One dominant metaphor for online communities (whether online games, social networks, or the Internet in general) is an outward branching web. This assumption helps fuel the narratives that online games like World of Warcraft are escapist fantasies where gamers pretend to be someone they’re not, that social networks allow us to be connected to a diverse range of people and opinions, that powerful online search tools help people find new information.
But the most ironic lesson from the digital age is that the exact opposite might be a more accurate description—online communities are powerful tools for identity management, especially in reinforcing who people already are and how they see and think about the world.
In my own research into online gaming, we’ve found that gamers behave in virtual worlds in ways that align with their offline personalities: for example, gamers who are more extraverted prefer group activities online, while gamers who are more introverted prefer more independent/solitary activities. And the same is true for choosing what genres to play: gamers who are more extraverted prefer games that focus on stimulation (whether this comes from guns, explosions, fast-pacing, or competition). In this sense, games aren’t about pretending to be someone we’re not, but more about identity management—helping us become more of who we already are.
This identity management isn’t just about the choices we make; it’s also about how online communities shape what we see in the first place. The emergence of big data analytics in selectively presenting information to us in search engines and social networks is creating a “mirror effect”. Psychology research has shown that people have a preference for faces/ideas/stories that evoke familiarity and similarity. Black box machine learning algorithms that try to predict more engaging content often bake these cognitive biases into their rules.
Thus, on Facebook, we’re more likely to be presented with information we already agree with, and the further away someone is from us in our social network, the less likely their information is presented to us. The flip side is also true: when we express our opinions on Facebook, they are more likely to be presented to people who we are highly similar to and who already agree with our opinion. This facilitates ideological balkanization because it creates illusory majorities and marginalizes opposing viewpoints.
These same algorithms also impact information retrieval. For example, Google personalizes search results based on your own search history—it prioritizes sites and sources that you have preferentially clicked on in the past. Over time, Google’s algorithms learn your ideology and returns information that aligns with your ideology.
The myth of digital exploration and adventure is concealing a much more sobering emerging phenomenon where cognitive biases and digital code conspire to transform online communities into powerful identity management platforms. After all, it’s much easier to algorithmically target consumers when they are predictable, and people are much more predictable if they don’t change.