Sport relies on artificial intelligence: this is how Sevilla will use it to sign | Technology

Sevilla striker Isaac Romero celebrates his goal during the match against Girona last Sunday.David Borrat (EFE)

Sevilla’s latest signing is not a common sight on the pitch: he doesn’t run, defend or finish and his physique comes from another realm. And yet, it is gaining ground and positioning itself as a key future player. This is artificial intelligence (AI), whose application in sport, according to the research firm Mordor Intelligencewill grow by 28.72% until 2026. Teams like the English Liverpool, the Spanish Valencia, the German Schalke 04, Club América from Mexico or many teams from the main leagues of the United States are already they use it for dedicationsanalyze athlete performance, develop training models, treat injuries and identify effective playing patterns.

To enter this league of technological teams, Sevilla has joined forces with IBM, one of the multinationals that, along with Microsoft and Google, are leading the race for artificial intelligence. The club and the technology company have developed a tool, based on a sports database of 200,000 reports, which allows players to be selected based on their characteristics through interaction with the machine through natural language. “There are other clubs that have more resources and that forces us to do things differently,” explains Elías Zamora, head of Sevilla’s data department.

The tool uses generative artificial intelligence that, unlike conventional systems that search restricted sources and return limited results, can create never-before-seen content from complex databases with pre-training and machine learning formula (machine learning). The result is more precise and can be profiled or refined through direct dialogue with the machine.

During the presentation of the tool, called Scout Advisor and designed to help select players based on reports from Sevilla recruiters, the artificial intelligence demonstrated that it can go through the club’s 200,000 reports in seconds and produce a list of footballers according to the required requirements. characteristics, with the identification of the team in which they play, their price, the expiration of the current contract, the percentage of similarity with the desired ideal and an average score based on all the parameters used.

Fernando Suárez, director of IBM programs, explains that signing is one of the most complex actions of a club because it combines a high investment with a high degree of uncertainty. “So far, this has been based on observing and analyzing a limited amount of data. “AI fills this gap.”

The result is cold. “Leadership or charisma are not detected by the machine,” admits Suárez. The final decision is therefore human. “It’s a source of information. Artificial intelligence would not have told us: “sign (Kylian) Mbappé when he was 13,” jokes Zamora, referring to the sought-after PSG striker.

But Emilio de Dios, head of the Seville scouts, believes that it is a tool that has completely changed management: “We went from sailing to electricity. Reduced to seconds what previously required 500 hours of research. “This makes our final selection work easier and will allow us to choose better players.” The first effective test will come during next season’s transfer campaign.


But the application of artificial intelligence is not without risks, also in sport. Alberto Carrio Sampredro, professor at Pompeu Fabra University and author of a legal and ethical framework to be used in this area, warns in your inquiry: “Athlete manipulation and competition are probably the most pressing. But there are others, such as the radical modification of the governance of sport and sports competitions, which will be entirely mediated by this disruptive technology. »

Javier Pérez Triviño, also from Peompeu Fabra, details in other work the main threats of AI for athletes: the loss of autonomy over their performances, the inequality of those who do not have access to technology, the loss of the human element in sports practice, the disruption of the comparison of merits and the loss of enthusiasm of the competition. “AI can be accepted in sport, but, in any case, measures must be established that distinguish the types of improvements and that guarantee respect for the primacy of natural talents in sporting success and equality” , he declares.

Fernando Suárez, technology program manager at IBM, is cautious: “Not just any type of AI is valid. » In this sense, the Seville program leverages the company’s WatsonX platform to ensure that the system is “open, accurate, scalable and governable.”

Despite the caution, AI in sports, as in other fields, is growing vigorously. Liverpool was one of the first to use the services of DeepMind (Google) to help the coach make decisions.

Global, La Liga Technology And Microsoft They came together to develop projects related to football, basketball, rugby and tennis. The technological Sparta Science uses AI to monitor athletes’ physical hotspots through a model that collects up to 3,000 pieces of data about athletes’ bodies. The Cleveland Cavaliers (NBA) basketball team and the Colorado Rockies baseball team use it and claim to have reduced their players’ injuries by almost 40%.

The programs of Second spectrum They make it possible to evaluate individual and collective performances, as well as to advise coaches on strategic decisions. The NBA has adopted systems to improve spectator information with automatic graphics.

In Formula 1, the association with the latest technological advances comes from the origins of competition and corporate artificial intelligence like AWS (Amazon), Dell and Oracle have become widespread for designing career strategies.

You can follow EL PAÍS Technology In Facebook And X or sign up here to receive our weekly newsletter.

Subscribe to continue reading

Read without limits