This weekend I played Secret Hitler! The games’ designers – Max Temkin, Mike Boxleiter and Tommy Maranges – had intended to create a social deduction card game for players 13 years and older. But it’s not only a physical game! There is also a virtual version simulating the physical gameplay (with some nuanced UI/UX differences). Regardless of modality, as reflected by the gameplay design, the designers wanted to curate a gameplay experience that has just as much of the “deduction” element as the “social” element in a social deduction game.
To understand the balance between both components, I want to highlight the formal elements of the game. Players are either assigned to the “Liberal” or “Fascist” party and there is a single unknown Hitler amongst all of the liberals. Although Hitler is unknown, everybody in the same party knows each other and the fascists are cognizant of who their leader is. With each round, the elected “President” chooses a “Chancellor” and it is put to a vote. If the vote fails to reach a majority, typically the next player in line becomes the “President”. Otherwise, a “Chancellor” is elected, and the “President” who has chosen three policy cards (namely, “Fascist” or “Liberal” policies) filters a single policy card out. The “Chancellor” is given these cards, and must choose between the two cards which may be: Fascist/Fascist, Liberal/Fascist, Liberal/Liberal. You win if your party gets political control (liberals enacting five policies versus fascists enacting six policies OR electing Hitler “Chancellor” after three policies).
After analyzing the game mechanics, I can see how the designers were intentional with the flow of information from round to round. Although it wasn’t really structured, there was some element of conditional probability in action where paying attention to the patterns and trends in behavior of each player was critical to decode their loyalty. For example, in our game, “I don’t trust him. He’s been really aggressive when it comes to the Chancellor vote” was screamed to emphasize how that person may just be a fascist. Why? Because players socially tend to employ more momentum and energy when it comes to pushing through with disruptive, fascist policies. There is a sense of urgency from the fascist party, especially to reach the point where electing Hitler becomes an end-all be-all. Another example of this pattern recognition came about when someone yelled “You ONLY voted against Chancellor nominations known to be fascist.” Such accusations aren’t exactly random; players must engage with the logic puzzle in the way they know how; weaving together bits of fragmented information into a cohesive logical conclusion. But no logical conclusion is truly definite, as the social dynamics of individual behavior (i.e, lying to create a false impression, bluffing, or deceptive/dramatic actors as covered in class) still reign.
In other words, players can make sound, calculated judgments based on potentially misleading information. It is a beautiful balance of “social” and “deduction”, but to improve the game it may be worthwhile to incorporate visual markers of loyalty. For instance, imagine there were a graph of how people’s perceived loyalties were and the group came about to discuss such as in games like Mafia. This would not only help organize information, but it may lead to a greater consensus and strategic gameplay. Of course, it might just equally cause chaos and introduce greater bias into the gameplay. Additionally, in the several times I played it, I found that due to the imbalance of roles in certain scenarios, there is an inherent bias for the Fascist team. In games with a larger number of players beyond 5 or 10, the likelihood of the Fascist team to enact policies, and eliminate Liberal players is extremely high due to the power dynamics (i.e., Hitler’s powers) and probabilities of drawing fascist cards. It’s almost as if their power compounds as a function of the number of players. This definitely makes the game feel lopsided, where deductive reasoning and paying attention to social behaviors may simply not be enough. A potential suggestion for improvement would be to employ some form of information balancing, where a certain number of identities are revealed in favor of the liberal party when they are at risk of being eliminated. Would be a fun way of introducing come-backs from the underdog 🙂
All in all – a beautifully crafted game!
