It’s uncomfortable.
It’s uncomfortable, and it’s embarrassing. Anyone who wants to take advantage of the best in current biomechanical analysis in baseball must undergo the rigors of the motion capture process. That means getting down to your underwear, getting sprayed with adhesive that won’t come off for weeks, applying the sensors, and throwing in front of the peanut gallery (and all the expensive cameras) in the lab that day.
It’s a credit to today’s baseball players that they don’t mind being embarrassed or uncomfortable in front of people. Of course, they probably throw harder than 60 when they strap on the tools of knowledge.
Maybe that’s why they don’t mind it. Unlike me, who was lucky enough to undergo this process at Driveline Baseball’s facility in Kent, Wash., this offseason, they’re really good at baseball — and this will help them get better.
Advertisement
“It wasn’t like, ‘this is weird,’” said Phillies minor leaguer Tristan Garnett at Driveline Baseball. “It was more like ‘holy cow, these guys are really going to break me down to a scientific level with no bias at all, to get to the truth.’”
Think for a second about the metrics current baseball technologies produce.
A radar gun measures the velocity of a pitch once it hits the catcher’s mitt. A Rapsodo pitch-tracking device measures a spin rate after the ball crosses the plate. A Blast Motion sensor, clipped to the bottom of the bat, measures rotational acceleration after a player swings it.
These are all outputs.
“The thing that’s been coming is the science of, ‘How do (the players) do it?’” Dr. Glenn Fleisig, the director of the American Sports Medicine Institute, said a few months ago in a phone conversation.
Fleisig was referencing the search for inputs: What does a player do that leads to the pitch that registers 100 mph?
We are now starting to know. New motion capture technologies are allowing teams to utilize biomechanical analyses for health, velocity and deception purposes.
“Teams have woken up to this,” Fleisig said. “The field of baseball has woken up to this. It’s the next competitive advantage. That’s what’s happening.”
The current biomechanics revolution in baseball traces its roots to the first nascent efforts in the 1980s, and it is similar to that period in this respect: Technology spurred growth. Nearly four decades ago, the inception of video cameras made high-speed motion capture possible. Utilizing multiple cameras as well as an algorithm — called direct linear transformation — researchers could compute a three-dimensional image. That ability led to the first modern-day biomechanical analysis process in the 1990s, an effort called markered motion capture.
Advertisement
In markered motion capture, pitchers would strip down to their underwear. They would spray on the adhesive, strap on the reflective markers and step up to the mound and unleash pitches. A dozen or so infrared cameras, focused on those reflective markers, would record frame-by-frame movements.
“It was awesome,” Royals senior director of performance science Austin Driggers said in a recent phone conversation. “Problem is, it was constrained to the lab.”
Pitching on the mound inside a facility is not the same as pitching in a playoff game in front of thousands of fans. While a lab setting does allow for controlled conditions, it does not always reflect performance in the real world, or, as scientists prefer to call it, “the field.”
Jimmy Buffi, the co-founder of biomechanics company Reboot Motion, knew this about a decade ago as he was finishing up his PhD in biomedical engineering at Northwestern. At the time, he knew he wanted to use biomechanical analysis to help baseball players remain healthy and improve. But he didn’t feel it could succeed with the uncomfortable and tedious nature of markered motion capture, so instead he accepted a job with the Dodgers in 2015.
That same year, Jeff Passan reported the Tampa Bay Rays were going to install Kinatrax, which introduced new, ultra-high-speed cameras with extreme computing power, good enough to gather three-dimensional biomechanical data. No adhesive spray nor reflective markers were necessary.
Experts call this markerless motion capture. It remains the new frontier within baseball biomechanics — capable enough for Buffi to found Reboot in 2019, impressive enough that those in the business began to realize the gap between lab and field research had been bridged. Teams have built labs with specified ways of capturing biomechanical information. Coaches can utilize technologies such as ProPlayAI in bullpens.
Advertisement
“Markerless motion capture is probably the closest thing to a disruptive technology that we’ve had recently in baseball sport science,” Driggers said. “This is a really big deal.”
What do biomechanical analyses offer if done right? Take Phillies pitcher Garnett and Cardinals outfielder Lars Nootbaar, who have used the motion capture lab at Driveline to improve themselves.
First comes the assessment, with the sensors attached.
That produces some hard data that compares how the pitcher or hitter moves against the general population of professional and amateur hitters and pitchers that come through the facility.
“I’m not used to seeing my entire skeleton being broken down, and every single frame of my pitching mechanics,” Garnett said of the experience. “It was eye-opening. It was cool to see that despite my deficiencies, where they were, I was still able to keep the velos that I was throwing, I threw 94.9 when I got here. They said that was great because my ceiling is much higher. I learned that my hip-shoulder separation is not where it should be, and it has to do with my front side, because it broke very early.”
Nootbaar had similar results, which popped when his trainer, John Soteropolus, took a look at the skeleton breakdowns.
In both cases, the players were assigned corrective drills to fix their deficiencies. For Nootbaar, it was the step-over drill, where the batter begins with his front leg behind his back leg in order to emphasize that hip-shoulder separation. For Garnett, it was something called the janitor drill, where he shows his back fully to the plate before turning and throwing home.
The brilliance here is not in the training process — which has been similar throughout time — but in the ability of these labs (and teams) to then test each drill to see if it is actually getting the desired results.
Advertisement
Dr. Mike Sonne, a former writer at The Athletic among his other pursuits, has developed a markerless motion capture application that pairs with any modern phone with a decent camera to produce these same skeleton figures and biometrics. With his app, ProPlay Ai, they were able to show that the janitor drill was effective.
“At the Baseball Development Group, they had an issue for a pitcher opening up too early, so his trunk was rotating too early into foot plant — his trunk was already facing home plate,” Sonne said. “That’s a commonly held injury factor because you can’t use your trunk energy anymore because it’s already dissipated. Our BDG guys found that this drill could delay the timing of the trunk rotation, using ProPlayAI, and now they can use that with pitchers that open up too early. And that’s important for BDG, which isn’t the biggest lab, they couldn’t afford a full mocap situation, and this gives them information with some reliability and that helps them a lot.”
The most difficult thing about player development has been having any sort of precision in the process. Consider that to know that a player has improved, one normally has to know the true talent of the player (which has some error), has to know that the process followed actually did improve that player (more error), and then know the post-coaching true talent of the player (even more error).
Now, with the proliferation of biomechanics and motion capture analysis, player development can be more precise. In these two examples, the relationship between hip-shoulder separation and the speed at which a player can move his trunk is a known thing. Now, coaching that up is as simple as assessing that is actually the problem, and then using drills that have been proven to address that particular issue.
The implications are massive.
Articles like these — and reader interest in them — have actually helped push the state of biomechanics to where it is in baseball today, incredible as that seems. They may even have something to do with how many teams adopt it in the future.
“If players think they are missing something, they complain, ‘What the fuck, why don’t we have that?’” said Driveline co-founder Kyle Boddy. “Collective intelligence and awareness is important.”
Advertisement
Baseball executives are people, too, human beings swayed by commonly held interests. But they are people who need to know what’s possible before they can ask for it. And now that it’s becoming clear there are so many different ways to analyze the way a player’s body moves, most teams are taking part in some fashion or another.
There’s the version of markerless motion capture represented by the Hawk-Eye system in each stadium, one that can provide entertainment-grade biomechanics — not necessarily good enough for rigorous analysis, but helpful in popularizing the effort. And then there’s the upgrade to Hawk-Eye pro that a handful of teams have undertaken. Does that mean only a handful of teams are all the way into biomechanics? Not really.
Players working at Driveline’s facility outside Seattle (Eno Sarris)“Nine or ten teams have Kinatrax at the big league level, they don’t need to upgrade to Hawk-Eye pro. Tampa and Boston are at the cutting edge of Kinatrax, the Cubs and Tampa have been using it forever, too. You just don’t hear about it, why would you hear about it?” Boddy said.
Some teams were early adopters and are going to stick with the systems they have in place, others are late adopters going with Hawk-Eye pro. A friendly assessment might say that about half to two-thirds of baseball has bought into this movement on some level or another.
That said, not all MLB organizations nor private player development outfits are utilizing the markerless motion capture. The obstacles to entry — and challenges that come with using it — begin with cost.
Ben Brewster, a former minor-leaguer who founded Tread Athletics, an online-based pitching development outfit, said the “really good” systems still cost anywhere from $250,000 to $2 million.
“If we could get our hands on a system we knew was accurate, and it wasn’t $500,000, we would,” Brewster said in a recent phone conversation.
Advertisement
Even if teams or deeper-pocketed private facilities determine the cost is worth it, there’s a further question for the teams themselves as they try to actually use all this data: How many of those have taken their findings from the research side and then seamlessly integrated them into their coaching, and even their player acquisition philosophies?
A biometrics evaluation. (Produced by Baseball Cloud for Premier Pitching Performance)Answering that question is, in large part, why Buffi founded his company. A coach, he said, should not have to have a PhD in biomechanics. The onus, he believes, is on those who understand the physics to inform what they’re seeing scientifically to the coaches.
“I spend a lot of time meeting with coaches explaining, ‘This is how the specific physics work. This is what these reports are capturing. Here’s how to use it,’” Buffi said. “My favorite meetings are player development meetings where it’s a pitcher or hitter and we go, ‘Hey, we want to build a plan with this pitcher or hitter.’ We generate an analysis. The medical staff is in the room. The strength and conditioning staff is in the room. We have the analysis. And it’s, ‘Let’s go through the analysis and find where a player is efficient and inefficient. Then, let’s see what assessments we can do to find the root cause of the inefficiencies and work together to build a plan to address them.’”
More landmines lie within the next step. Say the analysis gleaned from a markerless motion capture assessment identifies inefficiencies. From there, a coach has to decide: Do we go ahead and address the issue? Or, does some of that inefficiency actually relate to what the player is trying to do?
It’s tricky,” Driggers said. “On one hand, you’re trying to make players better, and there are infinite examples of coaches making mechanical improvements to pitchers and them performing better. It’s obviously possible. But there are also cases in which we’ve improved a pitchers’ mechanics by our technical model, but it contributes to injury because the pitcher’s body has adapted to those suboptimal mechanics. Measuring mechanics is getting easier, determining what’s good for each player is still really difficult.”
That’s why, as much as motion capture analysis has become democratized, especially with technologies such as ProPlayAI, many believe there’s still a major gap between the ways some clubs and private outfits are using this tool.
The technology may be new, better and faster. But implementing improvements based on biomechanical analysis still require advanced coaching expertise and high-level communication among staff.
Advertisement
“The publicity is, ‘Yeah, we just installed this lab.’ Or, ‘Oh my God, that team or org is so progressive!’” Buffi said. “Little do you know, when you get inside the organization, the data is being collected and just sitting in a lab. The data people are staring at it. The coaches are doing what they do. And they’re not really talking to each other.”
But then there are those who are operating in a different manner. Those who are setting the standard for what the future of this space might look like.
What comes next seems straightforward. It’s a matter of figuring out how to infuse all the different aspects of running a baseball team with the findings gleaned from limb-tracking technology.
“I think the next frontier in (player development) is not in applying more technology,” agreed Brewster. “All of this tech has gotten here really quickly. Everyone has slow-motion videos, Rapsodo, Trackman. A lot of orgs now use biomechanical analysis. The next frontier is learning how to apply it, implement it in the most optimal way.”
Driveline works with both hitters and pitchers at its facility outside Seattle. (Eno Sarris)“Some teams are just at The Best Buy saying, “Can you pick me up a biomechanic,” Sonne said, laughing. “The next era is when the mathematicians and the statisticians get their hands on the biomechanical data together. This is the next wave, where human movement data is just as important as the ball flight data.”
Being able to measure the physical changes in a player before and after drills selected by their coaches is one thing. It’s another to then use those measurements to create benchmarks with which to judge the efficacy of the coaches and the organization in developing players. It’s a third thing to actually then make organizational changes based on hitting those benchmarks. It’s maybe a fourth thing to understand your organizational strengths through this lens and then acquire players who will benefit the most from your ability to help them move better.
“All this is, biometrics answer the question: Does this coach know what the fuck he’s talking about?” said Boddy. “That’s important for the coach, the coach’s boss, but it’s really important for the player. The player can see that there’s an objective number that judges the system. Marketing, sure, but in the way that’s the most honest.”
Advertisement
Marketing? But talk to players and you understand what he means by marketing.
“Coming here, I made a joke to my family that I worked out with Luke Skywalker in California, and now I need to go work out with Yoda,” Garnett said of Driveline. “Science never lies. It’s not ‘it worked for me,’ which, cool, but I’m not built like you. If they tell you 1 + 1 = 2 and explain why 1 + 1 = 2, it’s kind of hard to deny it.”
Is there even more to be learned from the technology? So far, most of the energy has been put into moving optimally to create velocity, and to lessen the stress on the elbow to prevent injury. League-wide velocities have increased almost every year since baseball started tracking them, so there’s been a lot of success there. At the same time, injury rates seem higher than ever.
“Refer back to the fact that everyone is throwing harder,” Boddy says about that fact. “There’s a local maxima. There’s no world where everyone is healthy and throwing 100 percent. When you seek the highest performance, there are going to be injuries. We don’t have the injuries we had 20 years ago.”
“Julio Urías had the shoulder capsule tear, that injury 20 years ago, that’s it, no more pitching for you and that’s it,” Boddy continued. “We’ve done a good job of preventing that injury, which is a good thing. Return from Tommy John surgery is better than ever. Spiral fractures of the humerus … nobody has had this injury in over a decade, and they used to have it all the time, you’d get a spiral crack down the bone, and you couldn’t lift your arm for a year. Our understanding of how to keep people from catastrophic injury has improved. Maybe it’s inevitable that a certain number of people will get injured.”
One of the last frontiers for biomechanics will come in linking the way pitchers move to actual results on the field. Deception — how different the pitcher is from what a batter might expect — has long been thought of as something we can’t put a number on. But with this sort of technology, we’re getting closer to being able to quantify how pitchers hide the ball before release, or put funky, unexpected movement on their pitches.
Over at Reboot, they defined deception by looking at expected movement and spin from different spine tilts and which pitchers had movement and spin deviated from that expectation. They came away with four important findings, which Evan Demchick wrote in this post summarizing their effort.
Advertisement
• There is a relationship between body rotation plane tilt and ball spin, which allows us to calculate a “predicted spin.”
• This predicted spin is similar to what batters are implicitly “calculating” prior to ball release.
• Pitchers that score high on the internal Reboot deception metric tend to outperform.
• Even if deception is not as strong of a predictor as velocity or spin rate, it still matters. And you have to measure what matters.
That last bullet point is really a great summary for the past, present and future of biomechanics in baseball. There was the data acquisition phase in which everything was measured and tested by the early adopters in the field. Now we’re in the phase of trying to figure out what measurements matter most and how to get those measurements to the right people in the right format. Next? Democratizing the movement to give more players access; dreaming up the remaining measurements that haven’t yet been found; and building entire organizations fueled by those findings.
So where is the future of biomechanics in baseball headed? It’s not exactly clear, but one thing is for sure: We’ll measure it, even if it means doing it in our underwear.
Read more: The future of baseball
(Top illustration by John Bradford / The Athletic; Images via Getty Images: Irfan Khan / Michael Zagaris)
ncG1vNJzZmismJqutbTLnquim16YvK57kmprbW1kbXxzfJFrZmlqX2eCcMDHnmSfraSqv6Z5wKebZqiimsCmutNmpp9lkpbApq7ApaNmoaNitq95waKmpp2Tna6vtcKsZg%3D%3D