An Analysis of Video Games Using the Dimensions of Human-Agent Interaction

Abstract

Designers of human-agent interaction techniques may benefit from an analysis of existing video games that include aspects of human-agent teaming. Many popular multi-player video games have been designed to integrate multiple human and computer agents in pursuit of a common objective and can serve as a testbed to explore novel interaction methods in human-agent teams. A guiding framework of human-agent interaction was created to bridge best practices between video game and real-world domains. The framework was used to analyze games on five main dimensions: 1) Levels of Automation, 2) Levels of Interaction, 3) Control Mode, 4) Teaming, and 5) Interaction Timing. Two video games, Final Fantasy XIV and Mass Effect, were assessed to identify human-agent interaction paradigms, and ramifications for real-world applications for human-agent teaming. This research draws on interaction design principles, human-agent interaction theory, and existing video games to offer human-agent team designers potential examples of successful interaction paradigms.