Research Papers

Marlin Myrte, Sebastian Buzzalino, Thierry Aymerich

‍

This paper introduces a probabilistic framework for identifying basketball player archetypes that can be applied across leagues, including those without access to advanced data. While player segmentation is not new in basketball analytics, we propose a fundamentally different approach based on Archetypal Analysis, which identifies extreme and interpretable player profiles and models players as probabilistic mixtures of these archetypes, explicitly capturing hybrid roles and stylistic flexibility. We demonstrate the method on both European league data and an enriched NBA dataset, showing that the resulting archetypes are coherent and directly actionable for scouting, roster construction, and meaningful cross-league player comparison. Designed to work with widely available play-by-play and box-score statistics, the framework is accessible to teams across leagues and resource environments, and is accompanied by a detailed use case illustrating real-world applications for scouting, roster construction, and talent identification decisions.