Mafia 2015: Sign-up thread (5 Viewers)

Ford Prefect

Senior Member
May 28, 2009
10,557
In the game thread, the game in which we had two g families competing with each other and the Cs was brought up, it's be sick to do that again for the next round we run, when it's run'
 

Hængebøffer

Senior Member
Jun 4, 2009
25,185
I haven't read what Hoori is saying, but she is right about protecting the D. One thing I remember the experienced players have said several of times is that the BG should always protect the D, if he knows who it is.
 

Osman

Koul Khara!
Aug 30, 2002
61,496
Its not complicated, allways protect the D unless you are bluffing and making sure they think you are protecting D 100% for sure (who they dont know) and then protect yourself knowing that. I did that few times as special role player, either D, lawyer, Jailer etc, knew the D and BG, intentionally let in a G in our inner circle msn discussions, work him for few days, convinced him I was BG (the times I was not too), and made sure he got the gist in last minutes I was protecting myself. Generally misinformation purposes it can be useful, but otherwise, as a general rule, allways protect the D, no question. Unless the D is utterly useless player you cant work well with, but generally not the case.
 

king Ale

Senior Member
Oct 28, 2004
21,689
@Alen, I need ideas for determining the payoffs because we need them to model the problem. For example, do you agree with my way of giving weights (in terms of value for the Cs/Gs) to each scenario? Does it make sense to you, that in terms of utility for civilians, if we give a 4 to the scenario where both D and BG are alive and no civilian has been killed either, the utility of both D and BG are alive and a civilian has been killed should be 3? Or as another example, do you think if the utility of a dead detective is 4 for the Gs, the utility of a dead BG should be 2 and utility of no one's dead scenario should be -1? That's why I think anybody who likes Mafia and has played this game can contribute to determining the payoffs.

And there's no Janna versus Osman here. Whole point is: is there a strategy a rational/experienced/smart player should follow most of the time?

- - - Updated - - -

Why are you assuming that?

In this game, I don't see how you could be sure that the G's knew who the D was. Or are you just trying to find out in general?
For the sake of symmetry. Dealing with asymmetric information is not easy here. Besides, that's the most interesting case actually. Gs figured out that Fred was the detective (they claim they did). Thing is that Ocelot did not know that Gs knew but I'm assuming he did (BG should always assume that Gs know the D, imo).
 

Völler

Always spot on
May 6, 2012
23,091
@Alen, I need ideas for determining the payoffs because we need them to model the problem. For example, do you agree with my way of giving weights (in terms of value for the Cs/Gs) to each scenario? Does it make sense to you, that in terms of utility for civilians, if we give a 4 to the scenario where both D and BG are alive and no civilian has been killed either, the utility of both D and BG are alive and a civilian has been killed should be 3? Or as another example, do you think if the utility of a dead detective is 4 for the Gs, the utility of a dead BG should be 2 and utility of no one's dead scenario should be -1? That's why I think anybody who likes Mafia and has played this game can contribute to determining the payoffs.

And there's no Janna versus Osman here. Whole point is: is there a strategy a rational/experienced/smart player should follow most of the time?

- - - Updated - - -



For the sake of symmetry. Dealing with asymmetric information is not easy here. Besides, that's the most interesting case actually. Gs figured out that Fred was the detective (they claim they did). Thing is that Ocelot did not know that Gs knew but I'm assuming he did (BG should always assume that Gs know the D, imo).
Ok, then I get what you're doing. Your payoff ratings seem about fine.

The answer just seems so obvious. The G's would have to go for the BG, I assume, unrealistically often for it to make sense for the BG not to protect the D.
 

king Ale

Senior Member
Oct 28, 2004
21,689
Ok, then I get what you're doing. Your payoff ratings seem about fine.

The answer just seems so obvious. The G's would have to go for the BG, I assume, unrealistically often for it to make sense for the BG not to protect the D.
It's not really that obvious. With these payoffs there is no dominant strategy. You, either as a G or as a C, should mix your strategies. For example, go after the BG 60% of the time, a normal C 25% of the time and the D 15% of the time (I'm just throwing numbers, I left my calculations at home lol but it was something along those lines). So basically, Gs should be dominantly going for the BG but of course they have to mix this strategy by going after normal Cs and the detective every now and then. It's trickier for the BG but if I'm not wrong (I'll check again), the BG should protect the detective 66% of the time.
 

Völler

Always spot on
May 6, 2012
23,091
It's not really that obvious. With these payoffs there is no dominant strategy. You, either as a G or as a C, should mix your strategies. For example, go after the BG 60% of the time, a normal C 25% of the time and the D 15% of the time (I'm just throwing numbers, I left my calculations at home lol but it was something along those lines). So basically, Gs should be dominantly going for the BG but of course they have to mix this strategy by going after normal Cs and the detective every now and then. It's trickier for the BG but if I'm not wrong (I'll check again), the BG should protect the detective 66% of the time.
But do you think we'll see a scenario like this often enough for it to make sense actually doing the math? I mean, sure, mixing up the strategy would make sense if you're going to run into this scenario often, but I just don't see why that would be the case.
 

king Ale

Senior Member
Oct 28, 2004
21,689
But do you think we'll see a scenario like this often enough for it to make sense actually doing the math? I mean, sure, mixing up the strategy would make sense if you're going to run into this scenario often, but I just don't see why that would be the case.
Well, it indeed depends on how many times you run into the same scenario in either one single game or in multiple iterations of the game. A few years ago, I used to play this game multiple times per day perhaps for two weeks. There I adopted the strategy of ALWAYS protecting the detective as the BG even though I didn't do any math to figure that out. Now, it's obviously not an always versus never kind of a strategy (it shouldn't, you should mix) but it's still close enough (66% versus 33%). I learned it by experience. What I ind really fascinating about game theory is that it is very well mapped onto the true and natural state of things.

- - - Updated - - -

But the problem in this game, @king Ale, is you said to different things. That's why people were after you.
That's not true. I stayed true to what I believed and said. We didn't know if Ocelot was contacted, and there was always the possibility of a desperate G contacting him. So I told him he should protect himself. Even when Matt came up with his ridiculous story, Ocelot still seemed to be oblivious of who the D truly was so I assumed he didn't know the detective yet and I said in the thread that the detective cannot be really this dumb. When Fred was whacked I was shocked because Fred is an experienced player so he must have contacted Ocelot. I asked "how come he was unprotected?", was it because Ocelot didn't know him or because Ocelot knew him and didn't protect him? Had it been the first case, well it would have been all Fred's fault for not having contacted Ocelot but it turned out that Ocelot knew so it was in my opinion his fault.
 

Völler

Always spot on
May 6, 2012
23,091
Well, it indeed depends on how many times you run into the same scenario in either one single game or in multiple iterations of the game. A few years ago, I used to play this game multiple times per day perhaps for two weeks. There I adopted the strategy of ALWAYS protecting the detective as the BG even though I didn't do any math to figure that out. Now, it's obviously not an always versus never kind of a strategy (it shouldn't, you should mix) but it's still close enough (66% versus 33%). I learned it by experience. What I ind really fascinating about game theory is that it is very well mapped onto the true and natural state of things.
But if you run into the problem several times in the same game, it's level thinking (what do they think I think etc...) and not math that should be decisive.
 

Hængebøffer

Senior Member
Jun 4, 2009
25,185
Well, it indeed depends on how many times you run into the same scenario in either one single game or in multiple iterations of the game. A few years ago, I used to play this game multiple times per day perhaps for two weeks. There I adopted the strategy of ALWAYS protecting the detective as the BG even though I didn't do any math to figure that out. Now, it's obviously not an always versus never kind of a strategy (it shouldn't, you should mix) but it's still close enough (66% versus 33%). I learned it by experience. What I ind really fascinating about game theory is that it is very well mapped onto the true and natural state of things.

- - - Updated - - -



That's not true. I stayed true to what I believed and said. We didn't know if Ocelot was contacted, and there was always the possibility of a desperate G contacting him. So I told him he should protect himself. Even when Matt came up with his ridiculous story, Ocelot still seemed to be oblivious of who the D truly was so I assumed he didn't know the detective yet and I said in the thread that the detective cannot be really this dumb. When Fred was whacked I was shocked because Fred is an experienced player so he must have contacted Ocelot. I asked "how come he was unprotected?", was it because Ocelot didn't know him or because Ocelot knew him and didn't protect him? Had it been the first case, well it would have been all Fred's fault for not having contacted Ocelot but it turned out that Ocelot knew so it was in my opinion his fault.
But you know it doesn't work like that. You have to be careful what you write. Just look at my brain fart when Ocelot protected himself. And you know the chance that the D contacted him was very likely.
 

king Ale

Senior Member
Oct 28, 2004
21,689
But you know it doesn't work like that. You have to be careful what you write. Just look at my brain fart when Ocelot protected himself. And you know the chance that the D contacted him was very likely.
It wasn't a brainfart, I'd make the same comment everytime I play. It was right after we got to know about Ocelot's role, I wanted him to be cautious.

I was a suspect because 1) Fred tends to overestimate my G game. 2) I defended you (you and Joker were my sure C's, if I was a G I'd definitely try to recruit either of you). 3) Matt said I'm the recruiter (how silly of people to not conclude that it makes me a C 1000 times more than it makes me a G). 4) Klin said I'm a C (so as I said, when a G says I'm a G then I'm a G. When another G says I'm a C then I'm again a G).

So I had to remind people of Jasper's mistake. Didn't work either.

One last thing: Cs should have noticed that after Fred, Ocelot was as clueless as everybody else.

- - - Updated - - -

Yeah, Hoori tried to be all smart and ended up goofin.
Show me one instance of my trying to be all smart.
 

Hængebøffer

Senior Member
Jun 4, 2009
25,185
It wasn't a brainfart, I'd make the same comment everytime I play. It was right after we got to know about Ocelot's role, I wanted him to be cautious.

I was a suspect because 1) Fred tends to overestimate my G game. 2) I defended you (you and Joker were my sure C's, if I was a G I'd definitely try to recruit either of you). 3) Matt said I'm the recruiter (how silly of people to not conclude that it makes me a C 1000 times more than it makes me a G). 4) Klin said I'm a C (so as I said, when a G says I'm a G then I'm a G. When another G says I'm a C then I'm again a G).

So I had to remind people of Jasper's mistake. Didn't work either.

One last thing: Cs should have noticed that after Fred, Ocelot was as clueless as everybody else.

- - - Updated - - -



Show me one instance of my trying to be all smart.
I didn't say you made a brainfart, but you know people is going to use something like that against you.
 

Users Who Are Viewing This Thread (Users: 0, Guests: 5)