You're listening to TED Talks Daily, where we bring you new ideas to spark your curiosity every day.
I'm your host, Elise Hu.
What does AI have to do with sports equity?
I have to admit it's not a question I've thought about before now, but in this talk, sports scientist Richard Felton Thomas shares how he's helping to level the playing field across gender, geography and background, and reimagining how athletes are discovered and supported using tools like AI and biomechanics.
I'm going to start by getting you to visualize sporting greatness.
And what do you see?
It's probably athletes like LeBron James, Caitlin Clarks, Saquon Barkley, Simone Biles Ronaldo, Messi.
You see, we tend to think of athletes from quite a small subset of countries.
So I'd forgive you for thinking that the best talent in the world just comes from those places.
That's not strictly true.
Because talent existed long before their sporting greatness.
And many things have to happen along the way for them to realize their potential.
Now, one thing that's commonly overlooked is that they got an opportunity and were visible.
You see, talent exists everywhere.
It's finding the talent that can be the challenge.
So how do we do that?
Now, typically, that's scouting.
It's every young athlete's dream to be scouted.
You see the old guy up in the bleachers, he's writing his notes, and late you get the call.
He wants you to try out.
It's a thrilling fantasy for so many people.
There's only so many scouts.
Now I'm from the UK, where Chelsea Football Club have one of the most prestigious and well-funded youth academy programs in the world.
But each Premier League team or each Premier League scout can only see about 2000 players per year.
But millions play the game.
And even from those 2000, it's already super limited by the players' geography, their cost factors, their access factors.
Lots of young people are now taking that into their own hands.
Social media clips of their best plays online.
But now we've just replaced a human with an algorithm that's not designed for talent ID to decide who gets seen.
What do we do about that?
It's simply not possible for Chelsea Football Club to see every talent in the world.
Or is it?
See, there's technologies now like computer vision, AI and deep learning that are helping us bridge that gap.
Now, personally, I'm not a scout or a coach.
My route was biomechanics, which is the science of motion.
And like many people in my profession.
We work in sports laboratories or with clubs with a remit of improving athlete performance or reducing injury risk.
And I was working in my lab that one day my now founder and CEO, Darren Perry's, walked in with his son Reef, who was struggling with an injury.
And we started analyzing him with all this wonderful equipment.
And we got talking about the scouting problem.
And he noted just how unfair and biased scouting could be.
He'd seen it himself.
Entire futures could be decided in one day by one person who has an opinion.
He noted how completely devoid of data youth scouting can be.
And he said to me what if we took all this lab protocols, all the data, all the equipment and put them into a set of standardized smartphone drills so any kid anywhere in the world could be tested fairly and equitably?
For me, it was a brilliant vision to a genuine problem.
So I joined his team, AI.io. where we build AI-based solutions across all of sport.
Now, the first thing we built was AI Scout for the problem at hand.
And here's how it works.
Simply speaking, a kid downloads an app for free and they record themselves doing these predefined drills directly from the phone.
We use computer vision AI in the cloud to analyze that.
We analyze 22 key body segments.
We can even do things like turn the 2D video into inferred 3D.
And bringing all that together, We get things like what direction did they run?
How did they turn?
How high did they jump?
The speed, the symmetry, the coordination, and so on.
But collecting that raw level data is just one part of the problem.
So you've got to be able to interpret that data.
And you've got to be able to score an athlete based on that data.
And most importantly, you've got to make it specific for those that are looking for talent.
Specificity is key here.
For example, in football soccer, for the Americans in the audience Each team or scout or coach kind of look for a different thing depending on what they need at that given moment in time.
So some teams or coaches might want athletes with power and pace.
Others might want coordination and technique and great body movement.
So we have to make our scout, we have to be tailored on that club by club basis.
Now to answer these questions and really to build out that product.
We partnered with two Premier League teams Burnley Football Club and Chelsea Football Club.
And we started by just asking directly if we could analyze any kid in the world from a smartphone and give you football-specific metrics.
What do you need to know to make that relevant to you and to use it?
And interestingly, they said, comparable, benchmarkable, reliable data.
And, above all else, both themselves and the kids they'd be analyzing would need to understand where the data's coming from.
There'd be no ambiguity.
So we started developing these predefined drills with them.
They were 10-meter sprints.
They were counter-movement jumps.
They were dribbling through cones, they were passing, they were shooting.
This is nothing new.
This was just genuine things that scouts would normally look at to make an assessment of a player.
And then we sat down with those scouts and looked for a ton of video.
Do you prefer player A or player B?
Now it turns out the art of scouting is actually incredibly complex.
So much of what an experienced scout does is really intuitive.
So whilst they knew they preferred player B, but they couldn't articulate it to us.
So we had to sit with them, ask the questions and turn their insights into something we could use to score.
We created the algorithm.
And with that algorithm, we just pumped thousands of videos through it.
So we started to make benchmarks and standards across age and gender.
Because of course, you can't analyze a 13-year-old to a 22-year-old.
The 22-year-old's almost always going to be bigger, stronger, faster.
You need to compare 13-year-olds to 13-year-olds.
So you can really see who stands out.
And talking of standing out, I'll never forget when we were first developing the app, we recruited 50 college kids from the UK.
They were honestly kind of average footballers, except this one guy, Ben.
They all did our drills and he was head and shoulders above the rest, like just really, really good.
Like he was 17.
How has this guy not been scouted before?
And the crazy thing is he lived just minutes down the road from Chelsea FC training grounds.
He wasn't in some remote village.
He was literally down the road from the world's best academies.
The system didn't see him, but we did.
I'll cut a long story short here, but this is where we knew what we were building was working, because he got a trial for Chelsea Football Club.
He scored in his under-18 debut, and he later signed for another Premier League club and even represented his country.
That wasn't enough.
We needed to test this in those remote places.
This should be for everyone.
Now theoretically, because all the heavy lifting is done in the cloud the analysis of the video, the processing, the scoring, the modeling.
It means that, whether you're from London or Mumbai, if you've got access to a smartphone, this thing can work.
And we're lucky to partner with Reliance Foundation in India.
They have an amazing program where they send scouts out every single year and find the best 11-year-old talent to give them five-year scholarships to play sport and have free education.
They're the best of those generally going on to play professional sport in India.
Same problem as Chelsea.
A few scouts, see a few thousand people, potentially millions are eligible.
So they turned to us to try and find some of those hard to reach kids in difficult locations.
And they put out a call on WhatsApp to their audience.
New to us to do this on WhatsApp, but the parents and students were asked to download an app and do the drills inside the phone.
And trial for Alliance Foundation directly from AR Scout.
Tens of thousands of kids now do this every single year.
And the best ones, based on their data, get sent to an in-person talent ID day, or the scouts make the decision who gets those scholarships.
It's a great example of how we're augmenting the scouting process, not just replacing their processes.
The best example of success, whilst we've had many, there is one player who actually downloaded the app from a shared community phone, had never played organized sport and got himself a five-year scholarship.
On top of that, only last year, the IOC and their partner at the time, Intel.
They reached out to us about the upcoming Youth Olympics, which is in Senegal.
They had a little concern that actually the Senegalese national teams didn't have enough talent to favor all the teams that are going to be in those youth Olympics.
Now here's the great thing about the technology.
You want to play football?
You can try for a football club through our app.
You can also do the opposite.
Based on your strengths, it can tell you what you might be good at.
So if you've got great acceleration, great reactive strength, it could be rugby sevens or futsal.
If you've got great upper body power and hand-eye coordination, it could be a baseball or a softball.
And that's exactly what we did.
We put the app into tablets.
We gave it to military leaders and school teachers and they just stood there and they recorded the kids in their class.
A few days into that, thousands of kids later 40 are now being trained ahead of those Youth Olympics in things like wrestling and athletics and football.
Now, where does this go from here?
So the app today runs through partnership programs, talent ID, initiatives with clubs, with federations.
Sometimes they're brand partners on the app where hundreds of thousands of kids trial and do these drills.
Hundreds of those had successful outcomes and are playing professional sport.
And where do we go from here?
Actually, we go global with this, right?
So it's now multi-language.
It's starting to roll out into the app.
Multi-cloud is coming.
So it can be cloud agnostic, which means we can put country-specific things into the app in any region that we need to work with.
And, right here in the US, proud to announce that we've did the MLS, or Major League Soccer as we know it here, and rolled this out to the MLS Next program.
45,000 kids right now are using the app three times per year.
Pre-season, mid-season, post-season.
So we can track and monitor changes over time.
And the great thing is, the scouts and the coaches get all the data in real time via our control center.
It's not a black box but something they can trust and learn from the data and see where that data is coming from.
And where else do we go?
It's probably not a big leap for you all to imagine how we start to move into things like at-home healthcare and medical.
But we're also starting to create the kind of movement libraries for American football, for basketball, for baseball, for cricket.
Because the underlying movement primitives the cut, the deceleration, the jump, the throw, the strike.
They translate well across quite a lot of sports.
Because one thing in sport is always true.
Talent is universal.
And brilliance exists in every corner of the globe.
Now, with technology and your very own smartphone, you can make that talent visible and level the playing field.
Thank you.
That was Richard Felton Thomas at TED Sports Indianapolis in 2025.
If you're curious about TED's curation, find out more at TED.com slash curation guidelines.
And that's it for today.
TED Talks Daily is part of the TED Audio Collective.
This episode was fact-checked by the TED Sports Research Team and produced and edited by our team Martha Estefanos, Oliver Friedman, Brian Green, Lucy Little and Tanzika Sangmarnivong.
This episode was mixed by Christopher Faisy-Bogan.
Additional support from Emma Taubner and Daniela Balarezo.
I'm Elise Hugh.
I'll be back tomorrow with a fresh idea for your feed.
Thanks for listening.
