Ryan Alexander of Atlanta United came on the People Behind the Tech Podcast to discuss understanding the demands of the team, player profiling and brain training.
A Data & Innovation podcast brought to you in collaboration with
“Understanding how the physical demands and fitness is going to be interpreted on the field as it is going to relate to the technical and tactical execution of a certain style of play.”
Alexander, the Director of Sports Science at Atlanta United, was speaking to John Portch and Joe Lemire on the People Behind the Tech podcast ahead of the new MLS season, which began in late February.
He also spoke about the club’s groundbreaking work with i-Brain Tech, a neurofitness training aid that has transformed their skills and cognitive training and led to players having “higher levels of conversations with their technical coaches”.
Elsewhere, Alexander explored:
Listen above and subscribe today on iTunes, Spotify, Stitcher and Overcast, or your chosen podcast platform.
The Bees’ men’s under-18s coach discusses her career journey and the lessons she’s learned along the way.
A podcast brought to you by our Main Partners
“There’s no shortcuts, even though everyone seems to want them these days,” she tells Henry Breckenridge and John Portch on the Leaders Performance Podcast, which is brought to you today by our friends at Keiser.
“Time on the grass, working in different environments, working under different people as leaders will help you to understand what it looks like for you.”
Bedford, who has also coached with underage women’s teams at the Football Association [FA], Leicester City Women in the WSL, and served as an assistant coach at Arsenal Women, talks enthusiastically about her first six months Brentford.
In her new role she is a pioneer. One of the few women coaches operating at the top level of the men’s game.
Elsewhere in this episode, she delves into the importance of her mentors, who include Mo Marley, the current Head Coach of England Women’s under-23s.
Bedford recalls a time at an FA training camp when, at Marley’s side, she encountered the senior women’s England squad. She says: “Every senior player that walked passed her gave Mo a hug and I was like ‘I want to be Mo, I want to have that impact’. But actually, the more I worked with Mo, whilst I love her to bits and still have tremendous respect for her, how Mo leads is not how Lydia leads.
“You learn loads of things, good and bad, from people that you work under and then you find your own way.”
Elsewhere on the agenda, Bedford spoke about:
Henry Breckenridge Twitter | LinkedIn
John Portch Twitter | LinkedIn
Listen above and subscribe today on iTunes, Spotify, Stitcher and Overcast, or your chosen podcast platform.
Gary McCoy discusses Peak AI and its ability to track personality traits and emotional welfare.
A Data & Innovation article brought to you by
That was the question longtime performance coach Gary McCoy would ask his players each day, listening to the answers but also reading body language for subtle cues.
McCoy had success blending those daily check-ins with objective monitoring data from a series of tracking technologies, overseeing what’s likely the only injury-free professional baseball season with Taiwan’s Chinatrust Brothers in 2019.
In his latest venture, however, McCoy is the CEO of the previously stealth startup Peak AI that asks that same question, or a similar variation, and then leverages psycholinguistics and natural language processing to gain objective data for cognitive load — the amount of information that working memory can hold at one time — and emotional welfare.
Put simply, Peak AI seeks to identify traits and states — that is, to understand a user’s ingrained personality traits, which are largely invariable, and his or her daily state of mind, as influenced by physical and mental stresses.
“Psychology leads physiology,” McCoy said, outlining his assessment model that begins with psychological-emotional well-being before addressing physical systems, technical ability and then tactical use of the athlete. “It starts there, but I never had anything to measure it.”
For ongoing monitoring, users upload a video recording of about 30 seconds, and then Peak AI analyzes both the word choice and manner of speaking to determine one’s frame of mind. For more static attributes, such as a person’s need for group affiliation or attitude toward risk, any audio sample will work, even publicly available interviews. The AI is able to account for varying languages and accents, as well as identify attempts to trick it, such as reading a script.
McCoy, who lives in Arizona, eight time zones away from Chief Innovation Officer Walter Farfan, completed daily assessments through London-based Peak AI. One day, after saying what he thought was the usual response, McCoy received a call from a worried Farfan asking what was wrong. McCoy hadn’t said anything about it, but the intonation and timbre of his words triggered an alert. Turns out McCoy’s dog of 19 years had died the day before.
Mental wellness is the overarching mission, Farfan said, who was invited to give a talk on the subject at Buckingham Palace. Some early applications in the sports realm include individualized coaching, helping athletes reach their potential by appealing to their intrinsic motivation, rearranging clubhouse locker assignments to improve team culture, and scouting prospects to evaluate if they are a roster fit. It’s then easy to extrapolate its use from performance to other business operations, such as the mental health and culture of executive teams.
Peak AI completed seed funding rounds in July and now has a staff of six full-time employees and about a dozen contractors. The company charges $99 per month per athlete for the product. The plan is a limited rollout this year, followed by a broad deployment to teams in 2024.
Among the pro teams to have trialed or adopted the product from Peak AI include EFL Championship and Premier League sides Southampton and Brentford, Ligue 1 power Paris Saint-Germain, Formula One teams Red Bull and Aston Martin, and the NHL’s New Jersey Devils, who are the first to do so in North America.
After an assessment, Farfan said, “I’m going to speak to you about utilizing what you have internally, what you’re born with, what you’ve inherited — these unbelievable skill sets within your personality — to bring the best out of you, rather than trying to make something you’re not. And I can’t change your personality. But if I knew [what it was], I can use verbalization to say, ‘Well, I now see how you see the world.’”
Personality is roughly 70% inherited and 30% shaped by experience, Farfan estimated. And metadata analyses of large cohorts are helping unlock characteristics that underpin certain achievements, which could range from hitting home runs in baseball to excelling as an outdoor athlete in the cold climates of Scandinavia.
“We allow our clients to port in different youth players — and whatever they deem success looks like — and that builds a dataset for them to go and build a lens to shine on a group of youngsters to see which one of you 100 people have this specific trait,” he said.
Farfan and his longtime business partner, CTO Mike Blaster, have been collaborating for a decade on the study and automated analysis of language. They applied those techniques to sentiment analysis of social media posts and the development of marketing campaigns. They launched Trace Data Science in 2017 to parse ingrained human behavior from milder interests and to map cognitive load. The company worked out of Google’s London office from 2018 to 2020 as part of an incubator program.
Peak AI is a rebrand to reflect a change in mission. Farfan said external validation work is ongoing at the University of Georgia and Portsmouth University. Some prior studies have indicated the system’s ability to predict intrinsic motivation, cognitive load and personality traits at rates of 90% or higher.
McCoy said this sensitive personal information will be protected with military-grade security and be in accordance with HIPAA and all similar international medical privacy laws. A seasoned sports tech executive, McCoy was an early employee at Catapult whose signature GPS devices tracked stress on the body, at Whoop whose wearables assessed the body’s response to and recovery from that stress, and at Zone7, whose AI algorithms sought to predict injury risk.
The one piece missing was cognitive evaluation, especially not something that could be administered with so little friction. One persistent conundrum, McCoy said, was that “physically in an athlete, exhaustion and boredom present the same way. How do you know what to do?”
“We think sports psychologists will be the ones who have now a very accurate and effective tool, but they can prioritize, ‘OK, I need to attend to this person,’” he said.
“We’re teaching them how to be chameleons and how to interact with [every player]. It’s team culture. You hear that word all the time in sports clubs, right? But no one’s got a measurement tool for it. This, I think, will be the first measurement tool for culture.”
Dr Daniel Laby is talking to John Portch and Joe Lemire about his vision training with Liverpool and England star Trent Alexander-Arnold in 2021.
A Data & Innovation podcast brought to you in collaboration with
He says: “If the question is: ‘are you worse than you should be for your sport?’ And knowing what each sport needs, if you have that information you can say how someone needs to train.”
Red Bull commissioned the project having been impressed by Dr Laby’s consultancy work in the NFL and his burgeoning collection of World Series rings having worked for three decades in MLB.
“So if Trent did well, which he did in certain areas [I would say], ‘Trent, you did great. We don’t have to give you glasses’ but if [instead I said] ‘Trent, your ability to monitor multiple targets at the same time isn’t what it needs to be compared to what it should be for someone on average of your level, we have to train that’; and that’s what we did with Trent.”
Dr Laby tells The People Behind the Tech podcast that the first goal is to help athletes to correct to the required level for their sport, which will differ depending on the discipline.
This was just one aspect touched upon during the conversation. Others include:
Listen above and subscribe today on iTunes, Spotify, Stitcher and Overcast, or your chosen podcast platform.
Crystal Palace and Royal Antwerp have developed data storage and visualisation systems that increase athlete availability, enable smarter recruitment, and ensure more efficient workflows.
An article brought to you by our Partners
Dr Cedric Leduc, a sports scientist at English Premier League club Crystal Palace FC, is sharing his experience with practitioners working in athlete monitoring.
It is a natural enough recommendation for a sports scientist to make but the case Leduc makes to the Leaders Performance Institute is compelling. “If you aim to work in a sports club as a practitioner,” he continues, “one of the key things when considering your own learning and development journey is to integrate some of those data skills that will help you to understand what is actually required by an organisation when it comes to data and technology.”
Why a data infrastructure is important
Leduc, who has been at Crystal Palace for almost two years, addresses that question on a daily basis. “How can I refine a thousand possible metrics on the market into a presentation or construct of what I am seeking to present?” As he sees it, there are two viable approaches; firstly, by calling upon his own experience and relationships with colleagues, coaches and athletes; and secondly, by running statistical analyses that enable those selections to be made in an objective way.
Crystal Palace and Royal Antwerp track Player Availability using Smartabase, which enables performance staff and coaches to make informed training and performance decisions. Image: Smartabase
“A combination of both works,” he says. “When you have to turn things around quickly, you might actually use your experience, but when you have time and access to a good historical database, you might be able to run those analyses. There’s a trade-off between short-term actionable points and more long-term objective decisions.”
Leduc and Crystal Palace use human performance optimisation platform Smartabase as a data storage and visualisation solution for all players and staff, from the academy to the first team.
“In a way, building the system from nothing was an advantage because you can build it the way you like and set up the structure” – Peter Catteeuw
To operate without such a system puts a club at a disadvantage, as Dr Peter Catteeuw, the Head of Performance at Belgian Pro League champions Royal Antwerp FC, explains.
“When I arrived at Antwerp in 2017 there were no records of injuries, records of tests with the players, no records of training sessions and so on,” he tells the Leaders Performance Institute. Mindful of how well Smartabase had served him in his previous role at Racing Genk, he began to use their technology at his new club, building a monitoring system for a second time with the help of Smartabase’s agile customer success team.
“In a way, building the system from nothing was an advantage because you can build it the way you like and set up the structure,” he says. “It’s still changing every day and getting better, helping players from the academy to the first team. The team’s management and administration is also coming onboard.”
It is a product of the latitude afforded to Catteeuw and his ability to scale the system. “We needed time to build the system to our own needs and it continues to develop. On the other hand, we can implement changes immediately.”
Smartabase enables the collection of both objective and subjective sources of data relating to athlete wellness. Image: Smartabase
Of advice he would give a team who are starting from scratch, Catteeuw says, “You can easily start with a smaller group within the club, say the academy, and then progress through the organisation as you build the system.”
How accessible and actionable data can improve workflows
Alignment and accessibility are critical for new members of staff. When Leduc arrived at Crystal Palace, his first question was: where is the data and can I access it easily? “Then you start to realise that you have multiple data sources like in any sports organisation,” he says. “What’s important, if you want to make practical use of that data, is to first make sure they are stored in one secured place so that it can be easily accessed – then you can turn that into something actionable.”
“We have a holistic view of the players; what they did in training, what they are doing outside of training, how well they are recovering every day and if they are ready to train or take on more load or not, and if they are ready to play games or not” – Peter Catteeuw
At Royal Antwerp, Catteeuw welcomes the ability to tailor the club’s data management platform to his wishes thanks to Smartabase’s hands-on approach to customer success. A response within hours is the norm. A solution often follows in a day or two. “Most systems are fixed but Smartabase gives you the tools to create your own club system to enable you to work the way you like with physios, strength & conditioning staff, the technical coaches and management. Most companies only make it if it’s interesting for other teams.”
Injury Risk Profiling is an essential area of Catteeuw’s work at Royal Antwerp. Image: Peter Catteeuw / Royal Antwerp FC
Leduc has witnessed the benefit first-hand at Crystal Palace. “A new player signed this summer and the head physio asked me if he can integrate the profile of that new player so that he can start adding notes,” he says. “Another example from pre-season was the request to implement a new technology, integrating its data with their Smartabase storage system.” The organisation was able to facilitate the club’s request. “They are very reactive in trying to understand your needs and not simply relying on what already exists.”
At Royal Antwerp, Catteeuw was able to make the API work in linking the sleep tracker Whoop and Smartabase. “Now it’s up to me to pick the right data, the data we want to see, and make clarifications if necessary,” he says. “I will make the first simple dashboards for ourselves, the medical staff or the coaches to have a quick view every day. In the next days, I will try to combine data we have now from Whoop with the players’ wellness questionnaires and with all the training and game data we collect so that we have a holistic view of the players; what they did in training, what they are doing outside of training, how well they are recovering every day and if they are ready to train or take on more load or not, and if they are ready to play games or not.”
Agility is critical to data-informed decision making
There is the imminent possibility that this process will lead to red flags with some of the players. Perhaps they have not slept well on a consistent basis. This will, however, not lead to an overreaction from Catteeuw and his colleagues.
“We don’t have to take drastic action right away. These alarms just let us say ‘let’s first talk to the player and see what’s going on’ and then maybe check with the physios. Is there something else from the medical staff? Is there something from the training pitch that also raises an alarm?”
Catteeuw recalls an illustrative example from last season when Royal Antwerp used NordBord, ForceFrame and ForceDecks in strength testing. “In every first training session after a match, we ran tests. The data gives us a signal i.e. it’s too slow for these players, the difference between left and right is too large. We won’t pull them out of training immediately but we’ll check the player and see if there’s anything too serious to let them train. But most of the time it means we maybe have to adapt a little bit of training or we need to get an additional session in the gym.”
Access to the initial sources of raw data has enabled Leduc at Crystal Palace to streamline some of his processes. “The initial data collection with a given technology can be pushed into Smartabase in the right format,” he says. “I can then push it to get the right visualisation or run some analysis on it in a very straightforward way. You limit human interaction, which decreases the risk of errors. Having access to the original data enables you to be very agile with the data you’re collecting.”
However you use your databases, the important thing is to understand the needs of your organisation. As Leduc says: “Do you need a storage or visualisation solution? That will depend on your organisation.”
Crystal Palace and Royal Antwerp use Smartabase to track player soreness on a daily basis. Image: Smartabase.
The data landscape is changing and the days of teams failing to track even basic performance metrics are largely consigned to the past. In addition to Crystal Palace and Royal Antwerp, Smartabase clients include both Arsenal and Nottingham Forest in the Premier League, Stoke City in the English Championship, to AS Monaco in the French Ligue 1, Ajax in the Dutch Eredivisie and SL Benfica in the Portuguese Primeira Liga – all clubs looking to make a real difference both in training and in competition by developing a data infrastructure that enables coaches, practitioners and the players themselves to make faster, smarter and better informed decisions.
