• We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.

TheArchMede

General
11 Badges
Jun 16, 2004
2.410
137
  • Europa Universalis III
  • Europa Universalis III: Chronicles
  • Europa Universalis III Complete
  • Europa Universalis III Complete
  • Europa Universalis III Complete
  • Europa Universalis: Rome
  • Rome Gold
  • Victoria 2
  • Victoria 2: A House Divided
  • 500k Club
  • Europa Universalis: Rome Collectors Edition
Or as Einstein put it "Gott wuerfelt nicht".

I have seen a lot of dice rolling by the GM during the last couple of games I played, and I think the games were worse for it.

Dice rolling is a long established part of the big game so I think this belongs in a thread about dice rolling, rather than in a thread on a particular game.

The essence of the game is that players have to act on incomplete information, but this should be because they can't trust one another not to lie and backstab, not because the GM is rolling dice. Werewolf isn't backgammon where the skill lies in calculating the probabilities and how you manoever with the roll of the dice. It is about whether you can lie, con and backstab the other players, or catch them out in their attempts to lie con and backstab you. Every time the GM rolls the dice during the game, it reduces the opportunities players have to lie and cheat one another. The GM should be creating opportunities for players to get things wrong by misjudging whether other players are telling the truth or lying. He should not be taking them away by deciding who lives and who dies on the roll of a die.

It is the difference between chaos and randomness. In chaos, if a player knows exactly what the situation is, the outcome is completely predictable, but if they are a little bit wrong, the outcome ends up completely different. Randomness is when the player cant know the outcome, even if they read the other players perfectly, because the outcome depends on the die roll the GM makes.

From the outside, for a non-player, or the GM, or a player that got themselves killed early, there is little to see of the difference. Both random and chaotic games look unpredictable and good. For the player that has survived into the last couple of days the difference is very plain. A random game is insipid, it lacks bite, win or lose there is a vague feeling of disappointment about it. In a word it is BAD. A chaotic game is very different. Is that a wolf pretending to be a dodgy villager so that he doesn't get lynched, or a villager acting dodgy so he doesn't get hunted? There is a whole game full of information and you have to do your best to read it and decide which it is. You misread something crucial and you lose, or you let slip something that looks like a giveaway but is a fake and the opposition buys it and you win. Either way it is exciting, the game has a bite to it. In a word it is EVIL. Werewolf games ought to be EVIL, but of late they have been BAD and worse than BAD.

When the GM is rolling dice but the players don't know it, it is even worse. As a player you are then making decisions based on what you think the other players were thinking, but actually what happened was nothing to do with the players, it was the dice the GM was rolling. This is truly terrible for the game. The GM should never, ever under any circumstances whatsoever, decide something during the game by a die roll without informing the players in advance that is what is happening.

An example.
A GM decides that there should be some unpredictability about which pack hunts when. This is good. It might even be EVIL if it was set up right. The GM decides to do it by rolling a die every night. This is BAD. The GM doesn't tell the players that this is what is happening. This is worse than BAD.

The EVIL way to set it up would be to have the packs play a game of evens and odds. Each pack sends a number to the GM with its hunt order. If the sum of the two numbers is even, then the even pack gets the hunt. If it is odd, then the odd pack gets the hunt. The packs can try and outguess each other on whether to put an even or an odd number on their hunt order. They can try and do a deal so they each of them hunts on alternate nights. They can try and do a deal and then renege on it and to get an extra hunt at the expense of the other pack. The GM announces that this is how hunts are being organised to everyone. The villagers can look at what an outed wolf being bandwagoned puts in the thread and draw conclusions about what sort of deals the wolfs might be trying to make or break with one another. All this adds to the EVIL possibilities of the game and can't happen if the BAD GM rolls dice instead.
 
Personally, I like randomness, if it is made clear where randomization is being used.
I think it is not overused yet (maybe the random hunts are a bit too much), but adding much more to it would be bad.
But that said, I am a fan of randomness, so its a bit biased opinion. But removing it completely would be as bad as overusing it IMHO.

Oh, and I really like your idea about the odd/even game of the wolves, that is definitely an idea worth trying! Adding game theory elements to the game like this would certainly be a good thing!
 
TAM, I completely agree with you on all points, and your idea is quite nice. I would like to see that in play once.
 
I agree on all points so far, but what do you think of the spies dodgy scan returns, or a percentage chance to fail for a scanner?

Those are chances of whether someone succesfully scans or not. Consider for example the 10%-each day rule for the witnesses. That is simply a chance that improves, thereby placing the witness for a dilemma when to scan. However, that is not something that is crucial for the village, nor something that the village needs to know to base their information on. A scanner scanned or not, and the information is shared with someone else or not. Whether a scanned failed is not relevant.

And more important, those rules are known. We know there are chances for succesful scans and stuff.

However, hunts are of an all different matter, see TAM arguments.

Honestly I don't see much difference. The evens odds idea essentially works out as a di roll anyway so long as neither side is choosing terribly.

That is not the GM who interferes, but the players themselves. It is up to the players themselves to use the system or not. The GM has no part in the process thereafter, except as an algorithmic computer (as in: to compute)
 
I don't see how it would be different if the packs choose a number. It is still all random.
 

Because the wolves can observe a pattern, they can know whether their hunt was succesful or not. They might even try to contact (via-via) the other pack and try to coordinate their number-choosing. And they might break the deal, or not. It makes for an interesting added play for the wolves (or the messenger, who can "screw" messages as well)
 
Because the wolves can observe a pattern, they can know whether their hunt was succesful or not. They might even try to contact (via-via) the other pack and try to coordinate their number-choosing. And they might break the deal, or not. It makes for an interesting added play for the wolves (or the messenger, who can "screw" messages as well)
How would the wolves get in contact?

EDIT: And if they do, why would they negotiate their hunts and not kill the messenger?
 
How would the wolves get in contact?

They might not need to get in direct contact, but it can be done through, for instance, the sorceror, or a cultist, or maybe some other role.

EDIT: Who know what other stuff can be done by such a messenger. I don't know if it all works out nice, but it adds a nice twist, but with no part for the GM. Everything that happens is done by the players themselves.
 
They might not need to get in direct contact, but it can be done through, for instance, the sorceror, or a cultist, or maybe some other role.

EDIT: Who know what other stuff can be done by such a messenger. I don't know if it all works out nice, but it adds a nice twist, but with no part for the GM. Everything that happens is done by the players themselves.
If I was a wolf in a pack and the sorc came and asked us if we want to coordinate hunts, I would not. I just don't see that as a viable option. Wolves are not supposed to work together, they are supposed to rip each other apart, and if one pack is winning most of these "random" hunts they would not be willing to negotiate.
 
Because the wolves can observe a pattern, they can know whether their hunt was succesful or not. They might even try to contact (via-via) the other pack and try to coordinate their number-choosing. And they might break the deal, or not. It makes for an interesting added play for the wolves (or the messenger, who can "screw" messages as well)

But if they knew the other wolves, wouldn't they just out them?

Those are chances of whether someone succesfully scans or not. Consider for example the 10%-each day rule for the witnesses. That is simply a chance that improves, thereby placing the witness for a dilemma when to scan. However, that is not something that is crucial for the village, nor something that the village needs to know to base their information on. A scanner scanned or not, and the information is shared with someone else or not. Whether a scanned failed is not relevant.

And more important, those rules are known. We know there are chances for succesful scans and stuff.


By "fail", I meant a bad return, not no return at all.
 
If I was a wolf in a pack and the sorc came and asked us if we want to coordinate hunts, I would not. I just don't see that as a viable option. Wolves are not supposed to work together, they are supposed to rip each other apart, and if one pack is winning most of these "random" hunts they would not be willing to negotiate.

Villagers are not supposed to become a wolf and later change back.

But I was merely thinking along. It's TAM's original idea, so he might have thought about it deeper than I did, and might have better answers.
 
Packs A and B are hunting. A hunts if sum is even, B if odd.
It is a non-cooperative game, with the following pay-off matrix:

Code:
      |           B
      |  0                 1
-------------------------------
    0 | (1,0)           (0,1)
      |
A     |
      |
    1 | (0,1)           (1,0)
      |

Which has the advantage of having no equilibrium points

EDIT: cretin forum can't display things properly, I'll try to arrange it somehow

EDIT2: I hope it is clear what I meant, couldn't do it more nicely, sorry

EDIT3: It now looks nice, thanks for the tip Rysz! :)
 
Last edited:
Villagers are not supposed to become a wolf and later change back.

But I was merely thinking along. It's TAM's original idea, so he might have thought about it deeper than I did, and might have better answers.

I know, hence why I'm changing the thief trait as a one time thing.
 
Packs A and B are hunting. A hunts if sum is even, B if odd.
It is a non-cooperative game, with the following pay-off matrix:
Code:
      |           B
      |  0                 1
-------------------------------
    0 | (1,0)           (0,1)
      |
A     |
      |
    1 | (0,1)           (1,0)
      |

Which has the advantage of having no equilibrium points
Lay-out works better with CODE
 
If I was a wolf in a pack and the sorc came and asked us if we want to coordinate hunts, I would not. I just don't see that as a viable option. Wolves are not supposed to work together, they are supposed to rip each other apart, and if one pack is winning most of these "random" hunts they would not be willing to negotiate.
I certainly wouldn't trust anyone offering to do it, if they're not a villager trying to infiltrate then they could be a cultist from the other team,

And its not in a sorcs interest to keep both packs alive, he would want to destroy one and favour the the other, even more so in this case as he could still make sure there was a hunt every night
If a pack no longer exists, what number, if any, does it give?