By John Portch

When Ben Crockett served as Director of Player Development at the Boston Red Sox, he tended to be an early adopter of technology. But it didn’t always work out for the best.

At the 2025 Leaders Sport Performance Summit in London, he shared a decade-old story about a pitch-tracking technology that had been excitedly introduced by the club’s performance team. “It was a tool to help provide the coaches with more information that would help them in their coaching efforts and provide feedback,” said Crockett, who now serves the Red Sox as a Special Assistant of Basketball Operations.

The project proved to be a flop despite the technology doing exactly what it was supposed to do. Today, Crockett understands where his team went wrong. “We didn’t teach it enough; and we actually installed somebody to help operate it, which we thought would be helpful.” However, “in some ways it actually took away the ownership of our coaches for a solution that was ultimately going to help them.”

The Red Sox pitching coaches carried on doing their own thing; the merits of the technology were irrelevant to them. They didn’t trust, understand or see how it would help. “That was a good lesson in terms of what not to,” added Crockett, who shared the stage with Troy Thomson, the Football Operations Manager at the Brisbane Broncos, who had a cautionary tale of his own.

“Quite often, someone over-promises and under-delivers,” he said. “When you finally get the hardware in, it doesn’t quite have the capabilities you need.” It is no wonder that coaches can be wary of new datasets.

Ben Crockett (centre) speaks onstage at the 2025 Leaders Sport Performance Summit in London alongside Troy Thomson (left).

Over the course of half an hour onstage at the Kia Oval, Crockett and Thomson explained why data is valuable only if it can positively influence decision-making and behaviour. Their efforts to refine their approach point to five common principles: transparency, education, alignment, ownership and storytelling.

Transparency – information only matters if people can see it

Data is most useful when athletes understand the metrics, contribute to the resulting discussion and take responsibility for acting upon the information unearthed.

Transparency at the Red Sox, as Crockett explained, is embedded in a system of “continuous feedback loops within session, after session, after games” between coaches and the players, who develop a “continued understanding of where they stand on chasing specific measurable goals that have been agreed upon between the player and the team.” He added that “the athletes today are much more open to getting that type of information”.

Thomson also noted that Broncos players actively seek feedback on their performance metrics. For example, they now “want to finish the session and come and see what their line speed metrics are,” he said.

Education – data must be understood before it can influence decisions

Data can begin to influence decision-making when people understand both the numbers themselves and the behaviours those numbers are intended to shape. As Crockett noted, “it isn’t a one-time presentation; ‘here’s what everything means’. It’s ongoing.”

At the Broncos, Thomson emphasises the importance of helping players understand what data reveals about their performance. The coaches’ role is one of showing rather than simply telling.

He said: “The coaches also work really hard from a visual point of view as well, so being able to show the players technically how they’ve been working is a really important part of the process.”

Alignment – everyone working towards the same goals

To reach a collective understanding of what success looks like and how it is achieved, teams must first establish what matters most. Data can be a useful tool in that regard.

When Crockett discussed the challenge of coordinating a baseball operation that includes Major League, Minor League and player development programmes, he stressed the importance of “clearly defining key metrics that drive value for us at the Major League level and drive value throughout the industry.” Once those metrics have been identified, the task becomes ensuring everybody is working towards the same outcomes. Crockett described how the Red Sox seek to “help align our player development staff and personnel development staff with the things that were most important to driving that value”.

It is a similar story at the Broncos. “We’ve got a whole heap of different stats that we use; so post-game, there’s some standardised stats that we know if you get those processes right, the result will look after itself,” said Thomson, whose team has put in a lot of work behind the scenes. “To be able to bring the whole collective leadership group on the journey was really important.”

Ownership – data belongs to all stakeholders

Crockett and Thomson both argue that data becomes valuable when the people making decisions feel responsible for using it. Crockett’s pitch-tracking debacle is a prime example of what happens when that ownership is absent. As he recalled: “It ended up becoming a bit more of a ‘well, that’s their information and the pitching coach’s; and I’ll do what I’ve just been doing the whole time’.”

To prevent such scenarios, as Thomson explained, teams must create environments in which practitioners feel comfortable sharing ideas, challenging assumptions and seeking new sources of information. He said: “It’s really important to have a psychologically safe environment where we’re challenging each other on a daily basis to be able to find new metrics or new data sets that can actually enhance the programme.” At the Broncos, the software eventually “became part of a routine for us”.

Storytelling – data is most useful when it explains performance

Behavioural change in a data context depends on the story your numbers tell. Thomson repeatedly stressed the importance of using statistics to tell a story because it “adds to the coaching”.

Not that every stat holds the same weight; coaches have a responsibility to “share the relevant data that tell that story to achieve your outcomes”.

The Broncos’ momentum-tracking software is a case in point, as Thomson explained. While initially unsure how useful it would be, the tool was valued because “it allowed the coaches to be able to present the feel of the game”. By helping coaches explain something they previously struggled to articulate, the data became far more powerful as a catalyst for decision-making.

What to read next

How to Use Data with Confidence in your Coaching

Main Image: Getty Images

By Ethan Joyce

I regret to inform you that I am not very clutch.

Normally, it’s my colleague Joe Lemire who experiments with new tech for our recurring Sandbox Series. But this week, it was me dipping my toe into the test zone — and I recoiled with a wounded ego.

When it comes to human performance, clutch is still a polarizing topic. Is it even real? The analytical stat heads say no, but anyone who’s seen a game-winning shot or a career-defining performance by a star athlete may beg to differ. So, what if you could evaluate people based on their likelihood to perform in high-pressure situations?

NTangible has figured out how to measure it.

The platform, which provides mental scouting assessments for athletes for multiple sports, was created by President/CEO Dan Connerty. Its testing, which features open-ended and multiple-choice answers, uses AI to analyze things like thought process, perspective and wording.

Connerty has been on this startup adventure full-time since 2024, with NTangible now in use in 40 NCAA programs (including Boston College and Florida Atlantic). Last year, the company became the official mental fitness partner of RFK Racing to evaluate and support its pit crew members.

Image: SBJ Tech screenshot

Based in Canada, Connerty spent most of his time in the baseball development space as a former collegiate and pro athlete. He was the principal director of Baseball Development Group from 2017 to 2019, and during that time, he got to know industrial psychologist Ed Levine. And it started this ongoing pursuit.

“At the time I go, ‘We’re going to reach a point where we cap out what we can measure physically — it’s going to happen,’” Connerty recounted to me. “Mental has always been this black box … nobody wants to touch it.”

Enter the Clutch Factor score, which evaluates mental performance in clutch scenarios. In roughly 20 minutes, I ran through a handful of pictures depicting various scenarios. (I did a basketball test since it’s the sport I played the most.) One was an image of a player shooting a free throw, the next was a ball handler staring down a full-court press, and others were similar in-game scenarios. I had a chance to pick my perspective of the play, who I would be and what I’d be planning to do next. I also shared my past experiences of higher-pressure situations.

When I finished the test, I was given a score of 663. That figure is out of 1,000, which placed me in the average category. Connerty said the scores can be quite predictive of college success (NTangible can test athletes as young as 13). The athletes who score in the 750-799 range are twice as likely to play a Division I sport. Above 800? We’re talking a potential all-conference or All-American honoree. Example: Texas freshman P/UT Hannah Wells produced the highest Clutch Factor score in the country. She helped the Longhorns win their second consecutive College World Series softball championship earlier this month.

After sharing my lack of clutchness, I could tell you about my days as a newspaper college football reporter. I was constantly combating the harshest deadlines, filing stories that squeaked in just early enough to make the next day’s edition, some of which I won national awards for. But I won’t do that.

The good news for me (and other users)? Because my clutchness can now be measured, I can also improve it, according to Connerty.

“Our psychologists have said, at minimum, if you wanted to redo it, six months,” he said about retesting. “Usually, this should be an annual test just to see where you are. If you think of the Clutch Factor score as your capacity to perform at your top range, not every day are you going to be there.

“This is something where your score can fluctuate based on experiential parameters.”

That’s totally it! I had an off day. I’m sure of it. Case closed.

This article was brought to you by SBJ Tech, a Leaders Group company. As a Leaders Performance Institute member, you are able to enjoy exclusive access to SBJ Tech content in the field of athletic performance.

Main Image: Getty Images

By Rob Schaefer

Main Image: Getty Images

By Rob Schaefer

Main Image: San Francisco Unicorns

Main Image: NBA / Theia

By Joe Lemire

The SBJ Tech Sandbox series is where we share our experiences testing products, gear, solutions and more in the sports tech space.

I first dunked a basketball during a JV game on my 16^th birthday. Following a steal and a breakaway, I rather gently placed the ball through the rim, but the varsity team happened to be watching, earning me what was some scarce social currency at that point in my life.

Over the next few years, as I got stronger, my vertical inched up further, and I could dunk with a little bit of style — no one was inviting me to dunk contests or narrating my And1 mix tape, but it was still an epic feeling.

You certainly wouldn’t know it now.

The NBA recently invited me to participate in the same quarterly biomechanics screening that all its athletes started doing this season — the first league-wide program in pro sports. It’s a series of jumps and lateral moves, with some closed-eye, one-legged balancing mixed in.

The video may double as my audition for Blue Man Group — the cyan skeleton is part of the Theia software processing to evaluate joint angles and motion. But it also shows my diminished explosiveness as I move along the aging curve.

Each task has been carefully selected by a collaborative group of NBA and NBPA movement experts to gain insight into individual athlete movement patterns and hopefully build normative standards that can be used to help prevent injuries and improve performance.

One note the league makes clear is that biomechanics usually can’t answer questions in isolation — rather, it’s a notable piece of a larger performance puzzle.

“Biomechanics tells you how someone is moving, but it doesn’t tell you why,” said Courtney Chaaban, NBA Sr Manager of biomechanics and engineering. “So I really like to look at it alongside of lots of other information: how strong they are, what their range of motion is, how they’re moving on the court, what their demand is on the court, what their injury history might be.”

I’ve detailed my injury history in these Sandbox stories before — broken bones in my feet, sprains of muscles and ligaments, etc. — and I’ll share that my knees were oddly sore the day of my evaluation. (Of course I have an excuse ready… )

My single-leg jumps were paltry, just 3.7 inches for my left and 3 inches for my right. My two-legged jumps were a bit better. The countermovement jump (CMJ) — in which I started on the force plates, bent down and then leapt straight up — was 11.8 inches, good for 56^th percentile in the sample of non-NBA players. My drop vertical jump, in which I step down onto the force plates and immediately spring up, was 13.2 inches, at the 49^th percentile.

I was hoping for better. Image: NBA

There were diagrams showing how much I use my ankles, knees and hips to absorb forces as I jump or land. Of particular note was my landing strategy: I know my hip flexors are weak due to a previous Springbok Analytics scan, and my knees were achy that day, so it’s little surprise that my ankles worked overtime to mitigate force.

“It’s not to say it’s good or bad,” Chaaban explained. “Maybe that’s just how you move.”

I didn’t skip all of leg day, just the hip and knee lifts. Image: NBA

While recently chatting with Warriors superstar Stephen Curry about a different topic, we started talking about his recovery routine and the notion of career longevity, when I volunteered that I underwent the NBA’s biomechanics protocol.

“You did all the testing?” Curry asked with a bemused smile. I shared the verdict of my data, which is that I’m very much a suburban dad these days. He laughed and then added, “There’s a lot more data available, but the real art is to figure out how to make that useful, right?”

My testing was done at the Nets’ first-rate facility in Brooklyn, but the experience is intentionally universal and would have been the same in Oklahoma City or Sacramento, too. All the same tech vendors are involved: Qualisys motion capture cameras, Bertec force plates, Theia biomechanics software and Breakaway Data for player and team access to the data.

My primary differential: I received a pdf report (rather than the Breakaway Data dashboard) that largely mirrored what the players get but had a few alterations, mostly because I had no historical data from other assessments. My datapoints were compared not to other NBA players but to a handful of team and league staff who underwent the tests.

The reports convey what each metric represents, which is helpful, such as noting that deceleration during the CMJ is “similar to stopping when coming around a screen.”

That CMJ deceleration was my highest asymmetry: my right side produced 20.5% more force than my left. Asymmetry is generally thought to be a problem and a possible injury risk, but there also need to be some allowances for individuals having their own movement signature.

What Curry also told me about applying such high-tech evaluations is the need to surround himself with “as many experts as possible to understand what the data means and then figure out what that is going to do to impact my routine.”

So, kind readers with advanced degrees and extensive pro sports experience, I ask: Who’s going to help me dunk again?

This article was brought to you by SBJ Tech, a Leaders Group company. As a Leaders Performance Institute member, you are able to enjoy exclusive access to SBJ Tech content in the field of athletic performance.

Main Image: SwingVision

By Joe Lemire

By mounting their iPhones above the court, tennis players of all levels can use the SwingVision app to collect a trove of data and video from their matches. Add an Apple Watch and competitors can press a button and challenge a line call, receiving an animation on the wrist showing whether the ball landed in or out.

More recently, SwingVision has added a beta feature that yells “Out!” if a shot lands beyond the court boundaries. That audible line call takes a split second for the iPhone to process the image and make its determination, a rapid reaction only possible by the computing power of the chips Apple puts in its smartphones, bypassing the need to transmit to the cloud.

“That’s really only possible on device,” said SwingVision CEO Swupnil Sahai. “Because we’re processing 1080p at 60 frames per second, that’s a lot to process. And if you were to send that to the cloud and then have it process it and then run it back, you definitely can’t do that in half a second. And you certainly can’t do that on any court in the world.”

The proliferation of cloud providers and their processing have transformed sports tech for the better, democratizing access to storage and AI models, but there remain key use cases when the cloud isn’t the answer — when requiring speed, managing costs and operating with limited connectivity.

“This is a bit of a trend in my view, is that you see something that really works well in the cloud, but you need it speedier,” said Mike Davies, Fox Sports EVP of Technical and Field Operations.

Davies shared that observation in the context of his network’s IndyCar coverage. Fox and Indy Motor Sports Productions contracted with Lumen Technologies recently to install 10-gig fiber-optic cables at every track to support on-location servers. That enables faster rendering of data overlays on the driver’s eye camera and the use of a product called XtraMotion, which uses generative AI to turn regular footage into high-frame-rate video.

Sahai was a Tesla engineer before founding SwingVision, and the automotive manufacturer processes its Full Self-Driving cameras entirely with on-car hardware. “So that’s how I’ve always thought about this problem, is it should be done on device,” he said, calling it a philosophical decision.

“We just drew a line in the sand from Day 1: We’re doing this on device, we’re not going to use the cloud at all,” Sahai said, adding, “It just felt like it’d be way too expensive for a consumer app to run it in the cloud.”

Financial transparency can help. Gamecode, a soccer analytics startup, includes an upfront estimated cost for each predictive model run on the cloud.

Striking the right balance

Many sports tech firms offer a hybrid approach. Pixellot, whose AI cameras stream and analyze up to 20,000 games every weekend, is an AWS technology parter. Pixellot CTO Gal Oz once told SBJ that his company couldn’t have scaled without the cloud and “making Amazon do the heavy work. It’s not magic. I like the phrase that ‘there is no cloud, it’s just someone else’s computer.’”

Even though Pixellot CEO Doron Gerstel recently quipped of AWS, “They love us,” he added that every camera has an on-premises server to reduce the cost and expedite the video feedback for coaches and athletes. By the time the game ends, highlights are waiting.

“Now if you are doing it over there on the cloud, one of the most problematic things is you are on queue — if you want to pass the queue, fine with them, pay more,” Gerstel said. “The question is how much attention I’m losing if you’re not getting it immediately after the game. We are dealing with audience that is all about here and now.”

Gerstel added that he believes cloud computing prices will go down in time, and remaining affordable and timely is critical for the user. (Cloud pricing can be very dynamic based on time of day, volume of data, speed of turnaround, etc.)

GameChanger’s app was used to track more than 8 million baseball and softball games last year and is receiving an upgrade to stream in 1080p HD this year. Its President, Sameer Ahuja, said his group has invested in technical talent to optimize the efficiency of databases to minimize cloud reliance and cost. New innovations around video compression help retain quality while uploading faster.

And Ahuja previously shared that, internally, they often say “a bet on GameChanger is a bet on Apple and Samsung” because the smartphone providers are powering more on-device processing. That echoes a point from Sahai about SwingVision which, to date, is only available on iOS devices, but soon will expand to Android.

“A big reason why we haven’t gone to Android so far is because the compute wasn’t there,” Sahai said, saying it will finally be possible later this year. “We’ve started development for that, but even then, it’s only the very latest Androids, whereas on iPhone, you can go all the way back to iPhone 11.”

SwingVision’s electronic line-calling has been used in some junior tournaments as it applies to the ITF for certification that would enable use at higher levels. The audio alerts are, for now, a consumer feature to replicate the Hawk-Eye Live experience enjoyed by professional tennis players. But emitting that seemingly simple audio alert necessitates a huge behind-the-scenes engineering effort.

“Orchestrating all those different models, getting those to run as fast as possible,” Sahai said, “[required] a lot of experimentation and iteration for us throughout several years, honestly, of basically building our own architectures and getting it to run efficiently.”

This article was brought to you by SBJ Tech, a Leaders Group company. As a Leaders Performance Institute member, you are able to enjoy exclusive access to SBJ Tech content in the field of athletic performance.

Main Image: Getty Images

By Joe Lemire

The Unrivaled season is a three-month sprint, but the size of its performance staff and tech stack are robust enough for a yearlong schedule — and they are making an impact beyond their own league.

Susan King Borchardt, who has been a fitness consultant to Sue Bird, Kelsey Plum and Breanna Stewart, joined Unrivaled as its Director of Performance and Recovery. Borchardt’s first call was to Lindsey Elizondo, an athletic trainer and physical therapist with the Magic, who joined as Director of Medical.

“We crafted this idea — her coming from the NBA, me working with some of the higher-level individuals privately in the WNBA — and just saying, ‘This should be the standard,’” Borchardt said courtside before the semifinals in Brooklyn. Phantom went on to defeat Mist for the championship 84-70 in Florida on March 4.

Their department has a headcount of about 25 and is supported by an extensive array of technologies, such as Plantiga insole sensors, Kinexon wearables, Springbok Analytics MRI scans and VALD devices for the weight room. Katie Buria, Head Athletic Trainer at the Dream, is Unrivaled’s Director of Sports Science managing the relationships with the tech providers.

“What we’ve seen is that will raise the bar everywhere else that we go,” Elizondo said. “So if we set that bar here, that’s going to feed upward into the W, it’s going to feed into the NWSL, it’s going to feed into all of these other sports and force the hand of this is what it takes to perform at this level.”

The athletes have full ownership of their data and can choose whether to share it with their WNBA teams and personal coaches. Unrivaled didn’t have exact figures available but indicated that the majority of players opted into using the tech, with Borchardt calling it “part of our culture.”

“This is a very player-driven league, and they care about player safety and wellness. The treatment has been top of the line,” said Breeze G Paige Bueckers, praising both the staff and the tech, such as “the data and what we look like on the court with our Plantigas or our [Kinexons] or our chips that measure our asymmetries — little stuff like that goes a long way in basketball.”

Plantiga has made women’s sports a priority, with its CEO Quin Sandler saying of the work with Unrivaled, “Together, we’re not just collecting movement — we’re helping build the future of applied sports science for women.”

Her teammate Cameron Brink, who suffered an ACL tear in June 2024, had her first Springbok scan — which uses AI to turn an MRI into 3D muscle data — while rehabbing with Unrivaled last year and then again before this season.

“Once you have a big injury, you want to learn how to optimize everything, so all these tools really help,” Brink said. “It is really reassuring to get MRIs done like that and just a full body scan. From here you can see how you’ve grown and improved and what areas you still need to work on.”

Elizondo noted the unprecedented position Unrivaled is in by collecting all of this data since the league’s inception. By having access to data for every player, the performance group can identify trends and understand the demands of the elite 3-on-3 game, which are different than 5-on-5. That helps them train the players with more specificity and develop more targeted treatment and recovery protocols.

Tracking longitudinal data enables Unrivaled to share baseline metrics for each player, and for rehabbing players, it helps reinforce the value of the hard work. Borchardt saw it firsthand with one athlete recently.

“For that athlete to have those two visual pictures to be like, ‘Wow, this is actually really working’ — that was so motivating,” she said.

Unrivaled CEO Alex Bazzell emphasized the league’s player-first, development-minded vision as a reason why it has invested so deeply in the personnel and products to make the performance team what it is.

“These players are thinking long term, and the expectation is not just your compensation but resources of off the court,” Bazzell said. “Our staff and who we hire, those are the ones that the players need to trust. You have to build trust in a 10-week period, too. That’s not easy.”

This article was brought to you by SBJ Tech, a Leaders Group company. As a Leaders Performance Institute member, you are able to enjoy exclusive access to SBJ Tech content in the field of athletic performance.

Main Image: AFP via Getty Images

By Ethan Joyce

Turns out at least one medal outcome was wrong at the Milan Cortina Olympics, and officiating and scoring startup Owl AI founder Jeremy Bloom says he can prove it.

The women’s snowboard slopestyle finish prompted immediate criticism for the judges who awarded Japan’s Mari Fukada the gold (with a score of 87.83) and New Zealand’s Zoi Sadowski-Synnott (87.48) the silver — with notable reactions from NBC analyst Todd Richards and others.

The frustration focused on the inconsistency of scoring on jumps — specifically Fukada’s high scoring on the 720s (two full rotations) in her routine, while Sadowski-Synnott’s attempted more difficult 1080s (three full rotations) as part of her set. And it hit the exact reason why Bloom founded the tech startup to begin with — and in this case, the difference between silver and gold after a rejudging from Owl.

This graph shows the data sets utilized by Owl AI to determine that the women’s snowboard slopestyle at the Milano Cortina Games should have likely flipped the gold and silver finishes for the event. (Graphic used courtesy of Owl AI.)

“I don’t think much needs to be said about how important these moments are for the athletes,” Bloom told SBJ. “They’ve worked their entire lives to just make an Olympic team, let alone put down the best run and either miss the podium or have the wrong color medal for whatever reason.”

To check the medal finish, Owl AI created a data set to look back at the past five years of judging for both 720s and 1080s. What it discovered: The two scores Fukada earned in Milan Cortina on her 720s (7.7 and 8.3, respectively) stood out drastically. The scores on Sadowski-Synnott’s 1080s were also near the peak of scoring for the moves but still very in line historically with the judging from which Owl pulled.

Owl rejudging still slates Kokomo Murase as the bronze winner. But it flipped the finishes of Fukada (with an 86.67 adjusted score) and Sadowski-Synnott’s (89.13).

“This is where we really think we could augment the judges,” said Josh Gwyther, Owl AI’s CEO. “Because if you, as a judge, are putting in a score and immediately saw this deviation, you might be like, ‘Okay, let me see, Is there something there?’ And maybe notate that anomaly.”

Both Bloom and Gwyther have often shared that Owl is not meant to replace judges but to be a helpful tool to set everyone (judges, athletes, fans) up for success around an event. But because judges can vary in their experience, background, and biases, Owl becomes guardrails of context. For example, Bloom pointed out, 720s won the men’s version of this event in the 2006 Olympics. It’s no longer a cutting-edge move, which should be reflected in the score.

Because Olympic moments happen at a four-year frequency, it adds extra pressure to make these decisions as correct as possible.

“On one hand, you want to honor those who won,” said Bloom, a former world champion and two-time Olympic skier. “You want to say, listen, on this day, given this criteria, that’s your gold medal. And any sort of commentary beyond that shouldn’t sort of diminish you because that was their day.

“But at the same time, if that athlete or any athlete would agree that there’s a lot of subjectivity to these sports and everybody would like to push for a more objective world.”

This article was brought to you by SBJ Tech, a Leaders Group company. As a Leaders Performance Institute member, you are able to enjoy exclusive access to SBJ Tech content in the field of athletic performance.

Main Photo: Airo

By Joe Lemire

Speedskating team pursuit debuted at the 2006 Olympics, with three-skater teams racing in sync, taking turns leading the pack to absorb the wind.

At times while looping the course, because the skates don’t have brakes, one athlete might incidentally bump a teammate from behind. Back in 2018, Shane Domer, US Speedskating Chief of Sport Performance, wondered if those pushes were beneficial and how to optimize the number of exchanges at the front of the line.

Domer contacted the Chair of the governing body’s Sports Science Commission, aerodynamics expert Ingmar Jungnickel, to build a projection around these ideas. Jungnickel concocted what he called a “napkin math kind of model” that immediately showed a savings of about a second and a half, Domer recalled.

A week later, however, Jungnickel called back and told Domer, “Shane, I think we’re doing this thing all wrong. What if we don’t exchange at all?”

Jungnickel proposed that the two trailing skaters, who benefit from the energy savings of drafting, use that to push the leader forward rather than take the time to sprint out in front, likening the concept to bump drafting in NASCAR.

“The model showed that pushing was so clearly advantageous that you shouldn’t sometimes push,” Jungnickel said. “You just should abandon taking turns at the front altogether, and this should replace the old strategy.”

After years of testing — both through advanced computational fluid dynamics models and on-ice training — that revolutionary technique has propelled the American men from also-rans to both the podium and the record books. The US, which finished eighth at the 2018 Olympics, won a bronze at the 2022 Games and gold at the 2025 world championships while setting the world record in the event.

Internally, it’s called Project Slippery Fish, but to the world, the technique has come to be called the American Push.

Jungnickel had worked with Olympic cyclists in his native Germany, as well as with Tour de France teams, but at the time of this discovery, he was leading an innovation team at Specialized bicycles. Given the success he had in speedskating, Mike Levine, USOPC Senior Director of Performance Pathways and Innovation, suggested he apply for a grant from the donor-backed Performance Innovation Fund.

That funding enabled Jungnickel to start a sports tech R&D consulting firm, Inspire Gold, which then built an AI-powered aerodynamics spinoff — Airo — that replicates a wind tunnel by creating digital twins of athletes that can be manipulated in 3D to determine the best posture and formations. US Speedskating was the first client, but Jungnickel said he has also worked with national teams in ski and snowboard, cycling, luge and triathlon.

Without the grant, Jungnickel said he likely would not have started the company. Now, the core IP remains proprietary to US Olympians, but related use cases are helping support Inspire Gold. Levine emphasized that many Olympic sports don’t have technological support because they lack a large enough commercial market.

“Airo is selling the technology to bike fitters and bike shops, but there’s a speedskating version and the ski version that we will never sell to anybody but the US Olympic Committee,” Jungnickel said. “That’s our core business model: Essentially develop technologies that help Team USA win, and then commercialize them and long-term fund these businesses.”

The 26-member Performance Innovation Advisory Committee is chaired by Apple’s Eddy Cue and includes members from disparate backgrounds, such as team executive (the Spurs’ RC Buford), athlete (NFL lineman Kelvin Beachum), investor (Goldman Sachs’ David Solomon), business analytics (KAGR’s Jessica Gelman) and medicine (Texas Children’s Dr Jeff Shilt). The fund has raised about $50 million to date.

“It’s really a talented and generous group who provides us this risk capital, strategic guidance and network connections to invest in and execute bold ideas that can create competitive advantages for Team USA and elevate the performance, health and wellbeing of Team USA athletes,” Levine said, adding that the scope is “agnostic. We’re not defining what lanes we’re playing in.”

Founding committee member Geoff Yang, the managing director of Redpoint Ventures, explained the goal in 2015 was to combine “data, applied technology and ingenuity” to support and identify talented Olympians and Paralympians.

“The United States is home to the most innovative technologies in the world,” Yang said, “and Team USA should be a leader in applying those technologies.”

There are four main allocations:

• An international competitive analysis that uses machine learning to develop medal expectancies, helping drive high-performance planning;
• Athlete 360, a custom management system available to NGBs for free;
• Monthly online forums and an annual performance innovation summit “to create the collisions and spark a shared learning and cultivation of ideas,” Levine said; and
• Grants, such as the one utilized by US Speedskating and VAMA, a motion capture technology used to great effect by USA Track and Field.

“Without the funding to get these projects going, we lose steam on the innovation side,” Domer said. “Some of them fail, and these guys are OK with that. And that’s awesome because that helps us create the lack of fear of failure that we’ve had in the past.”

The speedskating team happens to be full of engineers who understand the underlying concepts. One skater, Emery Lehman, even spent time as an intern with Airo’s athlete engineering project. The willingness to experiment from the athletes and coach Ryan Shimabukuro has been critical.

“I give him and our athletes a lot of credit because they didn’t know what the payoff would be,” Domer said, “and to commit to doing this thing — well, as a speedskater, you’re not taught to skate pushing someone. So it took a lot of work to get them to adopt the technique.”

This article was brought to you by SBJ Tech, a Leaders Group company. As a Leaders Performance Institute member, you are able to enjoy exclusive access to SBJ Tech content in the field of athletic performance.