Motor-difficulty evaluation framework of 2D platformer games’ levels

The continual growth of the video game industry has led to a significant increase in the scale of development of new projects, requiring considerably more resources to complete. In an attempt to counteract some of the resource requirement inflation, we propose a motor difficulty evaluation mathematical model as a development tool to avoid difficulty targeted playtesting. The proposed model deviated from other currently researched approach by evaluating difficulty in a deterministic manner and without player data. We define the notions of reachability and danger of a game level based on the properties and possible actions of the player’s avatar. Reachability and danger are used to describe two difficulty evaluation metrics: one based on the relative area coverage of reachable to danger, and one based on the perimeter of overlapping areas. The metrics are put to test in levels taken from the game Super Mario Bros. and the results of both metrics are compared to each other as well as with the Nishikado difficulty. We underline current limitations of the model and their effect on the results and provide elementary solutions. We conclude by proposing a possible timeline of future research to expend our model into a general commercial applicable product. La croissance continue de l’industrie du jeu vidéo a conduit à une augmentation significative de l’échelle de développement de nouveaux projets, nécessitant beaucoup plus de ressources pour être achevés. Dans une tentative de contrecarrer une partie de l’inflation des besoins en ressources, nous proposons un modèle mathématique d’évaluation de la difficulté motrice comme outil de développement afin d’éviter les tests de difficulté ciblés. Le modèle proposé se différencie des autres approches actuellement étudiées en évaluant la difficulté de manière déterministe et sans données sur les joueurs. Nous définissons les notions d’accessibilité et de dangerosité d’un niveau de jeu en fonction des propriétés et des actions possi