GenAI Conclave in Kochi: How AI is helping football clubs in scouting, recruiting players

The fiercely competitive world of professional football is ever-evolving, with clubs constantly seeking new methods to stay ahead. Recently, Artificial Intelligence (AI) has revolutionised player scouting and recruitment. By analysing extensive datasets and video footage, AI systems can pinpoint talented players with greater speed and accuracy than traditional human scouts. This cutting-edge technology is proving to be a game-changer, allowing clubs to make more informed decisions and secure top talent.

****

Gonzalo Montiel, Argentina’s 2022 World Cup hero known for his decisive spot-kick against France in Qatar, transferred to Sevilla from River Plate in 2021. Long before Lionel Scaloni had his eye on him, Sevilla’s scouting team recognised him as a River Plate player with immense potential. Their expert scouts compiled a thorough report showcasing his impressive statistics:

"Montiel averages 43.84 passes every 90 minutes, reflecting his active role in the team's build-up play. He delivers 2.26 crosses per 90 minutes, with 35.44% accuracy. Additionally, he makes 1.51 progressive runs per 90 minutes, often receiving the ball in open space, advancing with a dribble, and ultimately delivering a cross into the box," read an excerpt from a scouting report scribbled in Spanish on Montiel.

How AI helps football clubs

For Sevilla FC to focus on Montiel, it required an avalanche of paperwork. Their elite scouting team, consisting of 20 to 25 scouts, generated as many as 40 reports on a single player, necessitating 200 to 300 hours of meticulous review. In 2021, the club created a dedicated data department specifically to help management make better business decisions. It has now grown to be the largest data department in European football, developing its own AI tool to help track player movements through news coverage, as well as internal ticketing solutions.

Speaking on the sidelines of the two-day International Conclave on GenAI in Kochi, Kerala, Dr. Elias Zamora, Chief Technology Officer (CTO) of Sevilla FC, elaborated on his team’s use of generative AI to build a team with a focused vision aligned with the club's goals. He explained how they accumulated thousands of aspects from scouting reports prepared by their scouting department to create a comprehensive and strategic approach to player recruitment.

"In the past, scouting primarily relied on quantitative and categorical data. This included metrics such as the number of minutes played, goals scored, and passes made. Players were assessed based on how they performed in these categories compared to their peers and competition," explained Dr. Zamora.

"We have one of the largest scouting databases in the world, filled with expert opinions from scouts and football analysts who write about player performances in specific matches. This has resulted in nearly a million reports, each containing two to four paragraphs of detailed analysis. It's impractical for a human to read and extract all the information from these reports. Therefore, from a technology and data perspective, we aim to extract information not only from quantitative and categorical data but also to glean valuable insights from the rich, descriptive content of these reports," added Dr. Zamora.

AI vs manual scouting

The stage for AI’s intervention in scouting and player database management was set back in the late 2000s. Manual scouting is fraught with shortcomings. Relying on subjective opinions and incomplete data, this traditional approach is not only time-consuming and costly but also susceptible to human error and bias. Scouts often travel extensively to evaluate players firsthand, yet they face challenges in making comprehensive comparisons between prospects. The inefficiencies in the transfer market have led to inflated fees for unproven talents. AI scouting offers a solution, addressing these limitations by providing more accurate and data-driven evaluations.

According to Dr. Zamora, the use of generative AI is not a random command to create a team from scratch, but a specific need-based search.

"Is it a requirement-based search, like 'we need a midfielder', so we are looking for a midfielder, or are you just randomly finding talents and including them? You need to provide input, such as, 'I am looking for an offensive midfielder with distribution and organisational capacity'. You introduce what you are looking for, and then you receive a number of players and reports detailing the characteristics you are seeking. So, it is not just a random or fully automated process where you click and see what appears. You have to identify what you are looking for," explains Dr. Zamora.

Concerns over AI manipulating match results

However, Dr. Zamora avoided directly addressing concerns about using GenAI for predicting or manipulating match results.

"Our goal with GenAI isn’t to predict match outcomes but to evaluate and identify the best talents based on the characteristics we are seeking through our scouting reports. Sometimes people try to use technology for everything, but it’s more effective to focus on the specific questions for which the data was intended," he said. However, he acknowledged that technology and methods for simulating match scenarios already exist, especially in betting.

Although there are concerns that AI might reduce players to just data, its ethical use can benefit clubs, players, and fans by bringing talent to the forefront, and more and more professional clubs have started engaging artificial intelligence to build their next-generation teams.