“I Keep Sweet Cats In Real Life, But What I Need In The Virtual World Is A Neurotic Dragon": Virtual Pet Designs With Personality Patterns

Anonymous

Keywords: virtual pet, machine learning, personality perception, game characters design

TL;DR: Our paper explores the design of virtual pets and their personalities, utilizing the Five Factor Model and machine learning techniques to create voxel-style pets that are preferred by users with similar personalities.

Abstract: Virtual Pets serve as companionships and meaningful in-game narratives in the metaverse. Players have unique personalities and personality preferences for their pets. However, the design of virtual pets often relies on designers' individual experiences without considering the virtual pets' personalities. We designed the virtual pets' visual appearances by following the design guidelines from the Five Factor Model (FFM) in voxel format. We conducted a study to investigate people’s perceptions of virtual pets’ personalities and appearances through two user studies. Our findings suggest that voxel-style virtual pets better-represented agreeableness than realistic pet pictures. Additionally, users prefer virtual pets (voxel-style pets generated with machine-learning techniques) that share similar personalities. The study's results provide valuable insights for game designers and researchers for future pet game design and understanding of how people perceive virtual pets based on their appearance and behavior.

Track: HCI/visualization

Revision: No