Many people have played Werewolf (aka Mafia) with large groups of people, at parties or social events. Many of us have probably played it with some non-gamer friends. But can werewolf be a gamer's game as well?

I think the recent popularity of online Werewolf games on boardgamegeek shows that the answer is yes. But much of making the game interesting is in the rules and role selection that one uses.

A brief overview: Werewolf is played in night and day phases. In the night phase, the werewolves get to kill (eliminate) one player from the game. Some other players get to use special powers at this time as well, for example, the Seer gets to look at someone and see if they are a werewolf or not. In the day phase, players discuss who in the village they think is a Werewolf, to be lynched. Eventually a lynching is decided and that player is eliminated. Once either all the Werewolves have been killed, or the werewolves outnumber the villagers, the game ends. All players on the winning tam win, even if they are dead.

While large werewolf games can be fun and interesting, in my mind, they have several problems:

1)Game length. The more players, the longer the game takes. Also, since players are eliminated form the game, those eliminated early will be waiting for a long time for the next game.
2)Randomness. With many players, the lynchings occur mostly at random. There are just so many people to choose from, most people don't know anything, and you have to decide on somebody. If the group spends a long time trying to make a good guess each time, then the game will just never end.
3)Relative lack of strategy. The werewolves just say 'I'm a villager' like the 10 other villagers. The seer has to keep quiet for a while, because even if he finds one werewolf, there are more, and the seer will be killed after revealing this, leaving the rest of the players to stumble about, accusing at random. There is not a tremendous ability to use logic and deduction in large werewolf games.


To make Werewolf more of a gamer's game, I recommend playing with smaller numbers of players. 5, 7, and 9 work the best, since with an odd number of players alive during the day, there cannot be ties in lynch voting, if everyone votes. Key attributes of these smaller games are:

1)Most players have special roles. To make the game exciting for everyone, most players will have a special role, with a power they can use at certain times during the game. This also increases the strategy potential, because not only are there more things that players can do, there are more roles that they can claim to be (truthfully or as a bluff), for various strategic purposes.
2)A higher proportion of 'evil' roles to good roles, compared to the larger game. Because of all the special roles in the game, things can break down if you just keep to the roughly 25% of the players as werewolves that works well for larger games. Instead, this will be more like 33 to 40%. However, some of the evil players will not know each others identities, just as the good players do not know each other's roles. This means that for the Werewolves, they are not just trying to stay alive, they are trying to find their teammates, and adds a lot of strategic potential. Also, it means that the evil team cannot simply vote as a block together, because they could be confused as to who each other is.
3)Each round is far more critical. Generally in a small game, there will be little margin for error for the village team. This means that they must make a very educated decision on who to lynch. This removes most of the randomness, and makes the game much more about logic, deduction, and bluffing. That they have several villagers with powerful roles enables them to make a good pick, much of the time.

One more key difference in small games is that unlike large games, it works best for roles to NOT be revealed when a player dies. While in a large game, this reveal is important for players to have any clue what is going on (and to help find more werewolves when one has died, by looking at who voted for who). However in a small game, it is critical to not reveal roles upon death, in order for people to be able to claim various special roles as a bluff.


5 player werewolf:

The smallest number of players that the game works with is 5 players (plus a moderator, who can hand out roles, do the first night, and then go back to playing some other game).

The roles for this game are:

1 Werewolf. The game is over when The Werewolf dies, or if only 1 werewolf and 1 other player remain.

1 Sorcerer. The Sorcerer is on the Werewolf team and wins if the Werewolf wins. The Sorcerer looks like a Villager to the Seer. During the night, the Sorcerer looks at one player, and the moderator tells them if that player is the SEER or not.

The Werewolf and Sorcerer do NOT know each others identity, and are not permitted to make any sort of signals during the night (when the other had their eyes open and could see it).

1 Seer. During the night, the seer looks at one player and the moderator tells them if that player is the Werewolf or not. The Seer cannot tell if the player is the Sorcerer.

1 Hunter. The Hunter is a Villager with the following ability: If at the end of the game, the only 2 remaining players are the Hunter and the Werewolf, then the Hunter kills the Werewolf and the village wins.

1 Villager. But being the only player without a power, could be seen as a power or special role in itself.

This setup yield tons of opportunities for role claims and deduction. For example, the Sorcerer will probably claim to be either the Seer (to make it a tossup which person to believe as a seer, and to contradict their information), or possibly the Hunter (but in a way that they convince the Werewolf that it is not true).

The Hunter can win the game simply by not being killed, killing the werewolf in the end. Often, they will try to convince the Werewolf that they are the Sorcerer in some way, so the Werewolf will keep them alive, only to get a big surprise at the end. To do this, the Hunter might even make a false Seer claim! For example, if the Hunter were to say they were the Seer, and that a person was the Werewolf, who was not, the Werewolf would know they were lying and would probably think they were the Sorcerer.

The Seer generally needs to come out with their information on the first day. By day 2, it tends to be too late for the information. The game tends to hinge upon day 1. In that day, players will make role claims, try to sort them out, and then decide who to lynch (by majority vote). Here, the two evil roles are trying to lynch a villager, while the three villagers are trying to lynch an evil role. All players need to deduce who each other is.

Once the first lynching happens, day 2 will see three players remaining, if the Werewolf was not killed. In the most interesting case, both the Sorcerer and Hunter are still alive. Now, both of these players attempt to convince the Werewolf that they are evil. Its pretty funny to see the transformation. Day 1 you claim innocence, and Day 2 guilt. Here, the Werewolf must decide who is the true Sorcerer!

Even the basic villager has a critical role. Often, things will come down to two teams of competing role claims, and the odd man out (many times the villager), must choose correctly which team or player is telling the truth. Since 3 of 5 votes will decide the lynching, each players vote is critical. And since this one lynching will tend to determine the game, deduction is very critical.


In very advanced strategic plays, the Werewolf can claim to be the Seer, hoping to been seen as a possible Seer/possible Sorcerer. Then, if the group decides 'well, we have two seer claims, one is a seer, and one a sorcerer, so lets pick who the wolf is among the other three', then the wolf is not among those being chosen from! Players can also make 'trap' role claims, saying that they are a certain role, seeing the reaction of others, and then retracting.

The possibilities for people claiming various roles are nearly endless. Generally, many role claims will come out in day 1, which will then eventually be resolved down to a series of 'one of these two is this, the other is probably that'. There is a great deal of potential to use logic to figure things out, if someone has made a mistake and claimed something that doesn't really work. Also, players gain information, because if someone else claims your role, you know they are not telling the truth, and are probably on the opposite side. One key point is that the first person who claims a particular role tends to be somewhat more believed than the second, because they did it with less information about what was going on, and thus if it is a lie, they took greater risk that it is a mistake that just gives them away. Many times, immediately upon day 1 starting, this will cause several people to make Seer claims. It gets the game going right away.

The game is excellent, and very deep and strategic, and well suited for gamers.

For a 6 player game, simply add one more villager role to the 5 player game. This tends to simply lengthen the game, by making it more difficult to get an agreement on a lynching.

The 5 and 6 player games like this are very balanced, and wins will be split roughly evenly among the evil and good teams, generally depending on who played better and was more convincing. As a final note, I think the game works better if players are allowed to 'whisper' to each other, and are allowed to leave the room and talk to each other. It is also interesting to allow players to agree to write down their role on a piece of paper, and simultaneously reveal it to each other or to the group. (They can lie, of course, this does not prove anything, it instead forces a person to decide what their claim will be, without already knowing everyone else's).


I will talk more about 7 and 9 player games next week. In the meantime, try a game of 5 player Werewolf! It has a ton of depth packed into a relatively short playingtime, usually 30-45 minutes if people take their time to make a good decision, but dont take forever.