Abstract:

Fun is the force that drives play in games. Therefore a science of fun is needed to engineer games that are engaging for players, including the design of software agents that play against and along with human beings in computer games. EVE? is a computational theory, based on a psychological progression of Expectations (E), Violations (V) and Explanations (E?) that in turn evoke emotional responses known as fun or flow. Here EVE? is developed for the case of gambling games, specifically slot machines, deriving a Bayesian-information measure of aesthetic utility that is not modeled by the economic utility of previous proposals like Prospect Theory. The derivation shows how aesthetic utility can be measured by entropy and how fun can be seen as a form of learning. EVE's contribution lies in going beyond classical economics and computational intelligence to analyze the aesthetics of enjoyment and engagement - towards a science of fun.