Agent-based modelling is a method that involves creating an experimental model that is parameterized according to the needs of one or more designated issues (Thom, 1979). ABM has now become a well-known and well-developed method for studying and documenting ancient periods in archaeology, such as the Mesolithic or Prehistory (see Lake, 2014 for the history of modeling in archaeology). ABM allows the testing of previously formulated hypotheses but also enables the proposal of new ones in an inductive approach; it is particularly used to better understand the mechanisms underlying a phenomenon. For example, ABM can simulate the environmental context (climate, fauna, vegetation) of a region during the Upper Paleolithic and integrate human agents to better understand the game exploitation strategies of prehistoric human groups (see Mithen, 1997). However, despite its longevity and the numerous studies it has generated, multi-agent modeling is still criticized today for reasons related to its accessibility as well as the very principle of modeling, particularly the modeling of human behavior, its irrationality, and the economic and/or environmental determinism that emerges from it (Doran, 2000; Bailey, 2007). To address these challenges, we propose approaching modeling from the perspective of video games, which would allow the integration of a "real" human factor into multi-agent models through a multi-player simulation where players control human avatars typically modeled by a computer. Video games could then be used in the same way as ABM but with the added dimension of real-time human decision-making irrationality. Such a project would require a sufficient number of players and thus necessarily involves concepts of citizen science, which are not very common in archaeology. However, citizen science is already widely practiced in the natural sciences, sometimes even through video games (see MMOS; GAWP). Archaeology, which is already accustomed to working with the public during excavation stages and mediation operations, should therefore be able to appropriate these new tools and the methods they entail.