In his paper The Role of Emotion in Believable Agents (1994)Joseph Bates describes a ‘believable agent’ as ‘one that provides the illusion of life, and thus permits the audience’s suspension of disbelief.’ [pp.1]

Bates contrasts the work of artificial intelligence designers on believable agents with that of character animators, arguing that ‘it is the artists who have come closest to understanding and perhaps capturing the essence of humanity that Bledsoe (Woody Blesdoe, 1986)? and other AI researchers? ultimately seek.’ [pp.2]

Bates believes that it is portrayal of emotion that allows an audience to engage with a virtual agent, and takes his research from the work of classic Disney animators, some of the first people to attempt to generate an emotional response from an audience with entirely human-created characters.

He quotes animators Frank Thomas and Ollie Johnson from their book The Illusion of Life (1981): ‘From the earliest days? it has been the portrayal of emotions that has given the Disney characters the illusion of life??.’ [pp.2]

I intend to investigate emotional engagement with believable agents through virtual pet design. By designing a virtual pet I aim to investigate what causes an audience to emotionally engage with a digital character and what practices designers can use to develop engaging digital characters.

Developing a virtual pet has a number of advantages. First, it allows me to work with a smaller palette of emotional responses than a ‘human-like’ character, with a virtual pet only having to respond in a believably animal way rather than with a full spectrum of human emotions. Secondly, it allows me to work at a reduced visual fidelity as the character does not have to appear believably human-like. Thirdly, it provides a specific narrow field for competitor research, allowing me to focus on past and existing virtual pet products.

The project will take into account character animation practice and social psychology theory surrounding player engagement with games, as well as researching existing virtual pet designs to find out where they succeed and where they fail in player engagement.