It offers neither the narrative escapism of Cyberpunk 2077 nor the anxiety antidote of Animal Crossing: New Horizons. Instead, the game makes me stressed, paranoid, and frustrated. And yet, since last fall, I’ve been playing Among Us with locked-down friends across the country every week. It’s a game of suspicion, and somehow, thus one of excitement and connection, which is a bit of a paradox—but one that can be explained by some insights into human psychology.
A social deduction game, the essential rules of Among Us are straightforward. Players are either categorized as innocents or as spies. Most are innocents, and their goal is to deduce who the spies are before the spies kill them all. In Among Us, the innocents are Crewmates and the spies Imposters, and the intrigue takes place on a spaceship prone to sabotage. Launched in June 2018, it was a surprise hit of 2020; US representatives Alexandria Ocasio-Cortez and Ilhan Omar played it on a Twitch livestream in October, and in November it broke half a billion monthly players.
It’s a game that, though periodically silent, demands that you talk. Perhaps it’s no wonder, then, that Among Us took off in a year of experiments in remote society, as we all ran out of things to say over Zoom. But what explains the gravitational pull of this social deduction game? Why do players frequently make bad decisions? And why did my own friends collectively vote to eject me, while I was diligently refuelling the engines?
Well, for one thing, humans are generally terrible at identifying lies. For Chris Street, a cognitive psychologist and expert in lie detection at the University of Huddersfield, social deduction games are variants on the classic guessing game: Which hand is the coin in?
What at first appears to be a simple “this one, not that one” decision invites a spiral of second-guesses. “What if they are bluffing, but they know that I know that they are bluffing?” says Street. “Social deduction games frame the penny problem in a more structured world, where there is useful information to be gleaned if you pull at the claims around the table and find out which unravel.”
Where liars can use any number of persuasive tactics to build trust, from pretending to complete tasks in Among Us to denouncing other Imposters, figuring out who is a spy need not be simply a case of refining your bullshit detector.
“I think we all hope for some hidden secret ability to root out the truth by detecting subtle behaviors and tells,” says Street. “Reality is less forgiving. Across many research studies over the decades, our best estimate of people’s lie detecting ability is ever so slightly better than guessing, with an accuracy of 54 percent, when 50 percent could be achieved by guessing at random.”
While there is research suggesting that liars give off behavioral cues—one influential paper claims that bluffers tend to tell simpler stories with narrower vocabulary and more negative emotion words—Street ascribes this skein of psychology largely to the realm of TV. “Popular fiction tells us that there may be subtle indicators of deception in the liar’s behavior,” he says. However, “when we lie, we do not give ourselves away so easily. If our lies were so readily detectable, we would likely choose not to lie in the first place.”
Liars may not give off obvious cues, but innocent players aren’t completely rudderless. When we pick up on lies, it’s because we recognize information in the speaker’s claim that is contradicted by the howling of other players, or because we caught them leaping from one the vents on the Among Us spaceship. In Street’s Adaptive Lie Detector theory, he suggests that people adapt to rely on context to guide their credulity. And when we play a social deduction game like Among Us—or Street’s preferred title, The Resistance—we’re already apprehensive. Street owns 160 board games and many of them are social deduction games. Yet given his somewhat pessimistic view on human lie detecting abilities, I’m not surprised when he says his expertise fails to furnish an advantage on the tabletop.
Another reason Among Us players make poor decisions is that the design of social deduction games confounds the resources in our brain. “They mostly challenge our wits,” says Celia Hodent, an expert in game user experience and author of The Psychology of Video Games. “More specifically, they challenge our attention. We need to focus on what’s going on, use our memory to connect the dots, while also engaging our logical reasoning and communication skills.”
While an Impostor can successfully frame another player, we are capable of making mistakes without their trickery. “Our perception is subjective, our attention resources are scarce, and our memory is fallible,” says Hodent. “We humans have a tendency to believe that we make rational decisions most of the time, when we are constantly influenced and misled by the numerous cognitive and social biases we have.”
Observers of social deduction games are regularly privy to information that escapes players, possibly because their attention is not being challenged in the same way. As a result, they have a front row seat to the formation of bad decisions. Playing Among Us, I have marveled at the mistakes that unfold after my assassination, as the remaining crewmates perceive each other’s behavior as “sus” instead of targeting the actual Imposter. In one of my games, I watched Sergio murder me! Someone even voiced their suspicion before they died. But for Will and Daisy, their respective denials only serve to entrench their misgivings about each other.
“We have a tendency to focus on, search, and remember information that validates our preconceptions,” says Hodent. “Once we have such intuition, it becomes very hard to consider information that is going against our belief.”
Daisy and Will vote against each other without interrogating Sergio. In the end, says Hodent, “Sergio, just like a skilled magician, just needed to take advantage of humans’ implicit biases and brain limitations to keep the obvious unnoticed or unremembered.” Such occasions are simultaneously infuriating and beguiling. Moreover, those shortcomings ensure that a surplus of a crucial ingredient, surprise, greases the progression of every game.
We also use heuristics, or rules of thumb, when playing a social deduction game. “In Among Us, one of the heuristics that people have at their disposal is crewmate color,” suggests Jon Roozenbeek, a postdoctoral fellow at the Cambridge Social Decision-Making Lab.
In December, developer Innersloth revealed that some colors were more popular than others, with red, black, and white the most popular and lime, brown, and green the least. Meanwhile, as this Reddit post demonstrates, colors generate their own stereotypes. “Basically, not all crewmates are created equal,” says Roozenbeek. Another heuristic is the way players’ reputations can precede them. A spy in one game is not automatically a spy in the next, but we find it hard to shake that notoriety.
Online social deduction games like Among Us and precursors like Town of Salem have made the genre accessible like never before. You can play the game with strangers, but it feels more tense when played with friends. Perhaps this is because you can’t bully your friends like you can anonymous netizens, but also because, as Roozenbeek suggests, “you trust your friends, so them not believing you when you’re actually telling the truth can feel like your friends are calling your trustworthiness into question.” Those social ties lend jeopardy, then, to the deception. So it’s not just confusion but anxiety and quiet flashes of fury that supplement the experience. These complex feelings arguably give the social deduction game its voltage. And nothing feels so foul in Among Us as being suspected of a murder by your pals when you are, in fact, completely innocent. We’ve gathered that humans are often mistaken. Why does it still hurt to be on the receiving end?
For Matthew Lieberman, professor of social cognitive neuroscience at UCLA, it’s because humans are bad at keeping things in context. “I do think that people can take these things more personally in games than they probably should,” he tells me over video chat. “One of the most famous things in my field of social psychology is that humans are notoriously bad at appreciating the ways in which the situational context is guiding behavior.” This is known as the Fundamental Attribution Error.
“We tend to think that the way people are behaving is always reflective of who they are as a person. And so when they do something that harms us in some way, even if it’s in the game context, we attribute a little more than we should of that to who they are as a person,” says Lieberman. Social deduction games excel at generating social tension. When your friends vote you out, it feels like betrayal. “People take it personally when they get accused and they feel like they haven’t done anything wrong,” says Lieberman. “That’s because the whole game is set up on finding the most mundane things suspicious.”
The classic social deduction game is Mafia. Created by Russian psychology student Dimitry Davidoff in 1987, its slight rules, which are organized around assassination, interrogation, and elimination, create an electric atmosphere of mutual distrust. Tension arises from the limited evidence that players have to work with, while spies plead their innocence, possibly by framing someone else.
Amendments to the game regulate and give context to those tensions. From Moscow, Mafia spread quickly among students and travelers. In the late 1990s, interactive fiction luminary Andrew Plotkin switched assassins for werewolves and turned Mafia into Werewolf. A more recent iteration is Secret Hitler, which changes the script by setting the game in 1930s Germany. It charges a majority of players to defend liberalism without conceding to measures proposed by clandestine fascists.
Lieberman regularly plays Secret Hitler with his family and two others in his Los Angeles bubble. Tabletop games, which require people you know gathering in one room, “do tend to be a bit richer,” he says.
Indeed, Hodent suggests that Among Us has grown so much because it generates a breadth of social interaction in the face of lockdowns. “One of the key elements to engage people in an activity is to consider their intrinsic motivation for doing so,” says Hodent, who formerly steered UX on Fortnite at Epic Games. “Activities that give us the feeling of competence, autonomy, and relatedness are typically intrinsically motivating. While Among Us can satisfy all of these human needs, it’s particularly efficient for satisfying our need for relatedness, especially at a time when we are all required to stay physically distant from one another because of a pandemic.”
Crewmates need to cooperate to identify the Imposters, while the latter secretly compete against the former. The heart of the social deduction game lies in the meaningful social interactions this conflict produces. Interactions that I, for one, have been craving. We can’t yet all gather around a table, but we can set a date for a round of Among Us. And I’ll keep looking forward to getting shot from the airlock—even when I’m innocent.
- 📩 The latest on tech, science, and more: Get our newsletters!
- 2034, Part I: Peril in the South China Sea
- The digital divide is giving American churches hell
- The Sims made me realize I’m ready for more in life
- Here’s what learning to juggle does to your brain
- A case against the Peeping Tom theory of privacy
- 🎮 WIRED Games: Get the latest tips, reviews, and more
- 📱 Torn between the latest phones? Never fear—check out our iPhone buying guide and favorite Android phones