A New Way to Study Pitching Injury

BauerDL
Indians’ starter Trevor Bauer prepares to collect data at Driveline Baseball.

Kyle Boddy spent years getting it wrong. “There were years of inconclusive results that led to more questions,” Boddy told me about his past work at his Driveline Baseball facility in Kent, Washington.

He had the best intentions. After years of day jobs, and coaching youth baseball with some competitive weightlifting sprinkled in, he started writing at The Hardball Times and studying injuries with Josh Kalk, now a member of the Tampa Bay Rays’ front office. They had some success using neural networks on PITCHf/x data in order to spot injuries earlier than usual.

In the end, though, the Seattle-based mechanics analyst wanted to take a look at pitcher development under the same data-based lens that he and Kalk had used to spot existing injury.

So he built a biomechanics lab, complete with high-speed cameras and objects of known size. (That object, known as the Cube, is a square box built of tubing that helps calibrate the cameras so that the video created is all comparable.) It was a lot of work with an uncertain reward. “We got a lot of great kinetic data,” said Boddy of that time. “Then we realized that there was a huge amount of noise.”

Helping Boddy with the realization was Dr. Murray Maitland in the Department of Rehabilitation Medicine at the University of Washington. When approached with analysis based on limb movement and pitchers’ physical tendencies and the link to injury, Dr. Maitland smiled and dropped what might have been a bombshell to Boddy that day. “Just because the joint or limb moves in this direction doesn’t mean the underlying muscle is doing that,” said Maitland in Boddy’s recollection. “The movement could be due to inertia, it could be due to whatever. You can’t infer muscle activity.”

That’s when Boddy realized how much noise there was in the supposedly rock-solid findings floating through baseball. When the American Sports Medicine Institute releases a paper that says that the internal rotation of the elbow is 7000 degrees per second, that “sounds like a really scientific thing,” said Boddy. But it’s using high-speed cameras, and the data is limited by the frame rate on the technology they’re using. Even at 160 frames per second, there’s a “huge amount of movement” between frames, pointed out Boddy. The error bands are up to 1500 degrees in both directions — that’s one standard deviation, according to the private pitching coach.

That’s the type of thing that frustrates anyone working with baseball and biomechanical data, for sure. But there was one last thing from Dr. Maitland that gave Boddy hope. The rehabilitative-medicine professor pointed him in the direction of EMG sensors in order to find out what is actually happening in each muscle group as a pitcher throws the ball to home plate.

The problem was that there were no cheap, publicly available EMG sensors that Driveline could use for data collection. Boddy saw a Ted Talk by Alexander Gray about creating just such a product, though, and realized what he needed wasn’t far from the market. “I bothered him for a while and finally he made a Kickstarter and I funded it on the max level. Six months later he shipped out units and they worked great.”

With EMG sensors in hand, Driveline had the capacity to understand exactly what was happening under the hood. And yet the progress is still slow.

One main finding has allowed Boddy and his team to check their pitchers at every point along the training process to understand how much stress they’re putting on their arms. “The biggest thing we saw with people who were about to be hurt or had been hurt was that forearm flexor activity was extremely low comparatively to the norm,” Boddy said.

Using the EMG sensor during a grip test, they set up a baseline reading. Then they had pitchers throw hard, throw strikes, with normal spin rates, and do the test again, for a healthy dynamic baseline. You test again five days later after rest. Any departure from that second baseline indicated that whatever they were doing was not good for them. Even with the same velocity, a pitcher might not be activating his muscles efficiently because of an underlying injury.

BuckelSpin
Rangers’ righty Cody Buckel throws as Driveline Lead Trainer Matt Daniels watches his spin rates.

From that discovery flowed others. “Spin-rate changes are actually one of the best predictors of injury,” pointed out Boddy, a fact that was confirmed as a known within many parts of the baseball community by multiple sources. At Driveline, pitchers have their spin rates monitored constantly for the earliest signs of unhealthy fatigue.

But, using the EMG sensors, the team also developed some best practices for their pitchers. Generally, Driveline is “applying general motor-learning principles into baseball and then monitoring it using the EMG sensors.” That means breaking skills down into chunks that help the pitcher’s brain and body map the healthy mechanics.

WeightedBalls
Driveline’s weighted balls are at the core of the program.

Most of those drills involved weighted balls, and once again that’s because of well-documented findings by both Kyle Boddy as well as Dr. Coop DeRenne. Boddy found the University of Hawaii researcher’s work, which indicates that weighted balls are good for velocity and have no impact on health — and immediately tried to replicate it. He actually fell into a situation where he had a control group — young pitchers who weren’t allowed to do the weighted-ball program — and the results were just as impressive as DeRenne’s.

MaxVelo-Analysis1

By continuing to collect data along the way, Boddy is still learning as he goes. By using force plates — basically just mats which are capable of measuring force — he and his associates recently discovered that “the back-leg push-off effect is very minimal for producing velocity.” Said Boddy: “The biggest thing that correlates between the two is the force going into the ground on your glove leg — the better you block the force, the more velocity you get. There’s also no evidence that stride length has anything to do with it. Just force into that front leg.” EMG evidence backed up the force-plate analysis. So they developed an exercise — called a rocker drill — that helps the pitcher concentrate on blocking that force on the front leg.

They can also now bring back in the kinematic analysis while checking the ramifications of that analysis along the way with EMGs and spin rate readings.

For example, one pro pitcher new to Driveline has an incredibly fast arm with a huge amount of torque. “The guys like Trevor Bauer, Cody Buckel, guys that have trained for a long time with weighted balls and do drills, they see that same force over a longer period of time,” Boddy says. So the lab at Driveline is trying to figure out how to flatten his curve. Like a car, having a “crumple zone” that spreads the stress over a longer space is good for the elbow. Take a look below at different shoulder internal-rotation angular-velocity curves for pitchers to get an idea of what Boddy means.

CbZOADxVIAA6AbL

There’s more to discover, but Boddy is more interested in small findings and tweaking existing deliveries than trying to revamp mechanics completely. He’s mostly dealing with adults, anyway. “The tissue has adapted in such a way to compensate” for their unique deliveries, Boddy says of his clients. “If they were 12, we could coach them out of issues, but the best way is to not coach them at all. Let them have fun.”

And as hard as it is to uncover actual findings around which he can wrap his hands, at least the people at Driveline Baseball have a few ways to measure directly what is happening inside a pitcher’s body as he throws the ball. Those measurements are fun, too — pitchers gather around as spin-rate readings are collected, and there’s a leaderboard of weighted-ball velocities on the wall. When they play command games to see who can hit the zone the most, the loser gets shot in the back with a paintgun.

Boddy seems happy to act as a resource and a gentle guide while the players have fun, because there were the harder years. “Throwing takes years of skill to acquire, and if there’s no quantitative way to test how well they are doing, it gets frustrating,” Boddy remembers of the early days. Now, though, it’s about asking questions and letting the data answer.



Print This Post



Graphs: Baseball, Roto, Beer, brats (OK, no graphs for that...yet), repeat. Follow him on Twitter @enosarris.


Sort by:   newest | oldest | most voted
Kyle Boddy
Member
2 months 15 days ago

Thanks for coming out, Eno!

dan
Member
dan
2 months 15 days ago

I love reading Kyle’s stuff, and have been doing so for several years now. Very funny that this article was posted just after the Jered Weaver article:

http://www.fangraphs.com/blogs/another-year-of-the-jered-weaver-experiment/

Somebody get these two in a gym together.

Kyle Boddy
Member
2 months 15 days ago

I used to make fun of Jered Weaver like a lot of people, then I realized a lot of high school kids get destroyed throwing 2-6 MPH faster than him. And any other big leaguer who sat 78 MPH and went out there would probably just be embarrassed. Weaver is actually somewhat competitive throwing ~80 MPH. Not saying he’s good but he does get actual big leaguers out from time to time.

That’s pretty impressive.

mike sixel
Member
mike sixel
2 months 15 days ago

This was a really interesting read. Thanks,

PXF
Member
Member
PXF
2 months 15 days ago

Thanks Kyle and Eno — this was really interesting!

Bobby Ayala
Member
Member
2 months 15 days ago

Really excellent stuff. Analysis on this level is the future of fangraphs, keep it up!

mZimm
Member
mZimm
2 months 15 days ago

Is Trevor Bauer a Duke fan? Did Stroman lend him a shirt? Did he not get enough UCLA or Indians gear? So many questions…

Kyle Boddy
Member
2 months 15 days ago

He is a well-documented Duke fan.

Matthew Murphy
Member
2 months 15 days ago

Excellent work — both by Eno and Kyle!

chipmunknunchucks
Member
chipmunknunchucks
2 months 15 days ago

Would like to know more about how pitching velocity is, in theory, more dependent on the front leg landing than the back leg push-off.

That’s pretty significant if true and certainly flies in the face of decades of pitching instruction.

Danny Sader
Member
Danny Sader
2 months 15 days ago

What exactly do they mean by “blocking that force on the front leg”? Trying to be lighter upon landing?

chipmunknunchucks
Member
chipmunknunchucks
2 months 15 days ago

That’s where I got confused, as well.

Is it like a Buddhist monk thing? Strive to be both violent and calm at the same time?

Tom Dooley
Member
Tom Dooley
2 months 15 days ago

Great stuff. Thank you.

DSTT
Member
DSTT
2 months 14 days ago

“Spin-rate changes are actually one of the best predictors of injury,” pointed out Boddy, a fact that was confirmed as a known within many parts of the baseball community by multiple sources.

———

What kind of changes are we talking about here?

Jeff Long
Member
2 months 14 days ago

<3

hbar
Member
hbar
2 months 13 days ago

This article was interesting, but it maybe missed some parts (or links) for unfamiliar readers. I’m guessing that an “EMG sensor” measures something with muscle activation? The article didn’t explicitly define it, or how many such sensors a subject would wear. Is there crosstalk? What about muscles under other muscles?

wpDiscuz