AI scout: this startup is using automation to find football’s next star

Football clubs are always looking for ways to gain an edge over rivals. A UK startup claims to have found a way for AI to do just that and is already working with Premier League clubs to find football’s next big star.

Aiscout, an app developed by London-based startup AI.io, aims to harness the power of artificial intelligence for identifying talent in professional football.

Footballers looking to showcase their abilities to top clubs can record themselves performing various drills and share them on the Aiscout app.

The app uses AI to track and analyse the intricate movements of the player and the ball. That performance data is then added to the wider system, where it can be compared with benchmark data based on previously added information.

The startup, which has been funded by its founder and “large sports tech funds”, said it is generating revenue and plans to scale up in the first quarter of next year. It has already partnered with some big names in professional football, most notably Chelsea FC, Burnley FC and the entirety of the MLS league in the US. Data from those organisations has contributed to the benchmark data that new entrants are compared against.

Scouting teams using the app can then access the information of players who have participated in the trials.

While data has been used in professional football as a way to enhance the quality of player assessments for some time, AiScout claims that football teams are still limited by the finite number of players they can observe.

Richard Felton-Thomas, COO and director of sport science at AI.io, told UKTN that the scouting teams at top clubs generally see around 2,000 players a year – a drop in the ocean compared with the number of players out there.

“Being able to see that vast volume of players, it’s just impossible from purely human scouts,” Felton-Thomas said.

“There’s only so many weekends you can go and watch fixtures,” he added. “[Scouts] have to be really selective about who they watch. But they’re going to be selective about who they watch from subjectivity.”

Additionally, he said that using an AI-powered scouting method can massively speed up the timeline of bringing a young player into a team’s academy.

Felton-Thomas said these youth players “only play once a week, so if you want to watch them 10 times, you’re 10 weeks into this thing before you’ve even done much of validation”.

‘The human element’

The incorporation of AI is growing in both the sports and business world. And with that comes concern that people will be replaced in key decision-making in favour of artificial intelligence which is subject to its own drawbacks, such as programmed bias.

While the team behind AiScout are clearly backing the integration of automated technology into decision-making, they maintain that the “human element” will always be integral to them.

“You still need that human interaction. You still need that human input. What we should not be doing is deciding careers purely by AI,” Felton-Thomas explained.

“It’s a source of information, it gives us more confidence in the decisions we make as humans, not to say we’re going to determine people’s lives and careers, by a machine ultimately telling someone the decision to make.”

For Felton-Thomas, maintaining the human element is a key point to consider as governments develop their AI regulatory regimes.

Read more: Premier League: How data has become a football game changer