World Cup Teams Compete for Leadership in AI

This year’s tournament teams will also leverage a tailor-made AI agent powered by Lenovo, a FIFA initiative to level the competitive playing field. Whether this will be sufficient remains to be seen.

“The data is fine-grained, multi-agent, and adversarial. What we do in sports mirrors the complexities of autonomous vehicles—you’re looking at trajectories,” adds Lucey. “For a single team, the player order alone creates 10 factorial permutations. Adding the opposition multiplies this complexity exponentially.”

Even smaller nations are finding creative methods to harness technology. Curaçao, a Dutch Caribbean island with a population of about 159,000, became the smallest nation ever to qualify for a World Cup by utilizing their own data and technology for “diaspora tracking”: mapping parentage, identifying eligible players, and employing geospatial data for scouting and trial organization.

“Only one of the 26 Curaçao players was actually born on the island,” reports Alex Stewart, chief executive of the data-driven sports consultancy Analytics FC. “The majority were born in the Netherlands.”

Another growing application of data and AI within national federations is in manager selection. Analytical tools can evaluate realistic squad options and identify managers whose tactical abilities best align with them. Teams can also use AI to shape squad compositions in advance of a tournament, informed by group-stage opponents.

England is employing AI for penalty analysis, knowing that a shoot-out could determine their fate. What once required five days to analyze every penalty taker for an opponent can now likely be completed in just five hours, as shared by the Football Association’s head of performance insights and analysis with the BBC.

Marcelo Bielsa, formerly the manager of Leeds United in the Premier League, once stated that his staff dedicated around 300 hours to analyzing an upcoming opponent. “We can automate this now,” Lucey points out. He presents a video showcasing red and blue dots moving on a pitch in pursuit of a yellow ball. Analysts can query how often specific plays have resulted in shots or goals, unveiling layers of new information each time.

“This moment can be compared to early web access,” states Jan Wendt, cofounder and CEO of PLAIER, an AI platform collaborating with clubs and national teams. British Airways and Amazon both developed websites in the internet’s infancy; one became an information and ticketing platform, while the other revolutionized commerce globally, Wendt observes. Similarly, AI is transforming both routine tasks and entire industries, including sports franchises.

However, acquiring AI tools and the skilled personnel to manage them can be costly. Not all nations possess the needed resources. Wendt advocates that smaller countries should consider partnering with established external firms, such as his, as a more efficient alternative.

Another issue is that increased data can complicate an analyst’s role. Their task involves distilling extensive information into concise insights useful for coaches or players.

“You wouldn’t want to flood a coach with a 47-page dossier on the opposition’s fullback,” Stewart remarks. “While there’s more information available, the analyst’s job becomes tougher as they must sift through it to extract essential insights.”

Technology can enhance match analysis and preparation for teams that previously struggled to compete with more affluent nations employing large scouting and analytics teams. But does this mean smaller nations might now find it challenging to compete against large teams brimming with computer scientists and analysts?

Will the data disparity between wealthier and poorer nations grow so pronounced that it disrupts competition in a tournament where smaller countries already face significant odds?

FIFA is sufficiently concerned that it has developed a customized AI agent, Football AI Pro, making it available to all nations participating in this tournament for the first time.

The agent offers a ChatGPT-like interface where coaches can input questions to gather intelligence on their next rivals. Matches are recreated in 3D, facilitating analysis from previously impossible perspectives. Every detail is quantifiable, from player movement to defensive actions and scoring chances.

“We aim to provide technology to all teams to ensure everyone can access and utilize it simply without needing additional experts, especially since not everyone can afford them,” explains Johannes Holzmüller, FIFA’s director of innovation.

Whether this will bridge the gap between a country with minimal data resources and the England national team, which employs in-house software developers, data scientists, and analysts backed by external AI solutions, remains up for debate. “That’s the least we can do,” Holzmüller adds. “We recognize that some teams leverage technology and data more effectively than others.”

The future of data, AI, and soccer is about predicting outcomes. “Long-term forecasting is the next frontier,” Lucey states, believing they will eventually be capable of counterfactual analyses recommending which players to rest for optimal success.

Will FIFA need to regulate nations to only use FIFA-approved AI tools?

“That’s a major question,” Holzmüller acknowledges. “Regulation is not for today, but [AI] will undoubtedly play a significant role in the future.”