then, Aidan has displayed his passion for sports analytics in his newest book, "The Stats Game," where he and his twin brother illuminate statistical tools and debunk myths in sports analytics, in his victorious Diamond Dollars Case Competition project, and in the Resnick Player Profiles, an interactive dashboard that visualizes modern baseball statistics. At Pitcher List, Aidan strives to create content that both builds on preexisting discoveries in the analytics revolution of the 21st century and introduces new methods of analyzing baseball.

One response to “Exploring Optimal Fastball Traits Through the Approach of Luis Castillo”

  1. Andrew says:

    Really interesting article, but I wish you explained the pitch metrics a little or included a diagram so I can understand what they mean. I’d say I’m pretty familiar with math and baseball terminology as a fan, but I had to look up induced vertical break, and reconcile the sign differences between vertical approach angle and attack angle to really understand what you’re saying. I get the geometry, but I’m still not sure why the VAA of a descending ball is defined as negative, when it applies to the vast majority of pitches. I didn’t even bother with the spin axes. I’m looking forward to reading more articles with this kind of analysis, but hopefully presented in a more accessible way.

Leave a Reply

Your email address will not be published. Required fields are marked *

Account / Login