Derek Holland Throws a Unicorn Pitch

In 2018, Derek Holland was given a second chance to revive his baseball career.

Holland was a former star for the Texas Rangers, helping them clinch two AL Pennants in 2010 and 2011, but he had fallen on hard times.

The 2017 season represented a massive struggle for the southpaw. Statistically, he was one of the worst pitchers in the Majors. But in 2018, he recovered, turning in a quality campaign for the Giants. His batted-ball metrics bounced back and he solidified himself as a useful backend arm.

Fangraphs‘ Jeff Sullivan explored Holland’s resurgence, noting that he changed his mound position mid-season. While he couldn’t conclude that’s exactly what led to Holland’s rebound, it’s tough to ignore the fact that virtually all of his metrics trended in the right direction, post-move.

Then, things took a turn for the worse.

2020 was a nightmarish year for all of us. For Holland, hell came a year early.

His ERA ticked up almost two and a quarter runs as opposing hitters made harder and more consistent contact against all of his offerings. His performance warranted a permanent move to the bullpen, a first for the veteran. While his 4.40 ERA was nothing to write home about, his .263 wOBA and 8/1 K/BB ratio were two bright spots in an otherwise dismal season.

He was promptly DFA’d mid-season and landed with the Cubs for the remainder of the season. In free agency, he latched on with the Pirates in time for the 2020 season.

Holland’s 2020 was interesting. Baseline stats: he posted a 1-3 record with a 6.86 ERA/6.14 FIP, finishing the campaign worth -0.3 fWAR. It certainly wasn’t the platform year anyone, much less Holland, desired.

Derek Holland’s 2020 Baseball Savant page

Despite the high ERA, I don’t think Holland’s career is finished. A couple of his pitches feature one-of-a-kind qualities, traits that a data-driven team could certainly take advantage of.

The intrigue doesn’t lay in his statistics; it lays in the physical properties of his pitches.

Holland’s always been a sinker-heavy pitcher, relying on it approximately 60% of the time while with the Rangers. But since his time with Texas, he’s incorporated a more progressive arsenal, mixing in his offspeed pitches more often.

In 2020, Holland’s arsenal consisted of a sinker, a fastball, a curveball, and a changeup.

From a movement standpoint, there’s some promise in his sinker. Adjusting for velocity, its horizontal movement was 9% above-average. There’s some work to do with his location of the pitch, but nothing that can’t be fixed. When Holland was in San Francisco, his sinker’s location trended towards the edges of the strike zone. When he was with Pittsburgh, the pitch was located more in the middle of the plate and got tattooed as a result.

His changeup was a nice fourth pitch during his time as a starter, but relievers can rely on their two top pitches and succeed. Going forward, I’d recommend scrapping that pitch entirely.

He also threw a four-seam fastball (located on the previous graph towards the lower right) 2.3% of the time, but for this piece, I’m going to focus on the bigger picture of his arsenal.

What makes Holland unique is that he’s an expert at exerting active spin on the majority of his pitches. (Writer’s note: Going forward, I’m going to be using the terms “Active Spin” and “Spin Efficiency” interchangeably)

He was the only pitcher in the Majors in 2020 to record spin efficiencies of 98% or better on three of his pitches.

Despite each pitch recording a spin efficiency of 98% or greater, they each didn’t fare well. Usually, certain pitches, like fastballs, sinkers, and change-ups, perform better if they have a higher spin efficiencies, but in this case, the poor movement profile of each delivery contributed to opposing hitters being able to make consistent, hard contact.

Granted, a high spin efficiency on a pitch isn’t everything, but if utilized properly, it can certainly behoove the overall profile. But as we see, there’s one pitch that doesn’t adhere to “throw everything with a high spin efficiency” rule that Holland imposed on himself.

That’s his curveball.

Well…. the thing is, it’s not really a curveball. It’s actually a slider.

Baseball Savant stopped tracking Holland’s slider at the end of the 2019 season. Instead of thinking that his slider did a Houdini, I’m of the belief that Savant’s new-and-improved tracking system is responsible for the current classification of Holland’s slider/curve hybrid. Furthermore, the spin efficiency on Holland’s pitch resembles a slider, more than of a curve.

I faced the same dilemma when I explored Jacob deGrom’s cutter/slider hybrid.

But as long as Derek Holland calls it a curveball, it’s a curveball. Who am I to argue with a 6’2″, 233 lb Major-League vet?

Anyway, his curveball had a 14% spin efficiency, a mark that was the lowest in the Majors in 2020. An extremely low spin efficiency on a pitch, coupled with a poor movement profile, designate the pitch as a “gyro” ball. For those who are interested in learning more about gyro balls, Ben Clemens wrote a really informative article on Fangraphs.

Gyro balls are thrown with gyro spin, which doesn’t impact the pitch’s movement (also called bullet spin). Most pitches that are thrown with more active spin (the opposite of gyro spin) fare better as they influence the pitch’s movement, retain the shape of the pitch better and generate more spin.

But, some pitchers, including Nick Anderson and Derek Holland, throw pure gyro balls.

The aforementioned Clemens looked into Anderson’s curve/slider hybrid, one that’s enjoyed great success and noted its unique features. Anderson’s pitch, much like Holland’s, has minimal movement, but unlike Holland, has enjoyed great success.

That’s not the only unique trait of Holland’s curve.

Baseball Savant’s latest feature Spin Direction gives us more data about a pitcher’s offerings by analyzing their spin and deception. This feature is now available thanks to Major League Baseball’s move to Hawkeye Technologies, the company in charge of their new tracking system. This data gives players, analysts, and coaches more of an opportunity to delve into the different spins exerted on a pitched ball.

Previous tracking technologies had only been programmed for the system to calculate the ball’s “inferred” axis, given a variety of inputs. This calculation would only occur when the pitch was near the plate, a limitation of the former system. Now, thanks to Hawkeye, we’re able to calculate the spin-based movement of a pitch when it’s being released from a pitcher’s hand. The delta (a more fancy word for difference) of the calculation of the pitch’s spin axis in front of the plate (the inferred axis) and the calculation of the spin axis upon release (also called observed axis) is the basis of Savant’s latest feature.

In layman’s terms, Savant’s new feature is telling us how much the pitch has moved during its flight, but instead of quantifying just movement, we now have a way of quantifying a pitcher’s deception. For those still confused, Mike Petriello and Eno Sarris wrote excellent articles on this topic on and The Athletic, respectively.

When looking at Savant’s Spin Direction leaderboards, I noticed something strange with Holland’s curveball.

Holland’s delta with his curveball was widest in the Majors last season, showing that the “observed” movement he released his pitch with is very different than the spin-based movement that arrived at home plate.

The 2 hour, 30 min deviation between his “Observed” and “Spin-Based” axis is the widest among all MLB curveballs thrown in 2020.
Active Spin vs ∆ of Inferred and Measured Movement of a CB (min 100 pitches)

This is just pure speculation on my part, but his wide delta could be a factor of the large amount of gyro spin that’s applied on the pitch. That being said, his wide delta could also be attributed to the variety of angles he released his curve with last season.

As for how a wide delta helps Holland, I’m really not sure. All I know is that the ball’s axis is different from the time he released it to the time it arrives the plate, which can confuse opposing hitters. But to what effect that helps Holland, I’m not really sure. This information has only been available for a little under a month and all of the pertinent research has yet to be published.

To be quite honest, I think I’m in a little bit over my head. Do I know exactly how to fix Holland’s problems? No, not exactly. All I can recommend to keep up the impressive spin efficiency on his pitches and perhaps locate his sinker more towards the corners.

But the tools are there. There’s the makings of an effective sinker, his changeup has 98% active spin, his curveball has some of the same qualities as Nick Anderson’s, and he has a funky arm slot. Add in the fact he got unlucky last season and it bodes well for another comeback.

Maybe there’s more to unlock with Holland. With teams getting wiser by the minute in terms of pitch design and pitch science, Holland could have another late-career renaissance in him. If he signs with a more forward-thinking team who could take full advantage of his arsenal, the tools are there for him to succeed.

Photo Credit: Christopher Horner of the Tribune-Review

Leave a Reply

Powered by

Up ↑

%d bloggers like this: