Challenge balancing for personalised game spaces

In this paper we propose an approach for personalising the space in which a game is played (i.e., levels) - to the end of tailoring the experienced challenge to the individual user during actual play of the game. Our approach specifically considers two design challenges, namely implicit user feedback and high risk of user abandonment. We contribute an approach that acknowledges that for effective online game personalisation, one needs to (1) offline learn a policy that is appropriate in expectation across users - to be used for initialising the online game, (2) offline learn a mapping from gameplay observations to the player experience - to be used for guiding the online game personalisation, and (3) rapidly converge to an appropriate policy for the individual user in online gameplay - employing the learned feedback model and a straightforward model of user abandonment. User studies that validate the approach to online game personalisation in the actual video game INFINITE MARIO BROS, indicate that it provides an effective basis for automatically balancing the game´s challenge level to the individual human player.