[QUOTE=Mtgman]
This is a good point, but the strategy would be structured so that pro-town roles don’t stand out, and anti-town roles do. How exactly needs more defining, but I think it’s possible, especially in a game of this size. We’ve had the germs of one throughout the various games we’ve played on the SDMB. Some components of it are the lying low and then telling all behavior of the investigative roles(compare and contrast the back-breaking effect of the Oracle in Werewolf versus the high-profile(and rapidly dead) Detective in MII). The Beat Cop sowed tons of confusion in MII, and almost certainly lead to two townie deaths directly(unintentionally, but still directly). Since our Apprentice is essentially a beat cop, until either the Oracle finds him/her or the Oracle dies, the Apprentice should hang tight and try not to give themselves away, even moreso than the Oracle. Similarly, the Crusader should not reveal himself/herself if possible, and the town shouldn’t try to micromanage the Crusader without having a reliable read on who the Crusader should wack and why(such as an Oracle saying X is Cult). One of the fundamental truths of large Mafia games is that the scum can’t win in a reasonable amount of time(which is essential when working against detectives) without help from the town. They can’t reduce the town’s numbers fast enough unless they can sway lynches(subtly, but still, it’s a necessary part of scum strategy if they hope to reduce the townies to a manageable level before the detective fingers them all). So if we can define a general strategy for the town then someone trying to steer the votes towards a lynch without a strong case behind it would stand out. Basically the town agrees not to move without strong evidence. We resist the bandwagon effect because that allows a charismatic, or persuasive, Cultist to guide the dunks.
These are just germs of ideas, and I would need help to develop them, but I think there can be a uniting strategy. Any such strategy would almost certainly need to incorporate the above elements. If we want to discuss more specifics, here is a couple suggestions.
Firstly, don’t be afraid of a no dunk. Especially in the information-free early game. We’ll almost certainly take down a townie because there are enough scum right now to subtly guide dunk bandwagons. The first Days are big ones for the scum because they have a huge advantage. But, in a game with information roles, they can’t race without the town’s help. If the town doesn’t dunk, then we’re only down one(maybe two, depending on how the Crusader plays) person at the end of the first day/night cycle. This puts pretty much all our eggs in the Oracle/Apprentice basket, but somehow it seems more civilized. Let the cops find the scum and then let society take them out. If the scum find the cops first, then we revert to lawlessness, but it doesn’t have to be our first resort.
With potentially two Oracles, this game is a double threat to the Cult. Imagine if the Apprentice roleclaimed now, and the Priest started protecting him/her and the Disciple starts protecting the Priest, since he/she knows who the Priest is. The Monks all claim, reducing the number pool of investigation targets for the Apprentice. Then the town votes no dunk and the Crusader skips his night kill. The Oracle could investigate the claimed Apprentice night 1, and then on Day 2 we’ve got a 60% confidence level in what the Apprentice pronounces(and of course the Apprentice simply says “Name - Pro-Town” or “Name - Anti-Town”, but we still don’t dunk. Night 2 the Crusader is due to kill, with 100% certainty. The protective roles won’t block this kill because they’re protecting the Apprentice and/or Priest. So the Crusader kills the person fingered by the Apprentice(if they’re a Non-Believer) or whoever they feel is the most scummy, if the Apprentice uncovered town. But basically what we’re doing is buying time for the Oracle to train the Apprentice and then break the Cult slowly with deliberate investigation versus mob mentality. We’d have probably six full days worth of time for the investigations if we don’t dunk and the Crusader sticks to an every-other-night schedule. This is more than enough time for our investigators to find enough names to break the Cult’s back.
Day 1 - No Dunk
Night 1 - Town goes to 22(if there are seven Cultists) or 21(if there are eight) a 2.5/22 or 2.5/23 chance of losing either the Oracle, Priest, or Disciple(the .5 is because the Disciple is protecting the Priest and his protection may fail). Crusader doesn’t kill.
Day 2 - No Dunk - Town gets a name/affiliation pair from the Apprentice. 20/28(or 20 minus number of Monks / 28 - number of Monks if they Monks are willing to go along) are the odds that we get a pro-town name, and we’ve got 60% certainty on it.
Night 2 - Town goes to 21, 20 or 19(depending on how many cultists and if the Crusader hit town or Cult) 5/20 chance the Priest, Disciple, or Oracle bites it, because there are two kills and the Crusader could hit any of them too. Crusader can’t risk doubling up on the protected Apprentice because that would negate their protection.
Day 3 - Town gets another Name - affiliation pair from the Apprentice. 18/26 chance of being town, 70% certainty. No dunk.
Night 3 - Town goes to 20, 19 or 18. 2.5/19 chance the Priest, Disciple, or Oracle bites it.
Day 4 - Town gets another Name - affiliation pair. 17/25 chance of being town, 80% certainty. No dunk.
Night 4 - Town goes to 18, 17, or 16. 5/18 chance the Priest, Disciple, or Oracle bites it.
Day 5 - Town gets another Name - affiliation pair. 16/25 chance of being town, 80% certainty. No dunk.
Night 5 - Town goes to 17, 16, or 15. 2.5/16 chance the Priest, Disciple, or Oracle bites it.
Day 6 - Town gets another Name - affiliation pair. 16/25 chance of being town, 80% certainty.
Now the Apprentice has investigated five times, with 60, 70, 80, 80, and 80% accuracy. Plus he/she knows who the Oracle is. If the Monks jump in and deflect potential investigations away from their fellow Monks, then assuming nearly worst case scenario where the Crusader has killed town every time and we’ve never found Cult with an investigation, we now have Monks + four nights of Oracle results + the Apprentice’s investigations. That’s probably about sixteen players, out of the twenty-three which would be alive on Day 6 under this scenario, that we know the orientation of with reasonable confidence(four monks, four Oracle results, the Apprentice(who knows the Oracle and can validate his/her claim) and the five Apprentice investigations. That should be enough to make a Cult win impossible. If we find Cult along the way(and we probably will) then we can eliminate them via the Crusader. That’s roughly twice as many known, and trustable, pro-town people as there might be Cultists left. From there we just have to find the Prophet, who can’t be investigated sucessfully, and possibly the convert. Worst case scenario at that point, start dunking non-Monks(since they can’t be converted, they’re trustworthy the whole way through.
So by minimizing the number of non-Cult kills we can reduce the chances of the Oracle and Apprentice being collateral damage.
The obvious way to respond to this for the Cult is to pretend to be the Apprentice and string us along while killing everyone at night, but if there is a way to pass encrypted data between the Apprentice and the Oracle, then this doesn’t work. Second best response would be for them to kill Monks, making the end game more difficult because we’ll have a bunch of trusted townies with a Prophet(and probably a convert) hiding among them, reading as “Believer”, but even that would be two Cultists among 16 townies, and those are long odds for the Cult.
There are lots of ways this could be deviated from, of course. Still, we’re humans and that’s another word for adaptable. If we lost the Oracle early we’d need to continue it for a couple more Days(at that point our Apprentice becomes an Oracle, or near enough) to get an overwhelming number of confirmed Town. Losing the Priest would hurt, having him/her protected by the Disciple should help some. If we find Non-Believers with the boosted Apprentice we could let the Crusader take them out, so we’re less likely to take out the Oracle, Priest, or Disciple with the pro-town night kill. Make the Cultists work for the power role kills.
And now that I’ve written all this I’m worried about my participation in this game. I’m a puzzle solver by nature and I’m not sure I can turn that part of me off to play the game versus try to solve it. The “use an algorithm” approach I mentioned a couple pages back I think would work well mathematically for the town, but it would kill the fun, and Blaster Master might never forgive me if I “broke” his game. The approach I just laid out would be more interactive and still have some strategic depth, but not nearly as much. On top of it all, I’m pretty sure I owe SnakesCatLady a dose of headache medicine by now.My first plan, using a transparent algorithm to generate random votes, would have negated the ability of the scum to manipulate their votes. So no one is ever “trusted” no matter how well they assimilate. Anyone can end up in the dunking chair, on any day. Monks can claim their way out of it, as can the Priest and Oracle(and whichever one ends revealed had better be the Priest’s target for protection for a while). I’m not arguing for this plan because it’s no fun, but it would almost certainly work. As for our epistemology, I never said we should try to do it without the assistance of power roles. They’re critical to our success. We should have a way to help them, like I mention at the beginning of this post. Let them lie low and do their jobs. Don’t allow bandwagons because that’s how Cultists end up helping with dunking townies.
Enjoy,
Steven
[/QUOTE]
First of all: Ugh, that made my head hurt.
Second: That’s no fun! Sitting back and letting a couple of people do all the work isn’t in the spirit of the game.
Third: I think there are other roles that you have forgotten that can screw with the whole scenario, namely the Alchemist and the Psychopath. If the Alchemist inadvertantly blocks the Oracle or Apprentice, your plan gets all screwed up. Or if the Crusader or Cultists activate the Psychopath, same thing.
Fourth: The Oracle or Apprentice (or Nairu forbid) both could have already given themselves away in the kerfluffle of discussion surrounding them. If so, and the Cultists have a bead on them already, then the whole plan is moot.
Lastly: I don’t think voting No-Dunk is good for the town. There are no voting records or associations to examine later in the endgame.
Me, I’m worried about all of the seeming major differences in a opinion this game has right now. Seems at least four or five separate ones. And to a player like me, who always feels the need to analyze everything, it gets me paranoid.
