How Manchester City use 650,000-player database to overhaul squad

Manchester City’s sophisticated data department operates under a single guiding question: What does it take to win? The club’s experts break this query down into three stages—identifying talent, evaluating ability, and assessing developmental potential—all aimed at boosting the statistical likelihood of Premier League triumph. While their models correctly identified Arsenal as favorites last season, the current campaign’s five wins from five opening games has reignited speculation that City will once again challenge for the title.

BBC Sport has been granted unprecedented access to these inner workings, revealing how the club leverages a database of 650,000 global footballers to navigate the modern transfer market. A core objective is mitigating risk in both purchases and sales by grading candidates on ability, overall characteristics, and how their current team’s system aligns with City’s requirements. The focus extends beyond current proficiency to project how a player might evolve within the Etihad Stadium environment.

This data-driven overhaul has been extensive since the 2023 Treble-winning campaign, with only four survivors remaining from that historic squad. However, sporting director Hugo Viana emphasized that data is merely one component of a broader strategy. “Data plays a role in our decision-making, but it is just one part of the overall picture,” Viana stated. “It cannot fully show a person’s character, mentality or behaviour.” Consequently, traditional scouting remains the primary filter, with data insights supporting rather than replacing human expertise.

The recruitment process begins with a detailed brief. For instance, full-backs are categorized into defensive, inverted, or wide playmakers. Filters are then applied to the massive database, narrowing the field to 20–50 candidates who meet specific positional and ability criteria. These names are cross-referenced with scout reports, which typically show significant overlap with data-generated lists. Further metrics are applied to prioritize the group, leading to approximately six players undergoing deep diligence on injury history, athleticism, and behavior.

The resulting “acquisition packs” include radar charts detailing on-field metrics, utilization rates over three years, and impact scores predicting both immediate and future contributions. An “action value score” also positions the prospect relative to peers in similar roles across the league. When Chelsea’s interest in Juventus’ Andrea Cambiaso faded due to tactical fit, City continued monitoring him, illustrating the method’s longevity.

The signing of £116m midfielder Elliot Anderson from Nottingham Forest exemplifies the system’s forensic depth. Despite Anderson being an obvious target and an English international, City ran him through their full algorithmic process. Initially, his metrics did not stand out as exceptional; his goal and assist numbers were modest because Forest operates as a transitional side rather than a possession-based team like City.

Data analysts adjusted for this context, modeling how Anderson’s performance would translate in a system where he would operate higher up the pitch with greater ball retention. His defensive error rates, for example, were re-evaluated considering the riskier passing options available to him at Forest compared to the shorter, safer outlets City provide. As his form improved throughout the season, scouts confirmed he matched the profiles of existing City midfielders, leading to a belief that the increased fee still represented low risk.

“Elliot Anderson is a great example of how the process works,” Viana noted. “The data was positive… but it was the insight from our scouts and wider performance team that helped us understand the player… and proved our belief that he was the right fit for the club.” This blend of quantitative analysis and qualitative scouting underscores City’s commitment to closing the gap on rivals like Arsenal through refined recruitment.