I'm definitely interested the idea Rym suggested of a Rym-only GeekNights special on how to run Mafia, but a blog post or some such would suffice. I'll probably end up running this at some point this summer.
I'd not read "I want my hat back" before, only heard you talk about it on the podcast. This video was the perfect introduction to it. The zoom in is just soooo perfect.
Okay. I'll write a generic Mafia simulator program that can handle different rules and numbers of players. I just want to make sure this hasn't already been done first. Has it?
Okay. I'll write a generic Mafia simulator program that can handle different rules and numbers of players. I just want to make sure this hasn't already been done first. Has it?
That doesn't include consideration of the Doctor (Scrym call it the "archangel" I think), though - nor would it handle any custom rules someone managed to come up with.
I guess this would be easy enough to write that it wouldn't matter much if it had been done before, though.
Of course, I'd rather a closed-form mathematical solution that's as generic as possible, but that would take much more effort.
Can someone find a more general solution involving both Doctors and Detectives?
Angel and archangel have always been doctors and sheriffs when I've run the game. I've also included:
Mafia boss: wakes up by them self and kills another towns person at the end of the night. The other mafia don't know who is the boss. I don't think I've ever played a game where a mafia boss hasn't killed one of his own mafia. This is only good for games with loads of people.
Hooded Hangman: a normal citizen, except if this person is selected by the towns people, there is nobody to hang him, so he doesn't die on that round. I don't think I ever played a game where the hangman wasn't immediately killed the night following the fact he was revealed. Also I don't think I've ever played a game with a hooded hangman card without most accused people saying "Don't bother voting for me, you can't kill me anyway."
Here's an important question - if the citizens win, the dead citizens still lose, right? I'm guessing that they still lose, which invalidates the strategy proposed here.
In that case, the strategy there is quite sound - the Detective finds out a sufficient amount of information, then announces it and gets themselves lynched on purpose - this reveals their card and certifies the info as 100% accurate, and the mafia gets screwed over.
In that case, the strategy there is quite sound - the Detective finds out a sufficient amount of information, then announces it and gets themselves lynched on purpose - this reveals their card and certifies the info as 100% accurate, and the mafia gets screwed over.
Except that assumption 1 completely changes the game. The game is supposed to be played without communication aids - you only have your body to work with.
Except that assumption 1 completely changes the game. The game is supposed to be played without communication aids - you only have your body to work with.
Can you elaborate on how this works? From what I've heard and read, the villagers communicate openly while only the mafia communicates in secret.
Except that assumption 1 completely changes the game. The game is supposed to be played without communication aids - you only have your body to work with.
Can you elaborate on how this works?
Yes. No writing or drawing is permitted. It's pretty straightforward, actually; the only visual aid other than your body is the card you're given, and those cards can only be revealed when you die.
If information is revealed during the game, you must rely on memory in order to access it.
You have to speak in order to be heard, and you might speak over someone or be spoken over; in other words, you need to take command of the group in order to get a point across.
You could devise a system of hand signals or other body language, but it needs to be done while the game is being played, and somehow communicated to the other people.
So the concept of a simultaneous communication method, or really any sort of note-passing or note-taking, totally alters the game.
All you need is to be able to speak - all the Detective has to do is say the following: "I am the Detective. My investigations have revealed that X, Y, ... are citizens, whereas A, B, ... are mafia. Lynch me to verify the accuracy of my words."
All you need is to be able to speak - all the Detective has to do is say the following: "I am the Detective. My investigations have revealed that X, Y, ... are citizens, whereas A, B, ... are mafia. Lynch me to verify the accuracy of my words."
Watch this.
Player A: "I am the detctive, and player X is Mafia..."
Player B: "BULLSHIT! HE'S LYING! I'M THE DETECTIVE AND HE'S MAFIA!"
And then the game devolves into arguing - which one do you believe? And if you ask for a revelation, they can just point at random people, because you can't know any better. It then becomes a matter of who is the more trustworthy player.
Which is why simultaneous communication changes things - you can't interrupt a written message.
Ah, but only the mafia would lie in that situation, so lynch either player A or player B and either way, the citizens benefit - you can even just pick randomly from the two options.
In other words, anyone announcing themselves as a detective is either a detective or mafiosi - never a plain ol' citizen - and so lynching them is always beneficial.
Furthermore, I would argue that whoever is first to say "I am the detective" is quite likely to be the detective - it's not in any mafiosi's interest to announce this first, because if they do so they're giving away free information, and asking to get themselves lynched.
On that note, I think in the scenario you've suggested the correct reaction is simply to remember what both player A and player B have to say, and then lynch player B, because you gain more by lynching a mafiosi than a detective, and player A is more likely to be the detective.
True, lying about being the detective is usually a losing proposition. So that specific strategy is still valid. However, it still essentially relies on luck; if the mafia kill the detective or otherwise convince the citizens to lynch him before he has a chance to collect enough information, the citizens are done for.
But yeah, having the detective divulge information and then be killed appears to be a sound strategy no matter what.
With skilled players, no strategy in this game gives 100% chance of victory. However, using this strategy increases the chances of victory by a significant amount.
I want to run Mafia in real time at my summer camp. Real time as in, the mafia meets at night and the meetings to discuss lynching and such happens during the day, all throughout the course of a week. Logistics is an issue, though, as everyone has a roommate, so I can't just pull someone from their room and reveal them to their roommate. Can anyone think of any solutions to these sorts of problems?
Alternately, convince me why running it in real time breaks the game. It's cool if I can do it, but if it'll totally suck I'll just run it the normal way.
I want to run Mafia in real time at my summer camp. Real time as in, the mafia meets at night and the meetings to discuss lynching and such happens during the day, all throughout the course of a week. Logistics is an issue, though, as everyone has a roommate, so I can't just pull someone from their room and reveal them to their roommate. Can anyone think of any solutions to these sorts of problems?
You could assign a single role to each room, rather than each person. That's probably the best solution.
Alternately, convince me why running it in real time breaks the game. It's cool if I can do it, but if it'll totally suck I'll just run it the normal way.
Well the first thing that comes to mind is the Dead Men Tell No Tales rule. I mean how fun would it be to not be able to talk for the entire week? Sucks to be the first person killed.
Allow them to talk, and just not talk about the game... something is going to slip. For a long-term game like that, I'd go with more of a dinner murder-mystery type plot.
Also, we could have possibly as many as 50 or 60 players per week, but I think that's solvable by pairing roommates together and playing two separate games concurrently, bringing the games to 15 or so players each.
30 or 60 players would be utter madness, but it could be a lot of fun as long as you got the balance right.
The only thing is that you need to make sure there's actually enough time to finish the game, I guess, and there wouldn't be unless you introduced some crazy rules.
Even one week might be a little too short for 15 players, depending on the character types.
The biggest factor is that communication is subject to confusion in actual play - so signal may become lost in noise.
I want to run Mafia in real time at my summer camp. Real time as in, the mafia meets at night and the meetings to discuss lynching and such happens during the day, all throughout the course of a week. Logistics is an issue, though, as everyone has a roommate, so I can't just pull someone from their room and reveal them to their roommate. Can anyone think of any solutions to these sorts of problems?
Alternately, convince me why running it in real time breaks the game. It's cool if I can do it, but if it'll totally suck I'll just run it the normal way.
This is an awesome idea. And yeah, it makes the mafia's job harder, but I think it's far more fun trying to slip out without being noticed.
The only thing is that you need to make sure there's actually enough time to finish the game, I guess, and there wouldn't be unless you introduced some crazy rules.
Even one week might be a little too short for 15 players, depending on the character types.
This is my biggest worry about a large game. Anyone know of some rules that exist for quick elimination of lots of players? Maybe do two or three lynchings and Mafia kills every day/night?
This is an awesome idea. And yeah, it makes the mafia's job harder, but I think it's far more fun trying to slip out without being noticed.
So you're saying not to even bother worrying about the roommates? Interesting.
I think what I'll end up doing is trying it the first week and then tweaking it as the weeks go on. I get eight chances here; chances are I can make it awesome after a few tries.
Comments
I'm definitely interested the idea Rym suggested of a Rym-only GeekNights special on how to run Mafia, but a blog post or some such would suffice. I'll probably end up running this at some point this summer.
I guess this would be easy enough to write that it wouldn't matter much if it had been done before, though.
Of course, I'd rather a closed-form mathematical solution that's as generic as possible, but that would take much more effort.
Can someone find a more general solution involving both Doctors and Detectives?
Mafia boss: wakes up by them self and kills another towns person at the end of the night. The other mafia don't know who is the boss. I don't think I've ever played a game where a mafia boss hasn't killed one of his own mafia. This is only good for games with loads of people.
Hooded Hangman: a normal citizen, except if this person is selected by the towns people, there is nobody to hang him, so he doesn't die on that round. I don't think I ever played a game where the hangman wasn't immediately killed the night following the fact he was revealed. Also I don't think I've ever played a game with a hooded hangman card without most accused people saying "Don't bother voting for me, you can't kill me anyway."
EDIT: Durr - it increases the chance of citizens being picked for lynching later, as compared to killing some other citizen.
Also, it seems that with Detectives there is actually nontrivial strategy involved.
If information is revealed during the game, you must rely on memory in order to access it.
You have to speak in order to be heard, and you might speak over someone or be spoken over; in other words, you need to take command of the group in order to get a point across.
You could devise a system of hand signals or other body language, but it needs to be done while the game is being played, and somehow communicated to the other people.
So the concept of a simultaneous communication method, or really any sort of note-passing or note-taking, totally alters the game.
"I am the Detective. My investigations have revealed that X, Y, ... are citizens, whereas A, B, ... are mafia. Lynch me to verify the accuracy of my words."
Player A: "I am the detctive, and player X is Mafia..."
Player B: "BULLSHIT! HE'S LYING! I'M THE DETECTIVE AND HE'S MAFIA!"
And then the game devolves into arguing - which one do you believe? And if you ask for a revelation, they can just point at random people, because you can't know any better. It then becomes a matter of who is the more trustworthy player.
Which is why simultaneous communication changes things - you can't interrupt a written message.
In other words, anyone announcing themselves as a detective is either a detective or mafiosi - never a plain ol' citizen - and so lynching them is always beneficial.
Furthermore, I would argue that whoever is first to say "I am the detective" is quite likely to be the detective - it's not in any mafiosi's interest to announce this first, because if they do so they're giving away free information, and asking to get themselves lynched.
On that note, I think in the scenario you've suggested the correct reaction is simply to remember what both player A and player B have to say, and then lynch player B, because you gain more by lynching a mafiosi than a detective, and player A is more likely to be the detective.
But yeah, having the detective divulge information and then be killed appears to be a sound strategy no matter what.
Alternately, convince me why running it in real time breaks the game. It's cool if I can do it, but if it'll totally suck I'll just run it the normal way.
Allow them to talk, and just not talk about the game... something is going to slip. For a long-term game like that, I'd go with more of a dinner murder-mystery type plot.
The only thing is that you need to make sure there's actually enough time to finish the game, I guess, and there wouldn't be unless you introduced some crazy rules.
Even one week might be a little too short for 15 players, depending on the character types.
I think what I'll end up doing is trying it the first week and then tweaking it as the weeks go on. I get eight chances here; chances are I can make it awesome after a few tries.