Verbal games

Game idea — Resistance is Useless

We’re off on hol tomorrow with T’s sister et al, and I was thinking about a game idea to try out on them.

I wanted a game for four (or similar number) players with the same sort of vibe as the Mafia game (aka Are You a Werewolf?), ie played out via chatting, and with elements of pretence, detection, etc. And for it to use minimal materials, have very few rules, and be acceptable to “family” groups as well as to gamers.

This is what I’ve got so far: see what you think.

The initial scenario is set in Occupied France and is called “Resistance Is Useless!”, and a wide range of others can be devised as required.

There are four roles: the Resistance worker, the innocent Innkeeper, the German officer and the downed English airman. Each player is dealt one of these randomly as their secret identity. They are then also dealt one randomly as their pretended identity. So eg. you might secretly be the English airman, but pretending to be the German officer, etc. (At the moment I don’t think it matters if you’re pretending to be the same thing that you actually are, but that might be wrong.)

At the end of the game, each player has to write down who they think the others secretly are, and this is then compared with the truth to see who got what right. Each role has its own victory condition — to identify one of the other roles, and to in turn be identified by one of them. So if you’re secretly the English airman, you want to identify the German officer, and to be identified by the Resistance worker. If you succeed in both of those, you win. Ties are broken by guessing other characters correctly, and broken again by not being guessed yourself, etc.

Play is triggered by event cards. These are drawn randomly from a small deck and are just descriptions of situations, eg. “Important German general wants to stay in the inn”, “Valuable painting of the Fallen Madonna has been stolen”, etc. Three are drawn during the game, each launching an unstructured scene of improvisation around the event. When people feel they’ve had enough of that event, the next one is drawn. When the third event has been played out, the game finishes and the players must write down their guesses as to who’s who.

So the challenge of the game is to drop clues to your identity that will be picked up by one (unknown) other player but preferably not by the other two. This is the aspect I’m not sure about, as basically I don’t know if it’s really possible. But then that is true of Mafia as well, I suppose. My feeling is that as people relax into the game (ie. drink) they will let things slip, and it’ll still be enjoyable if they let too much go.

Any thoughts / ideas / hideous brokennesses?

29 replies on “Game idea — Resistance is Useless”

I usually rely on my chum to spot theoretical brokenness, but this sounds in principle great fun to me – similar to stuff he and I have tried on different occasions. I usually find something‘s broken – but it scarcely matters if people are having fun, which I’m sure they will! I imagine you’ll get endless layers of bluff – hard to divine the truth, but good fun trying.

I think it’s flawed – though it might be entertaining once or twice.

All players have equal chances of picking up on any clues, whereas for the game to have a chance, there must be some things only known to particular players.

eg. (In keeping with the generally silly spirit of things), the downed airman may have been told by another resistance worker that all secret passwords involve fish, so he can identify the true resistance leader by talking about fish. (and the real resistance leader knows this)

The innkeeper may have been told that German officers are all drinking vodka this week because supplies of wine are thought to have been poisoned by resistance workers. (and the real German knows this)

Of course, confusion may result as players drink all kind of stuff and any fish joke may be innocently picked up and topped by other players, etc.

I have looked on for their 3-player mafia, but now can’t find it. There was a couple of interesting Mafia variant there.

I like this idea, would be fun to at least try… works in a sort of circle. Of course, everyone could win (everyone guesses everyone elses role correctly, no?).

You could have a game based on 4 possible roles:

The true resistance leader.
The french Quisling in the resistance.
The true gestapo leader.
The deep cover french resistance in the gestapo.

The True Resistance wants to find the mole, as does the true gestapo. The moles want to discover each other, so that their leader can find them.

Bung in some information (like the mole knows their leader, but the leaders only know each other, and not their moles….) and you could have some frolics.

Yes, in a circle was the idea. I like your opposed pairs of pairs too though. That would work well in an Internal Affairs police and mobsters kind of setting, and the opportunity to add in some extra information as you suggest would strengthen it. Hmm, I think I’ll give that a try on the holiday too.

I like it. It sounds like a lot of fun, but it is a bit broken as it stands methinks.

If the idea is for a bit of improvisational roleplaying, how exactly does the player pretending to be the German officer interact with those pretending to be the Airman or the Resistance worker? Shouldn’t he or she have them both arrested on the spot, if only to keep up appearances?

More game mechanically, Watervole is right – some private information needs to be introduced, or the players have nothing to work with. Not too much, though. The probability of a player guessing completely correctly is one quarter in the absence of any information, and this will rise rapidly as guesses become educated. Exchanging information with nods and winks is fun, so a significant fraction of games will result in everyone getting everything right if we’re not careful.

What about the following suggestion as the germ of a solution? (Unfortunately it’s a bit more work in present form.) Add the following information to everyone’s real identity card:

3 password phrases (e.g. “Gadzooks my feet hurt!”)
3 tactical insights (e.g. “Tanks travel much slower in winter”)

Resistance Worker:
3 password phrases (as above)
3 French facts (e.g. “De Gaulle has a big nose”)

3 French Facts (as above)
3 gastronomic observations (e.g. “Claret goes well with kippers”)

German Officer:
3 gastronomic observations
3 tactical insights

Here’s the real trick – each pair of lists overlaps only once, so for example the resistance worker and the airman will share only one password, and will not know which one. Now everyone has some things to drop into conversation, unsure as to whether anyone will pick up on them.

Of course, things get more than a bit chaotic when players realise they can not only make up convincing sounding phrases of their own in order to pretend to be fishing for information, but they can also pretend to have recognised a phrase that means nothing to them. At this point I think alcohol may be mandatory.

The real drawback to this ‘fix’ is that playing the game several times requires using new “real identity” cards, to prevent people learning the phrases. This needn’t be an infinite chore, though. If there are enough sets of real identities, and enough randomization within the lists, even perfect recall of all the possible phrases won’t significantly help you since guessing at phrases that aren’t on your card is (a) obviously suspicious and (b) unlikely to make an honest match with your associate.

Ho hum. Maybe that solution is a bit too involved, but at least it provides a trigger for some (comically stilted) conversation. Back to work. That was fun – hope the holiday is too!

Mm, interesting — it should work given a decent stack of real identity cards. Hopefuly as people played it lots and became familiar with those, they’d also become more adept at making plausible phrases up, and other types of faking, so when the first ones came round again it wouldn’t be a problem.

(Ideally you could combine them across sets such that there are three possible Resistance identities that you can use with each Airman identity, each sharing one password, etc. But that might be a bit tricky to administer to ensure consistency.)

After some reflection, I suspect this won’t quite work.

The mechanic that you’re missing from Mafia is the asymmetry between Town and Mafia with respect to what actions they wish to carry out. To summarise how things work in brief:

* Suppose Mafia maintain perfect cover, then every lynch is random and victory patterns perfectly match the odds.
* But suppose instead that the Mafia favour to some degree those lynches which they know to be Town members. Their odds of winning then rise rapidly…
* …unless the Town suspect this is going on and use this extra information to catch the Mafia from their voting patterns.
* But then the Mafia can deliberately vote against their own interests in the early game to obfuscate their identities in the late game.

And so on. But all of this gameplay springs from the fact that the Mafia have a solid game mechanical motivation to behave differently. (And they do. When the gamne is played well the Mafia do better than if they voted randomly.)

I propose the following tweak to the game as you have it currently: The player who is pretending to be the innocent innkeeper (henceforth PII) begins the game with knowledge of one other (random) character’s true player. This breaks the game’s information symmetry and will provide leverage for the players to try clever plans and deduce things.

In order to physically set this up the overt roles are first assigned. The secret role cards are then shuffled and one is dealt to the PII. The PII looks at it. The other secret role cards are then returned to him or her and the PII puts them out of sight (behind her back or under the table, for example) and places the viewed card in a known position. All players now watch as each person picks a card, with the PII choosing last. The PII now knows who has the viewed card, but nobody else does. (It could even be the PII!)

Doesn’t that give the PII a point one game in 4 (knows the ID he’s trying to guess), nothing one game in 4 (knows his own ID) and 1/6 of a point the other two (only guess from 2 instead of 3)?

How many games do you have to play with a given group before it doesn’t matter? Or are board gamers just very philosophical about such things (or only in it for the beer and pretzels), and don’t care so long as the free points before the game starts are even across all games ever played as opposed to across the games they play personally?

Obviously it’s for particular players to decide whether this effect harms their enjoyment or not.

Suffice it to say that there is a strongly established precedent for boardgames and card games where player positions are not perfectly symmetric and the positions are assigned randomly or rotated.

Alternative question:

Fischer and others have proposed some random chess variants. Does anyone play chess where black loses a pawn before the game starts with probability 1 in 4?

Not with people who want a game to play a few times and then move on. But maybe the likelihood of extended replay is beyond the scope of game design.

Possibly. A single hand is obviously very random – it’s rare to play only a few hands, so I’d suggest it’s too random for the purpose here.

Even if you play to a winner, thus having some chance of exploiting MVT, it’s still possible to win off a small number of lucky hands against a far superior player. That format is popular, though. So if randomly selecting the winner is considered an undesirable property of a game as I suggest, I guess that must not be too random for players’ taste.

I wonder how many rounds of this game you’d have to play to make the winner “less random” than a typical poker night. Of course the winner may not matter as much, since the game is more fun to play: “the scoring mechanics were bobbins but I enjoyed myself” is a perfectly acceptable review.

I wasn’t referring to randomness. The point is the asymmetry between players. Would you want to be Big Blind every hand? Didn’t think so.

In which case I don’t understand the basis of your original argument that asymmetry doesn’t matter provided it is randomly-generated (“as long as the PII is someone random each time”).

My claim is that randomly selecting the winner is bad. My further claim is that, since I imagine “Resistance is Useless!” being played maybe 2-4 rounds total, max player score 4-8, then giving away 1 free point with probability 1 in 4 each round has a good chance of randomly selecting the winner.

In the Texas Hold-Em example, ignoring the randomness of the cards and only considering the big blind disadvantage: if I’m playing a number of rounds close to the number of players, obviously I don’t want to be big blind every time. But also I don’t want big blind to be chosen at random every time, because in a lot of games someone will still get shafted. It just might not be me.

For a small number of rounds, I want a game with less handing out of random (dis-)advantage (e.g. we could be big blind once each, or the PII could have a more consistent advantage and we could play as PII once each).

For sufficiently many rounds, fine, randomly give people a million points if you like: it doesn’t affect the outcome. If you’re imagining RiU being played 20 or 30 times in an evening, then players have a reasonable chance of being a “lucky bastard” at least once. But that’s just my point about the winner being less random, which you say isn’t what you mean. So I’m at a loss.

This may be why I don’t generally play games. Or because I don’t.

If the number of plays is going to be low then it would be sensible to rotate rather than randomise the identity of the PII, certainly.

Beyond that, as I mentioned above, players will have to decide how much asymmetry they consider acceptable.

This may be why I don’t generally play games.

Different players prefer different levels of fairness. For some players the potential glory of winning from a disadvantage is a strong incentive. For others, an asymmetry is welcome because it’s their best chance to win. But on the flipside there are players whose enjoyment would be completely wiped out by the presence of unfairness.

Both as a designer and as an event organiser, understanding why your target audience plays things is critical. Fortunately I suspect Mo knows his players well enough to be confident of getting this right.

For this particular group, “because the baby’s soundly asleep and they’ve had a few drinks” is usually enough. But I think I will try it with rotating the PII rather than randomly, to start with at least, just in case 4 hands does turn out to be the right number.

An entirely fair point, and I get the impression that this is likely to be a game of length such that people are more likely to play once rather than several times in a row. (Sometimes I’ve been in a group that has played Mafia several times in a row, sometimes I’ve been in a group that has played it once and then played something different – or, more likely, broken into smaller sub-groups to play something different.) Possible (mutually exclusive) fixes include:

(a) All players score (e.g.) 8 points per thematically correct identification, except the PII who scores (e.g.) 6 points per thematically correct identification but starts off with, on average, +1/3 correct identification per game due to the information imbalance;

(b) Not all identifications score equally, so one character might score +8, +6 and +4 for the three other identifications, another might score +9, +6 and +3, a third might score +9, +5 and +4 and the PII might score +6, +5 and +4, but have a greater chance of scoring them;

(c) Score +1 per identification, but the PII loses all ties.

None of these are spot on, but I think they can be tuned close enough for jazz through playtesting. (When I say “identification”, I include the possibility, as desired, that you should earn credit for being identified by the character you want to be identifed by.)

The other thing to say is that this sounds easily fun enough to try at least once, but when so much of the joy of Mafia is that by introducing secret characters then every game can be different from the previous one, prescription of roles (and, more to the point, prescription of exactly four players per game) is a little inflexible.

Your (b) option is appealing to me, because it relates to some further changes that I made (and some furtherer ones that I subsequently thought of) during the holiday. Which I shall post about once I’ve cleared the post-hol backlog!

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.