Spyfall is a social deduction game created in 2014 by Alexandr Ushan. While it was initially a card game, this game is now available to play online. It targets a broad 13+ audience, and can be played among groups of 3 or larger. The game’s primary objective is for the designated spy to guess the location of the other players and for the other players to guess who the spy is based on the cues presented in the game. In this critical play, I argue that Spyfall creates a game of social deduction by incorporating the use of both symmetric and asymmetric information as core mechanics.
In this game, all players share the ability to ask questions to other players as an ongoing procedure. In doing so, shared (or symmetric) information is presented (i.e., What type of question was asked? How did the other player respond? Were there any suspicious behavioral cues?). However, the role that a player is assigned can influence how they may use that information to guess the role of others or hide their role as a spy. An instance where this played into the spy’s advantage was when the location was a supermarket, and a non-spy confirmed that you could go grocery shopping at this place. The spy used this information to determine the location and ask questions that shielded their identity.
On the flip side, asymmetric information contributed to deduction as well. While the non-spies know the location alongside the list of locations, the spy only has a list of the locations from which to try to discern the real one. These parameters highlight the boundary placed around the game. This asymmetric information emphasizes social deduction through players attempting to use this knowledge to outwit their opponents. For non-spies, they want to ask the right questions to the right individuals that do not obviously give away what the location is, but helps them gain evidence to know who to vote as the spy. For spies, they need to figure out the right questions to ask that do not make it too obvious that they do not know what the location is.
These game mechanics create a dynamic of cooperation among players where the non-spies attempt to piece together their perceptions of the presented evidence to figure out the spy. However, opponent-based conflicts can arise when a spy is able to shield their identity and appear as a non-spy effectively. During one round, I was convinced that my friend was the spy because (1) she was quick and vocal in supporting my efforts to accuse a different person and (2) she was asked the grocery shopping question and I was confident that she noticed the specificity of the question, figured out the location, and almost effectively appeared as a non-spy. Well, I was wrong. It turned out to be another player who was more quiet and appeared unsure of whether to support my, albeit bold, accusations. Here, I used asymmetric and symmetric information to cooperate with the wrong people, causing conflict and the non-spies to lose.
I thought it was clever for the designers to add a time limit to each round because it prevented the game from being overpowered in the non-spies’ favor. I also thought it was clever that players voted for the spy at the end of the round. In another iteration of the game (which we initially used), each player could accuse another of being a spy mid-game. However, given that my group played with 4 people, this mechanic could be abused by having each person accuse a new person and figuring out who the spy is by process of elimination. Therefore, I would recommend not including this mechanic in small groups.
