Your cart

Trevor Bauer, Driveline Baseball Pioneer Data-Driven Pitch Design

Baseball history is awash with stories of pitchers adding new pitches to their repertoire by the same rote method: a teammate or coach suggests a grip, the pitcher fiddles with the ball until he gets a good feel and then he debuts his new offering in a game. This is addition by anecdote, a happenstance way of expanding one’s arsenal while more analytically minded pitchers are relying on technology to make concerted plans to engineer complementary pitches.

“Pitch design is such an untapped resource,” Cleveland Indians starter Trevor Bauer said.

Bauer has trained for several offseasons at Driveline Baseball, a data-driven pitching incubator in suburban Seattle, whose facility is loaded with all the latest technology: Edgertronic high-speed video, TrackMan radars to Pitch F/X optical tracking cameras to Rapsodo, a radar-triggered camera system. Driveline pairs the gadgets and gurus powering that technology with its weighted baseballs to form the core of its research-based ballistic training.

While Driveline has gained renown as a velocity-boosting development program, its staff uses the tech at its disposal — especially Edgertronic and Rapsodo — to help pitchers invent or refine pitches with instant, objective feedback that Bauer calls “invaluable” in hastening and solidifying progress. Bauer, a mechanical engineering major at UCLA, is the ideal client who professes a passion for computing, engineering, robotics, science and, most infamously, drones outside of his baseball career.

“Anytime I can cross the two over and use the scientific method to improve myself as a baseball player, I try to do just that,” Bauer said. “I think it’s one of the reasons that I’ve made it to the point that I’m at now. I don’t think I would have had any chance of getting here just purely based on athleticism if I didn’t have that process. I wouldn’t be anywhere close to where I am.”

Bauer’s Cleveland Indians are fresh off an American League record 22-game winning streak with the 26-year-old having a career-best season, as judged by his strikeout rate (10.1 per nine innings, ranking fifth in the AL) and the advanced metric, Fielding-Independent Pitching, which approximates ERA but removes some factors out of a pitcher’s control.

Previously at Driveline, Bauer honed his two-seam fastball, throwing the ball with an axis tilted 45 degrees to maximize movement, which, for a time, had more horizontal tail than any other two-seamer in baseball (He spoke about making sure he established the correct laminar and turbulent flow.) Bauer also worked on his cutter, seeking to throw it harder while generating true fastball lift but with the opposite movement of his two- and four-seam fastballs (i.e. glove-side rather than arm-side run). Another time, he re-learned his curveball after overhauling his core pitching mechanics, adapting the pitch to a knuckle-curve after lowering his arm slot. Edgertronic clips of his changeup, two-seamer, curve and cutter show his process, which includes coloring part of the baseball to highlight the spin axis:

Trevor Bauer's Pitches

Cleveland Indians' Trevor Bauer teamed up with Driveline Baseball to use tech to design his pitches.https://www.sporttechie.com/trevor-bauer-driveline-baseball-data-driven-pitch-design/

Posted by SportTechie on Monday, September 18, 2017

Bauer often would catalog his inventory of pitches and identify a hole where the data says a pitch with a certain speed, spin and trajectory would be a productive addition. Last offseason, that addition was a split-finger as a variant to his changeup. He then went about designing the parameters of that pitch, starting with spin axis before divining the right velocity and spin rate to work in concert with each other effectively.

He starts by coloring portions of the ball with a black marker to enhance visibility of the ball’s spin axis, then sets up the Edgertronic camera directly behind his pitching hand at 1,000 frames per second. Bauer will throw the pitch over and over until satisfied the pitch is spinning the correct way, which for the splitter was as horizontal as possible. Then he transfers the work to the Rapsodo for velocity and spin rate.

“He’s trying to keep the spin rate as low as possible,” Driveline founder Kyle Boddy said of Bauer’s split-fingered fastball. “Zero would be great, a true knuckleball, but that’s not feasible, obviously. So it’s, ‘How can we drive it as low as possible?’”

Bauer reported that he reduced his splitter to about 800 revolutions per minute, slightly more than a third of his fastball spin, and got decent results earlier this season but never felt comfortable in his ability to command the pitch over the plate and below the strike zone as intended. As a result, he shelved it. Data at BrooksBaseball.net indicates that he threw 16 splitters in April and May but has thrown the pitch only three times since.

His usage of the curve, meanwhile, has skyrocketed to nearly a third of all his pitches since the end of May. Bauer said his original curved at 2,850 average rpm, a rate that fell to 2,450 after he changed his pitching mechanics; sometimes his index finger would protrude into the ball path and deflect the spin axis and hinder the spin rate. Using the high-speed camera, Bauer learned the knuckle-curve grip he uses now, spinning a 2,400 rpm pitch early in counts as a get-me-over strike and dialing it up as high as 2,800 for a late-count whiff.

“All the data says that for every 100 rpm’s you get on the curveball, swing-and-miss rate increases,” he said.

With so much in-game TrackMan data available from the majors and high minors, Driveline has researched millions of pitch pairs to determine the best velocity and vector combinations.

“These are the best ways to develop, based on what you already have, to make small changes to your current arsenal,” Boddy said. “We can make them a little more effective. That goes anywhere from pitch tunneling to sequencing and spin rate and type.”

*****

Velocity has always been Casey Weathers’ calling card. The No. 8 overall pick in the 2007 draft out of Vanderbilt, the righty could always dial up his fastball to the upper 90s. Weathers went viral two springs ago when, with the aid of a running start and crow hop, he hurled a baseball 107.8 mph in a drill at Driveline.

At the same time, Weathers’ lone secondary pitch — which he noncommittally called a “breaking ball” — was just “very OK.”

“I didn’t know what to even do with it,” he said. “Do I make it more like a slider? Do I make it more like a curveball?”

Seeking to move the pitch beyond the generic, Weathers conferred with the Driveline crew, who saw that the underlying metrics of the breaking ball more closely mirrored the profile of a slider that could be effective against high level hitters. Weathers set about to throw a classic gyroscopic slider, sometimes likened to a spinning bullet for the way the ball moves perfectly sideways with an axis directly approaching a hitter. Rapsodo computes spin efficiency based on a lack of gyroscopic spin; in other words, the pure slider Weathers was hoping to throw would register a 0 percent spin efficiency. His typical breaking ball had been between 20 and 30 percent, which he needed to lower below 12 percent to make it playable.

With the Rapsodo hooked up to an iPad and then to a big-screen television right next to the mound, the workflow was efficient, Weathers said. “Throw it. Was it? No. Throw it? Ok, I did it. What did I feel? How can I repeat that?”

“The Rapsodo just takes out all of the subjective questions and makes you objectively accountable to get the pitch to do what you want it to do,” he added. “I had many instances where I was like, ‘That one was nasty, that’s it,’ and then the machine comes back and is like, ‘No, that’s 30 percent. That’s not a good pitch.’

“If you go backwards to the old way of doing it, you might throw that one, be like ‘Oh, that’s filthy,’ and in the bullpen and to the eye test, and then you try to repeat that over and over again and really it’s not a very good pitch.”

Weathers admits that his own eye test scores most of his pitches as an incomplete; during his delivery, his head often yanks down sharply, negating any chance to follow the ball in flight. The Rapsodo and the data it produces, therefore, give him a “clear vision on what I need to do to try and keep [my slider] consistent,” he said. Driveline’s video guru Taiki Green overlaid video of Weathers’ fastball and slider onto one clip to reinforce the similar trajectory of the pitch while leaving his hand.

While pitching for Fargo-Moorhead of the independent ball American Association this summer, Weathers had 18 saves and 3.89 ERA while striking out 57 in 37 innings — a rate of nearly 14 strikeouts every nine innings, which was the league’s highest among anyone who threw at least 10 innings. Though still predominantly a fastball pitcher, Weathers said he threw more sliders in early counts than ever before.

“It’s adding at least a different element and given me the ability to show it early or use it late and hopefully keep the hitters a little bit more off-balance,” Weathers said, “rather than just [playing] that country hardball game where it’s just, ‘Hey, here’s four to six heaters, and we’re going to see what happens.’”

*****

Dean Jackson, who went to a small Arizona Christian University, touched 97 mph when he was 19 years old, but a major injury derailed the once top prospect’s linear ascendancy to pro ball.

“After my hip surgery, I couldn’t break 87,” he said. “So everything started to fall apart, and I had to look for alternative ways to get better.”

Jackson would build his fastball back up to the low 90s, but right-handed pitchers throwing in that velocity range are a dime a dozen in college and the minors. He had developed a good curveball but knew there was still room for improvement; after all, his bender would have to rank among the best thrown in the majors for him to even have a chance to reach the big leagues.

Get The Latest Sports Tech News In Your Inbox!

The Scottsdale, Ariz., native gained his appreciation for data from his parents — his mother, Debra Schaffer, is an IT consultant and his father, Charles Jackson, is an entrepreneur — and traveled to meet Boddy and Driveline after reading about them online. His first trip lasted eight days. His second was a six-month residency.

Though Jackson went undrafted out of college, his curveball caught the notice of an Arizona Diamondbacks scout and helped him land a contract to play in rookie ball. With the Missoula Osprey, he pitched well enough out of the bullpen, striking out nearly 12 batters per nine innings with a sub-3.00 ERA, that he landed on the Pioneer League All-Star team. He would get released after an elbow injury, leading him to tinker further with his curve.

Before going to Driveline, Jackson said his curve would register nearly 3,000 rpms of spin rate, which has been shown to be a key line of demarcation for effectiveness at the big league level.

Jackson had the opposite goal as Weathers, wanting to drive up his spin axis efficiency as close to 100 percent as possible. When he began using Rapsodo, his curve regularly registered between 60 and 70 percent. On the high-speed camera, he recognized that the seams oriented around the spin axis were wobbly, so he re-aligned his grip. With a couple weeks of adjustments, he consistently pinged past 90 percent efficiency and recalled one curve that reached 99.

“It was a very significant increase in a pitch that most people would have thought had already reached its peak,” said Jackson, who has worked as head of baseball analytics at his alma mater while recovering.

The process was streamlined by his understanding of the underlying process, aided by conversations with Bauer about fluid dynamics, of which aerodynamics is a subset. (Jackson acknowledges listening more than speaking. “I didn’t take the class,” he quipped, “Trevor did.”) Jackson soon realized that, as his spin efficiency improved, his spin rate actually declined somewhat, to between 2,700 and 3,100 rpm. He thereby came to understand that spin rate alone did not determine effectiveness so much as the relationship between rate and efficiency.

“Once you have objective data and you get it back very quickly,” Jackson said, “it doesn’t take a whole lot to actually get it to work.”

← Back to SportTechie