Via Jamie Madigan (Psychology of Video Games), I watched two GDC 2015 talks about anti-social behaviours in video games. Jeffrey Lin (Riot Games) on the science of shaping player behavior (GDC, 2015) and Ben Lewis Evans (Player Research) on how game design influence social behavior (GDC, 2015). Both touched on parts of Internet lore, the Greater Internet Fuckwad Theory (GIF; Penny Arcade, 2004). The GIF theory is often associated with the Online Disinhibition Effect (Suler, 2004), however Lin and Evans did not make any indications of applying Suler’s work. As a communication scientist, I recognized hints on a different theory: the Social Identity model of Deinidividuation Effects (SIDE).
The Greater Internet Fuckwad Theory
The comic gives us an intuitive formulation of how ordinary people can behave toxically to others online. An ordinary person is transformed into an aggressive online user through two factors. The first factor is anonymity, the quintessential element of the Internet. A person can come in to any site and barring any identifying information the user provides, no knows who their personal identity is. The intuitive explanation is that such anonymity allows the person to behave in ways they would not normally do with their exposed personal identity. The second factor is the audience, not as obvious as the first factor in causing toxic behaviors. In communication, you need two people, one to send and the other to receive. Hence, the audience is the receiver of what toxic behaviours the user sends to. The end result is the user acting as a “total fuckwad” as seen on any comment section, tweets, and places with an audience.
In dealing with toxicity in video games, the GIF theory was discussed as early as 2008 by Bill Fulton (2008) and he came to the same conclusions as did Jeffrey Lin and Ben Lewis: toxic online behaviors are reducing video games’ revenues. Unfortunately, we have yet to effectively regulate online toxic behaviours despite the GIF theory’s intuitive formulation since the time the comic was published. Paradoxically, the GIF theory’s “completeness” as the answer for everything wrong on the internet precluded further refinements (see rationalwiki, gtz, 2012).
My conceptual analysis of GIF theory reveals numerous theoretical holes. It lacks any explanatory power for positive behaviours (see reddit). The GIF theory cannot account for how some people don’t behave toxically, how some individuals of certain demographics are the targets of specific types of toxicitiy, or when such toxic behaviours occur. Conversely, I notice the GIF theory does not include any communication technologies in its formulation, so it could be applied outside of the internet. Hence, the GIF theory resembles to the classical theories of deindividuation more so than to the online disinhibition effect (Suler, 2004). Deindividuation theories’ basic formulation is the combination of anonymity and immersion within the group leading to the loss of one’s self turning into a “total fuckwad”, such “transformation” turns a group into a violent mob. These theoretical holes and its resemblance to deinidividuation theories suggest a look into the Social Identity model of Deindividuation Effects.
The Social Identity model of Deinidividuation Effects
Deindividuation theories were among others in explaining antisocial behaviours, such as flaming, during early computer mediated communication (CMC; email, SMS, etc.) research. The Social Identity model of Deindividuation Effects (SIDE) was developed in the early 1990’s to address CMC behaviours and as a response to the conceptual and empirical inconsistencies found in deindividuation research (see Postmes & Spears, 1998; Reicher et al., 1995). As its name show, the theorists approached CMC from a social identity perspective, a perspective informed by social identity theory and self-categorization theory. These theories explain why we behave in certain ways when in the presence of groups and SIDE model explains such processes within the context of anonymity, regardless of whether it is digital or physical.
The similarity between GIF theory and SIDE model is that both address the individual’s behaviours as a function of their relation with the audience under conditions of anonymity. SIDE expands on what happens under conditions of 1) anonymity of the individual and others, and 2) the anonymity to others. Second, the presence of groups in the social context and the individual’s relationship with the group influence what sort of behaviours the individual would perform and not perform. Thus, there are two aspects of the SIDE model, the ‘cognitive’ and ‘strategic’.
Cognitive aspect of SIDE: “Anonymity of…”
The anonymity of the individual and others mean that one cannot discern much about anyone’s personal characteristics because the computer mediated environment filtered out many identity cues. With the scarcity of identity cues, this meant for the individual that their sense of self is lost in the group, this is deindividuation. However, the SIDE model predicts the individual’s sense of self shifts rather than being “lost”. This shift depends on the visibility of identity cues in the social context and these cues are highly influential because of their relative scarcity. When there are personal identity cues, that is information unique to the individual, a personal identity is maintained. When there are social identity cues pertaining to a social group, the individual shifts from their personal identity to that social identity, this is depersonalization.
Based on self-categorization theory, the depersonalization process is a shift from a personal identity to a social identity. This social identity shift depends on what social cues are most salient to the individual in the current context. For example, being in reddit and communicating with other redditors leads the individual in identifying oneself as a redditor. With so little personal information among individual group members, depersonalization is also a process of stereotyping. Individuals self-stereotyping to their social identity and view other similar group members as more similar to themselves than focusing on personal differences. Dissimilar individuals from the main group are stereotyped and grouped in whichever characteristics they display. Following the reddit example, redditors interact stereotypically with a visibly tumblr user. As opposed to deindividuation theories’ predictions, individuals are more aware of their social identity, conformity to social norms are stronger which lead to a polarization of behaviours and opinions (Lee, 2007; van Hiel et al., 2007). It is important to note that depersonalization is not an “all or nothing” process towards a complete social “hivemind” identity, but is on a continuum that individuals still distinguish themselves among group members, nevertheless group identification is strong.
Prevailing social norms is another depersonalization factor. Social identities carry social norms of behaviours. Redditors, imgurians, youtubers, twitch, 4chan, tumblr, twitters among others have their own peculiar and often unspoken social manners. Individuals pick up these social norms as a means to fit in. An analogy is the social context of the restaurant, a waiter knows how to behave and what not to behave towards a customer and so does the customer. An extraterrestrial who never been to a restaurant would not know what being a waiter or a customer is like, so they observe others on what to do and not to do in restaurants, they imitate them or try different behaviours and see the results, the make mistakes and eventually learn what is normative (i.e. waiting to be seated) and antinormative (i.e. taking an already occupied table). A key distinction from deindividuation and SIDE is the latter posited that CMC behaviours are due to the interaction of anonymity and the influences of prevailing social identity and norms. Thus, the SIDE model explains both prosocial and antisocial behaviours.
Even if there are ambiguous social identities in a given social context, social norms emerge as a result of the interactions between individuals (Postmes et al., 2000). This emerges from inferring members’ common behaviours or attributes. For example, people start greeting each other with a certain phrase, if spread without resistance over time it will become a standard greeting form. Social norms can be shaped, changed or reinforced through social influence. Priming is one of them and used by Jeffrey Lin for League of Legends (GDC, 2013). In an experiment by Postmes et al. (2001), participants were primed by unscrambling sentence that contains ideas related to either efficiency or kindness, those ideas are salient in the participants’ mind which influence their behaviours in their next task, thinking of solutions for a dilemma. The participants with the primed norm were more likely to offer prime-related solutions to a problem. So, efficiency primed participants offered efficiency-related solutions. In Lin’s priming experiment, he primed players with different types of messages before the start of the game, either in the loading screen or as an in-game tip. The players with the prosocial prime were more helpful and reported less toxicity. Furthermore, you do not need to prime all groups members for this effect, you can count on social influence factors, such as social pressure, group identification among others to spread the primed norm. In Lin’s matchmaking experiment, the presence of a group (two friends) influenced other players’ behaviours to be like them. According to the SIDE model, social influence is stronger because of the individuals’ stronger group identification within CMC, in effect anonymity does not reduce visibility of the self, but rather increases visibility towards a group identity (Sassenberg & Boos, 2003).
In relation to video games, the SIDE model offers a theoretical framework why players think, feel and behave. As Lin pointed out in his GDC 2015 talk, players’ toxic behaviours and the views of these behaviours as normal/natural part of the online environment emerge as the result of cumulative communications between users, over times players internalize what is explicitly and implicitly tolerated within the social spaces. Fortunately, players can be nudged through social influence to perceive antisocial or undesirable behaviours as less tolerated. The SIDE model can help us explain why users of different social environments behave differently, helps us what to look for (e.g., platforms, games, geography, people, CMC affordances, rules, etc.), identify its impact, adjust it and replicate or adapt them to another social environments. The SIDE model give us a different understanding of identity in CMC, it is not so much a simple formulation of anonymity as a causal factor, but the nuances of identities in CMC that affects our behaviours. These nuances of identities can affect how an individual communicates to different others.
Strategic aspect of SIDE: “Anonymity to…”
In the early days of the internet, anonymity was thought to be an equalizer between groups, such as race, gender and other social standing, everyone is an equal (Dubrovsky et al., 1991). However, SIDE posited that everyone will act and think upon salient identity cues, thus everyone is depersonalized and stereotyped. They communicate others as a group rather than as unique individuals. The power relations between groups may be unequal and thus making the internet not so equal at all (Postmes & Spears, 2002). Such unequal power relations would have group members of different standing to communicate strategically.
The strategic aspect of SIDE refers to the power relations of an individual’s group membership and how that group stands in relation with other present groups in the social context. This relates to how our actions are accountable to the reactions of our own group and other groups. For example, students would not cheat on their exams because their teacher would punish such behaviour, a relation from a powerful group member (i.e. the teacher) to a less powerful group (i.e. students; Spears et al., 2002). Less powerful groups, especially those whose norms are in conflict with the more powerful groups, limit their behaviours in line with what the more powerful group accepts and rejects. Nevertheless, the expression of group behaviours depends on visibility of the in-group and out-group audiences. The less powerful groups can take advantage of anonymity to hide certain aspect of their identity or membership. Their anonymity to the more powerful group affords them the safety to communicate in ways they would been punished if they were identified. For example, victims of domestic abuse or harassment would feel safe in disclosing their stories without punishment from their perpetrators. Conversely, their anonymity to their own group would mean less visibility and less social support in the social context, giving greater power standing to the more powerful group. Nevertheless, group members who are identifiable to each other, but not the more powerful group, can provide a channel for social support to each other and social resistance towards the more powerful group (see Reicher et al., 1998; Spears et al., 2002). In the student-teacher example, students who are anonymous to the teacher, but identifiable to each other and who encourage support were more likely to endorse behaviours the teacher would punish, such as challenging the teacher’s authority. Although, some readers may think is a ‘us vs. them’ mentality, it’s a bit complicated than that. Thus, the social dynamics are quite nuanced on the individuals’ behaviour on the relative visibility of their social identity and the visibility of the audiences.
Olivier Klein and colleagues (2007) further elaborated the strategic SIDE as an identity performance, “the purposeful expression or suppression of behaviours relevant to those norms conventionally associated with a salient social identity”, that is communicating or not the identity visible to the audiences. They posited two general functions in identity performance: identity consolidation and mobilization.
Identity consolidation is to affirm an identity and its norms. To be part of a group, an individual needs to be recognized as such, this gives the sense of acceptance and belonging to the in-group. To fit with the in-group, they behave in ways that is socially accepted by the in-group, this is important among those with an insecure social identity. Some identities are costly to consolidate or sometimes such consolidations may be seen as illegitimate by the group. Consider the “fake” girl gamer stereotype. Female gamers who express their gamer identity face delegitimisation by other gamers because they belong to a another stereotyped group, women, that hold some undesirable and “essentialized” traits to the gamer group (see Bastian & Haslam, 2006). The gamer group would ask those “fake” gamer girls to prove that they are an exception to the stereotype (see Cote, 2015). They would “overconform” in order to publically make their identity consolidations clear, and for the girl gamer, such overconformity might be internalizing misogyny as one way to be acknowledged. Other girl gamers may choose to distinguish themselves rather than conform, where they challenge the gamer group’s negative views of the “fake” girl gamer stereotype. The second function is identity mobilization refers to engaging group members to preserve or enhance the standing of their group through collective action. As individual view themselves and others as a group, this provides a social power in shaping the social world. Consider how certain movements or groups were able to galvanize many people into action because they all share a common identity and such actions were important to that identity.
Much of the research on the strategic aspect of SIDE and most relevant to videogames is gender (Postmes & Spears, 2002; Lee, 2004; Lee, 2007; Palomares, 2004). Women’s knowledge and prior experiences with gamers would inform them about their status in the gaming community. The general gist is that women are not treated equally by their male counterparts, they face sexual harassment with the perpetrators facing little or no consequences from their fellow male gamers because of their bystanding silence, an implicit tolerance of these behaviours (Brehm, 2013; Cote, 2015; Salter & Blodgett, 2012). Women were more likely to remain anonymous, use gender-neutral identities or not reveal their gender, especially in male stereotypical environments (Flanagin et al., 2002; Spears et al., 2011). Such strategy allows these women to be treated equally among their unsuspecting male gamers. If their gender identity is revealed (see Burch, 2013), then a psychological spotlight makes them highly visible to the woman and to the audience, shifting from an intragroup to an intergroup communication, that is between gamers to between gamers and the girl gamers. How the audience communicate their intergroup relation can determine whether the women would feel safe in expressing her identity or not. The strategic SIDE would explain why some women prefer to play with their friends rather than with strangers as they feel safe in expressing themselves without sanction, their friends is their in-group (Cote, 2015).
There are other applications of the strategic SIDE to explore. An interesting dynamic to explore is live streaming (e.g., Twitch and youtube), the audience’s anonymity in relation to the streamer’s visibility may account why the audience can turn nasty due to the lack of accountability whereas the streamer may be careful in their self-presentation because of their visible accountability to their audience (see Sarkeesian, 2015). Developers are another group to consider as the relationship between developers and gamers are not equal as the former have to power to implement changes in the game, impose rules and enforce them. Some genres are social environments (e.g., MMOs, MOBAs) and its developers are already aware of their role in their players’ social environment, how involved they are in cultivating their environment may determine their products’ long-term success.
The Greater Internet Fuckwad Theory offered an intuitive yet simplistic understanding of antisocial online behaviours. However, it has not significantly contributed in combating such behaviours. The Social Identity model of Deinidividuation Effects gives us a theoretical approach that can offer game and social media designers ways in shaping, changing and reinforcing behaviours. Nevertheless, it has limits in what it can and cannot explain, for example sockpuppets. There are other CMC theoretical models to consider as well, such as Social Information Processing theory, Hyperpersonal model, and Media Richness theory (see Walther chapter for a more comprehensive look). As I hope readers appreciate the SIDE model’s utility in demonstrating the influences of social identities on computer mediated behaviours.
Joinson, A. (Ed.). (2007). Oxford handbook of internet psychology. Oxford University Press.
Sundar, S. S. (2015). The Handbook of the Psychology of Communication Technology (Vol. 32). John Wiley & Sons.
Walther, J. B. (2011). Theories of computer-mediated communication and interpersonal relations. The handbook of interpersonal communication, 4, 443-479.
Bastian, B., & Haslam, N. (2006). Psychological essentialism and stereotype endorsement. Journal of Experimental Social Psychology, 42 (2), 228-235. doi: 10.1016/j.jesp.2005.03.003
Brehm, A. L. (2013). Navigating the feminine in massively multiplayer online games: gender in world of warcraft. Frontiers in Psychology, 4 . doi: 10.3389/fpsyg.2013.00903
Cote, A. C. (2015). “I can defend myself”: Women’s strategies for coping with harassment while gaming online. Games and Culture. doi: 10.1177/1555412015587603
Dubrovsky, V. J., Kiesler, S., & Sethna, B. N. (1991). The equalization phenomenon: Status effects in Computer-Mediated and Face-to-Face Decision-Making groups. Human–Computer Interaction, 6 (2), 119-146. doi: 10.1207/s15327051hci0602_2
Flanagin, A. J., Tiyaamornwong, V., O’Connor, J., & Seibold, D. R. (2002). Computer-Mediated group work: The interaction of sex and anonymity. Communication Research, 29 (1), 66-93. doi: 10.1177/0093650202029001004
Klein, O., Spears, R., & Reicher, S. (2007). Social identity performance: Extending the strategic side of SIDE. Personality and Social Psychology Review, 11 (1), 28-45. doi: 10.1177/1088868306294588
Lee, E.-J. (2004). Effects of gendered character representation on person perception and informational social influence in computer-mediated communication. Computers in Human Behavior, 20 (6), 779-799. doi: 10.1016/j.chb.2003.11.005
Lee, E.-J. (2007). Deindividuation effects on group polarization in Computer-Mediated communication: The role of group identification, Public-Self-awareness, and perceived argument quality. Journal of Communication, 57 (2), 385-403. doi: 10.1111/j.1460-2466.2007.00348.x
Lee, E.-J. (2007). Effects of gendered language on gender stereotyping in Computer-Mediated communication: The moderating role of depersonalization and Gender-Role orientation. Human Communication Research, 33 (4), 515-535. doi: 10.1111/j.1468-2958.2007.00310.x
Palomares, N. A. (2004). Gender schematicity, gender identity salience, and Gender-Linked language use. Human Communication Research, 30 (4), 556-588. doi: 10.1111/j.1468-2958.2004.tb00745.x
Postmes, T., & Spears, R. (1998). Deindividuation and antinormative behavior: A meta-analysis. Psychological Bulletin, 123 (3), 238-259. doi: 10.1037/0033-2909.123.3.238
Postmes, T., & Spears, R. (2002). Behavior online: Does anonymous computer communication reduce gender inequality? Personality and Social Psychology Bulletin, 28 (8), 1073-1083. doi: 10.1177/01461672022811006
Postmes, T., Spears, R., & Lea, M. (2000). The formation of group norms in computer-mediated communication. Human Communication Research, 26 (3), 341-371. doi: 10.1111/j.1468-2958.2000.tb00761.x
Postmes, T., Spears, R., Sakhel, K., & de Groot, D. (2001). Social influence in Computer-Mediated communication: The effects of anonymity on group behavior. Personality and Social Psychology Bulletin, 27 (10), 1243-1254. doi: 10.1177/01461672012710001
Reicher, S. D., Spears, R., & Postmes, T. (1995). A social identity model of deindividuation phenomena. European Review of Social Psychology, 6 (1), 161-198. doi: 10.1080/14792779443000049
Reicher, S., Levine, R. M., & Gordijn, E. (1998). More on deindividuation, power relations between groups and the expression of social identity: Three studies on the effects of visibility to the in-group. British Journal of Social Psychology, 37 (1), 15-40. doi: 10.1111/j.2044-8309.1998.tb01155.x
Salter, A., & Blodgett, B. (2012). Hypermasculinity & dickwolves: The contentious role of women in the new gaming public. Journal of Broadcasting & Electronic Media, 56 (3), 401-416. doi: 10.1080/08838151.2012.705199
Sassenberg, K., & Boos, M. (2003). Attitude change in Computer-Mediated communication: Effects of anonymity and category norms. Group Processes & Intergroup Relations, 6 (4), 405-422. doi: 10.1177/13684302030064006
Spears, R., Lea, M., Corneliussen, R. A., Postmes, T., & Haar, W. T. (2002). Computer-Mediated communication as a channel for social resistance. Small Group Research, 33 (5), 555-574. doi: 10.1177/104649602237170
Spears, R., Lea, M., Postmes, T., & Wolbert, A. (2011). A SIDE look at computer-mediated interaction. Strategic uses of social technology: An interactive perspective of social psychology, 16-39.
Suler, J. (2004). The online disinhibition effect. CyberPsychology & Behavior, 7 (3), 321-326. doi: 10.1089/1094931041291295
Van Hiel, A., Hautman, L., Cornelis, I., & De Clercq, B. (2007). Football hooliganism: comparing self-awareness and social identity theory explanations. Journal of Community & Applied Social Psychology, 17 (3), 169-186. doi: 10.1002/casp.902