Random Party Game Idea

What makes a party game work

A party game is not a game that happens to be played at a party. It is a specific piece of engineering with a hard brief: eight people who did not all choose each other, some of them tired, one of them not really wanting to play, and a rules explanation that must be over in ninety seconds.

Press Generate and this gives you one. A sample of what is in here:

• Charades
• Pictionary
• Taboo
• Telestrations
• Gartic Phone
• Jackbox Party Pack

Look at that list and you can already see the shape of what works. Everybody is doing something. Nobody is waiting. And in every one of them, the fun is produced by the players rather than by the box.

Picking one

1. Press Generate and take what it gives you, which spares you the twenty minute conversation about what to play.
2. Ask for three and let the room vote, which is faster than letting the room suggest.
3. Type into Contains to search for a title you half remember.
4. Copy to keep a shortlist for the next time.

Where these games come from

Party games gathered and checked by hand, from parlour games older than anyone playing them to things that run on a phone. The tool picks from that pool in your browser.

The most copied party game began as a psychology experiment

In the spring of 1986, in the psychology department of Moscow State University, a student named Dmitry Davidoff invented a game.

He was not trying to entertain anybody. He was interested in a question about groups: what happens when a small number of people know something that the majority do not, and both sides are in the same room, talking. He wanted a way to watch that happen, repeatedly, under conditions he could control.

So he built one. Divide the players secretly into two groups. The small group knows who they all are. The large group knows only how many of the small group exist. Then alternate. At night, the small group quietly removes somebody. In the day, everybody argues, accuses, defends, and votes to eliminate a suspect. Repeat until one side wins.

He called it Mafia. He ran it in classrooms as a teaching tool, and it escaped. It went into the dormitories of Moscow University, and then into the student summer camps, and from there into the rest of the world, where it acquired new clothes and a hundred names. Werewolf. The Resistance. Secret Hitler. Blood on the Clocktower. Among Us, which forty million people played in a pandemic without any idea that they were running a Soviet psychology exercise on a spaceship.

It is one of the very few party games whose invention can be dated to a season and a building.

An informed minority against an uninformed majority

That phrase is Davidoff's, and it is worth understanding, because once you see the shape you will notice it in a dozen other games.

Ordinary games give everybody the same information and different resources. Chess is symmetrical. Poker hides cards, but the hiding is even. Davidoff's structure is different: the information itself is unequal, and the unequal side is smaller.

That produces something no other game mechanism produces. The majority has all the power and none of the knowledge, so it must extract knowledge from behaviour, which means from faces, from hesitation, from the shape of an argument. The minority has all the knowledge and no power, so it must survive by performing innocence convincingly.

The result is that the game is played entirely in the space between people. The board is the room. The pieces are your friends' faces. This is why it has been studied seriously by mathematicians, who have modelled optimal play, and why it turned up in a Russian textbook on nonverbal communication as a teaching exercise for reading people.

It is also why it is so uncomfortable, and why almost everybody has a story about somebody taking it too seriously. Davidoff built a machine for making people suspect each other, and it works.

Why elimination ruins an evening

Notice the flaw in the design he handed us.

Somebody dies on the first night. That person now has nothing to do. They have the whole evening, in a room where the game is the only thing happening, and they may not speak. Twenty minutes later a second person joins them. By the end, half the room has been sitting silently for an hour, and those are the people who will tell you afterwards that they do not really like party games.

This is the single most important thing to check before you commit a room to a game. Does it bench people, and for how long.

Most of the good games solve it in one of three ways. They keep the dead in play, as ghosts with some small residual power. They make rounds short, so being out means waiting three minutes rather than an hour. Or they remove elimination entirely and score instead, which is the design decision behind almost every party game invented in the last twenty years and the reason the drawing games have outlived the deduction ones at ordinary parties.

The general rule is simple enough to carry around. A party game should be judged by the experience of the person doing worst at it, not by the experience of the winner. The winner will have fun in any game.

Matching the game to the room

The mistake is to ask which game is best. The right question is which game fits these people, tonight.

People who do not know each other. Nothing that rewards knowing somebody's history. Drawing games are extraordinary here, because everybody is equally bad and the badness is the joke. Deduction games are cruel, because reading a stranger is impossible and reading a friend is the whole point.

A group with one very competitive person. Choose something where skill barely helps. This is not a compromise. A party game with a skill ceiling produces one person having an excellent evening and seven people watching them.

Mixed ages. Look for games where the winning move and the funny move are the same move. Charades survives Christmas because a grandmother miming a helicopter is both.

Late, and everyone tired. Rules must fit in one sentence. Anything requiring a role card, a moderator or a phase structure will die in the explanation, and the room will politely never play it.

A group that talks over each other. This is the room where Davidoff's game finally works, and it is glorious. Give a loud, argumentative group a hidden traitor and they will not need you again for three hours.

Questions people ask

How long should explaining the rules take?

Under two minutes, and if it cannot be done in two, teach it while playing the first round and tell everyone that round does not count. Nobody has ever learned a party game by listening to it being explained.

Are Mafia and Werewolf the same game?

Structurally yes. Werewolf is a retheming that spread in the English speaking world, and its innovations were mostly about giving the moderator better tools. The bones underneath are the ones from Moscow.

What is the best number of players?

For most party games, more people means more waiting, and the sweet spot sits lower than the box says. A game printed as four to sixteen players is usually best at six. The upper number on a box is a marketing decision.

Is it a problem that some of these games are about lying?

Some people find it genuinely unpleasant, and they are not being poor sports. Deception games ask you to be convincingly false to your friends and then to be forgiven for it, and not everybody wants to practise that. It is worth asking rather than assuming.

References

1. Mafia (party game), on Davidoff's creation of the game in 1986 at Moscow State University. https://en.wikipedia.org/wiki/Mafia_(party_game)
2. Braverman, M., Etesami, O., and Mossel, E. (2008). Mafia: a theoretical study of players and coalitions in a partial information environment. Annals of Applied Probability, 18(2), 825–846. https://arxiv.org/abs/math/0609534
3. History of social deduction games, on the line from Mafia to Among Us. https://impostergame.com/history-of-social-deduction-games/