Once You Eliminate The Impossible, Whoever Remains, No Matter How Improbable, Is A Werewolf

Social deduction games are one of those game genres that have a built-in audience almost separate from other hobby gamers. Entire clubs are formed around Werewolf and Resistance variants, and basic forms of the game like Mafia are found often on Internet forums and at camps. However, today I didn’t want to talk about the genre as a whole but about a trend that, to my delight, is slowly becoming more popular: Evidence.

Why is evidence so cool?

Games like One Night Werewolf are fun and tense, but they rely completely on the players’ ability to lie smoothly. Because there is no way to objectively determine someone’s identity before everyone has to vote, players have to rely entirely on intuition. This frequently devolves into guesswork, especially if players are exceptionally good at lying and/or exceptionally bad at telling that someone is lying.

For all my griping, this system obviously works, considering how popular the games are. However, with the addition of vague evidence, players can work on multiple levels. You can find out a traitor because they’re taking a long time to come up with something, catch them with a contradiction in the evidence, or even try to analyze if they’re being too vague. Evidence creates a layer of complexity that allows more play styles to work.

There are basically two categories of deduction games with evidence: Games where “something is there”, and games where “something is missing”. Let’s dive into both of them.

Something is there

My go-to example for this type of game is Deception: Murder in Hong Kong, which combines things like the Merlin mechanic from Resistance: Avalon (one of my favorite mechanics in board games) with the vague communication style of Concept. Each player has a pile of evidence in front of them, half of it potential murder weapons and the other half innocuous objects. The players have to guess not only who the traitor is, but also which pieces of lethal and nonlethal evidence they committed the murder with.

Central to this game is the lab technician. They are witness to the murder and the evidence, and are on the side of the detectives, but they can only communicate through unhelpful cards that say things like the victim’s build. In this way, it’s a little easier to determine who the traitor is by using the cards to eliminate evidence, but it’s vague enough to be difficult and the traitor (and accomplice, if playing with that role) can easily pivot the investigation towards someone else.

There are lots of ways to play Deception. You can focus solely on the lab technician, trying to suss out meaning from the fact that the victim was short and liked gambling. You can try to keep an eye out for people subtly trying to shift the investigation away from them and accuse them of being the murderer. You can even, as the murderer, intentionally make wrong accusations against yourself to throw the other players off the trail. It’s this freedom that makes the game superior to a lot of others in the genre.

Something is missing

In these games, the non-traitors have some piece of evidence, while the traitor doesn’t – and they have to desperately cover up their ignorance. I’ve played two games recently that use this mechanic in similar ways. In Spyfall, everyone is dealt a location except the spy, then asks questions just specific enough to figure out who is the spy without allowing them to figure out where they are. More off-the-wall is A Fake Artist Goes To New York, where everyone collaboratively draws a single picture, but one player (the fake artist) doesn’t know what they’re supposed to be drawing.

Something amazing about these kinds of games is how funny they are. The traitor can’t hide under a rock snickering about how nobody’s noticing them: The games, in one way or another, force them to participate and bluff that they have the relevant information. The acrobatics required on both sides lead to frequent bouts of nervous laughter – an atmosphere I find superior to the seriousness that simpler deduction games tend to create.

Fake Artist deserves special mention because the traitor’s awkward attempts to draw something they don’t know often leads to the end result being hilarious. Because the pictures are permanent and include a score sheet that says who the fake artist was, players can also look back on some of their funnier games and remember.


In part, my preference for deduction games with evidence is personal and nothing to do with a more objective evaluation. However, I think that these twists allow deduction games to be more fun and accessible for people who either aren’t good at or don’t like straight-up mindgames. (For the record, I’m OK at them if keeping quiet and being passed over is a valid strategy.) A game that expands the audience of a genre can only be good, and the polarizing genre of social deduction could use an wider audience.