Ortiz Sánchez, Ivan Dmitriy2021-07-202021-07-202021-07http://hdl.handle.net/10230/48239Tutors: Sergi Valverde Castillo, Salvador Duran NebredaThe mechanisms of biological evolution also apply to artificial phenomena such as culture and technology, and the evolution of video games through history has been shaped by the evolution of technology itself. In particular, the so-called speedruns, which consist in completing video games in the least time possible, have become remarkably popular recently. Since the evolution of performance in video games has never been quantitatively assessed, in the present study, we wonder whether there are universal patterns in the way speedrunning has evolved through video game history. Specifically, we aim to identify relations between performance improvement and the size and structure of the player community. Thus, a reliable dataset with the results of official speedruns has been manipulated and analyzed. First, we describe the dynamics of performance improvement and growth of the community since its origin. Second, we explore the effects of community structure with a game-player bipartite network framework and an infectious model of strategy and information propagation. Finally, we relate the model to the actual data and establish linkages between the properties of the network and the learning dynamics. Our results show how the growth of the community and the evolution of performance follow exponential descriptions and how the rank-ordered distribution of players according to their number of playthroughs follows a power law-like behaviour. A first minimal network model to describe the properties of the community is also provided. This study lays the foundation for a quantitative application of biological and evolutionary models to the video game field.application/pdfeng© Tots els drets reservatsinfo:eu-repo/semantics/bachelorThesisComputational modellingComplex networksInfection modelsCultural evolutionLearningVideo gamesSpeedrunninginfo:eu-repo/semantics/openAccess