AFAIK that's how SA is usually implemented, with a % chance of knowing who scanned you. Otherwise it could be certain, but only give it to one wolf. Which for 12/4/1 setup gives the same "odds" of scanning a wolf ending badly for the seer.
-
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.
AFAIK that's how SA is usually implemented, with a % chance of knowing who scanned you. Otherwise it could be certain, but only give it to one wolf. Which for 12/4/1 setup gives the same "odds" of scanning a wolf ending badly for the seer.
Basically the idea is that in an IRL game, the seer has to swing the vote towards the scanned wolf or away from the scanned villager openly without giving away that he is seer. SA would represent the chance of the wolves catching him. I'd probably shift the part where the wolves get told though as part of this simulation. If a seer scanned a wolf night 0, the wolves would get told after the night 1 deadline - as if they'd noticed the seer acting strangely on day one. If a seer also scanned a wolf on night one, well he better take a gamble on somebody quickly. He'd have already done more for his side than most IRL game seers, which is where the points system last page comes from.
But the wolves must use their brains to figure the seer, not get a free pass...
The only way I could see that implemented is if the wolves are unable to hunt the Seer. But I still don't like it, as it can be exploited.
The only way I could see that implemented is if the wolves are unable to hunt the Seer. But I still don't like it, as it can be exploited.
The problem is the standard 12/4/1 setup is already a % chance. If the wolves hunt the seer, they may then continue to play and possibly win if skilled enough. If they fail to hunt the seer, they lose. Fullstop. Unlike IRL werewolf, here the seer is free to communicate secretly, creating massively overpowered JLs. We need some way of stopping this, so that it can become about analysis - because right now too many players are taking free rides.
Taking away the seer is one way of doing this. But this may favour the wolves too much, as there is no body language on the internet. Throwing in a GA might help protect key players, but there's still the problem of the wolves refusing to act as a pack having no fear of discovery, making analysis impossible. So the final option to try is giving them a way of finding the seer, thus making the villagers more aware of their need to participate rather than sit around. I would prefer this was done with sorcery, but SA might better reflect face to face werewolf.
Wolves don't need to hunt the seer, just let the JL lynch the seer for them.
Seriously, though, everyone has the advantage of communication in an online game. Some players just know how to use it better than others. Most wolf packs I'm in either choose hunts randomly or never talk about it. The two games that I managed to win as a wolf, we had a lot of chatter going on planning and scheming. Most of the games I lose, there's talking but no real communication. One person goes "Hunt person A". A couple hours later someone else comes on and says "no hunt person B"
That should only be a % chance in the first couple nights. Decent wolves ought to figure out seers, and vice-versa. Lite demands analytical and behavioral analysis capabilities more than luck.
This is getting more complicated by the post. Remember j-L's ruling: Lite is simpler, few role, few traits.
Which is precisely what I proposed to start with
It's my favoured option if just wolves+villagers doesn't work, and if it still doesn't I'd prefer sorcery of some form rather than SA. I'm just mentioning it as another possibility.
There have been hundreds of games under a setup that was considered balanced for years. After a few months some new people are suggesting that the system was terrible.
Then again, maybe all those great old players were idiots who didn't know how to balance a game? I wouldn't know either way. I'm an idiot.
WW MATHS
A game with 12 villagers and 4 wolves gives the wolves near certainty of winning, assuming ideal play.
Upon the inclusion of a seer, those chances are brought back into a linear model. We account for how many villagers learn about each other as well as how many wolves get caught.
Statistically, if the seer dies right at the beginning, the wolves will, once again, have near certainty of winning, assuming ideal play.
Statistically, if the seer never dies, then the chance of winning for the village is dependent on how many of those seer scans got hunted. If none of them get hunted, the village has near certainty of winning, assuming ideal play. If all of them get hunted, the wolves have near certainty of winning.
If you want to look at the actual maths insanity, check the link and read the PDF. It's very VERY long and detailed.
It should also be noted that, in reality as opposed to "ideally". The chances of wolves winning goes down with less ideal play, as they begin to give themselves away.
So, what? Is the standard LITE model unbalanced? Not really. I'd say it is simply a delicate setup, that has the possibility to swing one way or another based on random chance, and the possibility to swing the other way based on a lack of ideal play.
Out of the last 10 LITE games (From Cluedo to Wimbledon, counting "Original Werewolf" as it would have resulted in a baddie victory). We have had 6 wins by the village and 4 by the wolves.
If you think the current system is unbalanced, then follow that link, read the proofs, and devise your own balanced setup that's under 20 players, without having dramatic shifts based entirely on luck.
Then again, maybe all those great old players were idiots who didn't know how to balance a game? I wouldn't know either way. I'm an idiot.
WW MATHS
A game with 12 villagers and 4 wolves gives the wolves near certainty of winning, assuming ideal play.
Upon the inclusion of a seer, those chances are brought back into a linear model. We account for how many villagers learn about each other as well as how many wolves get caught.
Statistically, if the seer dies right at the beginning, the wolves will, once again, have near certainty of winning, assuming ideal play.
Statistically, if the seer never dies, then the chance of winning for the village is dependent on how many of those seer scans got hunted. If none of them get hunted, the village has near certainty of winning, assuming ideal play. If all of them get hunted, the wolves have near certainty of winning.
If you want to look at the actual maths insanity, check the link and read the PDF. It's very VERY long and detailed.
It should also be noted that, in reality as opposed to "ideally". The chances of wolves winning goes down with less ideal play, as they begin to give themselves away.
So, what? Is the standard LITE model unbalanced? Not really. I'd say it is simply a delicate setup, that has the possibility to swing one way or another based on random chance, and the possibility to swing the other way based on a lack of ideal play.
Out of the last 10 LITE games (From Cluedo to Wimbledon, counting "Original Werewolf" as it would have resulted in a baddie victory). We have had 6 wins by the village and 4 by the wolves.
If you think the current system is unbalanced, then follow that link, read the proofs, and devise your own balanced setup that's under 20 players, without having dramatic shifts based entirely on luck.
I'm fairly certain that study does not take into account that here the seer is allowed to communicate privately. This massively unbalances the game. If we came to a gentleman's agreement that only wolves were allowed to PM, then it would be balanced under the old setup, as shown in that study and elsewhere. As it stands with the seer PMing his JL, the village has a 2:1 winning ratio, with all wolf wins coming after the seer was hunted prior to too much damage being done.
Actually, I just looked at their assumptions... and in no way does it resemble Mafia or Werewolf in any form that I've known it.
No problems here, unless they're trying to imply anonymity with there sudden random use of cryptography. Or the need for simultaneous posting.
Secret voting, and all players can communicate secretly but everyone knows how many are sent or received by each player? Relevance = 0.
No problems here, unless they're trying to imply anonymity with there sudden random use of cryptography. Or the need for simultaneous posting.
Secret voting, and all players can communicate secretly but everyone knows how many are sent or received by each player? Relevance = 0.
I disagree. The player who starts out with a few and goes to many would be the seer, while the wolves would maintain a steady amount throughout the game.
That would be a lot of information, and an interesting element. The wolves would have to try to not to communicate until other people are.
Of course if you meant relevance to us, then yea I agree.
Relevance to whether the existing setup is balanced = 0
not
Relevance to ways to make the existing setup balanced = 0
Which setup? Because I was pretty sure we'd already established 12/4/1 was 2:1 in favour of the village, which 5 times out of 6 was actually the JL rather than clever villagers.
No, we had not, as you are considering a low sample, in which the skill displayed by the wolves wildly varied.
It is just a little out of date
From what I've seen, most of the werewolf wins have come in entirely different setups to the 1/12/4 or 1/9/3 we've been using lately. It doesn't help that a lot of the old games haven't got the results outlined clearly in terms of role setup and players left, so it's taking time to look through them.
Edit: The last wolf victory listed in the past games thread with a setup like the current one, CLXXXX (why not CXC?), was after the seer was lynched day one, continuing the trend. Continuing to look...
Edit: CLXXVIII has a wolf win with the current setup, with a live seer, after a failed attempt to create a tie. 14 months ago was the last win against a live seer in this setup
I'm not necessarily saying it's unbalanced, although it could be, but I don't like how much weight is being placed on JLs. It's making players lazy.
Edit: CLXXIII is the next baddie win in the current setup. Stalingrad, seer hunted night 2.
Edit: CLXXII snoopdogg the snoopwolf only won because he hunted the seer the night he was scanned, and managed to create a tie of villagers.
Edit: CLXX a Stalingrad, seer hunted night 3.
I'm fairly happy the pattern is holding by now.
Coming up 15 months since the last time the wolves won without getting lucky on the seer early under a 12/4/1 or 9/3/1.
It seems only experimental games are played at the moment.
We have seen two Meso games : Both failures
We see a big game with subversion : Will be unplayable and uninteresting for the villagers, but he, it's a little bit more interesting for the cultists ...
We see a lite game without seer, with odds which favor the baddies greatly.
Unfortunately we don't see many classical games anymore, with set-ups like recently johho's balanced game.
No, it needs to be experimental nowadays, and the more drastic the better. And if it results in unbalanced or bad gameplay then that doesn't seem to matter ...
Ok, that's not my cup of tea ...
I wish you all the best.
We have seen two Meso games : Both failures
We see a big game with subversion : Will be unplayable and uninteresting for the villagers, but he, it's a little bit more interesting for the cultists ...
We see a lite game without seer, with odds which favor the baddies greatly.
Unfortunately we don't see many classical games anymore, with set-ups like recently johho's balanced game.
No, it needs to be experimental nowadays, and the more drastic the better. And if it results in unbalanced or bad gameplay then that doesn't seem to matter ...
Ok, that's not my cup of tea ...
I wish you all the best.