How Do Player-Avatar Relationships Mediate Physiological Responses To In-Game Events?
When looking at players' physiological responses to watching an avatar being beaten up in Wii BoxingTM, Ratan & Dawson (2016) found that they responded more when they felt emotionally connected to the avatar. So this physiological response to in-game events is affected in some way by our relationship to the avatar we are playing as.
Downs, Bowman & Banks (2017) describe what they call a "Polythetic Model of Player Identification", in simple terms this means several different aspects of player-avatar relationships interacting to produce an overall experience of connectedness between player and avatar. They split player-avatar identification into six categories:
- Physical Similarity: How much does the avatar physically resemble you?
- Value Homophily: How close are the avatar's views, background and outlook on life to your own?
- Value Homophily: How close are the avatar's views, background and outlook on life to your own?
- Wishful Identification: Do you wish you were more like the avatar?
- Liking: Do you like the avatar?
- Perspective-Taking: Can you take the avatar's perspective, sympathising with its experiences and motives?
- Embodiment: Does it feel as if you are acting through the avatar, rather than merely controlling it?
My MSc thesis research asked the following question: Which of elements of player identification are responsible for Ratan & Dawson's (2016) effect?
I investigated this using the game Unreal Tournament 2004TM; an action-packed first-person shooter in which players compete for "kills" and seek to avoid "deaths". My fellow technicians Foivos Vantzos, Ferenc Igali and I wrote a mutator for the game; a piece of code which modifies how the game works. This mutator forces the game output two pieces of data every 0.2 seconds for the duration of gameplay: The player's current health and their number of kills. Using this data I was able to identify timestamps of when players took significant (>20%) damage and when they killed an enemy.
Cross-correlating this with GSR data taken whilst playing gave a measure of how responsive players were to in-game events, both positive (kills) and negative (damage). I also asked participants to answer the same questions used by Downs, Bowman & Banks (2017) to determine in what way they relate to their chosen avatar.
By comparing the two sets of data, responsiveness to in-game events and polythetic factors of player identification, I hope to clarify what specifically it is about human-avatar connections which affect our physiological responses to in-game events.