A Mobile Game Controller Adapted to the Gameplay and User's Behavior Using Machine Learning

When playing games, the user expects an easy and intuitive interaction. While current controllers are physical hardware components with a default configuration of buttons, different games use different buttons and demand different interaction methods. Besides, the player style varies according to personal characteristics or past gaming experiences. In previous works we proposed a novel virtual controller based on a common touchscreen device, such as smartphone or tablet, that is used as a gamepad to control a game on a computer or game console. In this work we include machine-learning techniques for an intelligent adaption of the layout and control elements distribution, minimizing errors and providing an enjoyable experience for individual users. We also present different usability tests and show considerable improvements in the precision and game performance of the user. We expect to open a new way of designing console and desktop games, allowing game designers to project individual controllers for each game.