Call Us: US - +1 845 478 5244 | UK - +44 20 7193 7850 | AUS - +61 2 8005 4826

Meta-Analytic Synthesis

In this section we will cover the first set of tests we held for our agent. The purpose of this first set of tests was to have an idea of how much of an impact would emotions have when deceiving a human player. Would it make the player doubt the agent more or less? Since we are basing our work on modifying a previous attempt to tackle our problem explained in the related work we use the same setting with the COUP game. To measure said impact we will analyse the acceptance or doubts by the human players towards our agent’s actions, be they bluffs or truths. In order to achieve our goal for the tests we created three simple versions of our agent based on the work we mentioned in the related work as well, each of them with different characteristics regarding emotions. The first one would be called the True Emotions Agent, this agent makes use of FAtiMA’s appraisal process to be able to generate its own emotions and then express them unfiltered. The next one however, makes use of the same appraisal process but filters it to express emotions with a polarity opposite to the one felt, thus being called the Contrary Emotions Agent. Finally the last agent filters all emotions to be expressed and suppresses them becoming complete emotionless, mimicking the same behaviour as the base agent, earning itself the name No Emotions Agent.