Marco Beato, Gibson Moreira Praça, Antonio Dello Iacono, José Afonso, Antonio R. Gualtieri, Andrew Hulton, Naomi Datson, Peter Krustrup, Francisco Tomás González Fernández, Grégory Hallé Petiot, João Cláudio Machado, Piotr Żmijewski, Rodrigo Aquino, Michael King, Nuno André Nunes, Sigrid Olthof, Magni Mohr, Israel Teoldo da Costa, Filipe Manuel Clemente
BACKGROUND: Sided games are very popular in football (soccer) training, designed to integrate technical, tactical, and physical development through the manipulation of constraints such as player numbers, pitch size, and rules, among other variables. However, inconsistent terminology, methodological heterogeneity, and limited empirical guidance have hindered its methodological advancement across research and practice. OBJECTIVE: This consensus aimed to establish a unified framework and evidence-based recommendations for the design, monitoring, and application of sided games in football, providing conceptual clarity and practical guidance for coaches and sport scientists. METHODS: A multidisciplinary group of 20 international experts in football science and coaching developed the framework through a 3-stage process: (1) theoretical model development, (2) synthesis of priority areas based on an umbrella review of 12 systematic reviews and meta-analyses, and (3) iterative drafting and blinded consensus voting on statements using a 9-point Likert scale. RESULTS: Seven statements achieved consensus (scores 7-9), defining sided games in soccer, delineating important influencing factors (coach philosophy, environmental, task, and individual constraints), defining monitoring principles, and highlighting benefits and limitations. The consensus confirmed sided games as effective for developing aerobic fitness and tactical-technical behaviors but less effective for strength, power, and high-speed performance adaptations. CONCLUSIONS: This consensus provides the first framework for sided game training in football, unifying terminology, and guiding evidence-based design. Its implementation can enhance training coherence, optimize player development, and inform future research addressing existing methodological and population-specific gaps.