NFL Player Health & Safety Innovation Advisor Jennifer Langton sets out the ways in which AI, AWS and the NFL’s Digital Athlete Program has had a positive impact.
Main Photo: CNBC
A Data & Innovation article brought to you by
Lower-extremity injuries have become a major focus for the league, with the first two weeks of preseason training camp – a period of re-acclimation to the sport – as the period of greatest risk, yet for the first time ever, the NFL saw a reduction in leg injuries in consecutive summers.
NFL Player Health & Safety Innovation Advisor Jennifer Langton shared that finding on stage at CNBC Evolve: AI Opportunity in New York City in October. She attributed that success, as well as changes to the kickoff rule, to the league’s work with AWS on the Digital Athlete, in which data about every rostered player on every team is anonymized and analyzed. Positional benchmarks are shared league-wide to help inform player training and usage.
“When you can integrate and aggregate data across all 32 [teams] for all 53 [players], you have more power in the data that you are generating to model,” said Langton, who for years helped lead player health and safety efforts as an SVP in the league office before leaving her full-time position for personal reasons in August.
Other work Langton highlighted was the use of computer vision triangulated with the Next Gen Stats RFID sensors to calculate the severity of head impacts, which for the first time last year was distributed to offensive and defensive line coaches on a weekly basis so they can “put in injury prevention strategies to get the head out of the game,” she said.
The reformatted kickoff was a direct result of the league’s biomechanical consultants at Biocore collaborating with AWS to run 10,000 seasons’ worth of data on rule variations to determine the best combination of a rule change that would be safe but also encourage on-field excitement.
The NFL has crowdsourced innovations in computer vision and worked with AWS on collecting more accurate tracking data. The investment in data capture is paying dividends and, Langton noted, will expand in the future to full-body limb and joint tracking. It has been a challenge to get the necessary precision for actionable insights, particularly with the high rate of occlusion in a contact sport like football.
“With the new AWS deal, that’s the focus, to build that pose estimation so that we can get to that true Digital Athlete on quantifying body movement,” Langton told SBJ in a post-panel interview.
Much of the efforts to date have been in creating operational efficiencies. A half-dozen years ago, for example, staff would take four days to manually tabulate head impacts through painstaking film review. That’s now done in real-time. Similarly, injuries would be listed in the league’s electronic medial records database as happening only in a particular quarter, so officials would have to review game film to find the specific cause. Now, those injuries are automatically tagged with a clip of the play in question.
“The infrastructure and the data to fuse that together is power,” Langton said of the work with AWS. “If you can standardize them and then synchronize, then we can integrate and aggregate across the league.”
The acclimation period was instituted in 2022, with leg injuries down 27% in 2024 compared to the year prior, in 2021. Langton had noted that the league saw declines in consecutive years for the first time.
“The decrease in the lower extremity injuries that we saw in the preseason last year led to the savings of more than 700 games that players did not miss during the regular season,” Miller said. “And so those benefits of the fewer hamstring strains or soft tissue injuries pull through into the regular season. Those injuries don’t recur as often, and the fact that the players don’t suffer the injuries in the first place mean that they’re healthier for the regular season.”
The new dynamic kickoff helped encourage 70% of kicks to be returned in the preseason, up 15% from last year. The injury rate on those played declined by 32%, with Miller noting that player speeds — which are calculated by the Next Gen Stats RFID chips in players’ shoulder pads — were about 20% lower on average in the reformatted version of the kickoff, in which most players line up 5-to-10 yards away from each other.
“Because we eliminated some of the space and therefore decreased some of the speeds, that led to a substantial decrease in the injury rate,” Miller said. “In fact, we saw zero ACL injuries on the kickoff. We saw zero MCL injuries on the kickoff. And those huge time-loss knee injuries are going to substantially save a lot of players, a lot of time.”
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.
Springbok Analytics uses AI to create a tool with the potential to help all 450 players.
A Data & Innovation article brought to you by
A graduate of the NBA Launchpad, Springbok Analytics has been scanning players for more than three years and has grown its team partner list to 10. To date, performance and medical staffs had used the product to detect muscle asymmetries and fatty infiltrations into the tissue, both of which can be early signs of injury risk.
But teams didn’t gain value from Springbok’s normative database because, even though there is a large proportion of elite athletes to go along with recreational competitors, most NBA players are outliers for their height and ability.
Utah Jazz Director of Performance Science Barnett Frank said the new NBA database “allows us to really be a little more strategic with our information.”
“One of the biggest challenges I have in the space is always getting asked, ‘Well, what does that mean for an NBA player?’” he added. “There’s 450 of them. When we’re comparing them to the general research or what’s out there in the population, it’s really hard to make any specific conclusions for them. So knowing that it’s NBA-specific for us, that really gives me a little more juice, for lack of a better term.”
This feature has been requested by teams for a while, said Matt Brown, Springbok’s Director of Sales & Business Development.
“It’s the first time they’ll really use the comparison mode,” he said. “Now they’re going to have a better pathway forward of team-wide analysis, understanding how strength and development is working for their players, and what metrics that means, and what that looks like, and is there an attainable phenotype that they’re going after in comparison to other players?”
Brown added that other sport-specific databases are in the pipeline. A pro soccer database is next — consisting of players from MLS, the English Premier League and Championship and other European leagues — and slated for this fall. American football would follow that, primarily of college football players who participated in the NFL-funded hamstring injury research study. Similar datasets for women’s soccer and the WNBA are also progressing toward possible 2025 launches.
In 2023, Springbok Analytics was one of SBJ’s 10 Most Innovative Sports Tech Companies and also won Best in Athlete Performance Technology at the Sports Business Awards: Tech. Nominations have opened for this year’s awards, with the nomination window closing on Oct. 21. You can review the categories and make nominations here.
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.
The NYCFC custodian recently featured in SBJ Tech’s The Athlete’s Voice series where he discussed his career, education and forays into the business world.
A Data & Innovation article brought to you by
You can’t have a discussion about sports technology today without including athletes in that conversation. Their partnerships, investments and endorsements help fuel the space – they have emerged as major stakeholders in the sports tech ecosystem. The Athlete’s Voice series highlights the athletes leading the way and the projects and products they’re putting their influence behind.
* * * * *
NYCFC goalkeeper Matt Freese has started all 30 of his club’s matches this season in MLS and has a 73% save rate, stopping 102 of the 140 shots he’s faced. This is his second season with NYCFC after beginning his career with the Philadelphia Union. A native of Wayne, Pennsylvania, Freese signed a homegrown contract with the Union and, after two seasons at Harvard, made his MLS debut in 2019.
Freese, now 26, had an opportunity to begin playing pro soccer after graduating from the Episcopal Academy in suburban Philadelphia but elected to attend Harvard first. Though he left early, Freese completed his degree remotely, balancing Ivy League studies with professional soccer, which is something he actually considers an advantage for his athletic career. A curious mind and avid follower of sports business — and a reader of Sports Business Journal, he revealed — Freese wrote independent research projects on MLS franchise valuations and advanced analytics for expected goals.
On opting for college, not pro soccer, right after high school…
The first and most obvious [reason] is just the fact that I wanted to honor my parents’ wishes to go to college. When I got into Harvard, they pushed even harder. I was really fortunate and lucky for that to happen. My dad had gone there, and he really wanted me to make sure I got a degree. The really awesome thing about Harvard, or most colleges at this point, is if you go for a semester or two, you pretty much lock in the ability to go back and finish your degree at some point.
On a more personal level, I don’t know if I was ready to be an adult and live on my own outside of the college setting when I was 17, 18, signing a homegrown contract. Also, goalkeepers usually develop a little bit later, so there wasn’t as much of a rush, if that makes sense. Seeing now the way my career has unfolded and changed, maybe I’d make a different decision and start it earlier, rather than waiting that year.
Ira L Black – Corbis/Getty Images
On balancing school and sports…
I would [attribute] most of my on-field soccer career development to my off-field academic efforts. This was at a time when I was 19, when structure and schedule is so important for a 19, 20-year-old who’s now a professional athlete making good money and getting pulled to do things that that most 19, 20-year-olds are getting pulled to do. Having the structure, having a few hours of work every night after training, making sure I had to be on a good sleep schedule, it all really allowed me to focus on soccer and not get distracted with other things. It really grounds you. It humbles you.
The other thing that I really liked about it was that it gave me a de-stressor off the field. As a 19, 20, 21-year-old, you’re now competing for your career every single day that you’re playing, and it becomes stressful, and as a young guy, you don’t really know how to handle that. So when I got home, and I would be doing work, reading a textbook, doing whatever — my mind was able to get away from soccer, which is super important.
And then the third thing that’s also quite interesting is that there’s a lot of research out there that really supports cognitive development, especially at that age, and your ability to solve problems, lead and organize and be a team leader. A lot of that is correlated with academic and intellectual stimulation. As that was continuing to grow, as my brain was continuing to be pushed and grow, it allowed me to, in my opinion, learn more quickly on the field. Learning quickly, learning on the fly, is completely necessary for a professional athlete.
On writing an undergrad thesis on MLS franchise valuations…
It was my last semester. I had finished all my core requirements, and I was doing everything remotely and then flew up to take exams in the offseason. And so I was able to do two independent research projects as my last credits. The title of one was “The theoretical analysis of the rise of MLS valuations.” Since 2010, they just completely skyrocketed, and the whole point of what I was discussing is that demand was going up. The supply was very limited. It was very constrained for several reasons. The primary one is expansion is limited within the MLS.
Probably the bigger focus was just talking about how demand, from an ownership perspective as well as from a fan engagement perspective, is going up. The academy situation has really changed everything. People want to go see kids or teenagers from their hometown that they knew growing up. They want to go see them play. They want to see them succeed. The US team is obviously getting more and more attention year over year, and that impacts the way fans look at MLS games.
People want to buy into these teams. They’re becoming more and more profitable. Revenues are going up. Operating expenses are also going up, and salaries continue to increase, and transfer fees just always are rising. But in general, they’re just becoming more profitable and easier to operate.
On writing MLS papers while a player for the Union…
I was in my third year. I’d always go to this one coffee shop in Philadelphia and work on that paper. The other independent research project I should probably mention because it’s somewhat related was, I created an expected goals model using data from MLS over the last five, six years, which was also really cool.
Goalkeeper is a weird position [for analytics] because essentially the only one that matters is the post-shot expected goals model and how that relates to the goals conceded. Goalies are a little bit of an anomaly, but in general, yeah, I love looking at data. I love talking to our data analytics team in the organization about these things. I just think it’s really interesting. It can shape a strategy of a team to a degree.
It can’t completely take over what the philosophy of the team is, but it can point you in the right directions or show you what type of cross has the highest percentage of expected goal coming from the end result of, leading you to probably want to look at getting into the cutback scenario more than these long, high crosses. We’re a relatively younger team. Our height and our strength isn’t as much so fighting against these big center backs might not be as successful as getting into that cutback zone, which is something we’ve worked on a lot. This is not me driving that, by the way. [laughs] This is the coach, the data analytics team making those decisions, obviously.
Jeff Dean/Getty Images
On not looking too deeply into his training data…
I am into that, but I just trust our performance director and the medical performance side of things on the team. They handle all that, and they make sure that my dive count is not too high, my explosive [actions] count is not too high. I am hitting the numbers that they want, and I just trust them to do that. They’re very good at their job.
On his interest in sports business when he retires…
I do think about it. The clear priority right now is playing, and I want to play for a very long time and have a good career and get my name into the that top tier of MLS, goalkeepers. But at the same time, I also take a serious interest in what my post-playing career will be. I believe one avenue would be to stay in the sports realm, whether that’s on a business operating side, being on the finance or marketing side of an organization, or the sporting side —GM, Assistant GM, sporting director, that type of thing — is really fascinating as well.
And then there’s also the investing side. I have a background in investing as well. I took several classes and audited some MBA classes at Wharton when I was in Philly. So I’m comfortable and really enjoy that type of stuff. A lot of it also depends on how my playing career goes.
On his game prep…
As a goalkeeper, the routine really is everything. And I’ve become somewhat psychotic about my routine before every game. There’s a lot of research that has indicated that, for an athletic event, your sleep two days prior is actually more important than the sleep one night prior. So my routine really starts two days before the game. I try to get as much sleep, like 9, 10, hours two nights before, and then I usually do a series of meditations leading up to the game. I do the same type of film, just very serious about my routine.
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.
Data Analysts Julia Wells of the UKSI and Mat Pearson of Wolverhampton Wanderers deliver a series of practical tips to help address one of sport’s notorious blind spots.
An article brought to you in collaboration with
That is according to a straw poll of attendees at a recent Virtual Roundtable hosted by the British Association of Sport & Exercise Sciences [BASES] and the Leaders Performance Institute.
The perception is worse when it comes to analysis and recruitment, with over 60 per cent of attendees suggesting that their analytics and recruitment teams do not work closely at all.
Yet 63 per cent also believe that improved data and computer literacy across their staffs would directly impact performance.
The sense that there is room for improvement gave the session its title: ‘Mobilising Performance Analysis in Practice’. It was the second in our three-part collaboration with BASES called Advances in Performance Analysis and centred around two case studies.
The first was delivered by Julia Wells, the Head of Performance Analysis at the UK Sports Institute [UKSI], and the second by Mat Pearson, the Head of Performance Insights & Data Strategy at English Premier League side Wolverhampton Wanderers.
Five areas where data literacy can improve performance
Before Wells and Pearson delivered their insights, attendees were set a further task: ‘as a consequence of improving data or computer literacy, describe what you would see as being the most significant impact on performance’.
The responses were varied but five stood out:
How the UKSI are mobilising performance analysis work-ons in meeting common challenges in data analysis
The first session highlighted the four biggest challenges facing people who use data analysis in sport. To kick off her presentation in the second, Wells explained how the UKSI is trying to tackle those four challenges (plus another).
Challenge 1: integration
Go back to basics. That’s the approach of the UKSI, who have placed an emphasis skill development, support structures and a clear data strategy.
It goes like this: the relevant staff members are upskilled in areas such as collecting the right data, using the correct formats in the right places before the interrogation and analysis even begins. This is then supported by a clear data strategy geared towards performance planning. For example, roles such as the data & insight lead and the performance data lead are embedded within the organisation to better help those leading programmes with the direction and the integration of their data. Thus, the strategy can come to the fore and everyone can better understand what needs collating and why within the team.
Challenge 2: data collation
Wells described how easy it can be to stay on the “hamster wheel” of collecting data without taking the time to critically reflect and pause. Can you, for example, call upon efficient processes for collecting data and wade through the myriad datasets potentially available? She recommended asking “quality questions”: why are we creating the data, what is its purpose, what decisions is it informing, particularly in the coaching process? Teams should do this periodically and continue to plan, do and review. Wells also encouraged engaging in conversations with key leaders in the environment to discuss what to start, continue and stop. It’s important to intentionally carve out those opportunities as part of your performance planning.
Challenge 3: communicating data insights
Wells stressed the critical nature of human engagement in the process and regards communication is a highly technical skill, despite the views of those who might see it as a ‘soft’ skill.
She shared that the different performance departments within the UKSI work closely with the psychology team to help elevate understanding of self and others. Wells said, if we can better understand the people we work with, it will support how people can get the best out of each other. As part of this process, they’ve tapped into better understanding one another’s preferences in order to be more impactful in how they support each other.
Challenge 4: buy-in
It is not uncommon for senior stakeholders to not perceive the value of the work being done. This makes it incumbent on analysts to critically assess their impact and share the meaningfulness of their work. “It’s our job, and it’s our role to be critically analysing why and presenting that back,” as Wells said.
On that note, alignment to the sport’s strategy helps to provide a clearer connection. If this alignment and connection isn’t there, you’ll naturally get disconnection so it will be more challenging to get the buy-in.
In addition, relationships are just as critical when generating buy-in. Wells advocated inviting leaders and key stakeholders into your world and shadowing them. When they immerse themselves in better understanding the process you’ll find that it can quickly lead to them becoming a voice for you in wider conversations.
Challenge 5: data illiteracy
Too often, practitioners can suffer in silence when looking for solutions. In the latest Olympic and Paralympic cycles, Wells and her colleagues are seeking to increase data literacy across the board. They have introduced an internal online data community that provides access to resources, promotes connection, and leads to the sharing of good practice.
Wells’ team also put together a ‘Data Leadership Programme’ which is focused on pulling together the data leaders in the various sports with whom the UKSI work to look at opportunities, challenges and future direction. Courses, with titles including ‘Data Camp’, ‘Project Automate’ and ‘Code School’, were created to improve skills and processes for coaches and practitioners to help them be more efficient. In her mind, this has been crucial to enable people to be upskilled; and all support staff should be able to ask a good question and have the data skills to answer them.
How data analysis is supporting coaching and recruitment at Wolves
Pearson explained that he and his colleagues at Wolves are trying to align the club to ensure there is consistent evidence available and better identification of the trends impacting decision making from a data point of view.
He focused on two key areas: coaching and recruitment.
In the environment, the analysts are part of the multidisciplinary team. They are very much now voices in the room and, with it being a specialised discipline, all analysts must have an impact on decision making.
To that end, Pearson’s team have moved away from leaving the coaches to find the solutions themselves. Instead, analysts are encouraged to go and find solutions, present them to the coach, and then have good conversations to better find the optimal outcome.
Part of the challenge we can face, said Pearson, in particular with performance analysis at first-team level in professional football, is that many environments can be quite coach-led, which is in keeping with the nature of short tenures. The coaches will lean into their viewpoint as a way to exert their control. Therefore, education is important and, in particular, how you communicate with them to ensure the message lands. That said, Pearson observed that coaches in modern day football are more attuned to data and performance analysis and are much more data literate and comfortable with technology.
A key learning when integrating performance analysis and data work with coaching is to make insights as contextual as possible. If you provide insights to a coach that are out of context, you’ll lose them straight away.
Pearson told attendees that some of the biggest strides in performance analysis and the wider data team have been in the field of game modelling, recruitment and selection decisions, with the obvious caveat that subjective input is still valued immensely.
The team have worked to create objective measures against the game model. In better understanding this, it has provided an additional layer of information related to individual player requirements for the game model. These insights are helping to inform both selection for matchday but also the recruitment of new talent. When thinking about the recruitment process in particular, Pearson said this process has helped to educate scouts and other recruitment personnel in the attributes for which they should be looking.
Visuals have played a key role in this process too, particularly in being able to show what it looks like to play in this particular style that the coach or manager wants. They’ve worked to make the playing style more objective.
Natasha Patel of US Soccer and Simon Wilson of Stockport County discuss the influence of performance analysis on organisational strategy.
An article brought to you in collaboration with
That is according to a straw poll of attendees at a recent Virtual Roundtable hosted by the British Association of Sport & Exercise Sciences [BASES] and the Leaders Performance Institute.
We have collaborated with BASES on a three-part series called Advances in Performance Analysis and kicked things off with a first session, titled ‘The Influence of Performance Analysis on Organisational Strategy’.
Leading the conversation were Natasha Patel, the Director of Sporting Analytics at US Soccer, and Simon Wilson, the Director of Football at League 1 side Stockport County.
They began by leading a discussion of the biggest challenges facing people who use data analysis in sport. There were four that stood out:
Patel and Wilson, who began their careers in sport as performance analysts, shared a series of considerations rooted in clear principles, effective communication and strategic benchmarking when leveraging performance analysis to drive organisational success.
Establish key principles
Both Patel and Wilson continually referred to the importance of key principles. These, as Wilson explained, must outline how you are going to work and how data and analysis inform this; this allows for more creativity (and alignment) when you move through the layers. Patel, who worked at Premier League club Southampton across two spells, explained that from the beginning of her first spell, between 2011 and 2019, there was immediate buy-in from the technical director, who valued data and video analysis hugely.
Have a clear game model
A game model – a common requisite in football as well as other sports – can inform everything that follows, including data analysis. Patel said she better understood the coaches’ needs and how they want analysis delivered when there was a game model to follow. She and her colleagues were able to gain the buy-in of coaches when being intentional in spending time with them. This allowed the analyst to shine when they were able to take information from the coaches themselves and the athletes, turning it into digestible data and visuals that could help everyone. Similarly, Wilson explained how Stockport’s game model has informed their squad building and helped to generate a well-filtered target list of players who may improve the team.
Consider the end user
As Patel said, it is important to consider the end user and what performance analysis looks like to them. Once you have identified the end users, you can then work out how to get the best process for them and, subsequently, enable the trickling of information to help influence the end user, whether that be to help support or challenge their way of thinking. She referred to this as ‘stakeholder mapping’. In her second spell at Southampton, between 2022 and June 2024, Patel came to understand that each stakeholder had a unique information threshold and that more education could have been provided in-season for different stakeholders. This was a good reminder to Southampton that as performance analysis teams and departments grow and mature, so does the quantity and depth of insights.
Know the journey
Wilson, who has been with Stockport since 2020, shared that at the beginning of their current seven-year plan, they adopted a version of the Elo Rating System (derived from the world of chess), with support from a third party, to showcase the quality differences between clubs, leagues and countries. Wilson explained that the system provided objective insights into how much better the team needed to be and how they needed to grow to progress through the leagues. Engaging in this benchmarking exercise then informed the business case of how much to invest in players, staff, facilities and other infrastructure.
Patel spoke more specifically about the influence of performance analysis on player and athlete auditing and the amount of impact it has had in this space. When primarily operating in an academy environment, there are also decisions to be made around retaining and transitioning players. These metrics formed a core part of how decisions were made at Southampton, whether they were to challenge opinions and assumptions or to simply create more productive conversations. As a matter of course, Patel’s department collected athlete maturation data, leveraged the Premier League’s game-wide injury data and, finally, garnered insights from character profiling.
BreakAway Data’s new app aggregates health information from clubs, national teams and private consultants.
A Data & Innovation article brought to you by
The new product, BreakAway Pro, aggregates health information from all practitioners — all clubs, national teams, private consultants — through an athlete’s career where it can be displayed and compared against game stats, tracking data and training workload. It is available for all interested leagues and unions, with a custom-build for a first, unnamed partner almost complete.
Since launch, BreakAway has secured deals with the NFLPA, NWSLPA, WNBPA and Athletes Unlimited, among others. Its founders, Dave Anderson and Steve Gera, regularly heard from agents, athletes, investors and other stakeholders that adding EMR capabilities would be a helpful addition to the product.
“We didn’t know we needed to move this mountain in order to give all athletes access to their data, but this was the key piece and the key thing that was missing in sports that we’ve now got,” Anderson said, adding that the topline benefit of this fingertip retrieval is ensuring that what “costs them time, money and effort are now guaranteed and done quickly and swiftly.”
While the data infrastructure was largely in place, meeting the standards for EMR access required significant outlay from BreakAway — a 2023 SBJ 10 Most Innovative Sports Tech company honoree — to add higher levels of insurance, meet HIPAA compliance and build maximum digital security, including a revamp of its AWS storage. Anderson estimated that this project consumed about 75% of the company’s time, money and effort for most of the past year.
Athletes register using multi-factor authentication that is verified by government ID, and all records are stored in a secure server, with none of the information stored locally on a mobile device. Users can manage settings over who has access to what information, toggling permissions on and off as they change teams or seek additional opinions.
“Players have been advocating for better access to their data for a long time, and BreakAway was the first company to build a product specifically tailored for players,” Meghann Burke, NWSLPA Executive Director, wrote to SBJ. “They have set a new standard for what, how, and when information should be delivered. It’s no surprise that they continue to innovate in the digital space, providing players with functional and accessible data solutions.”
Anderson, who had a six-year career as an NFL wide receiver, recounted his own experience attending NFLPA-backed health and wellness testing at the Cleveland Clinic. When he returned to the same facility three years later for an additional checkup, the computer systems had changed, and the doctor couldn’t easily see his past records. Anderson had to bring his own paper copies, making him think, “There’s got to be a better way to do this.”
While that’s an acute pain point in elite sports, it’s also an issue for everyday people who change medical practices.
“We’re the first company that is daring enough to take it on. We built this for players, and let’s see how it works because this really doesn’t even exist in the normal world,” Anderson said. “It’s a huge build, and something hopefully that resonates well beyond just sports.”
Intelligence within the app helps provide context and comparisons to normative datasets. Visual tagging of joints and muscles is one of several ways to filter the information a user is searching for. BreakAway Pro also is agnostic to other EMR providers and supports all types of medical imaging as well.
“We heard from enough leagues and we heard from enough people that we were like, ‘All right, let’s just go all in. Let’s bet the farm on our company on this,’” Anderson said. “We claim to be the athlete data company and to have the app where they put all their information, and if this is the most important piece of information that they want, what are we doing here? It is the core piece that ties everything together.”
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.
Gemini has partnered with professional and college sport teams across the NFL, NCAA, European football and beyond.
A Data & Innovation article brought to you by
QPR soon turned its season around, however, after a change in leadership. The club won 10 times and added five draws over its final 19 matches to secure its spot in the second tier of English football another season. Most estimates suggest that relegation from the Championship to League One is a financial hit of more than $10 million (£7.9 million).
Right around the time of that upset over Leicester, QPR onboarded a new AI-powered predictive modeling tool, Gemini Sports Analytics, to make optimal use of the massive datasets they’ve compiled. Gemini is a “force multiplier,” CEO Jake Schuster has said, by simplifying the process of building machine learning algorithms catered to each club’s specific needs.
“What I really liked about Gemini was they didn’t have an ego in trying to solve every problem,” QPR Director of Performance Ben Williams said. “They created a tool where you can solve your own problems.”
Around the time of QPR’s on-field nadir, CEO Lee Hoos retired from that role, while remaining as chairman, and hired Christian Nourry as the new chief executive. Nourry was 26 and a managing partner at Retexo Intelligence, a data analytics and advisory business that worked with Real Madrid CF, AS Roma and the Mexican national team. (He became the youngest CEO in English soccer, with one European executive describing Nourry as “the Lionel Messi of the football business world,” according to the Independent.)
Nourry wanted to implement market-leading solutions to upgrade the club’s tech stack. QPR asked itself, according to Williams, “Are we able to interrogate that data optimally, to forge outcomes that are positive for the long-term future of the football club? Our answer to that was ‘no.’” That prompted the search that led the club to Gemini.
The very thesis of Gemini is to empower analysts, coaches or “anyone with a dataset,” as Williams put it, to take action with data. He noted that it can be used for everything from tactical match plans to traffic probabilities on bus trips to road matches.
Founded by Schuster, a longtime sport scientist, Gemini leverages the tech infrastructure of cloud and AI partners Snowflake, DataRobot and Databricks with data sources such as StatsBomb, SportRadar, Genius Sports, Sports Info Solutions, SkillCorner and Infinite Athlete.
As an example of what’s possible, Shuster explained that Gemini users can apply clustering algorithms to match stats and tracking data to create passing trees to identify how opponents like to create scoring chances of their own or concede them to others.
“The early lift was certainly centered around pre-match and post-match reports,” Schuster said. “So, opposition analysis — how do we approach this game? And then, post-match, what happened and what are the implications for future events? A big part of the early work with them was helping them automate those reports. And then the next step was approaching the summer transfer window.”
But it also remains an area of exploration, as QPR onboards more staff members over time.
“The power comes from our curiosity,” Williams said. “We’re in a phase of play and learn and discovery.”
Other Gemini clients include the NFL’s Indianapolis Colts, the SEC’s Texas A&M and Italian soccer club Parma Calcio, which just claimed a Serie B title to earn promotion back to Serie A. The Raleigh, North Carolina-based company also raised two investment rounds north of $3 million in the past year. There are now 27 sports franchise owners either directly invested in Gemini or through recent round-leading investor Will Ventures. QPR’s owners individually own minority stakes in two MLS clubs (LAFC and FC Cincinnati) as well as MLB’s Cincinnati Reds.
That financial backing has led to Gemini’s first customer success hire, former Arizona Diamondbacks Director of Operations Sam Eaton, and a budget allocation to hire a CTO, a role Schuster is actively recruiting. The company is also in the testing phase of some new generative AI features it hopes to roll out soon.
“The whole idea behind going with this tool was we can be really broad in our thought process of what we think helps our performance,” Williams said, “rather than be penned in by somebody else’s thought process of what is important to performance because they’ve created a tool that solves a problem that they once had.”
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.
In time, players on the ATP Tour will benefit from benchmarking data and the establishment of definitive norms in tennis high performance.
Main Image: the tests, which highlight the demands of being a professional tennis player, can be used to showcase the attributes of the ATP Tour’s next generation. (ATP Tour)
A Data & Innovation article brought to you by
Before the tournament began, one of the courts at the King Abdullah Sports City was strewn with gadgetry: force plates, cameras, flashing lights. The ATP had organized the first athletic combine in Tour history, called Basecamp, putting the young men through a series of NFL Combine-style drills such as the 10m sprint, vertical jump and a pair of agility tests.
For the actual tennis matches, the players were offered tracking devices — GPS trackers with conjoined heart rate monitors — with two wearing them and five others indicating they were interested to do so in the future after having more time to acclimate. In combination with the existing player and ball tracking from Hawk-Eye, with Kinexon’s data analysis platform, the ATP Tour produced a Physicality Index to measure the athletes’ exertion and effort.
All of the data collected was shared with the players and some of it was published in web stories and social media to tout their athleticism. BreakAway Data’s app was used to grant the athletes easy access to their data from matches, practice and Basecamp while ai.io’s mobile tech unit, aiLabs, provided the testing equipment.
One of the motivations for the adoption of the new tech and data from Basecamp is to support athlete wellness guidelines, such as informing mandatory rest periods between matches. (ATP Tour)
“We really want you to understand yourself a little bit better off the court and tell that story to the fans as well because I think tennis players are great athletes, but we’ve never really had anything to measure that,” said James Marsalek, ATP Tour Senior Manager for Strategic Projects & Event Operations. “All these different metrics will then help us tell a slightly different narrative of, ‘Actually, you know what, they are [great athletes] because they compare X, Y and Z on the scales with basketball, football, whatever it might be.’”
The activations around the Next Gen Final were the most acute example of a broader strategy from the ATP Tour. In 2023, Tennis Data Innovations, which is the joint venture of the ATP and ATP Media, mandated that every tournament court have player and ball tracking. In March, Marsalek said the Tour started offering raw tracking data to all players for free. In September, the ATP and TDI created Tennis IQ, an analytics platform accessible to all ATP Tour players.
Marsalek said the goal is to have video embedded and synced to the data by 2025, with integration and visualization of wearable and other biometric data on the road map as well.
“It’s trying to tell this full story where players have got this one-stop shop that has access to everything,” he said, adding, “We tried to level the playing field and provide access to all our members.”
Though the ITF, the international governing body for the sport, began permitting wearable technology in matches back in 2019, the ATP didn’t sanction it until its most recent board meeting in November. It remains contingent on the Tour platform supporting it, which Marsalek estimated should happen in the first quarter of 2024.
“The ATP have got a hugely ambitious and fantastic opportunity to make data not just relevant but really progressive for the sport,” said BreakAway Data Head of International Business Ben Smith, who formerly led research and innovation at Chelsea FC. “Tennis, with the ATP leading it, have got an opportunity to help the sport progress over the next two, three years in a way that is, I think, hugely exciting and will advance both the physicality and the quality of the sport in a way that fans should be really excited about.”
In a video summary of Next Gen Basecamp produced by the ATP Tour, Arthur Fils — who ranked first in every category — could be seen celebrating his wins, a testament to the competitive spirit even with something brand news.
Just as often, the players asked, “Is that good?” Officials from the ATP Tour, ai.io and BreakAway Data were able to share some benchmarks, but more definitive norms will be established as this combine testing grows. Flavio Cobolli, who finished top-three in three of the four tests, called it a “good experience” and was quoted saying, “I want [Carlos] Alcaraz to do this for sure.”
Alcaraz is perhaps the premier athlete on the men’s Tour right now, whose 2022 US Open title run scored highly on the USTA’s Physicality Index, and would surely be a devastating combine competitor. But while the NFL Combine and team pro days are a rite of passage for all top prospects to improve their draft standing, that incentive doesn’t exist in tennis. To induce the elite players to participate, attaching prize money or other reward is likely necessary. But the accompanying videos and data may well be a new sponsorable asset.
“We ought to be a little bit creative,” Marsalek said. “We don’t have the same sense of jeopardy as the NFL does, where there’s a lot on the line, so we need to make sure that our athletes enjoy doing it. If they don’t, there’s no content.”
All of the data collected from Basecamp was shared with the players and some of it was published in web stories and social media to tout their athleticism. (ATP Tour)
The other motivation for the adoption of the new tech and data is to support athlete wellness guidelines, such as informing mandatory rest periods between matches. In the ATP’s 48-week season, most players average about 25 tournaments and, with Masters 1000 events all expanding to two weeks, that increases the amount of travel time.
Marsalek emphasized that data will not become the sole determining factor in decisions, but it is intended to provide a balance with an athlete’s feel in a skill-based sport. The goal is to encourage the use of data but have a centralized process to govern it — which should aid all stakeholders in tennis, just as was evident in Basecamp.
“It’s genuine high performance. Yes, it’s really enjoyable and competitive, and so the athletes have a good engaging experience. But there’s also valuable insight that those practitioner teams will take and move into the training environment,” Smith said, before adding about the development of data-driven narratives for fans. “That’s just really good, interesting engagement that, I think, opens up tennis to a slightly wider market.”
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.
