Mazing

The Mazing project investigates how an agent's behaviour and the player's own performance and emotions shape the recognition of a frustrated behaviour exhibited by the agent. The Mazing testbed game sets a player to compete against an agent which externalizes frustration based on theory. The project collected gameplay data from users, along with an annotated ground truth about the player's appraisal of the agent's frustration, and applied face recognition to estimate the player's emotional state. Analysis of the data collected showed that the player's observable emotions are not correlated highly with the perceived frustration of the agent. This suggests that the player's theory of mind is a cognitive process based on the gameplay context. Later studies used the same game to explore the agent's believability based on the time-continuous believability annotations of a player and comparing those to gameplay metrics and to a binary classification between agents.