Evolution of Performance Analysis in Football: From Intuitive Observation to Big Data and Artificial Intelligence

This article presents a narrative review with a historical perspective on the evolution of performance analysis in football, from intuitive observation to the incorporation of Big Data and artificial intelligence. Based on academic literature in Spanish and English, a five-stage periodization is proposed: (1) 1863-1924, dominance of observation and intuition within a context of progressive normative institutionalization; (2) 1925-1959, consolidation of tactical analysis and professionalization, with the coach assuming an increasingly strategic role; (3) 1960-1979, emergence of audiovisual recording and a shift towards more systematic observation, alongside issues of incomplete information and observer bias; (4) 1980-2010, consolidation of video analysis software, performance indicator recording, and physical monitoring through GPS; and (5) 2011-present, expansion of large-scale analysis through the integration of event and positional data, and the application of machine learning for modelling and decision support. Despite technological and methodological progress, football retains structural limits to stable prediction due to its dynamic, non-linear, contextual, and strategic nature, in which teams continuously anticipate and adapt to one another. Consequently, advanced tools enhance the descriptive and diagnostic capacity of performance analysis but do not replace the expert judgement of coaches and analysts, which remains essential for interpreting evidence and translating it into actionable decisions for training and competition.