Arunava about Football A look at football & the world through my eyes!
Related Articles
When Fred Lipka moved to Major League Soccer in 2015, he did so with a timely passion and essential experience. Carrying years of work in French academies, Lipka hoped to shape what he believes can be an elite soccer league. But how does one reach that lofty goal?
Simple, says Lipka. It all comes down to development.
As MLS Vice President of Player and Youth Development, Lipka collaborates with academies and development programs strewn across two expansive countries (the United States and Canada), furthering homegrown talent reared at home. Lipka namechecks Canada star Alphonso Davies and US captain Tyler Adams as just two examples to emerge from MLS.
But Lipka’s mission and role require clear vision (and effective persuasion) that permeates far beyond individual talent. Cultural shifts in how Americans and Canadians approach the sport have been necessary, especially at the youth levels. Coaching development remains key, too.
And with an eye always on innovation, Lipka has eagerly searched for new ways to address unique challenges specific to the landscape (literally and figuratively) across North America. That’s where a novel partnership between Major League Soccer and London-based technology company ai.io comes in.
Announced earlier this month and available in December 2023, ai.io will partner first with MLS NEXT as a no-cost, accessible way for aspiring soccer players to participate in virtual trials from their backyard or their neighborhood park – wherever they have a camera and a ball – and then have the opportunity to be scouted by MLS. Through ai.io’s mobile recruitment tool, aiScout, a digital portfolio and assessment tool will be made available to anyone interested in using it.
Via technology that gathers and assesses the strengths of potential recruits, at no cost to the participant, MLS clubs can reach a wider breadth of potential talent, overcoming evergreen barriers of geographic hurdles to scouting players. It’s an extra tool for MLS clubs to supplement (not replace) their traditional scouting efforts and identify players that might slip through the cracks or just need a chance.
This new partnership is, most importantly, about accessibility. But, according to Lipka, this is also all about talent ID.
Talent Identification
Talent identification is why ai.io got started in the first place, as a tool built to serve the needs of both the player and the scout. Based in London and developed in partnership with Chelsea FC, the idea – according to aiScout’s COO and director of sport science, Richard Felton-Thomas – was born when the founder’s son was let go from a London-based academy, left floundering with no digital portfolio to properly quantify and then market his skill set to other potential clubs.
Consulting London-based Premier League clubs with that predicament, and a potential technology to solve it, Chelsea signed on to become ai.io’s research and development partner. They remain working in concert to this day, with new innovations for assessment and recruitment expanding through all levels of the club (including first-team scouting and assessment) and eventually, to include Chelsea’s women’s team as well.
As ai.io developed the technology in connection with Chelsea – collaborating with over 80 scouts, coaches, and sport science professionals – they cultivated an artificial intelligence-based platform to gather data on three key pillars.
“We have three core underpinnings, which is physical, so in the US you probably say athletic data a little bit more,” said Felton-Thomas. “These are things like your sprints and your jumps and your agility and your reactions, so the physical attributes that you have as an athlete.
“We have technical, that’s the second pillar. Technical is anything related to the ball really, what you actually do as a soccer player: passing, dribbling, shooting. And then we have the cognitive, psychometric side. So just looking into some of those cognitive skills around visual perception and visual processing.”
Players uploading their completed drills to assess the above pillars will have three attempts to perfect them before officially completing each one. This allows the player to receive and incorporate feedback from the application, which can include automated feedback as well as a customized, human component.
“So the player can go into their garden, a parent or friend will hold the phone, and they’ll just mimic the material they’re supposed to be doing,” said Felton-Thomas. “That video goes off to our cloud and runs all the analysis over the top, comes back down to the phone, and they’ll get a video back with tracking lines all over the body to show how the movement happens. And it’ll have scores and it will have some text feedback based on the scores. That is auto-generated based on a whole set of criteria.”
A human component for feedback through the app is also possible, and is among many customizations MLS may utilize as it rolls out the product to suit the needs and interests of the North American soccer landscape.
Other customizations could include what specific skills are being sought out for any given virtual trial, or the number of attempts players will have in completing each “drill.” For example, as Felton-Thomas talks through the three-attempts-per-drill model that is implemented in the UK, he notes Major League Soccer can shift that number to meet preferences while gauging potential talent.
“MLS, that comes down to their sport science, really, to say, this is what we want from it,” said Felton-Thomas. “They might be fine with a player doing five tests. They might want it all averaging out. Or they might just want to take the best one, or the last one. That’s part of the customization as well, to make sure that we do it the right way for whoever we’re working with.”
After years of working with Chelsea to develop the tool, ai.io has since gone on to partner with recently-promoted EPL side Burnley FC, and has also worked with a private academy in India.
For ai.io, Major League Soccer is the next big step.
