Human-Agent Team Game Analysis Framework: Case Studies
Event Type
Technical Groups
Human AI Robot Teaming (HART)
TimeThursday, October 7th10:48am - 11:06am EDT
LocationGrand Salon VII
DescriptionThis research details the development of a human-agent team (HAT) analysis framework specifically aimed at video games. The framework identifies different dimensions of interest related to humans and software agents working together. Video games have a variety of user-tested interaction paradigms that may offer useful insights into HAT dynamics, but it can be difficult for researchers to know which games are relevant to their research without a systematic method of characterizing HAT relationships. The framework was developed based on previous literature and gameplay analysis. This paper offers three case studies, applying the framework to the games Madden 21, Call to Arms, and Civilization V. Possible trends related to agent intelligence, team structures, and interdependence are discussed.