Note: In 2013, Health Games Research completed its work. This web site is now an archive and will not be updated. Please visit the web site of the Center for Digital Games Research at UC Santa Barbara to find current information about health games and the broader field of digital games, and to use the Health Games Research online searchable database.

The Proteus Effect

By Nick Yee, Ph.D.

Research Scientist, Palo Alto Research Center, Palo Alto, CA

October 2009

The creation or selection of an avatar that represents you is typically the first step in a computer game. Are you the sword-swinging warrior or the stealthy assassin? And it’s easy to think of avatar creation as a one-way process—we choose who we want to be in a virtual environment, but I’ve always been fascinated by the flip-side of the question. Do our avatars change how we behave in turn?

Studies in traditional brick-and-mortar psychology laboratories hint at why this might happen. For example, people asked to wear black uniforms behave more aggressively than people asked to wear white uniforms (Frank & Gilovich, 1988). As the researchers explained, “Just as observers see those in black uniforms as tough, mean, and aggressive, so too does the person wearing that uniform” (pg. 83). In another study (Johnson & Downing, 1979), participants in Ku Klux Klan uniforms delivered longer electric shocks to a “learner” (play-acted by a lab assistant) than participants in nurse uniforms. Both of these studies show that we subconsciously infer how we should behave by observing our own appearance.

In a series of studies, my colleagues and I tried to explore whether this same effect occurred with avatars in virtual environments (Yee & Bailenson, 2007). In the first study, I gave participants either an attractive or unattractive avatar and asked them to interact with a virtual stranger (who was controlled by a lab assistant). We found that participants who were given attractive avatars walked closer to and disclosed more personal information to the virtual stranger than participants given unattractive avatars. We found the same effect with avatar height. Participants in taller avatars (relative to the virtual stranger) negotiated more aggressively in a bargaining task than participants in shorter avatars. After the Greek God Proteus who was able to take on many different self-representations, we termed this the Proteus Effect—how avatars can change our behavior in virtual environments.

Of course, these findings beg the larger question of whether these effects linger when participants leave that avatar and the virtual environment. To examine this, we replicated the attractiveness study followed by an additional task in a different setting (Yee, Bailenson, & Ducheneaut, 2009). After the virtual interaction task, we told participants that they would participate in a different study about online dating. They were asked to create a profile on a mock college-dating site and then were shown 9 photographs of people of the gender they were interested in. They were then told to select the 2 people they would most like to contact and get to know better. We found that participants who had been given an attractive avatar in a virtual environment chose more attractive partners in the dating task than participants given unattractive avatars in the earlier task. This study showed that effects on people's perceptions of their own attractiveness do seem to linger outside of the original virtual environment.

There are several ways in which health games can leverage the Proteus Effect. Our earlier findings suggest that providing users with fit, athletic avatars in exergames may encourage longer and more engaged exercise sessions than if they were provided with normal-looking avatars or avatars that were modeled from their own bodies. On the other hand, some studies have found that users who are allowed to select their avatars become more engaged in the gaming experience (Lim & Reeves, 2006). One compromise that might achieve both goals is to allow users to select from a set of avatars that are equally fit and athletic. Of course, the Proteus Effect is only one of many mechanisms for using avatars to change behavior, and it is important to carry out more studies comparing and contrasting these effects as well as discovering what happens when they are combined.


Frank, M., & Gilovich, T. (1988). The dark side of self and social perception: Black uniforms and aggression in professional sports. Journal of Personality and Social Psychology, 54, 74-85.

Johnson, R., & Downing, L. (1979). Deindividuation and valence of cues: Effects on prosocial and antisocial behavior. Journal of Personality and Social Psychology, 37, 1532-1538.

Lim, S., & Reeves, B. (2006). Being in the Game: Effects of Avatar Choice and Point of View on Arousal Responses During Play. Paper presented at the International Communication Association, Dresden, Germany.

Yee, N., & Bailenson, J. (2007). The proteus effect: The effect of transformed self-representation on behavior. Human Communication Research, 33, 271-290.

Yee, N., Bailenson, J., & Ducheneaut, N. (2009). The proteus effect: Implications of transformed digital self-sepresentation on online and offline behavior. Communication Research, 36, 285-312.