AI Is Changing Soccer Analysis — How We Watch The Game Could Be Next

A few years ago, the highly-respected head coach at one of Europe’s top soccer clubs had a problem he couldn’t fix with words alone.

The coach, then in charge of one of the finalists in this season’s men’s Champions League, was meticulous about how his defenders should scan for danger. He wanted their heads, shoulders and hips perfectly positioned to anticipate attacking threats.

One central defender wasn’t getting it. He was struggling to grasp his coach’s instructions.

The solution the coach turned to wasn’t extra drills. It was data.

In one-on-one sessions, he used AI-powered 3D technology to show the defender precisely how his posture was off. In real time, to the naked eye, it was barely noticeable. But the coach insisted fixing it could gain valuable milliseconds to cut out danger.

The player improved. Shortly after, he was a key member of the first-team squad.

The technology came from ReSpo.Vision, a “deep tech” startup based in Warsaw, Poland. The company believes AI tracking and visualisation can transform not only how teams and broadcasters analyze matches — but also how viewers understand and experience the sport.

A New Layer of Insight

ReSpo’s technology uses computer vision and deep learning to extract 3D tracking data from single-camera video feeds, like TV broadcasts. The system detects more than 50 body points per player with centimeter-level accuracy, without the need for wearable technology or expensive camera setups.

Paweł Ośtęrreicher, a self-described “sports nerd,” founded ReSpo in 2021. A former management consultant, he had previously used computer vision to optimize industrial assembly lines.

“The first philosophical claim of a company like ours is that it is absurd to think that human perception is the deepest level worth going down to of football and sports analytics,” Ośtęrreicher tells me in an interview.

“There must be something deeper than this artificially set human perception level, which still explains the game, which still adds to the narratives, which still helps optimize your game or better prepare for opponents.

“The deeper we get, the more we unlock.”

An Underdog Cup Win, Driven by Data

Soccer data has typically focused on actions humans can record, like number of shots, passes or interceptions. Computer vision allows more complex data, like the speed of a shot, or a team’s pressing intensity, to be uncovered.

With soccer clubs and leagues increasingly drawn to data and AI analysis to gain an edge, ReSpo has provided its technology to clubs including UEFA Champions League winners and national federations including the Polish and Croatian FAs. A deal for the CONMEBOL Copa America 2024 saw the 16 competing teams access enhanced data from ReSpo.

A standout case is Wisła Kraków, a Polish second-division club that used optical tracking and analytics to optimize tactics and for match preparation. Using the data insights, Wisła secured an against-the-odds Polish Cup win in 2024.

The club used data to give detailed feedback to players, for example comparing a player’s pressing intensity to their personal average, the averages of teammates and even players from rival clubs.

Ośtęrreicher says coaches sought to “squeeze 10% or 20% more” from players.

“They were seeking better benchmarks and trying to make the players believe they could do it. (Coaches would say) ‘If this guy can do it, are you worse than him? Are you less physically able than this guy?’.

“It sounds like a very trivial thing, but it was using the data as an excuse to push your players.”

The tracking also enabled Wisła’s coaching team to have “very detailed opponent preparation”. They deeply analyzed the opponent’s three previous matches, focusing on the team’s response to scoring or conceding a goal.

“And then building a sort of a decision tree plan for a game. ‘If this happens, they will most likely apply this type of tactic, which we should counter,’” Ośtęrreicher says.

Legendary Goals Through The Player’s Eyes

ReSpo, which has 25 employees, also has ambitions to change how fans experience a match. Using its “Digital Twin” technology, 3D recreations of matches can be built, allowing fans to relieve moments — like Marco van Basten’s legendary volley for The Netherlands in 1988 — through the eyes of the player.

By the end of 2025, the company plans to have a post-match product where fans can explore a match a few hours after the final whistle. A live version, where fans can “step on to the pitch” in real time during a match, and broadcasters can integrate into production, is in development for 2026.

While this side of the business currently accounts for 5% to 10% of revenue, Ośtęrreicher says the commercial possibilities will ultimately see it bring in more than 50%.

In May, ReSpo raised €4.2 million ($4.79m) in a funding round. Company investors include Wayve co-founder Amar Shah and Polish soccer international Jan Bednarek.

Ośtęrreicher says “the way we watch football hasn’t fundamentally changed since the 1960s”, and new insights offer broadcasters — and fans — deeper levels of tactical insight.

But soccer is notoriously resistant to technological advances. Are fans ready for deeper data?

“I’m what can be called a reasonable optimist when it comes to that,” Ośtęrreicher says.

“I remember how much stick xG (expected goals) got when it was first introduced. I remember ex-players (pundits) slamming it on air. But now you have it in almost every summary. People casually discuss it in the studio.

“I will not assume that I’ll just throw a bunch of statistics and visualizations and everyone will love it. They will need a few years, or hopefully less than few years, to get used to it.

“But eventually, if it brings informative value, better understanding of the game, better excitement factor — then it’s going to happen.”