Mafia V: The Cult of Sekham [Archive] - Page 2

DiggitCamara

06-26-2007, 02:48 PM

Eh, I'm still not convinced. In the same discussion in the Pirates game, you two took the same positions, but you were scum and Blaster was not.

Our goal is to get people posting. To build up some relationships. So when one of us turns up eyeless in the morning, we'll have something to go on.

Discussing strategy is good, as long as it doesn't develop into real divisions.

Picking apart other peoples' posts is an easy way for scum to start bandwagons. Using trivial nits to choose votes is no more accountable than random--just as easy for scum to hide behind.

Basing suspicions/votes on what's in the game doesn't work well on the first Day. That encourages lurking and puts an undue burden on the talkative. I think we need to get everyone in here and talking. Voting aggressively (even randomly) does that. I'm at least willing to take some heat to accomplish that.

(snip)You're suspicion back on me would usually seem scummy to me. That's not a good way to dispel suspicion on yourself (attacking your attacker). But, I think in your case it's a the reaction of a newbie Townie. Be more careful in the future.
Overall, I'd have to agree.

Only on the suspicion part I wouldn't. In M3 most of the time, when I was pointed at and/or when I pointed at someone (incidentally, unvote ComeToTheDarkSideWeHaveCookies ) I got a pretty aggressive reaction. Most of the time I was pointing squarely at a townie. I think storyteller said something to that effect during the post-game discussions.

Pasta

06-26-2007, 02:48 PM

I tend to agree here with NAF, although I also agree that nit-based voting can be just as random but is often more scumvantageous.

The rules of M5 are complex enough that we should be discussing strategy. Strategy discussions are where the most content will develop.

ComeToTheDarkSideWeHaveCookies

06-26-2007, 02:49 PM

If memory serves, I was at least partially involved in that M3 "early game" random voting conversation as well, as a town player, and I leaned more towards Blaster's views both then and now.

Ymmv on all of what I'm about to say.

Mine is not an "all or nothing" perspective. There are subtleties that need to be examined.

Random voting vs random FOSing: I personally have a higher comfort level with random FOSing during the first Day than random voting. Why? Because a FOS cannot be the gateway drug to a bandwagon, whereas a vote can be.

Ice-breaking: Poking sticks at people prompts them to post and gives us all fodder for actual analysis. As of typing this post, we've had at least one allegedly random vote, and I guess what you might call an allegedly traditional vote ( :p at Diggit), neither of which bother me much, aside from the bandwagon potential that they carry.

Always remember to interpret all things with an extra dose of skepticism: It is indeed possible and probable that the scum will be voting and FOSing and trying to say that such things were random. It is also possible and probable that the scum will pounce on opportunities provided by allegedly random selections made by others, whether they were legitimately random or not.

All that being said, I think that the random topic discussion is healthy, as is (informed) use and interpretation of truly random votes by those who are not confident in any other selection method.

NAF1138

06-26-2007, 02:53 PM

Eh, I'm still not convinced. In the same discussion in the Pirates game, you two took the same positions, but you were scum and Blaster was not.

Well, I thought enough of myself at the start of that game that I figured the best way to hide would be to try and mimic my play in M3 as closely as possible.

Our goal is to get people posting. To build up some relationships. So when one of us turns up eyeless in the morning, we'll have something to go on.

This is my point. No we won't, not if all we have been doing is random voting. Odds are we are going to kill an innocent today. That's just the odds in this game. WW1 was the only game I have ever seen where scum got killed right off the bat. But that death will be meaningless if we just go around random voting people. And the person who is killed tonight is going to yeild us no additional info either if all everyone has been doing is random voting.

Discussing strategy is good, as long as it doesn't develop into real divisions. Agreed

Picking apart other peoples' posts is an easy way for scum to start bandwagons. Using trivial nits to choose votes is no more accountable than random--just as easy for scum to hide behind.
Basing suspicions/votes on what's in the game doesn't work well on the first Day.

Clearly I disagree. But I think as far as this goes we will just have to agree to disagree. I laid out my argument, you laid out yours. I am interested in what others think about what we have said.

That encourages lurking and puts an undue burden on the talkative. Then vote for lurkers, that is a legitimate vote at this stage.

I think we need to get everyone in here and talking. Voting aggressively (even randomly) does that.

I agree with all of that except the bit about random voting

I'm at least willing to take some heat to accomplish that.

Well, you will find me arguing against your line of thinking, but I am not going to give you heat just because we disagree.

NAF1138

06-26-2007, 02:57 PM

sachertorte

06-26-2007, 02:57 PM

On Random Voting: I would say that random voting is probably more useful than chattering about the merits of random voting. I can't see how random voting is all that hurtful. Possibly useless, yes, but not all that much of a cover for scum. now if someone starts random voting on day three or even during the later part of day one, then, yeah bad stuff.

Strategy discussions are where the most content will develop.
Does everyone agree with Pasta? I love discussing strategy and game mechanics, but I don't want to pull a chrisk from M2.

MadTheSwine

06-26-2007, 02:58 PM

Trouble is, many of us have already give places like mafiascum.net a thorough read-through and are well aware what they call "accepted scum tell". The undoing of a lot of it is that many of them are either crap, or would be rookie mistakes. Recall the "3rd vote" bit in M4, as an example.

Yeah, we've got precious little to go on at the moment, but we're less than one hour into a Day that doesn't end for five more days. That's a lot of time for people to make posts that are actually incriminating.

Like I said "scum tells aren't always scum tells",but I have seen the ones ring true that I have pointed out,especially with zuma in M4 when I FOSed him twice with "crappy" scum tells, before he role claimed as me .

Using the 3rd vote as an example isn't a good one, since I was the one making the 3rd vote (and knew I was making it) and felt confident I could defend that action, if I had to.(Not to mention, it did work in M1,unfortunately)

So no...I do not think they are "crap" and think some of these posts are already incriminating.

MHaye

06-26-2007, 02:58 PM

Random voting vs random FOSing: I personally have a higher comfort level with random FOSing during the first Day than random voting. Why? Because a FOS cannot be the gateway drug to a bandwagon, whereas a vote can be.I agree with this; I always feel that random voting is a mistake. As is pressure-voting - voting someone just to get them to say something specific. I never do them, and neither will I respond to them, even if it means I get lynched (or, in this case, drowned). I'll go along with FOSs though, because they can't be turned into a bandwagon.

The sheer size of the game offers some protection against bandwagons, but that does not change my opinion of random votes. Pleonast is looking slightly dirty in my eyes now - but then he seems to do this every game, so I should be used to it by now.

Still, there's plenty of time to consider action and reaction before voting.

Blaster Master

06-26-2007, 03:03 PM

Will all roles listed be present?

There will be at least one of every role listed.

Blaster Master

06-26-2007, 03:06 PM

All, I have started the Forbidden Thread (http://boards.straightdope.com/sdmb/showthread.php?t=426431). If you are alive, you are on your honor to not read. All dead, observers, and subs are welcome to comment as they feel appropriate.

Scuba_Ben

06-26-2007, 03:08 PM

Does everyone agree with Pasta? I love discussing strategy and game mechanics, but I don't want to pull a chrisk from M2.What happened in M2? I didn't follow that game; would you please post a recap?

sachertorte

06-26-2007, 03:16 PM

Speaking of voting, something I noticed in M3 that was driving me bonkers was the very low vote counts that were getting people lynched. Lack of consensus is bad. I don't know how we can force ourselves to reach a consensus in only 5 days, but letting a relatively small number (25%!) of votes determine the lynch was very bad for the town. Everyone needs to vote even if 'unsure,' otherwise we risk giving scum easy cover. So I state right now, I'll suspect anyone who gets to the end of the day with no registered vote. We still may end up with relatively few determining the dunk, but I can't see a way around that. Unless we as a town agree to chatter and discuss and cajole so that we can reach a true majority on each vote.

sachertorte

06-26-2007, 03:19 PM

What happened in M2? I didn't follow that game; would you please post a recap?

Oh, chrisk posted a strategy suggesting everyone post results of fake investigations, so that the real cop could post info in disguise. The strategy was flawed.

I have an idea I'd like to discuss, that could be flawed, but won't damage the town simply by discussing it.

Idle Thoughts

06-26-2007, 03:19 PM

As per usual, this topic always surprises me with how much I have to catch up on when I get on in the day.

NAF1138

06-26-2007, 03:19 PM

On Random Voting: I would say that random voting is probably more useful than chattering about the merits of random voting. I can't see how random voting is all that hurtful. Possibly useless, yes, but not all that much of a cover for scum. now if someone starts random voting on day three or even during the later part of day one, then, yeah bad stuff.

Does everyone agree with Pasta? I love discussing strategy and game mechanics, but I don't want to pull a chrisk from M2.

Um, so...how is talking about methods used for determining who gets a vote not a discussion of strategy?

MadTheSwine

06-26-2007, 03:23 PM

As per usual, this topic always surprises me with how much I have to catch up on when I get on in the day.

WHERE HAVE YOU BEEN? Lurking????

Scuba_Ben

06-26-2007, 03:23 PM

Oh, chrisk posted a strategy suggesting everyone post results of fake investigations, so that the real cop could post info in disguise. The strategy was flawed.It seems to me that the flaws are (1) the first person to post is almost certainly NOT the cop/detective/Oracle/Zadoc, and (2) nobody can tell which "investigation" was legit until the cop has been killed off. Do I have that right?

Queuing

06-26-2007, 03:25 PM

Re Random Voting/Unvoting/FOSing:

I don't really see the harm in random voting the first day. That said I really don't understand why someone would random vote only to later recant said vote in the very same real life day. What is that going to accomplish? The stated defense for random voting is getting people to talk as they are afraid that they will die. So why unvote so quickly?

HazelNutCoffee

06-26-2007, 03:25 PM

Gee... you disagree with me on the fact that I thought you made a scum tell? Imagine that.
I didn't say we didn't need to figure out how many N-B's and Cultists there are either,I said it was an accepted scum tell to bring it up.As far as figuring out how many there are..well,everyone can make a fairly educated guess, discussing it means nothing and gets us nowhere, as only the Cultists know for sure.
So on top of your original tell,I ain't to keen on revenge votes or word twisting either.
But Clockwork Jackal didn't make a revenge vote. All he said was that he was inclined to FOS you for saying that speculating on the distribution of players is a scum tell, which seemed to imply that you were discouraging such speculation as unnecessary. (I see that's not what you meant, but it initially appeared that way.) He didn't even point an actual FOS your way. If anything, it would seem that your vote for him is something of an over-reaction.

I do agree that discussing the distribution of players is not very fruitful this early on in the game, which is why I will withhold pointing a FOS at you for now.

sachertorte

06-26-2007, 03:26 PM

I should have quoted more of Pasta's post to be more clear.The rules of M5 are complex enough that we should be discussing strategy. Strategy discussions are where the most content will develop.
I'm talking analysis of the rules and game mechanics and exploiting them to the town's advantage. I don't mean to distract from the scum hunting and voting fun, but I think it is a worthwhile discussion. If everyone else disagrees, I'll let it go.

MadTheSwine

06-26-2007, 03:34 PM

But Clockwork Jackal didn't make a revenge vote. All he said was that he was inclined to FOS you for saying that speculating on the distribution of players is a scum tell, which seemed to imply that you were discouraging such speculation as unnecessary. (I see that's not what you meant, but it initially appeared that way.) He didn't even point an actual FOS your way. If anything, it would seem that your vote for him is something of an over-reaction.

I do agree that discussing the distribution of players is not very fruitful this early on in the game, which is why I will withhold pointing a FOS at you for now.

You are right.My apologies for saying revenge vote when it was just a revenge FOS.
I am satisfied with my vote, for now,however.

Fretful Porpentine

06-26-2007, 03:35 PM

Well, not to TOTALLY meta game here, but I think I can help with this one. Blaster said that he ran his numbers through my formula when he was putting the game together and that things lined up with that fairly well.

I am not good enough with numbers to back compute, but this is what I gave to him. Maybe someone else can look at it and figure out a general idea of how many people there are in each group.
Well, I'll take a shot at it. (Usual disclaimers apply, degree is in English, put too much faith in this at your own risk.)
The first thing you want to do is make sure that as close to 1/2 of your population is vanilla as possible.

The ratio I was told after this is, a doctor or similar type of character is worth 1 scum. A detective or similar is worth 2. A GF type character -1 scum (immune from detective). Add 1 scum for every 3 masons. Add 1 scum for every 6 vanilla. Vig and SK do not effect the scum count.

Total scum should not exceed 1/4 of the total players.

So with 20 players, if you have more then 5 (maybe 6) scum, your game is broken off the bat. This means that you can have 1 detective type, 2 role blocker types, and 2 night killer types.
OK -- so we're looking at a maximum of seven scum, maybe eight in a pinch. The weird part is that we have ... um ... something on the order of one and a half doctors and one and a half detectives. I'm going to figure three scum to balance the Oracle and Apprentice; could be one or two to balance the Priest and Disciple. There are probably approximately three Monks (I'd be surprised if there were as many as five), so I'm figuring another scum there. The number of vanilla townies is probably closer to twelve than six = two more scum; one Prophet = -1 scum. I'm not sure how the Avatar would affect the balance, but I can see ways his existence might benefit either side, so I'm going to guess that he's a null, like the Crusader and Psychopath.

However, there have to be enough nonbelievers for them to have a plausible shot at winning, so I'm guessing the numbers of scum and citizens are probably lower than they would otherwise be.

So, I'm estimating that the balance probably looks something like this:

Scum: 6, of whom two have power roles -- but I'm willing to buy seven.

Oracle, Apprentice, Priest, Disciple, Crusader: one of each
Monks: 3-ish
Citizens: 10-ish

Alchemist, Psychopath: one of each
Regular Non-believers: 4-ish

sachertorte

06-26-2007, 03:37 PM

It seems to me that the flaws are (1) the first person to post is almost certainly NOT the cop/detective/Oracle/Zadoc, and (2) nobody can tell which "investigation" was legit until the cop has been killed off. Do I have that right?
Oh I didn't mean to discuss chrisk's strategy specifically. I just didn't want to have the whole thing blow up in my face like his did. Which I seem to have managed not to avoid.

Here is my thinking: ignore if you so desire.
We have two investigative roles (I'm assuming there is one Oracle and one Apprentice). For the town, it is best to maximize investigative power by minimizing overlap of investigations. I propose that we divide the set of players into two groups and arbitrarily assign one group to the Oracle and one group to the Apprentice. That way we can be sure that investigations are disjoint. This method does nothing to reveal roles and gives sufficient coordination to avoid overlapping investigations.
In fact, I'll go as far as to say I propose the Oracle investigates players 1-15, and the Apprentice investigate players 16-30. Feel free to ignore me if you like. It was just a thought. We don't really need to discuss it much more and just let them decide on their own.

Now this line of thinking got me thinking about the Priest and Disciple as well. I thought about it and decided the benefit to town woefully do not outweigh the benefit to scum in employing a similar strategy.

FlyingCowOfDoom

06-26-2007, 03:38 PM

Mtgman

06-26-2007, 03:50 PM

Do believers care if non-believers live or die? Since A) Non-believers are capable of taking a win away from believers, and B) Scum show up in investigations as "non-believer", my initial thought is "no, not really...better them than me".

Off base or makes sense?As far as our investigative arm(the Oracle and his/her Apprentice(s)) are concerned, the Non-Believers are a superset of the Cultists. So rooting them all out should be their priority. Now the question for the town becomes, how do we sort the Cultists from the Non-Believers. This is an important exercise for three major reasons. Firstly, the Cultists are the ones trying to kill us. They should be the targets. Secondly, the Non-Believers count towards the town's side in the equation to determine a Cultist victory or not. Killing Non-Believers(as opposed to just those who investigate as Non-Believers, which includes most Cultists) advances the Cultist's win condition. Thirdly, while it is theoretically possible for the Non-Believers to take a victory away from the Believers, it's pretty damn unlikely. Let's not repeat the mistake of going for the Mason win in MIII by focusing on unlikely events when the most realistic threat is still out there.

On the other hand, the real value of investigation isn't who it implicates, it's who it exhonerates. Each verified Believer becomes a trusted townie. Given the ratios of Believers to Non-Believers, even with the confounding factors of the Prophet, and his/her convert(hmm, typing his/her made me think, what if the Prophet is a woman? How would she implement the conversion...) showing up as Believers and the pro-town Non-Believers being lumped in with Cultists, the odds are still heavy that the first few investigations by our Oracle and his/her Apprentice(s) will turn up Believers. These basically become globally known Masons/Monks once the Oracle is sniffed out(or snuffed out :() and their breadcrumbs are deciphered. These form a solid base of trusted players and can see the game through to the endgame. Think jsgoddess in the original Werewolf game. This is confounded somewhat by the Prophet and his/her convert registering as Believers, but we'll just have to remember, if we haven't found the Prophet and/or convert, when the endgame comes, even a "verified by the Oracle" notch on someone's belt loses it's protective power.

Also, lynching non-believers flat out kills the Psycopath. He/She will never go on their murderous rampage if they hang from a rope. This is a minor plus for the concept of hanging Non-Believers. It eliminates a potential loose cannon. Although, now that I think about it, odds are the Cultists will take care of the Psycopath for us. They'll know who he/she is because they'll know who they tried to night-kill when it failed(activating the Psycopath), so I don't think normal everyday townies need worry on this front. Killing Non-Believers also takes out the Alchemist, who can be of benefit to the town(blocking a night kill), but will more likely block one of the town's power roles(investigators or doctors) because there are more of them than Cultists(and the Alchemist only has about a 17-20% chance of blocking a Cultist kill anyway). On balance the Alchemist using his power aids the Cultists more than the townies.

So, some numbers. A general rule of thumb, borne out by mathematical analysis, is that you should have a number of scum roughly equal to the square root of the number of players. So in a game with 30 players, we should have 5-6 scum. More or less skews the probabilities of a scum or town victory significantly. I think we saw this in MII, which(no offense NAF) had too many scum. IIRC the spreadsheet NAF published showed he had 7 scum, which would have been enough for a game with 49 players! And the fact that there were two scum left at the end of the game(even allowing for good play by both Kat and storyteller0910) indicates there was an imbalance. So when the Oracle(et. al.) investigate(s) tonight the odds are they'll hit non-scum by about 4-1. This is fine. After the first Day/Night cycle we can start making "trusted" lists and this gives our investigative wing their first chance to leave breadcrumbs.

For the first lynch I suggest we all use random.org The numbers show that in a game with no investigators(which the first day basically is) the best strategy is random voting.

Enjoy,
Steven

On Preview: For the love of Nairu people! I knew it took me a long time to compose this post(interpspersed with work), but you've got more than full page in between when I started and when I previewed!

NAF1138

06-26-2007, 03:51 PM

Here is my thinking: ignore if you so desire.
We have two investigative roles (I'm assuming there is one Oracle and one Apprentice). For the town, it is best to maximize investigative power by minimizing overlap of investigations. I propose that we divide the set of players into two groups and arbitrarily assign one group to the Oracle and one group to the Apprentice. That way we can be sure that investigations are disjoint. This method does nothing to reveal roles and gives sufficient coordination to avoid overlapping investigations.
In fact, I'll go as far as to say I propose the Oracle investigates players 1-15, and the Apprentice investigate players 16-30. Feel free to ignore me if you like. It was just a thought. We don't really need to discuss it much more and just let them decide on their own.

I will be the first to admit that this kind of strategic thinking isn't really my strength, but I can't see any reason why we shouldn't employ this strategy. I only see up sides as far as the Oracle and apprentice are concerned.

Anyone else?

ArizonaTeach

06-26-2007, 03:52 PM

I am not getting dragged into a random vote discussion again (See case study M4 crew v ArizonaTeach, 2007 for details), but I will say that I myself was wondering about numbers. That's probably because it was such a big deal on the Hispaniola. Does the recruitment scenario enter into anyone's calculations here?

Fretful Porpentine

06-26-2007, 03:52 PM

It seems to me that the flaws are (1) the first person to post is almost certainly NOT the cop/detective/Oracle/Zadoc, and (2) nobody can tell which "investigation" was legit until the cop has been killed off. Do I have that right?
The really big flaw was that it made the detective way too easy to identify right off the bat. ((2) was actually a built-in part of the plan rather than the flaw -- the idea was that the trust lists would make it easier to ID the cop / detective's bread crumbs after the character had been killed.)

I don't see any problem with sachertorte's plan as long as we keep it flexible (there could, for example, come a time when we need a sure read on one of the players on the second half of the list, and it would then make sense to have the Oracle investigate this person instead of the Apprentice).

ComeToTheDarkSideWeHaveCookies

06-26-2007, 03:56 PM

I should have quoted more of Pasta's post to be more clear.
I'm talking analysis of the rules and game mechanics and exploiting them to the town's advantage. I don't mean to distract from the scum hunting and voting fun, but I think it is a worthwhile discussion. If everyone else disagrees, I'll let it go.

FWIW, I don't think everyone is going to be agreeing on any one particular point any time soon, if ever.

HazelNutCoffee

06-26-2007, 03:56 PM

Yeah. A dead player's words are a lot more meaningful than a living player's.

--FCOD
I agree - since we can always be 100% sure of a dead player's identity, but rarely of a live one's.

This may have already been mentioned before, but "random voting" on the first day doesn't mean everyone is random voting - scum will be sure who NOT to vote for, at least, and the Apprentice and Disciple know the identities of the Priest and Oracle, while the Monks all know each other, correct? Not much help initially, but in the long run I feel it will be (since some of these roles will almost inevitably turn up dead before the game ends), so I second sachertorte's suggestion that everyone ends up voting someone by the end of the first day.

I was wondering about the powers of the Oracle and the Apprentice - if/when the Oracle finds out the identity of the Apprentice, would it make sense for the Oracle to perform the ritual on the Apprentice until the Apprentice's accuracy reaches 80%? Or should the Oracle just ignore the Apprentice and concentrate on finding the cultists? And when would it make the most sense for the Priest to protect all of us at the price of being able to protect no one the following night? It seems that early on in the game, a few dead bodies are a necessary evil in helping us gather clues as to who the scum are.

Thoughts? *sips coffee*

storyteller0910

06-26-2007, 04:00 PM

Back, and reading. I'm going to concentrate a variety of responses into one post, to avoid drowning the thread in little posts. I want to start here:

I didn't say we didn't need to figure out how many N-B's and Cultists there are either,I said it was an accepted scum tell to bring it up.
As far as figuring out how many there are..well,everyone can make a fairly educated guess, discussing it means nothing and gets us nowhere, as only the Cultists know for sure.
So on top of your original tell,I ain't to keen on revenge votes or word twisting either.

1. He did not vote for you; there was no revenge vote. This looks to me like someone trying to get a vote in for no particular reason.

2. You win the First Internally Inconsistent Post award for this post, in which you tell Clockwork Jackal that he was twisting your words by saying that you don't think it's important to figure out how many N-Bs and Cultists there are in the same post where you explicitly state that you don't think it's important to figure out how many N-Bs and Cultists there are. Sure, you didn't state this in your initial post, but CJ inferred it and, based on the evidence above, it was a reasonable inference.

3. As he pointed out, CJ didn't say no one looked scummy, he said no one looked summy except for Hal.

Strong FoS MadtheSwine.

You're suspicion back on me would usually seem scummy to me. That's not a good way to dispel suspicion on yourself (attacking your attacker). But, I think in your case it's a the reaction of a newbie Townie. Be more careful in the future.

Disagree. I'm not sure why revenge votes have such a scummy reputation, but in my (admittedly limited) experience they are employed by the good guys exactly as often as they scum. The thought process of the typical townie seems to go: I know I'm innocent, and anyone who wants to hang an innocent is probably scum. And on to revenge voting.

I agree with this; I always feel that random voting is a mistake. As is pressure-voting - voting someone just to get them to say something specific. I never do them, and neither will I respond to them, even if it means I get lynched (or, in this case, drowned). I'll go along with FOSs though, because they can't be turned into a bandwagon.

This is going to be, I think, an unpopular opinion. I don't think that I think that bandwagons are necessarily such a bad thing - even if - and again, I know how this sounds, but it's worth saying it anyway - they end in a townie death. Bandwagons carry information. Who started it? Who jumped on it? Who jumped on it for little apparent reason? Who piled in at the end when the outcome was inevitable anyway?

I'm not saying we want a bunch of out-of-control bandwagons that lead to town lynches, just that we shouldn't spend so much time quaking in fear of a bandwagon that we refuse to vote, or spread our votes so thin that it's easier for scum to control a lynch.

And you know what? On that note, and on the basis of the admittedly modest evidence presented above, I'm going to start out my Day by voting to:

Dunk MadtheSwine

Hal Briston

06-26-2007, 04:00 PM

It seems to me that the flaws are (1) the first person to post is almost certainly NOT the cop/detective/Oracle/Zadoc, and (2) nobody can tell which "investigation" was legit until the cop has been killed off. Do I have that right?Huh...I forgot about the "fake detective" bit. Well, I wouldn't write it off as an altogether terrible idea...

(1) If the town was playing it smart, they would do it randomly...the real detective would occasionally post first.
(2) The flip side of this is the removal of the danger of the detective being murdered without ever getting a chance to spill the truth. Yes, he'll breadcrumb, but we all know that can get seriously wrong in its interpretation.

(And before anyone starts yelling "scum tell!", yes, I said "they" in reference to "the town". We're dealing in the theoretical here, not the actual town in this game.)

NAF1138

06-26-2007, 04:01 PM

For the first lynch I suggest we all use random.org The numbers show that in a game with no investigators(which the first day basically is) the best strategy is random voting.

Oh, for the love of... :smack:

Ok, maybe this is because I have a lot more experiance playing IRL where this sort of analysis is next to impossible to do, but this isn't a numbers game. Yes, you can use the numbers to help you, but the scum aren't robots. Statistically maybe a random vote helps us in the short term, I don't know, but we have to start playing for the endgame NOW. We can't rely on a detective to save our asses like they did in WW1 and Pirates. Even if we kill scum today with a random vote, what do we learn? If all everyone is doing is random voting WE LEARNED NOTHING? We are back at square one again tomorrow, and the scum have just gotten a free shot at taking out one of us.

Blaster Master

06-26-2007, 04:03 PM

I was wondering about the powers of the Oracle and the Apprentice - if/when the Oracle finds out the identity of the Apprentice, would it make sense for the Oracle to perform the ritual on the Apprentice until the Apprentice's accuracy reaches 80%? Or should the Oracle just ignore the Apprentice and concentrate on finding the cultists? And when would it make the most sense for the Priest to protect all of us at the price of being able to protect no one the following night? It seems that early on in the game, a few dead bodies are a necessary evil in helping us gather clues as to who the scum are.

Just wanted to clear this up since it looks like a mis-understanding of the rules. The Oracle needs only investigate the Apprentice once, thereafter the accuracy improves, regardless of whom they investigate, for each night the Oracle lives. I apologize if that wasn't clear in the rules.

fluiddruid

06-26-2007, 04:04 PM

Lack of consensus is bad. I don't know how we can force ourselves to reach a consensus in only 5 days, but letting a relatively small number (25%!) of votes determine the lynch was very bad for the town. Everyone needs to vote even if 'unsure,' otherwise we risk giving scum easy cover. On the surface I agree with this, but, on the other hand it gives scum a chance to justify their votes (or rather, avoid giving reasons for them).

Lack of consensus isn't ideal, but it's often a symptom of little information to go on rather than a problem itself. I've seen many times that people were convinced by the tiniest turns of phrase to vote for someone, hiding the legitimate scum coersion that was occurring.

IMHO, vote your head or your heart, but don't hide behind votes only you're voting for (which is basically the same as not voting) late in the day.

Mtgman

06-26-2007, 04:07 PM

On random voting versus non-random voting. There are two main types of analysis here. The largest body of mathematical analysis I've seen on how to vote on Day 1 was based on a lynch needing a plurality of votes, not a majority. It was also highly advantageous for all the votes to be revealed at the same time, so no factions got a chance to "skew" the numbers after seeing the initial spread. So people could vote with random.org and, because there's been no Night for the scum to stragetize during, we'd either get really random results(which is a strong strategy for the town), or we'd get the scum bandwagoning to direct the vote(which gives them away, and is also a fine thing for the town).

I don't think I've seen analysis for the particular setting we have right now, where it takes a majority to dunk on Day 1. So I'm up for discussion, but I don't think we should take it too seriously. We may get lucky and dunk the Prophet, but probably not.

Enjoy,
Steven

NAF1138

06-26-2007, 04:08 PM

On the surface I agree with this, but, on the other hand it gives scum a chance to justify their votes (or rather, avoid giving reasons for them).

Lack of consensus isn't ideal, but it's often a symptom of little information to go on rather than a problem itself. I've seen many times that people were convinced by the tiniest turns of phrase to vote for someone, hiding the legitimate scum coersion that was occurring.

IMHO, vote your head or your heart, but don't hide behind votes only you're voting for (which is basically the same as not voting) late in the day.

There we go, this here is sound advice. What's more, it is important to always state they why. Why are you voting for someone? A vote without a stated reason does us no good.

sachertorte

06-26-2007, 04:17 PM

I agree with the first half of Mtgman post. Investigations cannot distinguish cultists from non-believers. (Now who was it that said this game is skewed town? I think not).
For the alchemist we could employ a similar verification test used in M2 to verify a miller claim, but only if the Crusader has skipped the previous night kill and only with 75% success rate.
I also think Mtgman is astute in noting that the alchemist, on average will hurt the down by roleblocking.

I do not agree with Mtgman's assertion that a day one random lynch is optimal. I read the paper (I think) he did, and the analysis pretty much is stuck with random lynches. There really isn't any other way to do a mathematical analysis. This game isn't a math game, it's a social game. So the day one strategy is to talk. The end of day dunking may not end up better than random, but the chatter sets up social analysis for future days. The paper ignores the social aspect of the mafia game.
I also found the paper's analysis to be needlessly complex.

HazelNutCoffee

06-26-2007, 04:17 PM

Although, now that I think about it, odds are the Cultists will take care of the Psycopath for us. They'll know who he/she is because they'll know who they tried to night-kill when it failed(activating the Psycopath), so I don't think normal everyday townies need worry on this front.
Not necessarily. A failed night-kill could also mean the person was being protected by the Priest or the Disciple. It could also mean the Alchemist was successful in blocking the cultists.

Killing Non-Believers also takes out the Alchemist, who can be of benefit to the town(blocking a night kill), but will more likely block one of the town's power roles(investigators or doctors) because there are more of them than Cultists(and the Alchemist only has about a 17-20% chance of blocking a Cultist kill anyway). On balance the Alchemist using his power aids the Cultists more than the townies.
There are four power roles that help the town (Priest, Disciple, Oracle, and Apprentice), one that is supposed to be helpful but potentially dangerous (Crusader), and one that is neutral (Psychopath, if activated). Blocking a citizen or a non-believer, while not helpful, does no harm; blocking a cultist has at least a chance of preventing a nightkill. There are perhaps 5-7 scum in this game (and this is including the Prophet and the Avatar, I thing). It seems like that in the beginning of the game, at least, blocking a random someone has a higher percentage of either doing nothing, or doing something beneficial.

For the first lynch I suggest we all use random.org The numbers show that in a game with no investigators(which the first day basically is) the best strategy is random voting.
It seems like there will never be a consensus on this, but even for those of us who are regular citizens, I suggest that it might be more helpful if we posted some kind of reasoning behind our votes. That way, if we are murdered in the night, those surviving can re-read what we've posted and see if there was something in our words that prompted the cultists to hasten our death.

Just some thoughts.

HazelNutCoffee

06-26-2007, 04:20 PM

Just wanted to clear this up since it looks like a mis-understanding of the rules. The Oracle needs only investigate the Apprentice once, thereafter the accuracy improves, regardless of whom they investigate, for each night the Oracle lives. I apologize if that wasn't clear in the rules.
Ohh, so all the Oracle needs to do is perform the ritual itself - not necessarily perform it on the Apprentice? I understand now. Thanks for the clarification.

sachertorte

06-26-2007, 04:23 PM

And when would it make the most sense for the Priest to protect all of us at the price of being able to protect no one the following night?
my read on that ability is to avoid an endgame recruit. Theoretically, the Priest is unrecruitable and unnightkillable. Barring a lynch the Priest can use the protect everyone when it is likely that the recruitment is going to happen. In M3, the optimal recruitment time was when there was one mafia left. With the Priest mechanic, waiting becomes more risky since a recuitment when everyone expects it could be easily blocked by the Priest.
Now that recruitment is tied to a specific player, I don't anticipate an endgame recruitment anyway.

Oh, I know. The Priest should protect everyone to prevent a scum win. In other words in the endgame if going into night a scum kill wins the game, the Priest should protect everyone.

Idle Thoughts

06-26-2007, 04:33 PM

Okay! A new start and a new game. Time to dive in.

Just to let everyone know, I typically reply to things post by post..so as I'm replying to one, I have not yet read any posts after yet, but normally go back and make corrections as I learn more and more and get more and more caught up.

Ok, I'll kick this off with a strategy question:

Do believers care if non-believers live or die? Since A) Non-believers are capable of taking a win away from believers, and B) Scum show up in investigations as "non-believer", my initial thought is "no, not really...better them than me".

Off base or makes sense?

Well, as far as I can tell, believers win only if the cultists are dead, right? Not if non-believers are dead, so there's a better chance winning with them.....however like you say (and with your own experience as a group that could take a single-only win) it would open the door to a non-believer (but not cultist) win alone. So I think it's something that should/may be looked if the believers are being picked off left and right and not so many vanilla non-believers.

And, going on, basically what FCOD put better.

FINALLY!

I would like to start by FOSing Mad, for destroying the pirates chances in the last game, and being the first to post durring the day in this one.

AVAST!

er..I mean...HERETIC!

Are you joking? Sounds like you are so I'll give you the benifet of the doubt. :p But like I said before in the last game, I just hope that people don't use past game roles and stuff as a reason to look at someone suspicious.

Andddd....going on..I see Mhaye saw this too.

Andddddd, going on, I now see your post saying you were. Don't scare me like that, NAF! :smack:

I suppose it would depend on how many of them there are to start.

Anyone care to speculate on the distribution this time around? We've got a larger starting pool than ever this time, and three groups instead of two.

With 30 players, what do you think, six cultists, give or take one? How about the non-believers...more or less than the cultists?

Just going off of what I've done in hosted games, I usually make the scum about 1/5th or 1/6th the total number of players. But then you have to wager in the other roles that have potential to kill, plus the recruit. I'd say, with all of that put together, about five or six scum would seem most likely. As in, pure scum as opposed to non-believers.

Hmmm...well, I'll post now, even though there's no real information to go on right now. Except for one little thing (which I'll get to in the next paragraph) no one really looks scummy to me yet. I think I'm just gonna play as though I'm getting killed tonight, and just tell everyone my thought processes. OK?

Snipped and bolding mine.

See, this gets me a little. I've said this before too in other games. Why feel the need to say this and put it out there? It's sort of a given and just strikes me as something someone would say to subtly throw someone off. That and that nobody strikes you as being scummy yet. Well it's the first Day. What were you expecting? :dubious:

I have noticed two accepted scum tells so far(according to outside sources I have read), and as has been pointed out in other games, scum tells aren't always scum tells, but I think it's better than random votes at this point.
The first one is from you Clockwork, saying no one looks scummy.The second was from my good buddy Hal speculating on how many cultists and non-believers there are.
Course it is early,but those are my thoughts so far.

Good to see I wasn't the only one that caught it.

Then vote for lurkers, that is a legitimate vote at this stage.

Majorly snipped.

Maybe just me but I feel that's just as bad. Seems like giving an easy out to scum (who know who is what) to just gang up on someone who's not posting as much. If someone is lurking too much, I just feel a sub out would be the best course rather than a pile on.

And about the whole random voting thing (this is to everyone who was talking about it, not just NAF). Here's what I think and have said before. It's the first day. Yes it REALLY IS random for the good side and yes the for the scum it's not, but that doesn't matter because EVERYONE will be acting like it's random. So in the end, where does it get us? Nowhere. So while I don't think we should pussyfoot around and not vote at all, I also don't think it's wise to just throw votes out there, here and everywhere cause it could start a bandwagon really easy that scum might take advantage of or at least help along a bit.

WHERE HAVE YOU BEEN? Lurking????

Hahahaha. I like the light hearted play. :D

But Clockwork Jackal didn't make a revenge vote. All he said was that he was inclined to FOS you for saying that speculating on the distribution of players is a scum tell, which seemed to imply that you were discouraging such speculation as unnecessary. (I see that's not what you meant, but it initially appeared that way.) He didn't even point an actual FOS your way. If anything, it would seem that your vote for him is something of an over-reaction.

I do agree that discussing the distribution of players is not very fruitful this early on in the game, which is why I will withhold pointing a FOS at you for now.

I dunno... I said earlier that s/he just seemed iffy and it tripped me a bit too (although I didn't vote as MtS did)...and so far, I've learned to trust my meter trips.

FWIW, I don't think everyone is going to be agreeing on any one particular point any time soon, if ever.

Ladies and Gentlemen. The sanest post in here.

sachertorte

06-26-2007, 04:35 PM

Huh...I forgot about the "fake detective" bit. Well, I wouldn't write it off as an altogether terrible idea...

Heh. I agree it wasn't that bad. The problem was the strategy was partially implemented then abandoned; leaving the detective exposed. Had everyone joined in, the detective may not have been exposed so early. Either way, the gambit assures a dead detective in log(N) time. Not so hot.

ArizonaTeach

06-26-2007, 04:35 PM

HazelNutCoffee

06-26-2007, 04:38 PM

Oh, I know. The Priest should protect everyone to prevent a scum win. In other words in the endgame if going into night a scum kill wins the game, the Priest should protect everyone.
But if the Priest dies, the opportunity is wasted. Even if the Disciple is still alive, the number that the Disciple can protect is reduced to two. I was thinking vaguely it might make sense for the Priest to use this ability when there is a possibility of a maximum number of kills (such as after the Psychopath is activated, although we'd probably have to guess at that, or after the Avatar is lynched). A normal night will have a maximum of two kills (cultists and Crusader, who could potentially kill a citizen) while an unlucky night could have three or four (if the Psychopath is activated and the Avatar gets lynched).

Gah, my head hurts. Is the bar open yet? *looks hopeful*

ArizonaTeach

06-26-2007, 04:41 PM

Maybe just me but I feel that's just as bad. Seems like giving an easy out to scum (who know who is what) to just gang up on someone who's not posting as much. If someone is lurking too much, I just feel a sub out would be the best course rather than a pile on.Problem with that thinking is, that's the exact strategy the pirates used in M4. Auto intentionally lurked and committed the kills so that there would be no trail. I was hesitant to lynch lurkers in M4, but once lurking became an actual scum strategy I think the rules on that one changed.

Hal Briston

06-26-2007, 04:41 PM

Gah, my head hurts. Is the bar open yet? *looks hopeful*Oh, I'm sorry, weren't you aware? Adamsborough is a dry town.

Mtgman

06-26-2007, 04:44 PM

One note I would make to the general discussion. It's ok to talk a bit about the Apprentice, but remember, he/she is basically a beat cop. How many townies died in MII because they tried soo hard to make a beat cop a useful role? Blaster Master has thrown in a twist here with the "increasing chance of success" rules, but we've got about a 1/15 chance the Oracle will investigate the Apprentice early enough in the game to make this happen(one one of the first two Nights). Even then their feedback won't be really useful until they hit the "false readings come up as blanks" stage, which won't be for several Days at the earliest. We can save the discussion of that role until then, when there will be a smaller audience and we've got a little bit of a feel for each other.

NAF, I agree with you about real-world games. I wouldn't roll a dice for a vote on Day 1 in a live game. I'd be reading body language and tone of voice. Since all that is unavailable here, I'm thinking it comes out to a wash and random voting is as good an approach as any. The hundreds of games played on mafiascum.net have left a strong impression of random voting being the best Day 1 strategy for the town. A number of players have argued against it and tried to shoot it down, mostly because it turns off newbies when they're killed for essentially no reason on the first Day of their first game, but it's still pretty much the norm there.

Enjoy,
Steven

Blaster Master

06-26-2007, 04:47 PM

sachertorte

06-26-2007, 04:50 PM

But if the Priest dies, the opportunity is wasted.

Hello Mr. Risk, I'd like to introduce you to Ms. Reward.

Also, why would the Priest die? The Priest is not night killable and is not convertable. The only way for the priest to die is to get dunked. And I for one don't plan on dunking any priests.

Hal Briston

06-26-2007, 04:50 PM

So in a nutshell, the Apprentice needs to find an unobvious way to say "Yo, Oracle! Over here!" asap....

HazelNutCoffee

06-26-2007, 04:51 PM

Oh, I'm sorry, weren't you aware? Adamsborough is a dry town.
:eek: You mean, the cultists have already WON?!?! NOOOOOOooooooooo. . .

Is lurking being allowed as a strategy, or will lurkers be switched out?

Pleonast

06-26-2007, 04:51 PM

A totally randomized vote, as suggested by MtgMan, is worse than the Town not dunking anyone. We'd gain no information while probably killing a Townie. If we want to forfeit our turn, better to not dunk anyone. That way we won't kill a Townie. Never mind the logistics of getting a simultaneous vote on this board.

Random votes as advocated by me are a tool for bootstrapping the discuss in the early part of the first Day. The final vote decision needs to be something accountable.

I'm waiting for Autolycus to make his entrance. Then I can accuse him of being suspicious because (1) he's using a Cultist accent, (2) he's not using a Cultist accent, (3) he took too long to make an entrance, or (4) he jumped in too soon. Pick your poison! :D

Blaster Master

06-26-2007, 04:54 PM

The conversion (recruitment) is a one shot deal. If it is attempted, regardless of whether if succeeds or fails, it may not be attempted again.

Further, the ability to convert is lost upon the death of the Prophet. As with other night actions, if the Prophet is attacked at Night, by the Psychopath or Crusader, when they are attempting to recruit, the recruit will still be processed.

HazelNutCoffee

06-26-2007, 04:54 PM

Hello Mr. Risk, I'd like to introduce you to Ms. Reward.

Also, why would the Priest die? The Priest is not night killable and is not convertable. The only way for the priest to die is to get dunked. And I for one don't plan on dunking any priests.
All I'm saying is that the Priest can get killed by accident - for example, they can be night-killed if they miscalculate and protect someone else on the night the cultists decide to attack the Priest - before they get a chance to use their bless-everyone power. I'm all for taking calculated risks for the greatest reward - I was merely speculating as to how the Priest's bless-everyone power could be maxmized for greatest effect.

HazelNutCoffee

06-26-2007, 04:55 PM

The conversion (recruitment) is a one shot deal. If it is attempted, regardless of whether if succeeds or fails, it may not be attempted again.

By one shot, do you mean it can only be done once during the entire course of the game?

Malacandra

06-26-2007, 04:56 PM

Hello Mr. Risk, I'd like to introduce you to Ms. Reward.

Also, why would the Priest die? The Priest is not night killable and is not convertable. The only way for the priest to die is to get dunked. And I for one don't plan on dunking any priests.

Well, the priest's not killable if he's protecting himself, but I guess at least some of the time he'd be protecting someone else.

Malacandra

06-26-2007, 04:57 PM

Gah. Thread moving too fast, again!

MHaye

06-26-2007, 04:59 PM

By one shot, do you mean it can only be done once during the entire course of the game?Yes, that's what it means.

The prophet's role :Prophet - The blood of many great prophets of Sekham flows in your veins and you have spent your life studying to become a powerful leader. As such, you hold dominion over all the cultists (i.e., the tie breaking vote). Your power allows you to shield yourself from the prying eyes of the Oracle and Apprentice, and if you are targeted, will show up as a believer. Further, you also have the power to twist the minds of weak spirited individuals. One time, you may spend a night to perform a powerful ritual, twisting the mind of a non-cultist and forcibly converting him to a follower of Sekham; however, this ritual requires the assistance of all of your cultists, and will prevent a sacrifice to Sekham on that night. Further, because the recruit is the result of twisting his mind and not a change of heart, he will ALSO continue to appear as he would previously would to the Oracle and Apprentice. Because the ritual requires a powerful prophet, the cultists will be unable to perform it if the prophet is dead.(Bolding mine).

The Cult can only do it once.

Mtgman

06-26-2007, 05:01 PM

Wait, what? Is Mtgman saying we should all use random.org and reveal at the same time and go with the results? Is that seriously what he's saying?? Am I missing something here because that seems like a bad, bad, bad idea. I'm going to go back to my fallback whenever someone suggests a strategy: How does the town benefit from this? I see three drawbacks for the town:
1) There are still going to be those who delay even a minute or two that could throw the vote a certain way by casting a non-random vote and saying it's random. And even if that doesn't happen, if the votes truly are random then suspicion will fall upon those people who voted for the dunk unfairly.
2) The chances a pro-town power role could be revealed are small, but possible; in fact, probably at a greater percentage than a normal vote. If a power role is revealed, I damn well want a trail to who forced that to happen. Plus, will there even be time for a role-claim? Will that role-claim be more or less believable?
3) This negates any information that could be gleaned from today's conversations. There will be no trail to follow.

I hope I'm just understanding wrong, and it's against my nature to vote this early in a Day, so I'm not going to accuse or FOS Mtgman, but MAN, this seems like a bad idea of epic proportions.
From Mafia : A Theoretical Study Of Players and Coalitions in a Partial
Information Environment (http://arxiv.org/PS_cache/math/pdf/0609/0609534v3.pdf)(warning, PDF)The game without detectives – optimal strategies
In this section, we demonstrate that the game without detectives has a simple optimal strategy for both sides.
...
3.2 Properties of Citizens Strategy
Note that following this strategy will result in killing a random player as long as the citizens form a majority.
This follows since the number n has the uniform distribution as long as there is at least one citizen and since
when there is a majority of citizens the vote will result on killing player numbered n.
Note furthermore that the above protocol relies on the assumption that the players can announce their
numbers at the same time before they may observe other player’s announced numbers, since otherwise the
mafia may vote for a number such that the sum corresponds to a citizen.I've listed a number of reasons I'm skeptical of this analysis holding for this particular format, and I'm open to discussion about other strategies.

Also, upon re-reading the article, I found this passage.As we will see the different players have immensely different powers: For a large population of size R, a
mafia of order √R already has a reasonable chance of winning. And a larger mafia will surely win.
Interestingly, as soon as there is one detective present, the game becomes fair only when the mafia
consists of a linear fraction of the total population.So I'm thinking we're probably dealing along the lines of 7, maybe 8 scum. Hopefully 7 because we have a scum power role which can recruit and if they recruited then we'd be looking at 9 scum, which is a really high number for a 30 person game.

Enjoy,
Steven

HazelNutCoffee

06-26-2007, 05:02 PM

The Cult can only do it once.
:smack: Sorry. I read the rules, I swear.

NAF1138

06-26-2007, 05:03 PM

Oh, I'm sorry, weren't you aware? Adamsborough is a dry town.
The Horror THE HORROR

NNNNNOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO!!!!!

Idle Thoughts

06-26-2007, 05:06 PM

Problem with that thinking is, that's the exact strategy the pirates used in M4. Auto intentionally lurked and committed the kills so that there would be no trail. I was hesitant to lynch lurkers in M4, but once lurking became an actual scum strategy I think the rules on that one changed.

True, but in any case, I think that there shouldn't be anyone allowed to lurk that much...because if they are subbed out, at the very least the new person would know not to lurk that much and would give the believers more info to work with.

Hello Mr. Risk, I'd like to introduce you to Ms. Reward.

Also, why would the Priest die? The Priest is not night killable and is not convertable. The only way for the priest to die is to get dunked. And I for one don't plan on dunking any priests.

Excuse me? The Priest is not Night killable? Where are you getting that from? :dubious:

A totally randomized vote, as suggested by MtgMan, is worse than the Town not dunking anyone. We'd gain no information while probably killing a Townie. If we want to forfeit our turn, better to not dunk anyone. That way we won't kill a Townie. Never mind the logistics of getting a simultaneous vote on this board.

Snipped some.

I think it's half and half. Like I said in my post above..NOT voting at all is bad, I think. It gives up the right to kill a scum off. However we do have about five days to go over stuff and go back and forth with accusations and ideas and thoughts and inconsistences and all the like. So I think voting just based on random.org and in the first hour alone is a bit hasty and aggressive.

Votes like MtS gave, I think, are at least based on things. Using Random.org seems to easy without a good explanation behind it.

Hal Briston

06-26-2007, 05:08 PM

Ok, working off of my "the scum want the rest of us to be miserable" theory, we have HazelNutCoffee and NAF cleared. Good job, folks.

HazelNutCoffee

06-26-2007, 05:09 PM

The Horror THE HORROR

NNNNNOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO!!!!!
*pats NAF comfortingly on shoulder* There, there. I'm sure the secret role is actually the town's Underground Moonshine Dealer, here to save us all from insanity. All we have to do is find them. *looks around speculatively*

Hal Briston

06-26-2007, 05:09 PM

Excuse me? The Priest is not Night killable? Where are you getting that from? :dubious: If he continuously blesses himself, he's not.

Hockey Monkey

06-26-2007, 05:10 PM

I agree - since we can always be 100% sure of a dead player's identity, but rarely of a live one's.

This may have already been mentioned before, but "random voting" on the first day doesn't mean everyone is random voting - scum will be sure who NOT to vote for, at least, and the Apprentice and Disciple know the identities of the Priest and Oracle, while the Monks all know each other, correct? Not much help initially, but in the long run I feel it will be (since some of these roles will almost inevitably turn up dead before the game ends), so I second sachertorte's suggestion that everyone ends up voting someone by the end of the first day.

*snip*

Just want to mention for the first timers: Don't think that scum won't vote for other scum, or that the other roles won't vote for each other either. The last thing they want is a clearly distinguished voting pattern. That said, as more people die and roles are revealed, the voting patterns become very interesting indeed.

Blaster Master

06-26-2007, 05:12 PM

Is lurking being allowed as a strategy, or will lurkers be switched out?

I do think there's a fine line between lurking as a strategy and lurking because you're not participating. However, I will not set a rule of thumb because I think it encourages fluff posting.

I won't be explicitly checking either because it's too much leg work; if it's not explicitly obvious that someone is missing, does it matter? If I see a few complaints like "Hey, So-So isn't participating!" I'll investigate and see if it warrants poking. If so, I'll poke him, and if he doesn't respond to poking, I'll forcibly substitute.

Hal Briston

06-26-2007, 05:12 PM

Just want to mention for the first timers: Don't think that scum won't vote for other scum, Very true, but the presence of the Avatar makes that a very risky play for the scum this time around. They may be less willing to join up on a "well, they're gonna swing anyway" bandwagon.

MadTheSwine

06-26-2007, 05:13 PM

Back, and reading. I'm going to concentrate a variety of responses into one post, to avoid drowning the thread in little posts. I want to start here:

1. He did not vote for you; there was no revenge vote. This looks to me like someone trying to get a vote in for no particular reason.

2. You win the First Internally Inconsistent Post award for this post, in which you tell Clockwork Jackal that he was twisting your words by saying that you don't think it's important to figure out how many N-Bs and Cultists there are in the same post where you explicitly state that you don't think it's important to figure out how many N-Bs and Cultists there are. Sure, you didn't state this in your initial post, but CJ inferred it and, based on the evidence above, it was a reasonable inference.

3. As he pointed out, CJ didn't say no one looked scummy, he said no one looked summy except for Hal.

Strong FoS MadtheSwine.

I am surprised at you story

1.Already acknowledged my mistake

2.What you say is simply not true,there is nothing inconsistent about it . Jackal said I said something I did not say.Wether I agree with what he said I said does not matter.

3.Saying no one looks scummy is scummy, even with the "one little thing" thrown in.Scum tells are not gonna be by the book and that particular post seems like a variant of said scum tell.

Idle Thoughts

06-26-2007, 05:13 PM

If he continuously blesses himself, he's not.

Well, yeah but she seemed to just be saying/assuming that the Priest could never be Night killed.

If I were the Priest, I'd certainly be self-protecting a few times...but might ultimately protect another if anything came to light.

Blaster Master

06-26-2007, 05:15 PM

By one shot, do you mean it can only be done once during the entire course of the game?

Yes. If they attempt and succeed, they've got their convert; they will not be able to convert anyone else. Similarly, if they attempt and fail, they don't get a "do over"; they will not have a convert.

HazelNutCoffee

06-26-2007, 05:24 PM

Very true, but the presence of the Avatar makes that a very risky play for the scum this time around. They may be less willing to join up on a "well, they're gonna swing anyway" bandwagon.
Good point, especially since even the Avatar isn't aware of their own identity. I'm assuming that any cultists who vote unwittingly for the Avatar would incur the wrath of Sekham as much as any citizen would. (Although voting for one of your own seems like a last-ditch strategy at any rate.)

Mtgman

06-26-2007, 05:25 PM

In response to a rules query:

At the beginning of the game, the Apprentice starts with 50% accuracy. There are two ways the Apprentice can improve his power, through being discovered by the Oracle and gaining his tutelage or through the Oracle's death. As an example, let's say that the Oracle happens to investigate the Apprentice Tonight, the Tutelage bonus is immediate, and thus, the Apprentice's investigation Tonight will have an accuracy of 60%. If they both survive the Night and the next Day, the Apprentices accuracy will then be 70%. It'll look something like this.

Night Apprentice % Oracle Investigates
1 50% Player(1)
2 50% Player(2)
...
i 60% Apprentice
i+1 70% Player(i+1)
i+2 80% Player(i+2)
i+3 80% Player(i+3)
...
If the Oracle dies before he discovers the Apprentice, this power cannot be activated. If the Oracle dies during the training process, the Apprentice's accuracy is frozen (he doesn't unlearn).

Regardless of whether or not the tutelage is activated, when the Oracle dies, the Apprentice WILL receive the Oracle death bonus and it stacks with the Tutelage bonus. Thus, if the Apprentice was fully trained, he will acheive 100% accuracy on the night following the Oracle's death (he has to find out he's dead before he can loot his belongings).

I hope this clears things up a bit.That's hilarious. You took what is arguably the most confusing/difficult role in the game(cowardly reporter may be worse) and made it even more difficult. Now when going through breadcrumbs from the Oracle and Apprentice we'll have to guess and try to figure out if the Oracle ever investigated the Apprentice, and when it was(which is something even the Apprentice probably doesn't know). Then we have to apply weighted percentages to each of the breadcrumbs left by the Apprentice based on when during the course of his/her tutelage(and the timing of the subsequent death of his/her teacher) each person was investigated. Barring a "I know I'm going out tonight and here's who I investigated and when" from both roles I think we're going to have a hell of a time teasing any useful information out of the Apprentice's role at all.

Nicely done Blaster Master, the evil Dungeon Master in me salutes you.

Enjoy,
Steven

Zeriel

06-26-2007, 05:29 PM

Blaster Master

06-26-2007, 05:30 PM

That's hilarious. You took what is arguably the most confusing/difficult role in the game(cowardly reporter may be worse) and made it even more difficult. Now when going through breadcrumbs from the Oracle and Apprentice we'll have to guess and try to figure out if the Oracle ever investigated the Apprentice, and when it was(which is something even the Apprentice probably doesn't know). Then we have to apply weighted percentages to each of the breadcrumbs left by the Apprentice based on when during the course of his/her tutelage(and the timing of the subsequent death of his/her teacher) each person was investigated. Barring a "I know I'm going out tonight and here's who I investigated and when" from both roles I think we're going to have a hell of a time teasing any useful information out of the Apprentice's role at all.

Nicely done Blaster Master, the evil Dungeon Master in me salutes you.

Enjoy,
Steven

I apologize... I did want it to be a LITTLE bit convoluted, but not THAT convoluted. :smack:

You're right, I neglected to mention an important point that, while my intention, should not have been left simply to implication. The Apprentice WILL know when he's been investigated (how can he be under his tutelage and not know it).

NAF1138

06-26-2007, 05:32 PM

(cowardly reporter may be worse)

Nothing to do with this game at all, but I want to know...what is the cowardly reporter? I haven't heard of this one.

[quote]Nicely done Blaster Master, the evil Dungeon Master in me salutes you.[/qutoe]

Seconded. Beat cop has always had a special place in my heart, I especially enjoy this variant.

(btw, I think my next game is going to be a bastard mod game, if anyone is interested)

Zeriel

06-26-2007, 05:32 PM

And an immediate followup to say holy cow, Mtgman is right. The Apprentice is going to be hard to really usefully interpret results from.

Hockey Monkey

06-26-2007, 05:37 PM

Good point, especially since even the Avatar isn't aware of their own identity. I'm assuming that any cultists who vote unwittingly for the Avatar would incur the wrath of Sekham as much as any citizen would. (Although voting for one of your own seems like a last-ditch strategy at any rate.)

What I forgot to mention or make clear...scum will vote for other scum, but they do not necessarily leave their votes on their compatriots. Sometimes they do, but you are right...the presence of the Avatar complicates strategy for them.

Hal Briston

06-26-2007, 05:38 PM

And an immediate followup to say holy cow, Mtgman is right. The Apprentice is going to be hard to really usefully interpret results from.Perhaps the Apprentice makes it a point to investigate his targets twice (at least until he gets his level up into a decent percentage)? Yeah, slower...but the additional accuracy might be worth it.

Hal Briston

06-26-2007, 05:41 PM

(Although voting for one of your own seems like a last-ditch strategy at any rate.)(Almost missed this one..)

Not always. In M3, one of the identifiers was the fact that while there were still multiple scum, every single time one of them was lynched, all living scum were part of the voting block. Every scum that was lynched had every other member of team scum voting for them.

Doubt you'll see that this time around.

HazelNutCoffee

06-26-2007, 05:41 PM

Wow, this thing is moving right along.

At this point, aside from my apparent early scum tell (at least according to Clockwork Jackal), who's therefore receiving a halfhearted sorta-FOS from me) the only think I have to say is I agree with sachertorte--I can't see any downsides to the town from the Oracle and Apprentice following that exact strategy. Except for the big one--if the Apprentice is in 16-20 him/herself, it's suddenly a much less effective strategy. Can't be helped, I suppose.
Damnit, I knew there was something about that tactic that was bugging me.

Upon further reflection, I'm not sure how helpful having seperate lists for the Oracle and the Apprentice would actually be. After all the Apprentice's findings will always be uncertain until after the Oracle dies. Seems like there would be much more of an advantage in the Oracle finding the Apprentice ASAP, which is never going to happen in the event they are given seperate lists and the Apprentice ends up not being on the Oracle's list.

There's also a strong argument for making sure that the Oracle and Apprentice are investigating their respective halves in random order, because otherwise if they're going right down the list, the scum will be easily able to know when they've been confirmed. Or maybe that's a good thing, making them nervous?

If we agree the Oracle and Apprentice should do this, it might behoove us to remember who's up for investigation that particular previous night assuming it's a good idea to do it in order.

Any other thoughts? Obviously, the hardest thing is for the Oracle and Apprentice to stay safe while delivering useful information, which is something I'm still not sure of the best way to accomplish.
Seems to me like this would put both of them in quite some danger, since everyone would be aware of who was being investigated that night. The scum would keep a sharp eye out for those who pointed FOS at the correct person after the night said person was investigated, which would considerably endanger the Oracle, and to a lesser extent, the Apprentice.

Mtgman

06-26-2007, 05:43 PM

That's fine, but it still doesn't help us much. The only way it changes the scenario is we now don't need a prescient post from the Oracle saying when he investigated the Apprentice. To get a good reading on what the Apprentice's bread crumbs mean we still need to either deduce the timelines of when each investigated who. So one prescient post from the Apprentice(detailing the timeline) is all we need now. :)

Info on the Cowardly Reporter (http://mafiascum.net/wiki/index.php?title=Information_Roles)

Enjoy,
Steven

NAF1138

06-26-2007, 05:46 PM

Info on the Cowardly Reporter (http://mafiascum.net/wiki/index.php?title=Information_Roles)

Enjoy,
Steven

Thanks! My place of employment seems to have decided that mafiascum.net is no longer allowed to be viewed. That is new in the last month or so :mad: . I will take a look when I get home.

Hal Briston

06-26-2007, 05:56 PM

Thanks! My place of employment seems to have decided that mafiascum.net is no longer allowed to be viewed. That is new in the last month or so :mad: . I will take a look when I get home.Ouch...that sucks (still, better mafiascum.net then the SDMB!). Well, if you're impatient:

Cowardly Reporter: The Cowardly Reporter can check a player's house to find out if they are home (anyone doing something during the night is not home), or hide in the newsroom to avoid being killed (can not hide on consecutive nights).

Hal Briston

06-26-2007, 06:02 PM

Damnit! It's time for me to leave work! Oh well...back again tomorrow...

See what this damn game does?!

DiggitCamara

06-26-2007, 06:09 PM

Idle Thoughts

06-26-2007, 06:11 PM

Thanks! My place of employment seems to have decided that mafiascum.net is no longer allowed to be viewed. That is new in the last month or so :mad: . I will take a look when I get home.

No problem. There's a Cowardly Reporter role on that site I linked to on my boards. That werewiki site?

Or just google "werewiki".

sachertorte

06-26-2007, 06:14 PM

Thanks to all who explained my thinking in stating the Priest is not night killable. Yes, technically he is, but there really isn't any reason to protect someone other than himself. The only exception from this is possibly if the Oracle roleclaims and therefore needs protection. In that case I can see the Priest dying to protect the Oracle. But that's the only case I see the Priest dying. Sorry for being confusing.

There's also a strong argument for making sure that the Oracle and Apprentice are investigating their respective halves in random order, because otherwise if they're going right down the list, the scum will be easily able to know when they've been confirmed. Or maybe that's a good thing, making them nervous?

My intention was simply to establish disjoint sets for investigations. Order would be up to the Oracle and Apprentice. A set investigation list is very very bad. Scum would simply kill townies they know were investigated, thereby eliminating the usefulness of the Oracle. Part of the oracle's power is discovering town alignment without scum knowing who!

Yes, there is the possibility (50%) that the Apprentice would be in the incorrect set for the Oracle to investigate. On the other hand, if the Apprentice is in the Oracle's set, the probability of getting hit doubles :D . So mathematically, its a wash.

Personally, I wouldn't put too much credence into the paper Mtgman cites. The analysis was for plain games with only vanilla townies and vanilla mafia. The analysis provided there doesn't apply to this game.

DiggitCamara

06-26-2007, 06:18 PM

As far as our investigative arm(the Oracle and his/her Apprentice(s)) are concerned, the Non-Believers are a superset of the Cultists. So rooting them all out should be their priority. Now the question for the town becomes, how do we sort the Cultists from the Non-Believers. This is an important exercise for three major reasons. Firstly, the Cultists are the ones trying to kill us. They should be the targets. Secondly, the Non-Believers count towards the town's side in the equation to determine a Cultist victory or not. Killing Non-Believers(as opposed to just those who investigate as Non-Believers, which includes most Cultists) advances the Cultist's win condition. Thirdly, while it is theoretically possible for the Non-Believers to take a victory away from the Believers, it's pretty damn unlikely. Let's not repeat the mistake of going for the Mason win in MIII by focusing on unlikely events when the most realistic threat is still out there.

(snip)

Nope. The Non-Believers are a non-factor. There are 3 (guaranteed) Non-Believers as of now, whereas there are 7 (guaranteed) Believers as of now. Unless there are a rash of Believer killings/dunkings/lynchings (and you better believe there will be), they are in a known minority right now.

They don't know each other.

They can't disclose their identities fearlessly, since they'll be Crusader fodder right now.

And, like you say, there is no way to root them out (unless they are about to face The Dunk). So let's forget about them (right now). They're a non-factor to town.

Fretful Porpentine

06-26-2007, 06:22 PM

I'm waiting for Autolycus to make his entrance. Then I can accuse him of being suspicious because (1) he's using a Cultist accent, (2) he's not using a Cultist accent, (3) he took too long to make an entrance, or (4) he jumped in too soon. Pick your poison!
You know, it would be hilarious if the secret role were the Autolycus, and the person who drew it had to TRY to get themselves lynched :)

DiggitCamara

06-26-2007, 06:34 PM

And an immediate followup to say holy cow, Mtgman is right. The Apprentice is going to be hard to really usefully interpret results from.
Depends.

(snip)Apprentice (Beat Cop) – He Knows who the Oracle is. (snip)

Therefore: he can throw posts the Oracle's way all Day long.

Actually, a "random vote" might be an excellent strategy.

'nuff said.

Hockey Monkey

06-26-2007, 06:42 PM

Nope. The Non-Believers are a non-factor. There are 3 (guaranteed) Non-Believers as of now, whereas there are 7 (guaranteed) Believers as of now. Unless there are a rash of Believer killings/dunkings/lynchings (and you better believe there will be), they are in a known minority right now.

They don't know each other.

They can't disclose their identities fearlessly, since they'll be Crusader fodder right now.

And, like you say, there is no way to root them out (unless they are about to face The Dunk). So let's forget about them (right now). They're a non-factor to town.

I'm going to agree here that the non-believers are basically a non-factor as Diggit said. Your basic vanilla non-believer is the same role as a believer. The Alchemist doesn't seem to have much power. He doesn't even block reliably. Maybe he can use his "scientific knowledge" to make liquor for us instead of gold since this seems to be a dry town. (Harumph!) The Psychopath will only be activated if targeted. A 1 in 30 chance right now, so I'm not worried about him right now either. Their best strategy would seem to be to play as vanilla town and go for a town win. For some reason I was under the impression that there would be as many vanilla non-believers as believers. I see that others don't think so.

NAF1138

06-26-2007, 06:46 PM

3.Saying no one looks scummy is scummy, even with the "one little thing" thrown in.Scum tells are not gonna be by the book and that particular post seems like a variant of said scum tell.

This isn't really the post I wanted to quote, but I was having a hard time finding the exact post I wanted (not that I looked all that hard).

First off, I am going to go out on a limb and defend Mad here. I don't think he was being inconsistant. It doesn't look good that he got his facts wrong, but I am not getting a strong scum vibe off that particular mistake.

And I just finished a re read, and he is right. CJ's initial post is super scummy. I agree that even with the exception, saying that "no one looks too scummy to me right now" is a big tell. But on top of that CJ's first post has not only the one scum tell that Mad already pointed out, it also has the "awe shucks" technique being employed. A second scum tell.

This isn't enough to hang him (dunk him), but it is enough for him to get my vote for now.

Vote Clockwork Jackle

HazelNutCoffee

06-26-2007, 06:54 PM

Therefore: he can throw posts the Oracle's way all Day long.

Actually, a "random vote" might be an excellent strategy.

'nuff said.
I am unable to make the connection between the Apprentice knowing who the Oracle is, and the idea of a random vote being a good idea. :confused:

HazelNutCoffee

06-26-2007, 07:09 PM

First off, I am going to go out on a limb and defend Mad here. I don't think he was being inconsistant. It doesn't look good that he got his facts wrong, but I am not getting a strong scum vibe off that particular mistake.
It wasn't the mistake, but the over-reaction that made me initially suspicious. Clockwork Jackal didn't even properly point a FOS, but Mad accused him of revenge voting and put in a vote himself.

The only thing that's keeping me from casting a vote for Mad at this point is that Clockwork Jackal's first post IS actually suspicious; especially the way he half-heartedly nitpicks at Zeriel's use of pronouns. I withhold judgement, however, until he comes back to defend himself.

NAF1138

06-26-2007, 07:12 PM

It wasn't the mistake, but the over-reaction that made me initially suspicious. Clockwork Jackal didn't even properly point a FOS, but Mad accused him of revenge voting and put in a vote himself.

That makes sense. I guess I am just used to that being Mad's playstyle, because it didn't even register as an over reaction to me. Just struck me as Mad being Mad. Weird, I didn't realize I was doing that. It is surprisingly hard not to bring history of past games (good and bad) into this.

Something to watch out for I guess.

ComeToTheDarkSideWeHaveCookies

06-26-2007, 07:13 PM

Holy Nairu, people. I step away from my desk for a couple of hours, and y'all decide to snort all of the SDMB hamsters' meth.

About the investigative strategy...

How about they both randomly select targets until either consensus deems otherwise, or either of them decide to rogue because they're sick of our collective deafening "guidance"?

SnakesCatLady

06-26-2007, 07:21 PM

storyteller0910

06-26-2007, 07:41 PM

I will be the first to admit that this kind of strategic thinking isn't really my strength, but I can't see any reason why we shouldn't employ this strategy. I only see up sides as far as the Oracle and apprentice are concerned.

Anyone else?

Well, here's one possible scenario under which it would be problematic; it is far-fetched, and I am emphatically not saying I think this is what's happening, but it's not outside the realm of possibility.

Hypothesize that sachertorte is scum. Unknownst to us, but knownst to him, the random distribution of scum just happened to place all of the Cultists, with the exception of the Prophet, among players 16-30. We follow his plan. The Oracle is now completely hamstrung - (s)he'll never get a scum reading, even if (s)he survives 15 days. The Apprentice might, but we'll be disinclined to lynch anyone on the Apprentice's say-so alone, and there's a very good chance of a few players incorrectly cleared. Bada bing, bada boom, bada our investigators are essentially worthless.

But that's only part of the problem.

The way I see it, this game is entirely about information. At the outset of the game, the scum collectively have a lot of it and the town collectively has virtually none. That's the point of the game - can the scum leverage their superior information to overcome our superior numbers? The way to maximize our chances is to make sure that the scum gets as little additional information as humanly possible. What information can we keep out of their hands? Well, ideally, the identity of the power roles, of course. But also the way the power roles will act. By "directing" the power roles in their night activities, we're taking away one of the few secrets we collectively still have. In M2, one of the best things that happened to us (the Mafia in that game) was when the town decided to put the Vigilante on a leash. Before that, the Vig was a wild-card, an unknown quantity to us - we had no idea what he'd do from night to night, and therefore no way to adapt our strategy. Once the town leashed him, they handed us precious information about how he would behave, who he would kill, on a platter.

FOR THE RECORD: I believe that the power roles should act independently, following their own noses and their own hunches. Don't give the Cultists more information about their activities than they have already (or give the Cultists a chance to help guide the pro-town power roles during the Day, which they most surely have if we keep giving them direction.

Furthermore, all of this talk about the power roles is a huge, huge mistake IMO. First of all: we're sitting here talking about how the Oracle can find the Apprentice, blah blah blah; power roles have tells, too, and the scum will be looking for them as eagerly as we're looking for scum tells. Look at what we're doing right now: discussing the Apprentice in detail, giving him or her a chance to slip up and show his or her hand, instead of talking about who might be scum, and giving the scum chances to slip up. This is a horrible plan, because it clouds the discussion and benefits no one but the bad guys.

And on that note: by all that is holy, if we start spending forty percent of our Days generating page after page of posts analyzing the math associated with the Apprentice, I will lose it. It does not matter even a little bit if the Apprentice has a 53% chance of being right or a 65% chance of being right. Battling about minutiae like that will waste a lot of time and words, and it misses the point of the role. The Apprentice, like the Beat Cop in M2, is most useful as partial, not complete evidence. If a player comes up scum on an Apprentice reading, our best bet is probably to lynch that player, particularly if it's early in the game; we have a much poorer than 50% shot of nailing scum at this stage left to our own devices. At the moment, a reading of town from the Apprentice should probably not mean much. Later in the game, think of a scum reading as having exactly the same weight as a "scum tell;" it's not definitive, but it can be added to a case in the presence of other evidence.

Now, please, please, please, can we start talking about who might be scum instead of about how the Oracle can find the Apprentice, or what have you?

Pretty please?

storyteller0910

06-26-2007, 07:56 PM

I am surprised at you story

Why? Seriously, why? Do you feel you should be above suspicion?

1.Already acknowledged my mistake

Yes. And if you were scum you would have done exactly the same thing. "acknowledged your mistake - " but left your vote to stand.

2.What you say is simply not true,there is nothing inconsistent about it . Jackal said I said something I did not say.Wether I agree with what he said I said does not matter.

Of course it does! It's not as though you said, "I like turnips," and he turned that into your saying, "Britney Spears is the most talented singer in the world." You said that talking about the composition of the game is a scum tell - something scum would do. It is perfectly reasonable to infer from this that you think talking about the composition of the game is not something the town should be doing. Why else throw out a remark that is bound to chill discussion of that issue (no one wants to exhibit a "scum tell")? And if you don't think it is something we should be doing, it is perfectly reasonable to further infer that you think the subject is not valuable. Which is what CJ said. And obviously his inference wasn't so unreasonable, because it was correct. Flimsy.

3.Saying no one looks scummy is scummy, even with the "one little thing" thrown in.Scum tells are not gonna be by the book and that particular post seems like a variant of said scum tell.

"Scum tells are not gonna be by the book?" The only reason for anyone to think of this as a scum tell is because you're citing a (virtual) book as your source. If you'd care to make an actual argument as to why what CJ said was a scum tell, have at it, but if you're going to use the book as the only basis for your vote, best not to complain when someone points out that you're not really adhering to the book.

Because I'll hazard a guess as to why saying "no one looks scummy" is often seen as a scum tell - because saying that is so plainly an effort to avoid alienating another player. It's my "making friends" theory of Being Effective Scum; tell all the townies they look like townies and they'll trust you. What CJ did - no one looks scummy except Hal - is not just "one little thing;" it's fundamentally different, in that it DOES express suspicion of someone. Not indicative of towniness on CJ's part, but not the same as the scum tell you are citing.

DiggitCamara

06-26-2007, 07:59 PM

(snip)Now, please, please, please, can we start talking about who might be scum instead of about how the Oracle can find the Apprentice, or what have you?

Pretty please?

If you insist.

Next topic on the menu: What is the "secret role" all about? Do you think Autolycus got it? How can we find out what it is all about?...

d&r

Idle Thoughts

06-26-2007, 08:05 PM

sachertorte

06-26-2007, 08:10 PM

I didn't read anything in the rules about this being a dry town. I think Hal Briston just made that up...maybe because he's afraid of liquids....afraid of liquids....hummmmm .

*looks at Hal suspiciously*

What a wonderfully fluffy post. I can't honestly say that I've contributed much towards actually finding scum, but I think this type of post by SnakesCatLady is the type of post we should try and avoid. It feels to me like someone trying to fly under the radar, but not willing to not post at all. The fluff post isn't even followed by a post of substance. FOS SnakesCatLady

storyteller makes a good point about my possibly choosing which half to send the Oracle on a wild goose chase, but to that I would say two things support my theory:1) Blaster said the roles were assigned randomly, the probability of the set 1-15 having no cultists is quite small. 2) The oracle confirming town is not a waste. M1 was won by the overpowering number of confirmed town in the game.

In summary, I would like the Oracle and Apprentice to choose for themselves what they want to do. They don't have to follow the plan. Look at the arguments, pros, cons, and make the decision for yourself.

storyteller is also correct in pointing out that we're getting a bit off track, and I apologize. Like I said, I love analyzing games and talking about theories and strategies. It's going to be hard to keep that under control for me, because I find it fun. And ultimately this exercise is for fun. But to put things on a better track, I reiterate my FOS SnakesCatLady.

ArizonaTeach

06-26-2007, 08:16 PM

From Mafia : A Theoretical Study Of Players and Coalitions in a Partial
Information Environment (http://arxiv.org/PS_cache/math/pdf/0609/0609534v3.pdf)(warning, PDF)I've listed a number of reasons I'm skeptical of this analysis holding for this particular format, and I'm open to discussion about other strategies.Um, ok...so why bring it up in the first place? Are you at least backing away from the idea of random.org, because I gotta say, you still haven't addressed my points as to why I think that's a phenomenally bad idea that can only, only benefit scum.

You know what I don't need? You know what that is? I'll tell you what I don't need. A Day One smackdown between Mad and storyteller, because I am absolutely terrified about backing one horse at the expense of the other when it comes down to those two. My biggest concern is that one reason Mad is voting for CJ is because Jackal said I said something I did not saybut when it's pointed out that Mad said CJ said something CJ did not say, Mad dismisses it with My apologies for saying revenge vote when it was just a revenge FOS and Already acknowledged my mistake as if that just makes it go away. If it's enough to dunk CJ...?

Hal Briston

06-26-2007, 08:20 PM

Snipped.

What makes you so sure of these numbers? I'm curious.Because there are three Non-Believer roles and seven Believer roles.

HazelNutCoffee

06-26-2007, 08:25 PM

Fretful Porpentine

06-26-2007, 08:29 PM

Because there are three Non-Believer roles and seven Believer roles.
That doesn't necessarily follow, since we don't know the numbers of people with each role. Similarly, there are ten roles with special powers of some sort, and three without any, but that certainly doesn't mean that players with power roles are a clear majority!

SnakesCatLady

06-26-2007, 08:51 PM

What a wonderfully fluffy post. I can't honestly say that I've contributed much towards actually finding scum, but I think this type of post by SnakesCatLady is the type of post we should try and avoid. It feels to me like someone trying to fly under the radar, but not willing to not post at all. The fluff post isn't even followed by a post of substance. FOS SnakesCatLady

storyteller makes a good point about my possibly choosing which half to send the Oracle on a wild goose chase, but to that I would say two things support my theory:1) Blaster said the roles were assigned randomly, the probability of the set 1-15 having no cultists is quite small. 2) The oracle confirming town is not a waste. M1 was won by the overpowering number of confirmed town in the game.

In summary, I would like the Oracle and Apprentice to choose for themselves what they want to do. They don't have to follow the plan. Look at the arguments, pros, cons, and make the decision for yourself.

storyteller is also correct in pointing out that we're getting a bit off track, and I apologize. Like I said, I love analyzing games and talking about theories and strategies. It's going to be hard to keep that under control for me, because I find it fun. And ultimately this exercise is for fun. But to put things on a better track, I reiterate my FOS SnakesCatLady.

By all means, let's not have a good time and enjoy the game. Let's all wait around for someone to post something of substance.

You go first.

Queuing

06-26-2007, 09:01 PM

Wow, a guy leaves work, goes the gym and comes home, eats dinner, does some tech support for his parents, and BAM page 8? I would like to reiterate that I will be gone from this Friday evening until Monday evening. Please don't give me to much homework :).

Anyway, this post (http://boards.straightdope.com/sdmb/showpost.php?p=8720167&postcount=350) is one reason I was hoodwinked by storyteller in M2. So reasonable, damn you man! n all seriousness I agree with a lot of this post, ok all of it, particularly the whole math point, lol.

A totally randomized vote, as suggested by MtgMan, is worse than the Town not dunking anyone. We'd gain no information while probably killing a Townie. If we want to forfeit our turn, better to not dunk anyone. That way we won't kill a Townie. Never mind the logistics of getting a simultaneous vote on this board.

Random votes as advocated by me are a tool for bootstrapping the discuss in the early part of the first Day. The final vote decision needs to be something accountable.

I'm waiting for Autolycus to make his entrance. Then I can accuse him of being suspicious because (1) he's using a Cultist accent, (2) he's not using a Cultist accent, (3) he took too long to make an entrance, or (4) he jumped in too soon. Pick your poison! :D

This strikes me as odd, considering his earlier posts, he votes for USC Diver, post 221 (http://boards.straightdope.com/sdmb/showpost.php?p=8718947&postcount=221) and then unvotes him post 248 with no posts from USC Diver in between the 2 (http://boards.straightdope.com/sdmb/showpost.php?p=8719217&postcount=248) .

I just really don't get how Pleonast's vote for USCDiver was a bootstrap for discussion when the person he voted for never came in and said anything, yet he still unvoted rather quickly.

As far as I can tell the only thing that happened in between these actions was NAF vehemently arguing against any sort of random vote.

Pleonast is not the only one to have done this however.

here in post 225 (http://boards.straightdope.com/sdmb/showpost.php?p=8718968&postcount=225) DiggitCamara does the random vote thing for Cookies. In all fairness he says its tradition. Ok, maybe, I don't know about the tradition thing. I did not follow M3 or the pirates game at all. I was burned out (and somewhat annoyed with myself <again damn you story! lol>) from M2. Still its odd, particularly when he unovtes in 251 (http://boards.straightdope.com/sdmb/showpost.php?p=8719255&postcount=251) . Again with Cookies having said nothing.

To me these actions are a LOT more suspicious then MtS voting for Jackal. My FOS is pointed.

storyteller0910

06-26-2007, 09:04 PM

Hey, I don't mind fluffy posts. They keep the game somewhat more lighthearted. Plus, Hal's assertion that this is a dry town concerns me as well. Concerns me muchly. :dubious:

storyteller, didn't Clockwork Jackal say Zeriel was supicious, rather than Hal, in his earlier post?

Yes, yes he did. I have absolutely no idea why I thought it was Hal. Sorry, Hal. Maybe it's cause you frickin' killed me in the last game.

storyteller0910

06-26-2007, 09:07 PM

.
I just really don't get how Pleonast's vote for USCDiver was a bootstrap for discussion when the person he voted for never came in and said anything, yet he still unvoted rather quickly.

DiggitCamara does the random vote thing for Cookies. In all fairness he says its tradition. Ok, maybe, I don't know about the tradition thing. I did not follow M3 or the pirates game at all. I was burned out (and somewhat annoyed with myself <again damn you story! lol>) from M2. Still its odd, particularly when he unovtes in 251 (http://boards.straightdope.com/sdmb/showpost.php?p=8719255&postcount=251) . Again with Cookies having said nothing.

To me these actions are a LOT more suspicious then MtS voting for Jackal. My FOS is pointed.

This is a very good point. Pleonast, care to respond? Diggit?

HazelNutCoffee

06-26-2007, 09:07 PM

I went back to trace this whole Clockwork Jackal / MadTheSwine thing, and here's what I found:

-CJ says that everyone so far does not look scummy, except for perhaps Zeriel and her(his?) unfortunate choice of "town" versus "us."

-Hal speculates on distribution of roles amongst players.

-MtS states that such speculation, as well as CJ's observation of no one looking guilty at the moment, could be read as a scum tell.

-CJ points out that he DID post an example of what he thougt was scummy, and suggests a FOS towards MtS for discouraging Hal's speculation, when such speculation could be important.

-MtS accuses CJ of twisting his words AND of revenge voting, and votes for him. He also says that he never said role-distribution was not important; just that encouraging discussion over it is a scum tell.

-I point out to MtS that CJ never voted for him in the first place.

-MtS acknowledges this but keeps his vote.

I still think that CJ's original post looks suspicious, but I also still think that MtS seemed a tad to eager to vote for CJ, even after he realized that CJ never voted for him. CJ's supposed "twisting" of MtS's words doesn't seem that crafty to me, either - is it really that farfetched to read "X is a scum tell!" as an attempt to dissuade people from doing X?

So, two questions:
Clockwork Jackal, are you going to defend yourself? You already have two votes for you.
MadTheSwine, what, exactly, are you basing your vote for CJ on?

ComeToTheDarkSideWeHaveCookies

06-26-2007, 09:26 PM

I have gotten at least one first Day vote in every game that I've played (2 town, 1 scum), and the justifications have pretty much all been tongue-in-cheek. Actually, if I'm remembering correctly, at least one tongue-in-cheek vote popped up a few Days into one of the past games (when I was a townie). I freely admit that I have not done any post-mortem analysis as to the orientation of any of the folks casting such votes, and I really don't want to. Please don't make me. I've got my hands sufficiently full with this thread, so I'm loathe to go snorfing around in the old ones.

I will say (in keeping with the healthy skepticism that I advocated earlier), that while there are "traditional" precedents for voting for me in such a manner, that doesn't mean Diggit isn't trying to dress scummy motives up in traditional garb.

I don't know why, but in my minds eye the traditional garb I'm seeing is a kilt and tam.

USCDiver

06-26-2007, 10:26 PM

Whew... ok I'm here. Just wanted to say off the bat, I'm new to Mafia. I began following M3 and the Forbidden thread on about Game Day 6 or so. So I will admit I know little about complex strategies.

Sorry to be coming in so late, but I was working all afternoon and it took me 3 hours (!!) to catch up in the last 5 hours of strategery.

I agree with stimulating discussion as opposed to all out random voting, particularly when there is little information. Even if that means randomly voting for players to lure them out (thanks Pleonast, but I'm the springboard kind of diver not the SCUBA type; if that helps y'all keep me separate in your minds from Scuba_Ben).

My thoughts so far:
Too much discussion of what happened in previous games. I understand that there is a lot of experience to draw off of in those games, but I wasn't a participant in them and don't have the hours it would take to read through all of them. Also this is a different game with many new and different roles. So while I think it's important to discuss strategy at some level, I'd like to avoid too much 'let's do it this way because we did that in M1 and it worked!"

Regarding splitting the duties of the Oracle and Apprentice (which I don't think is a good idea in the end), I haven't run the numbers, but it would seem like a bad idea for us to split their populations evenly if they aren't even players. In other words, the Oracle is 100% and the Apprentice is (initially) 50% so it would seem like a better strategy to 'give' the Oracle 67% of the players to investigate and the Apprentice 33%. The Apprentice could investigate the same players multiple times and could make stronger conclusions based on the results.

In the end however, I agree with sachertorte that getting too bogged down in discussion of strategy now could be counter-productive. I'm enjoying watching y'all deconstruct each other's posts at this point and will hopefully pick up on some scum tells as the game goes on.

For now though (since I'm still bitter),
FOS at Hal Briston
for making a last minute deal in Haggle and moving from last to first and stealing my victory! I was hoping this was going to be a Mafia themed game so I could make some kind of Mafia related accusation, but I guess I'll have to say he must have made a deal with Sekham to win that game.

Keep in mind everyone, there’s still 4.5 days until the end of the Day.

Pasta

06-26-2007, 10:41 PM

Regarding the investigatory branch:
I agree that we should not try to openly control the oracle or the apprentice. There is a small risk of a wasted investigation if they act independently, but chances are good that they are both trying to investigate someone who seems worthy of investigating. In the rare case of overlap, I'll happily take the prospect of having two trails to follow since neither trail will be easy to identify (assuming they're hiding clues cleverly.) And, as was just mentioned by storyteller0910, limiting the choices of pro-town special roles gives the scum useful information that we could have easily held from them. (This seems an important enough point that I will likely trust storyteller0910 for quite a while now for bringing it up.)

Miscellany:
Here's an analysis regarding Idle Thoughts that I want to put out there. Folks should comment.

Post 292 (http://boards.straightdope.com/sdmb/showpost.php?p=8719654&postcount=292) is the source of the following quotes.

<snip>...there's a better chance winning with them.....however like you say (and with your own experience as a group that could take a single-only win) it would open the door to a non-believer (but not cultist) win alone.

<snip>...basically what FCOD put better.
This smells, I dunno, ... thin. It's like he wants to sound like he's churning over the possibilities while trying not to commit or take attribution. "I conclude A. However, not-A is good, too. Also, other people are really saying all of this, not me, so, uh, nothing to see here."

The next bit of #292 is fairly free of content, though it again smells like it's supposed to look like it has some. (Perhaps that's just your style, Idle Thoughts?)

Then, regarding the killing of lurkers:

Maybe just me but I feel that's just as bad. Seems like giving an easy out to scum (who know who is what) to just gang up on someone who's not posting as much. If someone is lurking too much, I just feel a sub out would be the best course rather than a pile on.
Idle Thoughts, could you explain what you mean by "easy out" to scum? The first two sentences ("just as bad" -- as what?) seem like a handwavy attempt to suppress the idea of killing under-the-radar players. It seems to me that scum lurking is an established strategy. While there will certainly be veteran scum who are skilled enough to post a lot without giving much away, I think some scum will certainly fly low. (And, if I may meta-game for a moment, it is also reasonable since they have no time off; they may naturally take it easier in the days.) I'm not sure why you want to switch out likely scum rather than kill them dead. (I'll admit I haven't thought through the numbers aspect of this.)

FOS: Idle Thoughts

Zeriel

06-26-2007, 10:50 PM

HazelNutCoffee, I'm a him. Either that, or MrsZer is VERY charitable in feeding my delusion.

Blaster Master

06-26-2007, 11:02 PM

FOS at Hal Briston

Not a big deal, but please try not to use Red or Blue for FOSing. In the future, it may lead to me miscounting votes which could be bad.

I think the nominal color for FOS has become Dark Orange, but obviously, that's up to you all.

That said, if you need my attention, like for a vote count, just to make it standard (as in the rule set), it'd be nice to know that Green means I need to respond.

HazelNutCoffee

06-26-2007, 11:22 PM

HazelNutCoffee, I'm a him. Either that, or MrsZer is VERY charitable in feeding my delusion.
I'll take your word for it. ;)

Lurking is not necessarily a scum tell, but it does make you look suspicious. Either way, it's probably not a good strategy - whether you are scum or townie, you certainly don't want to give everyone else a cause to doubt your faith. Plus, the whole point of the game is social interaction. It's no fun when people lurk!

I think what Idle Thoughts meant is that automatically suspecting lurkers make it easier for scum to lead bandwagons for drowning townie lurkers without raising suspicions.

Looking over my notes, the unofficial vote count so far looks like:
Clockwork Jackal: MadTheSwine, NAF1138
MadTheSwine: storyteller0910
USCDiver: Pleonast
ComeToTheDarkSideWeHaveCookies: DiggitCamera

I included the un-votes as well. Sorry, haven't been keeping up with the FOSs.

Blaster Master

06-26-2007, 11:25 PM

I've received a few PMs with various questions about the Apprentice Role, I will try to clarify the process.

First, a "correct" reading is whatever the Oracle would read as specified in the rule summary or the chart that panamajack posted.

Here is how I will process the Apprentice's investigation. Say he wants to investigate PlayerX. First, I flip a coin (it starts out unbiased at 50%, but if/when his accuracy improves, the coin will be biased to land on heads at the appropriate rate). If it comes up heads, I return exactly what the Oracle would read off of PlayerX. If it comes up tails, I will take all living players (including the Apprentice himself and PlayerX), stick them in a hat (metaphorically speaking) and pull out a player and the Apprentice will receive whatever the Oracle would have read off of him. If the player investigated is the Avatar, this process is skipped and he receives the appropriate response.

Essentially, this means that incorrect responses are more likely to come from the more prevalent roles versus being uniformly distributed like it was in M2.

Also, upon the death of the Oracle, the process changes such that when the biased coin comes up tails, the Apprentice will receive no reading.

Finally, if/when the Oracle discovers the Apprentice, even though he is under his tutelage, they will not be able to discuss strategy at night (he's too busy teaching him, afterall).

With regard to the Crusader, if he attempts a kill, regardless of whether if succeeded or failed, he will enter the next Night in a "just killed" status and thus he will only have a 50% success rate and he will also be able to opt out of killing.

To preempt a potential question, with regard to biased coin flipping, I thought Random.org had it, but apparently not. Thus, I'll simply generate a random integer in an appropriate range and set a threshold (eg, 60%, between 1-10, if it's 6 or less "heads" otherwise "tails").

Blaster Master

06-26-2007, 11:34 PM

Clockwork Jackal - ( 2 ) - MadTheSwine NAF1138
MadTheSwine - ( 1 ) - storyteller0910

Vote History

Pleonast - Vote - USCDiver - 221
DiggitCamara - Vote - ComeToTheDarkSideWeHaveCookies - 225
MadTheSwine - Vote - Clockwork Jackal - 244
Pleonast - Unvote - USCDiver - 248
DiggitCamara - Unvote - ComeToTheDarkSideWeHaveCookies - 251
storyteller0910 - Vote - MadTheSwine - 281
NAF1138 - Vote - Clockwork Jackal - 344

---------------------------------------------------------------------
I apologize that it's a little sloppy, I'm still working on getting it to format well with a simply copy and paste from my spread sheet.

Hockey Monkey

06-26-2007, 11:58 PM

Some general thoughts before bed:

On Lurking: Lurking is a legitimate strategy, and I will call people out for it. However, with this many people in the start of the game, lots of people may appear that they are lurking when in fact they aren't. We do have some inexperienced players in the game, and when I was new at this, I was a little afraid to post. Lurking will come into play moreso as we progress. I will be more suspicious of people who post much but say little, as I feel this is a form of lurking. Again, with this many people, it is hard to wade through fluff posts. I will be trying to post with more content and try to address several things in one post. I agree that we should be light-hearted and have fun. Jokes and smileys should abound. If you won't be around for a significant period of game time, please let us know.

On Random Votes: Already we have a lot to pick through and FOS people for. My vote today will be based on something I think is suspicious.

On Bringing Up Previous Games: I'm against using someone's style or posts from a previous game as a basis for a FOS or vote. I can certainly see where comparing someone's playstyle could be advantageous, but players like myself are developing their style, and unlikely to be parallel to previous games. My posting habits in the Pirates game changed significantly over the course of the game as I got my footing. That said, I will try not to even bring up previous Mafia games at all. As someone else noted, not everyone wants to wade back through other games.

General Advice for Newbies: Post like you are going to die. If you are thinking something and don't let the rest of us know, then we can't draw any conclusions when you die. Death for a townie is not to be feared. Above all, have fun. This game is a lot of work, but it's fun work. And when you can do it at work, it's even more fun. :p

Idle Thoughts

06-27-2007, 12:04 AM

Miscellany:
Here's an analysis regarding Idle Thoughts that I want to put out there. Folks should comment.

Post 292 (http://boards.straightdope.com/sdmb/showpost.php?p=8719654&postcount=292) is the source of the following quotes.

This smells, I dunno, ... thin. It's like he wants to sound like he's churning over the possibilities while trying not to commit or take attribution. "I conclude A. However, not-A is good, too. Also, other people are really saying all of this, not me, so, uh, nothing to see here."

Yeah, a lot of people put me on "FOS" last game for this too. Me, I don't get it. I'm actually trying to keep my mind open to all possiblities. How is that so weird to people? It's not that I'm trying not to make any solid thoughts right now..it's just that I'm making observances and thinking aloud and trying to keep all possibilities open.

Idle Thoughts, could you explain what you mean by "easy out" to scum?

Because if a lurker happens to be pro-good, how easy would it be just to gang up on them for the reason "well, they're lurking" and bump them off?

The first two sentences ("just as bad" -- as what?) seem like a handwavy attempt to suppress the idea of killing under-the-radar players.

Just as bad as what NAF was saying (it was who I was quoting, by the way) about how bad he thought voting randomly is.

And I remain in the belief that if there are under the radar players, it'd be better, I think, rather than vote for someone who may be busy or not able to take the time to defend themselves, to sub them out instead.

I'm not sure why you want to switch out likely scum rather than kill them dead. (I'll admit I haven't thought through the numbers aspect of this.)

Funny you say likely scum. What about likely town? As far as I can tell, lurking didn't become a play until last game.

Idle Thoughts

06-27-2007, 12:05 AM

storyteller is also correct in pointing out that we're getting a bit off track, and I apologize. Like I said, I love analyzing games and talking about theories and strategies. It's going to be hard to keep that under control for me, because I find it fun. And ultimately this exercise is for fun.

Snipped mucho.

Boy do I hear you. I do this too, too much sometimes. :(

Because there are three Non-Believer roles and seven Believer roles.

Okay. I read it and took it as there are at least three/seven of regular, in general types of each side. Not nessacarily ones with roles.

My thoughts so far:
Too much discussion of what happened in previous games. I understand that there is a lot of experience to draw off of in those games, but I wasn't a participant in them and don't have the hours it would take to read through all of them. Also this is a different game with many new and different roles. So while I think it's important to discuss strategy at some level, I'd like to avoid too much 'let's do it this way because we did that in M1 and it worked!"

Snipped.

Yeah, I don't know why this is but it just seems like people so easily do this all the time. Either that or they say things like "well town would NEVER do/say things like this!" and assume that every REAL pro-good player would act alike.

Not a lot or often, but it happened a lot in the last game, I noticed.

Blaster Master

06-27-2007, 12:19 AM

Here's the format I'm planning on using. If anyone has any suggestions to make it more useful for you, let me know and I'll see what I can do.
----------------------------------------------------------------

Clockwork Jackal (2) - MadTheSwine, NAF1138
MadTheSwine (1) - storyteller0910

Voter - Action - Votee - Post
Pleonast - Vote - USCDiver - 221
DiggitCamara - Vote - ComeToTheDarkSideWeHaveCookies - 225
MadTheSwine - Vote - Clockwork Jackal - 244
Pleonast - Unvote - USCDiver - 248
DiggitCamara - Unvote - ComeToTheDarkSideWeHaveCookies - 251
storyteller0910 - Vote - MadTheSwine - 281
NAF1138 - Vote - Clockwork Jackal - 344

Clockwork Jackal

06-27-2007, 12:30 AM

Blaster Master

06-27-2007, 12:56 AM

Hi guys,

I need to be subbed out. My boyfriend's mom was in some kind of car accident this afternoon. They said it was minor at first, but now they're saying something about punctured lung or lung damage or something, so my BF has flown down to NC to be with her, and I'm waiting here with the dog to hear more details. I'm a fucking wreck right now, and I don't think I'm in any shape to play this game.

Blaster, I hope you can find someone to sub in for me, thanks for letting me play, for however brief it was.

Thanks guys,

Clockwork Jackal

I'm sure I can speak for everyone here in hoping she gets well soon and that we're sorry you have to leave so soon.

I have already PMed the sub list. The first to respond will get the spot.

zuma

06-27-2007, 02:49 AM

Jeez.... 8 pages already? This may have been a mistake to embark on this :)

I jotted down a few notes

random voting: I'll just say I see nothing wrong with it to start the game out and start discussion, but I'm not going to get into any distracting arguments about it or judge someone one way or another based on it.

MHaye, post 258:

I agree with this; I always feel that random voting is a mistake. As is pressure-voting - voting someone just to get them to say something specific. I never do them, and neither will I respond to them, even if it means I get lynched (or, in this case, drowned). I'll go along with FOSs though, because they can't be turned into a bandwagon

Holy crap! Should we talk about the weather in here? Putting pressure on people through accusations and votes is how we catch scum. And you won't respond even if it means you get lynched? If you're town, you don't want to get lynched and should respond if you're in trouble. I'm not a fan of this cavalier attitude.

sachertorte: post 262

Speaking of voting, something I noticed in M3 that was driving me bonkers was the very low vote counts that were getting people lynched. Lack of consensus is bad. I don't know how we can force ourselves to reach a consensus in only 5 days, but letting a relatively small number (25%!) of votes determine the lynch was very bad for the town. Everyone needs to vote even if 'unsure,' otherwise we risk giving scum easy cover. So I state right now, I'll suspect anyone who gets to the end of the day with no registered vote..

Sing it, brother. I too will be suspicious of non-voters this game. It'd be way too easy for scum to sit back and let a few misguided townies lynch someone they know to be innocent, and have no blood on their hands towards the end. We need as much information as possible, and everyone needs to go on the record as to who they want lynched each day.

Mtgman's strategy of actually lynching someone randomly: Others have stated the obvious flaws in this strategy, and I don't think there would be many here who would actually subscribe to doing such a thing.

Storeyteller, post 350:

In M2, one of the best things that happened to us (the Mafia in that game) was when the town decided to put the Vigilante on a leash. Before that, the Vig was a wild-card, an unknown quantity to us - we had no idea what he'd do from night to night, and therefore no way to adapt our strategy. Once the town leashed him, they handed us precious information about how he would behave, who he would kill, on a platter.

Interesting, especially after all the crap I took regarding BM's "choice" of who to vig in M2. I think I'd rather have the Oracle and apprentice make the ultimate decision of who to investigate. I guess there isn't anything wrong with recommendations, but those could be scum-influenced.

Finally, I too would like to hear from Pleonast and DiggitCamera regarding the points Queuing brought up in post 360 (http://boards.straightdope.com/sdmb/showpost.php?p=8720379&postcount=360)

I just really don't get how Pleonast's vote for USCDiver was a bootstrap for discussion when the person he voted for never came in and said anything, yet he still unvoted rather quickly.

As far as I can tell the only thing that happened in between these actions was NAF vehemently arguing against any sort of random vote.

Pleonast is not the only one to have done this however.

here in post 225 (http://boards.straightdope.com/sdmb/showpost.php?p=8718968&postcount=225) DiggitCamara does the random vote thing for Cookies. In all fairness he says its tradition. Ok, maybe, I don't know about the tradition thing. I did not follow M3 or the pirates game at all. I was burned out (and somewhat annoyed with myself <again damn you story! lol>) from M2. Still its odd, particularly when he unovtes in 251 (http://boards.straightdope.com/sdmb/showpost.php?p=8719255&postcount=251) . Again with Cookies having said nothing.

To me these actions are a LOT more suspicious then MtS voting for Jackal. My FOS is pointed.

HazelNutCoffee

06-27-2007, 07:49 AM

Clockwork Jackal, I hope she's okay. We're sorry to see you go, but don't worry about it - it's just a game, and in the early stages at that. :)

*pours another cup of coffee and waits for someone to post something incriminating*

Scuba_Ben

06-27-2007, 08:09 AM

I'm still trying to figure out the analyses, and I am still stuck with three possibilities for this first Day:

1. Find a bandwagon and join it. (Not a good idea, I won't learn anything from it.)
2. Vote at random. (Not a good idea, I won't learn how to read posts from it.)
3. Vote for none of the above. (Very not a good idea, as people have already stated their willingness to hunt down people who don't vote.)

Hopefully by the time we reach the end of the Day, I'll have learned enough to do something slightly more useful.

Meanwhile, what worries me strategy-wise is, as previously noted, the Oracle and Apprentice will ID the Cultists only as Nonbelievers, not as confirmed scum. That's going to be a limiting factor.

In other news, a follow-up:I'm the springboard kind of diver not the SCUBA typeNobody's perfect :)

storyteller0910

06-27-2007, 08:20 AM

I'm still trying to figure out the analyses, and I am still stuck with three possibilities for this first Day:

1. Find a bandwagon and join it. (Not a good idea, I won't learn anything from it.)
2. Vote at random. (Not a good idea, I won't learn how to read posts from it.)
3. Vote for none of the above. (Very not a good idea, as people have already stated their willingness to hunt down people who don't vote.)

Hopefully by the time we reach the end of the Day, I'll have learned enough to do something slightly more useful.

I've said this before, in a past life, but it's worth repeating here:

Vote for the person you think is most likely to be scum.

Don't vote on a bandwagon because it's a bandwagon. Don't avoid a bandwagon because it's a bandwagon. Don't decline to vote unless you have a damn good reason. Don't vote for any meta-game reason.

If you are town, and if you want to be as helpful as possible, with a very few exceptions vote for the person you think is most likely to be scum.

sachertorte

06-27-2007, 08:38 AM

Hey, I don't mind fluffy posts. They keep the game somewhat more lighthearted.

I agree, but if that's all someone is posting -- it's bad. For example, HazelNutCoffee has fluff posts and discussion posts. I'm fine with that. As far as I recall, SnakesCatLady made one post this gameday, that says nothing. That looks suspicous to me. Sorry for killing SnakesCatLady's 'fun.' Maybe a nice swim will refresh her.

For those of you just joining us, it's not really 8 pages. The first four are pre-game chatter, it's only 4.

And... I have to violate my promise not to discuss the Oracle and Apprentice separation strategy; honestly, I think I'll explode otherwise. One of the flaws storyteller pointed out is that the town doesn't know my alignment, so I could be manipulating how the Oracle investigates to scum's advantage. Also, letting scum know which subset the Oracle is investigating gives them undue information. The problem is, how do we coordinate the Oracle and the Apprentice without telling scum the plan. Well, there is one bit of information that only the Oracle and Apprentice both have that can be used to coordinate without tipping anything off to the scum. Both Oracle and Apprentice know the identity of the Oracle.

Proposed:
Divide the set of players into two disjoint subsets: players 1-15 and players 16-30.
The Oracle will investigate the set of players to which the Oracle does not belong.
The Apprentice will investigate the set of players to which the Oracle does belong.
Voila!
This method achieves seperation without tipping off information to the scum. The only flaw I see now is that information will be revealed when the Oracle is killed or outed. Anyway, blah blah blah. Oracle and Apprentice, choose to follow or ignore on your own.

Actual game content stuff:
I'm quite curious about the MadTheSwine - ClockworkJackal discussion. In CJ's early post the phrase I think I'm just gonna play as though I'm getting killed tonight, struck me as odd, but I dismissed it because in watching other games statments like this were flashing at me and I was wrong nearly all the time.

FlyingCowOfDoom

06-27-2007, 08:39 AM

Thanks to all who explained my thinking in stating the Priest is not night killable. Yes, technically he is, but there really isn't any reason to protect someone other than himself. The only exception from this is possibly if the Oracle roleclaims and therefore needs protection. In that case I can see the Priest dying to protect the Oracle. But that's the only case I see the Priest dying. Sorry for being confusing.Having played as the Doctor in M4, I can tell you that there are other reasons for protecting another player (although certainly not on the first Night). For example, if one of the pro-town power roles is outed it may be wise for the Priest to protect that person. Of course, the Cultists might anticipate this and kill the Priest instead, so it can be difficult to decide. The Priest knows that the Cultists know, so he can clearly not protect himself. But the Priest knows that the Cultists know that the Priest knows, so he can clearly not protect the other player! IT HAS WORKED! YOU'VE GIVEN EVERYTHING AWAY! I KNOW WHO YOU'RE GOING TO KILL!!

Sorry, flashbacks...

First off, I am going to go out on a limb and defend Mad here. I don't think he was being inconsistant. It doesn't look good that he got his facts wrong, but I am not getting a strong scum vibe off that particular mistake.In general, I believe that it's scummy to defend another player on the first day. Only the scum have information about roles, and it's a known scum strategy to warm up to a townie by defending him/her. There's not really any good reason for a townie to defend someone else this early in the game. NAF, J'accuse!

Vote NAF1138

--FCOD

FlyingCowOfDoom

06-27-2007, 08:48 AM

Proposed:
Divide the set of players into two disjoint subsets: players 1-15 and players 16-30.
The Oracle will investigate the set of players to which the Oracle does not belong.
The Apprentice will investigate the set of players to which the Oracle does belong.
Voila!
This method achieves seperation without tipping off information to the scum. The only flaw I see now is that information will be revealed when the Oracle is killed or outed. Anyway, blah blah blah. Oracle and Apprentice, choose to follow or ignore on your own.I urge the investigators not to follow this plan. What if the Apprentice is not in the Oracle's set? He'll never learn. What if the scum aren't evenly distributed among the two sets? It might make it harder for the Oracle to find one. No offense to sachertorte, but it's just a bad idea. The investigators should investigate whomever the hell they feel like. I think the best plan would be that the Oracle and Apprentice investigate players from whom they get a strong scum or a strong town vibe. I think redundant readings are a non-issue. At this point the Apprentice isn't that useful anyway.

--FCOD

Zeriel

06-27-2007, 09:13 AM

In general, I believe that it's scummy to defend another player on the first day. Only the scum have information about roles, and it's a known scum strategy to warm up to a townie by defending him/her. There's not really any good reason for a townie to defend someone else this early in the game. NAF, J'accuse!

Only one TINY caveat: what if they're both masons (monks)?
Granted, I'd expect that a monk wouldn't take that risk of outing himself and a buddy unless there was imminent danger of an actual lynching of a monk. Therefore, it might be scummy unless it's later in the day and the someone getting defended has a decent plurarity. Even then, tread with caution.

In other news, you make some great points on the Oracle/Apprentice thing and now I'm not sure what to think--but I'm leaning towards the default of "Power Roles shouldn't be controlled by the majority in open discussion" just due to the point someone else made--that lets scum have a chance to influence them.

At this point, I'm only getting a few low-level scum vibes, but at least I've got a set of about five players I can suspect with varying degress--namely, Clockwork Jackal('s replacement), MadtheSwine, NAF1138, Pleonast, and DiggitCamera--Jackal for nitpickiness in pronouns to try to hang one on ME, Mad and NAF for the reasons FlyingCowofDoom stated, and Pleo and Diggit for random voting because honestly, I do think it hurts town when you do stuff without thinking about it.

That's roughly in the order I'd choose to vote on them right now, if I thought I had enough information to vote. I don't.

HazelNutCoffee

06-27-2007, 09:21 AM

Pleonast

06-27-2007, 09:33 AM

Random votes as advocated by me are a tool for bootstrapping the discuss in the early part of the first Day. The final vote decision needs to be something accountable.This strikes me as odd, considering his earlier posts, he votes for USC Diver, post 221 (http://boards.straightdope.com/sdmb/showpost.php?p=8718947&postcount=221) and then unvotes him post 248 with no posts from USC Diver in between the 2 (http://boards.straightdope.com/sdmb/showpost.php?p=8719217&postcount=248) .I mistook Scuba_Ben for USCDiver and retracted my vote because of that. I don't feel strongly enough about a random vote to reinstate it. I haven't made another random vote because it feels like this thread has enough discussions going. Seems consistent to me.

Also, bootstrapping discussion is different than poking someone with a vote (i.e., voting for someone to get a response from them). A poke-vote is not random, anyway.

USCDiver: springboard--Thanks! I'll keep that mental picture.

To preempt a potential question, with regard to biased coin flipping, I thought Random.org had it, but apparently not. Thus, I'll simply generate a random integer in an appropriate range and set a threshold (eg, 60%, between 1-10, if it's 6 or less "heads" otherwise "tails").Ah, I was visualizing you soldering little weights onto a fifty-cent piece.

sachertorte: clever plan! But I think it's best our power roles do not follow any predictable plan.

I feel like I should vote for someone. Hmm...
Vote MadTheSwine
For shaky reasoning. I'm not gung-ho about dunking Mad, but he's highest on my likely scum list.

sachertorte

06-27-2007, 09:34 AM

What if the Apprentice is not in the Oracle's set? He'll never learn.
Mathematically, this is a wash.
No seperation yields a 1 in 29 chance of the Oracle targetting the Apprentice.
Seperation yields 50% * 1 in 15 chance of the Oracle targetting the Apprentice.
They are roughly equal.

The investigators should investigate whomever the hell they feel like. I think the best plan would be that the Oracle and Apprentice investigate players from whom they get a strong scum or a strong town vibe. I think redundant readings are a non-issue. At this point the Apprentice isn't that useful anyway.

--FCOD
You are welcome to your opinion. Like I said beofre, I state the policy for the investigators to choose whether to follow it or not. I don't think we should direct them to follow it, and I don't think we should direct them not to follow it either. Let them choose.

I think overlap is a very bad thing, and I'm suspiscous of your casual outlook the the efficacy of the Apprentice and the deteriment of overlapping investigations. You even go so far as to direct both Oracle and Apprentice to target strong-town and strong-scum vibe people. In following this, they are even more likely to overlap. Furthermore, I would argue that the Oracle and Apprentice should do the oppose what you sugget and specifically target people who are, at the moment, unlikely to get dunked (i.e., not strong-scum) and unlikely to get night killed (not strong-town). If they investigate strong-scum strong-town, then there is a higher change that the investigation goes to waste if that person gets dunked/nightkilled.
The course of action proposed by FlyingCowOfDoom strikes me as very anti-town.

Vote FlyingCowOfDoom
I vote for FlyingCowOfDoom not becuase of disagreement over the seperation theory, but for the contradictory suggestions that investigators do their own thing and then encouraging them to target strong-scum and strong-town, which I feel hurts the town.

FlyingCowOfDoom

06-27-2007, 09:37 AM

Only one TINY caveat: what if they're both masons (monks)?
Granted, I'd expect that a monk wouldn't take that risk of outing himself and a buddy unless there was imminent danger of an actual lynching of a monk. Therefore, it might be scummy unless it's later in the day and the someone getting defended has a decent plurarity. Even then, tread with caution.A mason that is that foolish will be picked off quickly anyway. This early in the game, it's just scummy.

In other news, you make some great points on the Oracle/Apprentice thing and now I'm not sure what to think--but I'm leaning towards the default of "Power Roles shouldn't be controlled by the majority in open discussion" just due to the point someone else made--that lets scum have a chance to influence them.

At this point, I'm only getting a few low-level scum vibes, but at least I've got a set of about five players I can suspect with varying degress--namely, Clockwork Jackal('s replacement), MadtheSwine, NAF1138, Pleonast, and DiggitCamera--Jackal for nitpickiness in pronouns to try to hang one on ME, Mad and NAF for the reasons FlyingCowofDoom stated, and Pleo and Diggit for random voting because honestly, I do think it hurts town when you do stuff without thinking about it.

That's roughly in the order I'd choose to vote on them right now, if I thought I had enough information to vote. I don't.A fact about this game is that the town has NO information to work with on the first day. The only thing we have to go on is hunches, barring a major slipup by a scum. You will probably not get "enough information to vote" before the end of the day. You should vote for the person you think is the most scummy. Voting tends to make people nervous and more likely to make a mistake, so IMO it's good to vote early on for someone.

Don't forget that you can always unvote should you change your mind.

--FCOD

Captain Klutz

06-27-2007, 09:38 AM

I urge the investigators not to follow this plan. What if the Apprentice is not in the Oracle's set? He'll never learn. What if the scum aren't evenly distributed among the two sets? It might make it harder for the Oracle to find one. No offense to sachertorte, but it's just a bad idea. The investigators should investigate whomever the hell they feel like. I think the best plan would be that the Oracle and Apprentice investigate players from whom they get a strong scum or a strong town vibe. I think redundant readings are a non-issue. At this point the Apprentice isn't that useful anyway.

--FCOD
(bolding added)

Yes, depending on how players are distributed in the two sets then it may make some things harder. However, it is just as likely that they are distributed in a way that makes things easier.

Having said that, I agree that the Oracle and the Apprentice need to make their own decisions as to who to investigate. Sachertorte, it's a clever idea but we don't want the scum to be given any clues as to who has been investigated.

I'll also add that the real benefit of the investigations is to confirm players as town, especially in this game where they cannot distinguish between scum and non believers.

sachertorte

06-27-2007, 09:45 AM

sachertorte: clever plan! But I think it's best our power roles do not follow any predictable plan.

Thanks. I was all warm and fuzzy last night when I realized that there was a way to coordinate without tipping off scum or relying on a not-yet-trusted player's direction. I can't believe I didn't think of using the fact that the Apprentice knows who the Oracle is before.

It seems like there's quite a bit of resistance to coordinating the two investigators. Personally, I don't view the plan as predictable, but they can judge it for themselves. From scum point of view each player has the same chance of being investigated whether coordinated or not. Oh well, it probably doesn't really matter all that much. But I had fun discussing it!

Pleonast

06-27-2007, 09:49 AM

Queuing

06-27-2007, 09:52 AM

Quick question:

Why are we focusing so much on the apprentice/oracle? What purpose does this serve this early in the game? Bad idea or good idea, it matters not. It is out there, the 2 roles know about. The 2 roles can now choose whether or not follow it by themselves. All this discussion about helpful town power roles can do nothing but help the scum.

Pleonast, in the post were you voted for USC Diver here is your quote:

And, on the basis of starting discussion, I'll randomly choose
Vote USCDiver
He'll like being the first in the water.

Then you said this:

I mistook Scuba_Ben for USCDiver and retracted my vote because of that. I don't feel strongly enough about a random vote to reinstate it. I haven't made another random vote because it feels like this thread has enough discussions going. Seems consistent to me.

So you randomly chose the wrong person? huh?

Also, bootstrapping discussion is different than poking someone with a vote (i.e., voting for someone to get a response from them). A poke-vote is not random, anyway.

I may be a little slow, but I don't see the difference, nor do I understand this statement. Particularly since you did random vote for someone (by your own admission) to generate discussion.

As well what is the difference between a poke-vote and a random vote within the first 12 hours or so of the game?

NAF1138

06-27-2007, 09:54 AM

Hi guys,

I need to be subbed out. My boyfriend's mom was in some kind of car accident this afternoon. They said it was minor at first, but now they're saying something about punctured lung or lung damage or something, so my BF has flown down to NC to be with her, and I'm waiting here with the dog to hear more details. I'm a fucking wreck right now, and I don't think I'm in any shape to play this game.

Blaster, I hope you can find someone to sub in for me, thanks for letting me play, for however brief it was.

Thanks guys,

Clockwork Jackal

Clockwork, I am so sorry. I am sure that all our thoughts are with you. Don't worry about having to sub out sometimes life happens.

Take care.

since she isn't playing anymore I guess I should unvote Clockwork Jackel.

Who took over for her?

FlyingCowOfDoom

06-27-2007, 09:56 AM

Sorry Pleonast, but I composed this before you posted. :p

Mathematically, this is a wash.
No seperation yields a 1 in 29 chance of the Oracle targetting the Apprentice.
Seperation yields 50% * 1 in 15 chance of the Oracle targetting the Apprentice.
They are roughly equal.I'm not good at math, but that doesn't sound right to me. It's only 1 in 29 the first Night. After that, the pool of unconfirmed players keeps getting smaller, so the second Night it would be a 1 in 26 chance (if the Crusader succeeds), then possibly 1 in 23, and so on. If the Apprentice is in the wrong set, there's no chance of ever finding him.

I think overlap is a very bad thing, and I'm suspiscous of your casual outlook the the efficacy of the Apprentice and the deteriment of overlapping investigations. You even go so far as to direct both Oracle and Apprentice to target strong-town and strong-scum vibe people. In following this, they are even more likely to overlap. Furthermore, I would argue that the Oracle and Apprentice should do the oppose what you sugget and specifically target people who are, at the moment, unlikely to get dunked (i.e., not strong-scum) and unlikely to get night killed (not strong-town). If they investigate strong-scum strong-town, then there is a higher change that the investigation goes to waste if that person gets dunked/nightkilled.
The course of action proposed by FlyingCowOfDoom strikes me as very anti-town.The Apprentice's investigations are not reliable. No matter whom he investigates, there is a 50% chance it will be wrong. If it's wrong, the chances are he will get back a pro-town role. This makes it extremely dangerous to rely on the investigation. That's why I don't care about overlap.

but for the contradictory suggestions that investigators do their own thing and then encouraging them to target strong-scum and strong-town, which I feel hurts the town.Here's what I said:The investigators should investigate whomever the hell they feel like. I think the best plan would be that the Oracle and Apprentice investigate players from whom they get a strong scum or a strong town vibe.Here's what you said:Like I said beofre, I state the policy for the investigators to choose whether to follow it or not. I don't think we should direct them to follow it, and I don't think we should direct them not to follow it either. Let them choose.We're basically saying the same thing here. We both suggested an investigative strategy and we both said the investigators should do what they want. Why is it scummy for me but not for you?

--FCOD

USCDiver

06-27-2007, 09:57 AM

Mathematically, this is a wash.
No seperation yields a 1 in 29 chance of the Oracle targetting the Apprentice.
Seperation yields 50% * 1 in 15 chance of the Oracle targetting the Apprentice.
They are roughly equal.

Ugh, I'm so tired of this discussion, but it's what we're talking about at the moment and I'd like to contribute some.

While statistically it is a wash, the downside of the Apprentice having 0% chance of being investigated (by being in the wrong group) outweighs the benefit of not having overlapping investigations. In other words, there is a palpable benefit to ensuring that the Apprentice has a chance of being investigated as this increases his accuracy.

Queuing

06-27-2007, 09:58 AM

Ugh, I'm so tired of this discussion, but it's what we're talking about at the moment and I'd like to contribute some.

Some of us are talking about other things. You could have contributed that way :)

USCDiver

06-27-2007, 10:01 AM

Ok, have we (meaning all of us) gotten the Oracle/Apprentice tactics discussion out of our system now?

I'm going to vote for the next person who discusses in any way any of the non-scum power roles. It just does not help the Town to talk about these things.
Ack, a posting flurry while I was composing. I'll learn to preview before posting from now on. At least I wasn't the 'next person' so we won't be back to where we started all of this mess!

NAF1138

06-27-2007, 10:02 AM

In general, I believe that it's scummy to defend another player on the first day. Only the scum have information about roles, and it's a known scum strategy to warm up to a townie by defending him/her. There's not really any good reason for a townie to defend someone else this early in the game. NAF, J'accuse!

Vote NAF1138

--FCOD

snipped/color removed

I expected this. I know that it is considered scummy to defend other players, but honestly I don't get it. For the record, I defended Mad against storyteller's argument, not against all accusations of scummitude.

I know I know, "Mad could be scum and only scum would know if he is scum". Maybe he is scum, but that doesn't make storytellers argument any more convincing. I see it as the job of every player in the game to talk about other peoples arguments, otherwise we end up running in circles. If we just let any old accusation hang out there we don't end up talking about anything real.

This acutally answers Pleonasts question about, if we don't random vote what do we do this early in the game? We make arguments against other players, and we examine and anylize others arguments about other players. That is what this game is about.

Pleonast

06-27-2007, 10:10 AM

So you randomly chose the wrong person? huh?

I may be a little slow, but I don't see the difference, nor do I understand this statement. Particularly since you did random vote for someone (by your own admission) to generate discussion.

As well what is the difference between a poke-vote and a random vote within the first 12 hours or so of the game?A poke-vote is not random by definition--choosing to vote for someone in particular in order to get a response from them.

A random vote means choosing a player at random and voting for them. I use random votes at the start of the first Day to jumpstart discussion. Discussion in general, not necessarily from the person voted for.

sachertorte

06-27-2007, 10:11 AM

Here's what I said:Here's what you said:We're basically saying the same thing here. We both suggested an investigative strategy and we both said the investigators should do what they want. Why is it scummy for me but not for you?

--FCOD
I don't think we are saying the same thing at all. I want to coordinate the two investigators so that there is no overlap in investigations. The negatives of my plan are minimal. Overall I assert that the scheme helps the town.

You have stated your belief that the investigators should target strong-scum and strong-town. I feel such an action is deterimental to the town as these players are more likely to be dunked and nightkilled. Overall, your suggestion hurts the town.

Yes, we both gave suggestions, and yes we both use the words, let them decide for themselves, but your suggestion is a bad one. Now I'm willing to concede that you might think that your suggestion is town helpful. Could you explain your reasoning for wanting the investigators to investigate strong-scum and strong-town? I think I have been been painfully upfront about the reasoning for my suggestion, and was quite careful to minimize scum benefits.

FlyingCowOfDoom

06-27-2007, 10:12 AM

I expected this. I know that it is considered scummy to defend other players, but honestly I don't get it. For the record, I defended Mad against storyteller's argument, not against all accusations of scummitude.It's not that you defended against storyteller's argument that gets me, it's the way you phrased it. You could have disagreed with storyteller without explicitly defending Captain Pig. Saying things like, "I'm not getting a scummy vibe" are what catch my eye.

--FCOD

SnakesCatLady

06-27-2007, 10:18 AM

I agree, but if that's all someone is posting -- it's bad. For example, HazelNutCoffee has fluff posts and discussion posts. I'm fine with that. As far as I recall, SnakesCatLady made one post this gameday, that says nothing. That looks suspicous to me. Sorry for killing SnakesCatLady's 'fun.' Maybe a nice swim will refresh her.

I just can't seem to please you. It's the first 24 hours of the day. I posted a lighthearted post and pissed you off. Then I had nothing serious to post and that pissed you off.

And... I have to violate my promise not to discuss the Oracle and Apprentice separation strategy; honestly, I think I'll explode otherwise. One of the flaws storyteller pointed out is that the town doesn't know my alignment, so I could be manipulating how the Oracle investigates to scum's advantage. Also, letting scum know which subset the Oracle is investigating gives them undue information. The problem is, how do we coordinate the Oracle and the Apprentice without telling scum the plan. Well, there is one bit of information that only the Oracle and Apprentice both have that can be used to coordinate without tipping anything off to the scum. Both Oracle and Apprentice know the identity of the Oracle.

Proposed:
Divide the set of players into two disjoint subsets: players 1-15 and players 16-30.
The Oracle will investigate the set of players to which the Oracle does not belong.
The Apprentice will investigate the set of players to which the Oracle does belong.
Voila!
This method achieves seperation without tipping off information to the scum. The only flaw I see now is that information will be revealed when the Oracle is killed or outed. Anyway, blah blah blah. Oracle and Apprentice, choose to follow or ignore on your own.

Actual game content stuff:
I'm quite curious about the MadTheSwine - ClockworkJackal discussion. In CJ's early post the phrase struck me as odd, but I dismissed it because in watching other games statments like this were flashing at me and I was wrong nearly all the time.

Many players have commented that the Oracle and the Apprentice should decide who to investigate with benefit of orders from the town. Any decisions made for them are providing information to the scum, which doesn't seem to concern you.

Vote sachertorte

Hal Briston

06-27-2007, 10:18 AM

I know that it is considered scummy to defend other players, but honestly I don't get it.Yeah, that's always been a shaky premise to me as well. Hell, Clockwork Jackal found suspicion with Zeriel for this post (http://boards.straightdope.com/sdmb/showpost.php?p=8718926&postcount=217) and I came to his defense (http://boards.straightdope.com/sdmb/showpost.php?p=8718965&postcount=224). Yet, no FOS my way for it. Maybe it's because defending someone can simply be a matter of pointing out errors in logic.

Either way, unless someone is standing on the temple roof screaming "I know this person isn't scum!", I'd say the defense post isn't indicative of very much.

Pleonast

06-27-2007, 10:20 AM

Zeriel

06-27-2007, 10:22 AM

At this point, I'd think I'm agreeing with the camp that says we've said enough about the Oracle and Apprentice. If someone has a different plan or modification thereof, then maybe they should speak up, but now that people are going to start finger-pointing over it, it seems even more counterproductive to discuss it.

The only thing we know for sure about the Oracle and Apprentice is that they're both dopers--by definition, they're both smarter than the average bear and will be able to pick out a good strategy without too much debate from the peanut gallery.

NAF1138

06-27-2007, 10:27 AM

It's not that you defended against storyteller's argument that gets me, it's the way you phrased it. You could have disagreed with storyteller without explicitly defending Captain Pig. Saying things like, "I'm not getting a scummy vibe" are what catch my eye.

--FCOD

I get that. To be fair I was saying I didn't think Mad's mistake was scummy, something I think is different than saying that I don't think Mad is scummy.

sachertorte

06-27-2007, 10:35 AM

While statistically it is a wash, the downside of the Apprentice having 0% chance of being investigated (by being in the wrong group) outweighs the benefit of not having overlapping investigations. In other words, there is a palpable benefit to ensuring that the Apprentice has a chance of being investigated as this increases his accuracy.

But... but... but. Oh dear.
You can't say the Apprentice has a 0% chance of being investigated. There is a 50% chance that the apprentice has a 0% chance of being investigated. There is also a 50% chance that the apprentice has a 1 in 15 chance of being investigated! That's why statistically it's a wash. Oh Nairu, what have I done?!

I just can't seem to please you. It's the first 24 hours of the day. I posted a lighthearted post and pissed you off. Then I had nothing serious to post and that pissed you off.
Who's pissed off? What did I say that makes me sound pissed off? This is what I wrote in Post #354:
What a wonderfully fluffy post. I can't honestly say that I've contributed much towards actually finding scum, but I think this type of post by SnakesCatLady is the type of post we should try and avoid. It feels to me like someone trying to fly under the radar, but not willing to not post at all. The fluff post isn't even followed by a post of substance. FOS SnakesCatLady

(snipped)

storyteller is also correct in pointing out that we're getting a bit off track, and I apologize. Like I said, I love analyzing games and talking about theories and strategies. It's going to be hard to keep that under control for me, because I find it fun. And ultimately this exercise is for fun. But to put things on a better track, I reiterate my FOS SnakesCatLady.
How does that constitute pissed off? I'm suspicious of you; that's all. You know, the type of thing that happens all the time in this game. Now I'm more suspicious of you.

FlyingCowOfDoom

06-27-2007, 10:39 AM

Sorry again, Pleo, but sachertorte is asserting things that I strongly disagree with, and I have to respond to his question.
I don't think we are saying the same thing at all. I want to coordinate the two investigators so that there is no overlap in investigations. The negatives of my plan are minimal. Overall I assert that the scheme helps the town.

You have stated your belief that the investigators should target strong-scum and strong-town. I feel such an action is deterimental to the town as these players are more likely to be dunked and nightkilled. Overall, your suggestion hurts the town.

Yes, we both gave suggestions, and yes we both use the words, let them decide for themselves, but your suggestion is a bad one. Now I'm willing to concede that you might think that your suggestion is town helpful. Could you explain your reasoning for wanting the investigators to investigate strong-scum and strong-town? I think I have been been painfully upfront about the reasoning for my suggestion, and was quite careful to minimize scum benefits.You've decided that your plan helps the town, but that is only your opinion. I think your plan is harmful to the town. Is either of us right? Who knows. I certainly would never assert anything that isn't fact.

I would love to explain my reasoning. I've already stated that I feel the role is useless at this point in the game. Unreliable information is worse than no information. Let's say the Apprentice investigates a Cultist, but gets a wrong result. He'll most likely show up as...Believer. That's a very dangerous gamble to take. We cannot rely on the Apprentice's investigations. That's why I don't care about overlap. In my mind, the Apprentice's investigations don't count for anything, so there isn't any overlap.

The Oracle would obviously not investigate players that are likely to be lynched. That's just common sense. I'm not saying he should investigate players that are widely held to be suspicious. I'm saying he should investigate players that HE finds scummy. It makes more sense to me to investigate a player that is acting scummy than to just randomly choose targets. I think that doing so increases the chances of finding scum. And what's the worst that could happen? The Oracle finds a confirmed townie. That's actually REALLY helpful. Regardless of the outcome of the investigation, the town gains valuable information. Sure, it would be most beneficial to find scum, but it's almost as beneficial to confirm townies. If the Oracle can breadcrumb his findings, we'll potentially have a great deal of information to help us. AND, this strategy doesn't help the the scum, because they already know who the townies are. One thing the Oracle should not do is reveal specific roles. THAT would be giving information to the scum.

If the Oracle isn't feeling scummy vibes from anyone, it's still helpful to investigate someone with a townie vibe. Like I said, any information is good information. My plan has the advantage of not being a plan...it's a strategy. We want to give as little information to the Cultists as possible, and I just think it's a bad idea to outline a plan.

Again, these are my suggestions. The Oracle should do as he or she sees fit. All I'm trying to do offer a suggestion for consideration. I've explained myself to my satisfaction, and I won't be discussing the Oracle any further.

--FCOD

sachertorte

06-27-2007, 10:39 AM

Unvote MadTheSwine
Vote sachertorte
I can ignore FlyingCow and USCDiver, since they may not have seen my warning before posting. But, you are responding to a post made afterwards, one that even explicitly recognized my threat. Thus you cannot reasonably claim you didn't see it.
Of course I saw it. But I had to respond to FlyingCow.
And under no circumstances can I not respond to bad math (USCDiver).

Hal Briston

06-27-2007, 10:45 AM

Hi All...

I wasn't planning on posting this until tomorrow, but I'm going to be computerless for the next 24-48 hours. My company is taking back their laptop for some servicing and updating, which I thought was scheduled for tomorrow. Imagine my surprise when I just got a "why aren't you in the office today?" e-mail (I telecommute).

So, whoops...going to have to run to the office on my lunch hour and hand over my PC. Be back at some point before the weekend.

DiggitCamara

06-27-2007, 10:46 AM

Snipped.

What makes you so sure of these numbers? I'm curious.

Blaster Master said that there is at least one of each role in the game. From there I looked at the allegiance of said roles and came up with the least number of pro-town Believers/Non-Believers present.

Hal already said it but I thought you might want to hear my explanation as well.

DiggitCamara

06-27-2007, 10:51 AM

(snip)

Finally, I too would like to hear from Pleonast and DiggitCamera regarding the points Queuing brought up in post 360 (http://boards.straightdope.com/sdmb/showpost.php?p=8720379&postcount=360)
Cookies already answered it: she was voted for, during at least two of the past 4 games, during the first Day. Most of those votes made a reference (tongue firmly planted in cheek) to her name and her obvious allegiance to the Dark Side.

I saw an early vote for someone else and made a (bad?) joke. I retracted my vote because I have no real reason to vote for her.

USCDiver

06-27-2007, 10:58 AM

So you randomly chose the wrong person? huh?

Pleonast unvoted me because of comments made by Scuba_Ben. That's what he meant when he said he got us confused. You're misinterpretting what Pleonast said and taking his comments out of context.

FOS Queuing

Pleonast

06-27-2007, 11:00 AM

Sorry again, Pleo, but sachertorte is asserting things that I strongly disagree with, and I have to respond to his question.
Of course I saw it. But I had to respond to FlyingCow.
And under no circumstances can I not respond to bad math (USCDiver).No, neither of you really need to continue discussing this. It's seriously not good for the Town. At least pretend you're on the Town's side. I'd vote for you too, Flying, if I could vote for more than one person.

If you feel the urge to discuss numbers and power roles, think about some of the scummy roles. Does the Town need to take precautions against the Avatar? How should we handle recruitment by Prophet?

sachertorte

06-27-2007, 11:06 AM

Forget Pleonast, I voted for FCOD, of course a response is warranted. I don't really see Pleonast as the gatekeeper of discussion anyway. Go ahead and vote for me, I don't mind. I think our discussion is important, and I need to learn what other people are thinking, because obviously they aren't thinking the same way I am. And part of the reason I brought up the entire strategy was because I thought it would provoke good discussion. Which I think it did.

Let's say the Apprentice investigates a Cultist, but gets a wrong result. He'll most likely show up as...Believer. That's a very dangerous gamble to take. We cannot rely on the Apprentice's investigations. That's why I don't care about overlap. In my mind, the Apprentice's investigations don't count for anything, so there isn't any overlap.
I appreciate your caution regarding the Apprentice. However, I hold the investigation result as better than nothing. Furthermore, multiple investigations by the Apprentice on the same target will yield actionable data. I'm not sure if it's better to target the same person more twice or just once, but it's an option.

The Oracle would obviously not investigate players that are likely to be lynched.
I read your suggestion as stating exactly that. I still think that someone the investigator has with a strong-scum vibe will likely have a strong-scum vibe with the rest of the town, but I think we see sort of eye-to-eye on this now.

It makes more sense to me to investigate a player that is acting scummy than to just randomly choose targets. I think that doing so increases the chances of finding scum. And what's the worst that could happen? The Oracle finds a confirmed townie. That's actually REALLY helpful.
We're actually not that far off from each other. I agree that finding believers is just as good as finding scum. My point is only that by investigating a player that is acting scummy is more likely to be useless as the acting scummy player is more likely to get dunked. 'randomly' (as you put it) selecting investigations minimizes the risk of losing the benefits of the investigation by death.

If the Oracle isn't feeling scummy vibes from anyone, it's still helpful to investigate someone with a townie vibe. Like I said, any information is good information. My plan has the advantage of not being a plan...it's a strategy. We want to give as little information to the Cultists as possible, and I just think it's a bad idea to outline a plan.
Well, plan and strategy are pretty much the same thing, I don't see how you're drawing a distinction, and I don't see how a plan that targets scum-vibe people and town-vibe people is incompatible with coordination of disjoint sets. Also, I don't see any problem with investigating someone with no vibe at all.

I'm satisfied with FCOD's explanation. I mostly disagree, but I no longer feel the intent is anti-town.
Unvote FlyingCowOfDoom

ComeToTheDarkSideWeHaveCookies

06-27-2007, 11:07 AM

Let's see if I can summarize the current list of things that might get you voted for (in no particular order):

- discussing opinions or trying to build consensus regarding what the Oracle and Apprentice should do with their investigative powers, assuming that either of them would consider such suggestions

- defending players

- voicing suspicions based on posting style/fluffyness

To me, the playing field sure seems to be filling up with land mines that discourage conversation. Isn't it a bit early to have to hold our breath and tread so cautiously?

Picking up a couple of votes here and there doesn't necessarily equal the kiss of death, but can't some good old-fashioned FOSing get the same sort of point across?

FlyingCowOfDoom

06-27-2007, 11:10 AM

No, neither of you really need to continue discussing this. It's seriously not good for the Town. At least pretend you're on the Town's side. I'd vote for you too, Flying, if I could vote for more than one person.

If you feel the urge to discuss numbers and power roles, think about some of the scummy roles. Does the Town need to take precautions against the Avatar? How should we handle recruitment by Prophet?No offense, but I don't see who made you boss of discussions. I also don't see how our discussion is detrimental to the town. Someone voted for me and I think I should be allowed to discuss the reasoning behind it. Your reason for voting for sachertorte is pretty weak, if you ask me.

On preview, I see sachertorte and I agree on at least one point :D

--FCOD

Queuing

06-27-2007, 11:15 AM

Cookies already answered it: she was voted for, during at least two of the past 4 games, during the first Day. Most of those votes made a reference (tongue firmly planted in cheek) to her name and her obvious allegiance to the Dark Side.

I saw an early vote for someone else and made a (bad?) joke. I retracted my vote because I have no real reason to vote for her.

So in a game where votes count dearly you chose to use this power as a joke? And then, with no explanation, retract that vote as you had no reason to vote for her? All of this in the midst of an argument about random voting?

I don't want to make a big deal out of 2 seemingly random votes that were then quickly rescinded with explanations provided (however weak or strong) but IMO its a bad play regardless and I keep my FOS pointed.

On preview:

Huh, USCDiver? How did I misinterpret what Pleonast said? As I showed he said he was randomly voting for someone. That someone was you. Then he unvotes you, because he thinks Scuba Ben is you. Scuba Ben had thought it was odd that Pleonast had randomly voted for you and had FOS'd pleonast in return. This is where Pleonast became confused. Pleonast stated that his vote for a waterperson was completely random. He then gets confused after Scuba Ben FOSes him in return, and unvotes you.

This does nothing to explain the logic behind randomly voting for someone to generate discussion, another playing calling into question this move, a debate begins about the play of random voting, the person who Pleonast thinks he voted for (but who in actuality is just another water related name) FOSing him because of his actions, Pleonast immediately unvoting player after FOSing him and calling said player to task for a "revenge vote" which must have been done by a "newbie townie".

How did this vote generate any discussion from or about the actual player? It didn't.

Where am I misinterpreting Pleonast again?

sachertorte

06-27-2007, 11:22 AM

Pleonast

06-27-2007, 11:22 AM

No offense, but I don't see who made you boss of discussions. I also don't see how our discussion is detrimental to the town. Someone voted for me and I think I should be allowed to discuss the reasoning behind it. Your reason for voting for sachertorte is pretty weak, if you ask me.

On preview, I see sachertorte and I agree on at least one point :DDidn't you get the memo? I am the secret role: the Boss of Discussion. Obey me or else. :D

Yeah, you two can go on talking about whatever you feel like. All I can do is vote for you.

How did this vote generate any discussion from or about the actual player? It didn't.The point of random voting isn't to discuss the votee, but to generate discussion. Any discussion. The fact that you keep harping on my vote proves it was successful. :p

... in a game where votes count dearly ...Votes are cheap to give and take. It's the analysis of votes that's valuable.

FlyingCowOfDoom

06-27-2007, 11:24 AM

Yay FlyingCowOfDoom!
See, no hard feelings right?No hard feelings.

But I won't trust you until you're dead ;) .

--FCOD

SnakesCatLady

06-27-2007, 11:24 AM

Personally, I don't think a lighthearted or "fluffy" post is that much a waste. When people are making that type of post they are much more likely to let things slip than they are if they feel they have to watch each word they type.

And as far as discussing "assignments" for the Oracle and the Apprentice, in addition to possibly giving information to the enemy, it makes you (general you, not any one poster in particular) look very bossy. It is as much a "waste" as a lighthearted post, if not more so. If at some point during the game you feel a certain player should be investigated you can suggest it, and the Oracle or Apprentice can take or leave the suggestion as they see fit. But going on for endless posts deciding what they should do is not doing anyone any good.

sachertorte

06-27-2007, 11:30 AM

Does the Town need to take precautions against the Avatar?
Well, I have one idea that would be a bitch to implement.
Instead of traditional day time voting, we two seperate votes to end slightly before the end of day. The first determines who to dunk and is basically a proxy for the real vote (i.e. all voting accountability lies in this vote). Simultaneously, we hold a second proxy vote to determine who does the dunking. At the end of the day only the dunker casts a real vote, and the dunkee gets dunked with only the dunker at risk of being Avatar killed.

Messy eh?

Zeriel

06-27-2007, 11:32 AM

If you feel the urge to discuss numbers and power roles, think about some of the scummy roles. Does the Town need to take precautions against the Avatar? How should we handle recruitment by Prophet?

I don't really think we CAN deal with recruitment by the Prophet. The only fortunate thing about the prophet is that conversion can fail if it hits any of several power roles. Beyond that, we just need to be vigilant about midgame shifts to scummy behavior--after all the discussion earlier, I'm not going to be surprised whatsoever if they play for a conversion earlier rather than later since my not-re-reading brain tells me the consensus was that conversion should happen later.

The Avatar? Honestly, that almost has to be filed under shit happens. I just hope when we get him, we get a large number of scum who're jumping on to hide their scumminess and he kills one of them instead.

It seems to me that the scum have two power roles that are harder than average to strategize around, but they also don't have any who can hide (which is somewhat unusual). Given the estimates of 5-7 scum, 5-7 nonbelievers, and the rest town (this makes sense to me), that means scum have 2 Power /3-5 normal, NBs have 2P/3-5N, and town has around 7-10 power roles (assuming between 2 and 5 monks), that means Town ranges from the extremes of 7P/13N and 10P/6N. Given this, I find it distressingly likely the secret role is a non-believer or, Nairu forbid, a cultist--even the most positive estimates of town power-to-normal ratio is only at best roughly equivalent with the other two factions, and the worst estimates mean the town is not only shorthanded but power-role-heavy already. (best and worst, here, refer to "number of townspeople at game start, higher being better.)

I don't know if we can necessarily do anything about it right now but be not surprised when something bad happens.

Scuba_Ben

06-27-2007, 11:35 AM

Huh, USCDiver? How did I misinterpret what Pleonast said? As I showed he said he was randomly voting for someone. That someone was you. Then he unvotes you, because he thinks Scuba Ben is you. Scuba Ben had thought it was odd that Pleonast had randomly voted for you and had FOS'd pleonast in return. This is where Pleonast became confused. Pleonast stated that his vote for a waterperson was completely random. He then gets confused after Scuba Ben FOSes him in return, and unvotes you.

This does nothing to explain the logic behind randomly voting for someone to generate discussion, another playing calling into question this move, a debate begins about the play of random voting, the person who Pleonast thinks he voted for (but who in actuality is just another water related name) FOSing him because of his actions, Pleonast immediately unvoting player after FOSing him and calling said player to task for a "revenge vote" which must have been done by a "newbie townie".And, as I noted, was in fact done by a first-time player, just not the one Pleonast thought.

As storyteller pointed out to me in #381, I'd like to see votes based on estimates of who's scum, not on pure randomness; my problem is learning how to read posts to get a useful estimate. All I've got so far are other people's opinions. Consequentially, I suspect everybody. (But on the first Day, that's a good thing. Is it?)

Added on preview, to weigh in with my thoughts on the concept of the Oracle and Apprentice (not strategies for them in particular): It seems to me that it is in the Town's interest for the Apprentice to draw the Oracle's attention for an early investigation, to improve their accuracy. And it is in the Mafia's interest for the Apprentice to draw the Oracle's attention, thereby giving them an important target. Rock, meet Hard Place.

sachertorte

06-27-2007, 11:35 AM

I'm sorry my ideas and discussions are annoying everyone else. I'm learning quite a bit from them so at least they are doing me some good.

Queuing

06-27-2007, 11:36 AM

Votes are cheap to give and take. It's the analysis of votes that's valuable.

That is true, which I why I am "harping" on it. Well ok, the real reason is because I have no idea what else to talk about. I think my position of math being close to useless in this game is well known. As well I am completely onboard with the idea that power roles should decide what to do when they choose to. As well in no way should we try to guide the oracle/apprentice, IMO, as that could lead to a quick death of the oracle, much like what happened to CaerieD in M2. That must be avoided at all costs.

Further, the "bossy" idea. That doesn't bother me in the slightest. Pleonast, Storyteller, and myself (all who have suggested the Oracle/apprentice discussion be dropped) have clearly not been listened to. Which is the people who are having the discussion's prerogative. I don't see those posts as a waste, but I don't see any post as a waste. Feel free to discuss whatever you want. Personally I would like to see every power role discussed if any are going to be.

On preview: sachertorte, that idea is so very confusing and, IMO, a complete waste of time. To much in the way of logistics behind it. I like that you are giving out ideas however.

Scuba_Ben

06-27-2007, 11:38 AM

It seems to me that the scum have two power roles that are harder than average to strategize around, but they also don't have any who can hide (which is somewhat unusual).I thought the normal scum ID as Nonbeliever and the Prophet IDs as Believer, neither one IDing as Cultist. Did I copy the wrong reference chart?

Queuing

06-27-2007, 11:38 AM

Ok, Scuba Ben and USC Diver, you guys have to change your name or something. Or not quote posts directed at the other person. After reading what Scuba Ben wrote (quoting what I directed at USCDiver) I thought I made the same mistake that Pleonast made. Damn you wet people! :)

Scuba_Ben

06-27-2007, 11:45 AM

NAF1138

06-27-2007, 11:45 AM

Just FYI for the group:

A huge project was just dumped in my lap at work. The plus side is that with my job even huge projects won't last more than a day or two. The downside is, my participation is going to be dropped very low for the rest of today, and probably most of tomorrow.

I will be able to read along (I think) and maybe put in a post or two, but nothing lengthy.

Fortunatly there are 27 or so other players who can fill my void (Hal and CJ not being around either.) :D

Queuing

06-27-2007, 11:49 AM

What shall I do, go jump in a lake? ;)

So long as you don't dive into it...

ComeToTheDarkSideWeHaveCookies

06-27-2007, 11:49 AM

I'm sorry my ideas and discussions are annoying everyone else. I'm learning quite a bit from them so at least they are doing me some good.

There's that "everyone" word again. I do not think that word means what you think it means... :D

If I were truly twisted, I'd suggest a sub-game. Contextual movie quotes in periwinkle. Winner has the most quotes at the end of the game, and gets to take the squid (and aquarium) home for a week.

Zeriel

06-27-2007, 12:06 PM

I thought the normal scum ID as Nonbeliever and the Prophet IDs as Believer, neither one IDing as Cultist. Did I copy the wrong reference chart?
No, I misread. Stupid Prophet.

I'm honestly of the opinion that it's in the town's best interest to treat "non-believer" as "scum" unless proven otherwise. Dang, that REALLY confuses the issue with detecting the convert now that I think about it.

I think we're in for a bumpy ride.

DiggitCamara

06-27-2007, 12:07 PM

sachertorte

06-27-2007, 12:12 PM

As well in no way should we try to guide the oracle/apprentice, IMO, as that could lead to a quick death of the oracle, much like what happened to CaerieD in M2. That must be avoided at all costs.
Can someone please explain how talking about the Oracle gets the Oracle killed? I just don't see it. I see many people hold this belief, but I don't know why. All I see are statements that equate Oracle discussion with a dead Oracle without explaining the mechanics of such a conclusion. Would discussing how discussing the Oracle kills the Oracle somehow kill the Oracle too? What the hell am I missing? The only discussion that came close was storyteller's mentioning of directing Vig kills in M2. But that's different than setting up subsets to investigate, espeically subsets that scum don't know who is allocated to whom. There is no definitive information being disseminated. Or is the consensus that these are really the same thing and I'm just insane? I really want to understand.

DiggitCamara

06-27-2007, 12:18 PM

No, I misread. Stupid Prophet.

I'm honestly of the opinion that it's in the town's best interest to treat "non-believer" as "scum" unless proven otherwise. Dang, that REALLY confuses the issue with detecting the convert now that I think about it.

I think we're in for a bumpy ride.
Again: not so. Though the Non-Believers can win by themselves, they count against the Cultist's win condition. If we start treating them as "equal time foes", we'll be:

1. Losing votes against the Cultists
2. Giving their "special roles" (Alchemist) incentive to act against town

The Psychopath is a separate issue, of course. But (s)he doesn't come into play until (s)he is targeted. Until then not even him/her knows (s)he's the Psychopath.

DiggitCamara

06-27-2007, 12:21 PM

Can someone please explain how talking about the Oracle gets the Oracle killed? I just don't see it. I see many people hold this belief, but I don't know why. All I see are statements that equate Oracle discussion with a dead Oracle without explaining the mechanics of such a conclusion. Would discussing how discussing the Oracle kills the Oracle somehow kill the Oracle too? What the hell am I missing? The only discussion that came close was storyteller's mentioning of directing Vig kills in M2. But that's different than setting up subsets to investigate, espeically subsets that scum don't know who is allocated to whom. There is no definitive information being disseminated. Or is the consensus that these are really the same thing and I'm just insane? I really want to understand.

The problem is one of "tunnel vision". Usually during talks like this the person who has the role gets involved, some way or another (especially if there are newbies around).

Cultists, if they are smart, will start narrowing down a pool of names who could hold the Power Role being discussed (Oracle for purposes of this discussion). Hence, in most games it's assumed that the less the power roles are discussed, the better. And as a corollary, (s)he who starts the discussion is seen as scummy.

FlyingCowOfDoom

06-27-2007, 12:23 PM

Well, I have one idea that would be a bitch to implement.
Instead of traditional day time voting, we two seperate votes to end slightly before the end of day. The first determines who to dunk and is basically a proxy for the real vote (i.e. all voting accountability lies in this vote). Simultaneously, we hold a second proxy vote to determine who does the dunking. At the end of the day only the dunker casts a real vote, and the dunkee gets dunked with only the dunker at risk of being Avatar killed.

Messy eh?This is a bad idea, for the reasons DiggitCamara stated. The Avatar could actually help us out if we're lucky.

I thought the normal scum ID as Nonbeliever and the Prophet IDs as Believer, neither one IDing as Cultist. Did I copy the wrong reference chart?HOLD THE PHONE! I DID NOT REALIZE THIS!! Am I the only one that missed that part of the rules?I'm honestly of the opinion that it's in the town's best interest to treat "non-believer" as "scum" unless proven otherwise. Dang, that REALLY confuses the issue with detecting the convert now that I think about it.IMO this is a bad idea. Killing Non-Believers is only a teeny bit better than killing Believers. Let's not forget that for the Cultists to win, their numbers must be more than Believers + Non-Believers.

--FCOD

Queuing

06-27-2007, 12:28 PM

Can someone please explain how talking about the Oracle gets the Oracle killed? I just don't see it. I see many people hold this belief, but I don't know why. All I see are statements that equate Oracle discussion with a dead Oracle without explaining the mechanics of such a conclusion. Would discussing how discussing the Oracle kills the Oracle somehow kill the Oracle too? What the hell am I missing? The only discussion that came close was storyteller's mentioning of directing Vig kills in M2. But that's different than setting up subsets to investigate, espeically subsets that scum don't know who is allocated to whom. There is no definitive information being disseminated. Or is the consensus that these are really the same thing and I'm just insane? I really want to understand.

I will give it a shot.

By dividing up the investigations we are basically telling the scum who the Oracle will be investigating. Not necessarily when this investigation when this investigation will happen, but that it will happen. So the scum now know who the Oracle will be investigating. The scum also know who are scum. This gives the scum 2 pieces of information with which to find the oracle; who the scum are and who the oracle will be investigating. If the oracle then finds a scum, from the predetermined list, and begins to breadcrumb their finding the scum will know what to look for. Someone who is scum from the list the oracle will be investigating all of a sudden being mentioned as scum, or FOS'd or whatever. Any person who does this will instantly be suspected of being the Oracle. The scum will then kill that person. This will sooner rather then later mean the death of the oracle.

HazelNutCoffee

06-27-2007, 12:39 PM

Discussing what the Oracle should do can potentially get the Oracle killed simply because the Oracle's actions could become predictable. It gives scum another clue as to what to look for when trying to identify the Oracle. Maybe discussing Oracle strategy might be more helpful later on in the game; at this point it doesn't seem like it's going to do much good.

Well, I have one idea that would be a bitch to implement.
Instead of traditional day time voting, we two seperate votes to end slightly before the end of day. The first determines who to dunk and is basically a proxy for the real vote (i.e. all voting accountability lies in this vote). Simultaneously, we hold a second proxy vote to determine who does the dunking. At the end of the day only the dunker casts a real vote, and the dunkee gets dunked with only the dunker at risk of being Avatar killed.
How would this be helpful? Someone gets killed by the Avatar's death regardless. The only way this could even possibly do any good was if we knew who has what power role and make sure those roles do not vote for the person getting drowned. And we don't want our power roles have to role-claim any earlier than they have to.

Imagine that we choose the drowner, and that person happens to be a power-role. They either have to keep their mouth shut and pray that the one being drowned is not the Avatar, or role-claim and blow their cover, not to mention earn suspicion that they are simply a cultist trying to save their own neck.

I don't think we can really develop a working strategy for dealing with the Avatar as of yet. We just better hope we get lucky and the Oracle/Apprentice finds the Avatar and is in a position to somehow let the Crusader know who it is.

As for reading a defense of someone else as a scum tell, I agree that it really depends on the manner of defense. It's possible to defend an argument without defending the person. As I've already stated, I wasn't sold on NAF or Idle Thought's defense of MtS, which is why I am suspicious of both. NAF may be right - MtS's mistake may be an honest one. Then why did he keep the vote? That's what I'm asking MtS to explain.

sachertorte

06-27-2007, 12:43 PM

I will give it a shot.

By dividing up the investigations we are basically telling the scum who the Oracle will be investigating. Not necessarily when this investigation when this investigation will happen, but that it will happen. So the scum now know who the Oracle will be investigating. The scum also know who are scum. This gives the scum 2 pieces of information with which to find the oracle; who the scum are and who the oracle will be investigating. If the oracle then finds a scum, from the predetermined list, and begins to breadcrumb their finding the scum will know what to look for. Someone who is scum from the list the oracle will be investigating all of a sudden being mentioned as scum, or FOS'd or whatever. Any person who does this will instantly be suspected of being the Oracle. The scum will then kill that person. This will sooner rather then later mean the death of the oracle.
Thank you for taking the time to educate me.
I'm pretty sure I designed the system so that scum don't know which set the Oracle was investigating until they know who the Oracle is, by which point the whole deal is moot. (In case you missed it, I amended my idea to obfuscate which set the Oracle investigates. The Oracle investigates the set to which he does not belong, and the Apprentice investigates the other set) How is this scenario different from regular investigations? Without dividing up into investigation sets, scum still know who is scum and who isn't, won't breadcrumbs have the same effect either way? How does investigating from a set of 15 all that different from investigating from a set of 30? Am I asking too many questions?
I'm so sorry for being so obtuse.

HazelNutCoffee

06-27-2007, 12:49 PM

ComeToTheDarkSideWeHaveCookies

06-27-2007, 12:49 PM

I was typing out a long "thinking out loud" post trying to follow the list-splitting logic without incurring risk, and I just can't. I keep coming back to the fact that the Apprentice knows his/her own identity, and the identity of the Oracle, but has no way of sharing that information in anyway without risking exposure of him/herself to scum.

I suppose the one conclusion that I have come to is somewhat of a hair-splitting one. It isn't the Oracle that we risk exposing by trying to do this, but the Apprentice.

sachertorte

06-27-2007, 12:53 PM

Oh, and the whole Avatar idea was not suggested to be seriously implemented. My problem is the process is complex for little (possibly no) gain. The reasoning behind the idea was to take the randomness out of the Avatar's death-kill. Essentially the idea was to turn the death caused by the Avatar dunk into a second lynching. My feeling is that a directed kill is better than a random kill.
I'm not actually advocating we do this. I was just responding to Pleonast's query as to whether I'm capable of analyzing scum-role to the same extent as pro-town roles.
People seem to be hung up on the idea that we might lose another townie or power-role by doing a directed kill. Well, we face that risk everyday. Presumably the person we choose to do the dunking is rather high on the dunk list anyway, so what's the big deal?
Again, all moot points, but I thought the idea was rather clever. Use the rules to your favor is my motto today.

Pleonast

06-27-2007, 12:53 PM

Queuing

06-27-2007, 12:58 PM

Thank you for taking the time to educate me.
I'm pretty sure I designed the system so that scum don't know which set the Oracle was investigating until they know who the Oracle is, by which point the whole deal is moot. (In case you missed it, I amended my idea to obfuscate which set the Oracle investigates. The Oracle investigates the set to which he does not belong, and the Apprentice investigates the other set) How is this scenario different from regular investigations?Without dividing up into investigation sets, scum still know who is scum and who isn't, won't breadcrumbs have the same effect either way? How does investigating from a set of 15 all that different from investigating from a set of 30? Am I asking too many questions?
I'm so sorry for being so obtuse.

The main difference, as I see it, is that it still narrows down the options of whom the Oracle will be investigating. To the scum that set of 30 is only a set of 23-24 (or whatever). They will know if more of their people fall in one set then the other. Again this narrows done said set further. If it is decided that the Oracle only investigates 1-15 and say 4 of the 7 (or whatever the number happens to be) then to the scum that looks like a set of 11. Now from that set of 11 one person appears to be breadcrumbing, to be FOS scum, to never go after a town person. Well that person looks like an oracle to scum, and Bam dead person. Maybe dead oracle.

Now with 30 people to draw from the set goes up to 23. It doubles the number of people who are possible Oracle's to the scum. This is important.

One thing you have to remember the Oracle must breadcrumb to be effective. The scum know exactly who is scum. Therefore what looks like innocent FOSing, or whatever, to the believer will look much more suspicious to scum. Admittedly this is a problem that exists regardless of division of labour, however I want the problem to be as big as possible for the cult. Dividing up the labour, to avoid the chance of our investigative staff doing double work, also divides up the problem the scum face.

Does that help? Does that even make sense? :)

HazelNutCoffee

06-27-2007, 01:02 PM

Again, all moot points, but I thought the idea was rather clever. Use the rules to your favor is my motto today.
It's not working within the rules. Holding a vote to decide who does the dunking is actually implementing another rule - the only way it would work without additional rules is if the person did it voluntarily. What if the person refused to be the dunker?

No point in discussing strategy we can't use.

All this thinking is making me hungry. *wanders off in search of food*

ComeToTheDarkSideWeHaveCookies

06-27-2007, 01:05 PM

Reasons the Town should not discuss the Oracle (or any other pro-Town power-role):
1) We're looking for scum, not trying to find the Oracle.
2) Let the scum spend their own time thinking about the Oracle.
3) Discussion may inadvertently cause the Oracle to make a mistake.
4) It's best the Oracle keeps their tactics to themself at all times.
5) Because of all the above reasons, persistently talking about the Oracle is scummy.

As long as the town is careful, safe discussion can occur, imho.

By (carefully) discussing hypotheticals, we may actually prevent the investigators from making mistakes. Further, the more townies who do not actually hold investigative roles who participate in the conversation will muddy the waters and make it more difficult for the scum to key in on the actual investigators during said careful and hypothetical conversation.

I do not disagree that the investigators should keep what they intend to do close to the vest. But I disagree that discussion is equivalent to painting targets on their backs. It doesn't have to be that way if we're all careful, especially the investigators themselves.

sachertorte

06-27-2007, 01:07 PM

I would really like to drop this topic as it is of no help at this point in the game
Apologies again. While I understand your frustration at my hounding at this point which doesn't help the game, I want to learn something here. Or maybe even teach something.

Quite frankly, we could have had this discussion before roles even went out. Everything is general and doesn't rely on any particular role assignment, nor does it reveal any.

The Apprentice doesn't have to reveal anything to anyone. He already knows who the Oracle is. The Oracle already knows who the Oracle is. They use this common knowledge to organize. Organizational tactics are good. Armys coordinate because it is good. The apprentice and oracle can coordinate without revealing anything to scum. The Oracle, knowing who the Oracle is, chooses the set not containing the Oracle. The Apprentice, also knowing who the Oracle is, chooses the set that does contain the Oracle.

Please tell me how this reveals the Oracle's identity to scum.

MHaye

06-27-2007, 01:13 PM

Thank you for taking the time to educate me.
I'm pretty sure I designed the system so that scum don't know which set the Oracle was investigating until they know who the Oracle is, by which point the whole deal is moot. (In case you missed it, I amended my idea to obfuscate which set the Oracle investigates. The Oracle investigates the set to which he does not belong, and the Apprentice investigates the other set) How is this scenario different from regular investigations? Without dividing up into investigation sets, scum still know who is scum and who isn't, won't breadcrumbs have the same effect either way? How does investigating from a set of 15 all that different from investigating from a set of 30? Am I asking too many questions?
I'm so sorry for being so obtuse.The problem with your system is that as soon as the Oracle discovers a nonbeliever and starts dropping hints, the cult might pick up on it and know which group he is investigating. They therefore know what group he is in - the other one.

They've just narrowed the field of search for one of their major targets by 50% - more if there are a large number of Cultists in the same group. The Cult already have an advantage in information. Your idea gives them more information.

I'm against any form of steering any of the town power roles. it gives the Cult a chance to meddle in the choice directly. Also, how can you enforce it? The Town has no executive oversight of what the player puts in the PM to the GO.

I'm now seriously wondering about Sachertorte. It would be worse if he hadn't said 9when floating the idea) that the players were free to ignore his idea.

SnakesCatLady

06-27-2007, 01:15 PM

sachertorte

06-27-2007, 01:17 PM

The main difference, as I see it, is that it still narrows down the options of whom the Oracle will be investigating. To the scum that set of 30 is only a set of 23-24 (or whatever). They will know if more of their people fall in one set then the other. Again this narrows done said set further. If it is decided that the Oracle only investigates 1-15 and say 4 of the 7 (or whatever the number happens to be) then to the scum that looks like a set of 11. Now from that set of 11 one person appears to be breadcrumbing, to be FOS scum, to never go after a town person. Well that person looks like an oracle to scum, and Bam dead person. Maybe dead oracle.

Now with 30 people to draw from the set goes up to 23. It doubles the number of people who are possible Oracle's to the scum. This is important.

One thing you have to remember the Oracle must breadcrumb to be effective. The scum know exactly who is scum. Therefore what looks like innocent FOSing, or whatever, to the believer will look much more suspicious to scum. Admittedly this is a problem that exists regardless of division of labour, however I want the problem to be as big as possible for the cult. Dividing up the labour, to avoid the chance of our investigative staff doing double work, also divides up the problem the scum face.

Does that help? Does that even make sense? :)
Yes! This helps immensely!
I made an original proposition yesterday afternoon that contains the flaw that you have (thankfully) pointed out to me clearly.
This morning I amended the idea with a twist where we as a town don't arbitrarily assign the Oracle to 1-15, but use the common knowledge of the Apprentice and Oracle to coordinate secretly. I think this helps the problem you point out. I have a glimmer of an idea that the division may still provide the possiblity of isolating behaviors to help scum, but they would have to be much more complex and sophisticated than the case where we assigned the sets openly.
I still think the benefits of coordination outweigh the risks, but at least now I understand the counter-point-of-view.

sachertorte

06-27-2007, 01:20 PM

This may be a dumb question, but how can the Oracle and the Apprentice coordinate if the Oracle doesn't know who the Apprentice is? (I may have read the rules wrong; this game has a lot more roles and rules and win conditions than I have played before.)

The Oracle doesn't know who the Apprentice is, and the scheme doesn't require it. Both the Oracle and the Apprentice know who the ORACLE is. That's all that is needed for coordination. They coordinate based on the Oracle's position in the player queue. This is information that both have and need not communicate.

DiggitCamara

06-27-2007, 01:20 PM

(snipped)

Please tell me how this reveals the Oracle's identity to scum.
It doesn't. Not directly.

However:
1. It would constrain the Oracle's investigations. (S)he has certain interactions, ideas, etc. which would possibly leave him/her with a completely different investigation angle.
2. The breadcrumbs that (s)he leaves would, probably, make it easier to detect him/her. For instance:
a.)(s)he says so-and-so is suspicious
b.)so-and-so happens to be in subset I
c.)so-and-so happens to be scum (scum WILL know that)
If someone "happens" to point a finger a two scum from subset I, it would be a pointer for scum that that person is the Oracle

And it doesn't actually help town. The odds of both of them investigating the same player are one in 28. And since the Apprentice knows who the Oracle is, the odds are even lower (the Apprentice probably wouldn't investigate someone with whom the Oracle had high interaction).

In other words:
1. Power Roles will probably figure out how to use their powers without your input
2. Talking about power roles helps scum far more than it helps town (involuntary slips, etc.)
3. Try to find scum,

Fretful Porpentine

06-27-2007, 01:21 PM

This may be a dumb question, but how can the Oracle and the Apprentice coordinate if the Oracle doesn't know who the Apprentice is? (I may have read the rules wrong; this game has a lot more roles and rules and win conditions than I have played before.)
If I'm understanding sachertorte's posts correctly, the idea is that they both know who the Oracle is, and therefore one of them can choose to investigate the group the Oracle is in, and the other one can investigate the group where he isn't.

Upon reflection, I'm inclined to agree that the disadvantages to sachertorte's plan outweigh the advantages, but I don't see the harm in discussing ideas like that and turning them over -- the Oracle and Apprentice, after all, are perfectly free to ignore the discussions.

sachertorte

06-27-2007, 01:24 PM

The problem with your system is that as soon as the Oracle discovers a nonbeliever and starts dropping hints, the cult might pick up on it and know which group he is investigating. They therefore know what group he is in - the other one.
Well, um. If the Oracle starts dropping hints and the cult picks up on it, don't you think it would be easier to find the Oracle by looking at the person who dropped the breadcrumb?

fluiddruid

06-27-2007, 01:26 PM

ComeToTheDarkSideWeHaveCookies

06-27-2007, 01:27 PM

This may be a dumb question, but how can the Oracle and the Apprentice coordinate if the Oracle doesn't know who the Apprentice is? (I may have read the rules wrong; this game has a lot more roles and rules and win conditions than I have played before.)

Hypothetically, let's say that the 11th player on Blaster's player list is the Oracle. (In actuallity, 11 was my number for high school sports *queue that damn cheer in my head*)

Both the Oracle and the Apprentice know that the hypothetical player 11 is the Oracle, and can therefor split the list such that the Oracle investigates the half of the list that he/she is not on (players 16-30) while the Apprentice investigates the half that the Oracle is on (players 1-15).

And you can all disregard my post 445 (http://boards.straightdope.com/sdmb/showpost.php?p=8722323&postcount=445) . Due to a lack of braincells, I was adding the unnecessary complication that it was a bad thing if the Apprentice was on the same half of the list as the Oracle. :smack:

And upon preview, everything that Fretful said.

FlyingCowOfDoom

06-27-2007, 01:31 PM

No scum - certainly not one as experienced as NAF - would willingly associate themselves with another scum based on such a frivilous risk with little possibility to gain.Who said that NAF was defending scum? I certainly didn't. You missed my point. I never said scum would defend scum. I said scum would defend town. It's a common scum tactic to get said townie to think subconsciously, "oh, X is defending me...he must be a friend."

--FCOD

sachertorte

06-27-2007, 01:32 PM

The odds of both of them investigating the same player are one in 28.
That is true for Night One.
After the first night we need to be concerned about overlap that night and from previous night investigations. If the probablity of overlap was only 1/28 each night I wouldn't have bothered. But as each day passes, the chance of overlaps will increase.

Queuing

06-27-2007, 01:33 PM

I think this helps the problem you point out. I have a glimmer of an idea that the division may still provide the possiblity of isolating behaviors to help scum, but they would have to be much more complex and sophisticated than the case where we assigned the sets openly. I still think the benefits of coordination outweigh the risks, but at least now I understand the counter-point-of-view.

What are these benefits you keep speaking of? The only one I see is that by dividing it up into 2 groups there will never be a duplication investigation. Is that it? Are there any other benefits? If that is it then what is gained is very limited and duplication is unlikely regardless. The Oracle has 29 other people to choose from., the Apprentice 28. I will let some of you who enjoy Math to work it out. I believe the odds will be against this happening.

Therefore the benefits are slim. The risk is an easier to find Oracle. On the scales of worthwhileness, this plan is not.

sachertorte

06-27-2007, 01:36 PM

IMHO "defending" others is not a scum tell at all. I never defended scum as scum, I only ever defended townies (and did little defending at all, frankly). Scum know that creating associations is essentially how they are caught.

Do you really think NAF is stupid?

The logic of your post is wanting.

FOS FlyingCowofDoom

What is that Cow up to now? Oh wait, I agree with the Cow on this one. I don't think there was the implication that NAF was scum defending scum. He could be scum defending a known town to make himself look good.

The missing logic is fluiddruid's

DiggitCamara

06-27-2007, 01:43 PM

That is true for Night One.
After the first night we need to be concerned about overlap that night and from previous night investigations. If the probablity of overlap was only 1/28 each night I wouldn't have bothered. But as each day passes, the chance of overlaps will increase.
What would be your threshold then? 1/13? 1/20?

If the necessity should arise by then (that is: both players are still alive and well and have not outed themselves), the discussion should held then.

Of course, if they have a comprehensive list of living players and their investigation results at such a time (4, 5 Days?), they should out themselves and publish the investigation and that would render the scenario moot, anyhow.

sachertorte

06-27-2007, 01:44 PM

What are these benefits you keep speaking of? The only one I see is that by dividing it up into 2 groups there will never be a duplication investigation. Is that it?
Well, yes.

Okay. Maybe, I've gotten a bit carried away. I still think the risks are small (if not non-existant). But happily, I understand the counterview now, so I can sleep and stuff.

Fretful Porpentine

06-27-2007, 01:45 PM

What are these benefits you keep speaking of? The only one I see is that by dividing it up into 2 groups there will never be a duplication investigation. Is that it? Are there any other benefits? If that is it then what is gained is very limited and duplication is unlikely regardless. The Oracle has 29 other people to choose from., the Apprentice 28. I will let some of you who enjoy Math to work it out. I believe the odds will be against this happening.

Therefore the benefits are slim. The risk is an easier to find Oracle. On the scales of worthwhileness, this plan is not.
Well, there is another potential benefit -- it makes it easier for the town to sort the real breadcrumbs from the white noise if either the Apprentice or the Oracle is killed, assuming we know which half of the possible suspects the dead player was investigating. I don't think this outweighs the drawbacks we've already discussed, though.

sachertorte

06-27-2007, 01:48 PM

What would be your threshold then? 1/13? 1/20?
The probabilities go up, not just because there are fewer people, but because they can overlap from previous night investigations. In other words the Apprentice's Night 3 investigation can overlap with the Oracle's Night One investigation and vice versa. I admit that the numbers may not be as large as I had imagined, but they are not totally insignificant. I'd do the math, but factoring in the day and night kills makes the whole thing rather daunting.

fluiddruid

06-27-2007, 02:03 PM

Who said that NAF was defending scum? I certainly didn't. You missed my point. I never said scum would defend scum. I said scum would defend town. It's a common scum tactic to get said townie to think subconsciously, "oh, X is defending me...he must be a friend."Fair enough, I didn't think it through enough to understand you meant it that way. So I remove my FOS.

Nonetheless, in my experience, defending another player is not a scum tell. It's natural to want to defend someone if you feel that they are being criticized unfairly. Yes, it can be abused, but, frankly, if I were scum and I were NAF, I wouldn't be wanting to create associations so early; the less information to give, the better, from a scum perspective at this point.

The best (or, perhaps more accurately, easiest) way for scum to succeed in the early game is to simply stay off the radar altogether. Make weak or no assertions, vote little, suspect little, post innocuously and avoid suspicion. MIII showed the value in letting townies go about their business in the early game. If I would have changed one thing about that game it would have been to draw back more and let the town go about its dirty business. The odds are frankly on their side (i.e. nonscum being dunked rather than scum) so early and why give people a case to build on?

Blaster Master

06-27-2007, 02:04 PM

Kyrie Eleison was the first to respond to the Sub PM and he will be taking Clockwork Jackal's spot. I will be posting an updated player list shortly.

Blaster Master

06-27-2007, 02:18 PM

What role does the Avatar believe he is playing?

The Avatar believes he is a regular Cultist. Similarly, the Psychopath believes he is a regular Non-Believer.

Is the Apprentice told whether the results of his investigation are correct?

No, he will get a response functionally equivalent to "PlayerX is playing the role of RoleX." If the Oracle is alive, he will only be able to determine his accuracy through mathematical formulae (which I'd love to provide, but I'll let you all figure it out). Because an incorrect reading is a random sample of all living players, it's still possible that he could get a correct reading (if he happens to hit the same person or someone with the same role) and it's also possible he could get an obviously incorrect reading (if he gets his own role, or another role he knows that player can't be).

OTOH, if/when the Oracle dies, if he gets an incorrect reading, he will know, because the response will come back functionally equivalent to "You were unable to determine the role of PlayerX".

Blaster Master

06-27-2007, 02:22 PM

1 ArizonaTeach
2 USCDiver
3 Idle Thoughts
4 Zuma
5 Hockey Monkey
6 storyteller0910
7 NAF1138
8 sachertorte
9 SnakesCatLady
10 Malacandra
11 Mtgman
12 Kyrie Eleison (repl. Clckwork Jackal)
13 Hal Briston
14 Pleonast
15 DiggitCamara
16 Fretful Porpentine
17 Captain Carrot
18 Pasta
19 FlyingCowOfDoom
20 Scuba_Ben
21 Queuing
22 Zeriel
23 MonkeyMensch
24 ComeToTheDarkSideWeHaveCookies
25 Autolycus
26 HazelNutCoffee
27 MHaye
28 fluiddruid
29 MadTheSwine
30 Captain Klutz

MHaye

06-27-2007, 02:33 PM

Well, um. If the Oracle starts dropping hints and the cult picks up on it, don't you think it would be easier to find the Oracle by looking at the person who dropped the breadcrumb? :smack:

(I'm very fond of the headbanging smiley, but that'll do.)

I'm going to put it down to five hours sleep followed by a hard slog home on public transport.

Right now it's time to try reading the thread from where I left off last night, which is aroundabout post 350. Who knows? I might pick something up. With the help of the coffee.

storyteller0910

06-27-2007, 02:34 PM

Fair enough, I didn't think it through enough to understand you meant it that way. So I remove my FOS.

Nonetheless, in my experience, defending another player is not a scum tell. It's natural to want to defend someone if you feel that they are being criticized unfairly. Yes, it can be abused, but, frankly, if I were scum and I were NAF, I wouldn't be wanting to create associations so early; the less information to give, the better, from a scum perspective at this point.

The best (or, perhaps more accurately, easiest) way for scum to succeed in the early game is to simply stay off the radar altogether. Make weak or no assertions, vote little, suspect little, post innocuously and avoid suspicion. MIII showed the value in letting townies go about their business in the early game. If I would have changed one thing about that game it would have been to draw back more and let the town go about its dirty business. The odds are frankly on their side (i.e. nonscum being dunked rather than scum) so early and why give people a case to build on?

This is one of my pet subjects, so I'd like to respond (anything to get off the subject of the Apprentice. Yeesh. I'd FoS the whole lot of you, but disagreeing with me on strategy doesn't equal scum tell, something I will struggle to remember as the game goes on but am trying to keep firmly in mind for the moment).

There is no best way for scum to succeed, or rather, no best one way. Pick any given pattern of behavior, and chances are at least some scum will act that way, and some will not. Early in M4, Gadarene and NAF were filling page after page with posts, Autolycus was as silent as he's been in this thread so far, and tirial was relatively quiet and inoffensive as you suggest. If the scum are all acting in the same way, as your last paragraph suggests, then they are handing us the keys to the kingdom if we catch even one of them.

That said, a word on the "defending other players = scum tell" meme, which is valuable as a caution if nothing else. I'll disagree with your contention that creating associations is something the scum don't want at this stage; creating associations is a huge advantage in the mid- to end-game, and creating associations early is the only way to create them at all. If you're scum, and someone decides on Day 1 that you are trustworthy, you are golden.

What's the best way to make townies think you're trustworthy? Why, exploit their own fears of being suspected / lynched. Tell someone you think they're town and they will - not entirely consciously - put a little green check mark next to your name in their ledger.

Here's the thing: defending another player against spurious reasoning is not inherently scummy; it's playing the game properly. I don't have a problem with NAF's rejection of my logic against MadtheSwine because NAF disagreed with my arguments; it was the gratuitous "I'm not getting a scummy vibe" that raised my eyebrow. At this stage of the game, we have no reason to trust anyone.

Remember: scum know who is town and who is not. If Player X seems to be very fervently sure that a particular player is town, and that player IS town, we will often tend to mentally clear Player X. "If he were scum," we think, "why would he defend a townie like that?" But the answer is, because he knew he'd be proven right, and earn trust. Remember: in order to win, all the scum have to do is not all die. Just one, just one of them has to live. It is much more important for any given scum to lay the groundwork for future trust then to waste goodwill lynching a townie.

So in a situation like that, we should be looking very closely at why Player X defended the townie - what were his reasons? Saying that you think someone is town without any good reason should be no less suspicious than saying you think someone is scum without any good reason.

[/off soapbox]

Pasta

06-27-2007, 02:49 PM

The probabilities go up, not just because there are fewer people, but because they can overlap from previous night investigations <snip> but factoring in the day and night kills makes the whole thing rather daunting.
While I think the oracle/apprentice division is a bad idea, I couldn't resist running a quick Monte Carlo simulation of the problem. Hopefully this will put the discussion to rest.

Assuming:
- The oracle doesn't investigate him/herself.
- The apprentice doesn't investigate him/herself.
- Otherwise, each investigates a random person they didn't already investigate.
- Two random people die each day (turns out not to matter much for a while since N=30>>1.)

The expected number of investigatory collisions, accounting for past investigations, is:

After Night 1: 0.03
After Night 2: 0.14
After Night 3: 0.33
After Night 4: 0.60

Alternately, the probability that no collisions have occured is:

After Night 1: 0.97
After Night 2: 0.86
After Night 3: 0.70
After Night 4: 0.50

After four nights, there's still a 50-50 chance that a collision hasn't even occured. Presumably we are ready to move on.

Also, I share the sentiment that fine-tuned math analysis (while fun) isn't even in the the same league as thinking about how a person might play his/her role when it comes to weeding out the scum. To all the newbie townies: I encourage you to spend a few moments role playing. Pretend you're scum. No, really believe it. Pick a random list of 6 players or so as your scummates. I'll wait until you're really convinced your scum.... Okay. Have any scum been accused in the thread so far? Respond to it (offline, of course) as you naturally would. Would you do anything about it at all? Would you sit back? What might you post otherwise?

Rather than trying to turn noise into scumtells, try to figure out what straightforward scum would do, and look for that. (In the one game of Mafia I modded (elsewhere), this was the biggest fault of the town. Several (certainly not all) scum were playing super straightforwardly but the town saw all this noise and ripped itself to shreds.)

Of course, we're still certain to dunk our share of townies. But, that's the game...

Pasta

06-27-2007, 02:59 PM

Rather than trying to turn noise into scumtells, try to figure out what straightforward scum would do, and look for that.Also, this shouldn't be read an contradictory to storyteller0910's post above. There will be hard-to-find scum, so certainly be on your toes! But there will presumably be a couple of fish (to use poker terminology) at the table, too. I'm happy to get them first. The sharks will show themselves in due time.

MHaye

06-27-2007, 03:02 PM

MHaye, post 258:

Holy crap! Should we talk about the weather in here? Putting pressure on people through accusations and votes is how we catch scum. And you won't respond even if it means you get lynched? If you're town, you don't want to get lynched and should respond if you're in trouble. I'm not a fan of this cavalier attitude.I'm quite the fan of people talking. I even do it myself - sometimes when I'd be better keeping my mouth shut. Talk away.

The only thing I object to is the practice of saying " I vote player name to get him to do something." Whether that something be to talk, or to make a roleclaim, don't vote me just to make me do it. If you're going to put me forward for full-immersion baptism in Nairu's holy water, give some reasons and I'll try and answer them one way or another. Say something controversial or challenging and I'll reply. maybe I'll try and refute the statement, maybe I'll ask questions.

I simply won't respond to any form of pressure vote. Not in this game nor any other. I'm not stopping you from making pressure votes, or responding to them if you wish. I will not do so, is all.

Idle Thoughts

06-27-2007, 03:32 PM

It's a nice surprise to come on and not find five or six new pages, but only two. Wheew.

CJ, I hope everything is all right with everyone.
And speaking of CJ:

I am tempted to point one at NAF and Idle Thoughts as well, for the same reasons FCOD listed above. MtS's reasons for voting for CJ are rather shaky, IMO, and neither NAF nor Idle Thoughts has made any convincing arguments that support them.

Well, I didn't vote for CJ, now did I? However I did say, earlier, in this post (http://boards.straightdope.com/sdmb/showpost.php?p=8719654&postcount=292):

See, this gets me a little. I've said this before too in other games. Why feel the need to say this and put it out there? It's sort of a given and just strikes me as something someone would say to subtly throw someone off. That and that nobody strikes you as being scummy yet. Well it's the first Day. What were you expecting? :dubious:

Now I've hosted this game a bunch of times and it's been my experience that scum like to throw in subtle comments like this to make it appear that the person in question really doesn't know jack squat. They tend to do it more than others. At least that has been my observations. So to me, that's a good reason to put a little ping on my meter about someone.

And about all this Apprentice stuff, man..

But... but... but. Oh dear.
You can't say the Apprentice has a 0% chance of being investigated. There is a 50% chance that the apprentice has a 0% chance of being investigated. There is also a 50% chance that the apprentice has a 1 in 15 chance of being investigated! That's why statistically it's a wash. Oh Nairu, what have I done?!

This has been a huge thing so far..and mostly just with you. You seemed obsesseed with the role almost or at least learning all about it. I dunno, but I seemed to understand most of it just from the read up. Why the huge interest? Seems to me you've already garnered a lot of suspicion and even votes because of it..come on now. I can understand over analyzing. I do it myself. But even I'm starting to get fairly suspicious of you and FCOD a bit. I don't know what to think about it.

Blaster Master said that there is at least one of each role in the game. From there I looked at the allegiance of said roles and came up with the least number of pro-town Believers/Non-Believers present.

Hal already said it but I thought you might want to hear my explanation as well.

Yeah, it makes sense now. :smack:

As for reading a defense of someone else as a scum tell, I agree that it really depends on the manner of defense. It's possible to defend an argument without defending the person. As I've already stated, I wasn't sold on NAF or Idle Thought's defense of MtS, which is why I am suspicious of both. NAF may be right - MtS's mistake may be an honest one. Then why did he keep the vote? That's what I'm asking MtS to explain.

Ohhh, I must have misread you then. I thought your suspicions of me were based on finding CJ scummy.

As for "defending" MtS, I don't know or see that I really did anything of the sort. I DID say that "at least he has a reason that works for him" in an earlier post and also that I saw the same things he did, but it's all in the eye of the reader if that'd be considered defending someone.

Welcome, Kyrie, to the game.

USCDiver

06-27-2007, 03:39 PM

Well, yes.

Okay. Maybe, I've gotten a bit carried away. I still think the risks are small (if not non-existant). But happily, I understand the counterview now, so I can sleep and stuff.

See I think this is where you misinterpreted my earlier post. There IS a risk. If the Apprentice is NOT in the group that the Oracle is investigating, he has a 0% percent chance of improving his accuracy. The fact that the Apprentice CAN improve is an important nuance to his role. Even though the chance of that happening are 50%, I think that is too big of a risk to lose what could end up being two Oracles (if the Apprentice reaches 80% prior to the Oracle's death).

Kyrie Eleison

06-27-2007, 03:41 PM

Hello, all. I'm just checking in at the moment -- I see I've got quite a bit of reading I should do before I post anything substantial.

USCDiver

06-27-2007, 03:42 PM

Hello, all. I'm just checking in at the moment -- I see I've got quite a bit of reading I should do before I post anything substantial.

Welcome Kyrie

FlyingCowOfDoom

06-27-2007, 03:46 PM

Hello, all. I'm just checking in at the moment -- I see I've got quite a bit of reading I should do before I post anything substantial.Welcome to the game!

...unless you're a Cultist, in which case I hope you die soon.

:D

--FCOD

Idle Thoughts

06-27-2007, 03:49 PM

Oh yeah, that's another thing. Most of the two pages I needed to catch up on consisted almost of nothing but Oracle and Apprentice talk. Whatever happened to trying to find out who was on your team or not? :p

To HOPEFULLY get it back to that, I'll start.

Right now I'm suspicious of a few,...most notably Clockwork Jackal/Kyrie for CJs few posts earlier and sachertorte for seemingly wanting to investigate the roles as much as possible. Don't get me wrong, a lot of others have spent time talking about this too but I'm in the mind of those who think that it doesn't really hope town and doesn't really matter at this point. How much more do you need to learn about it? What else are you confused on? Anaylzying is good and all, but even I know sometimes it's too much. At you're looking at a world class analyzer here.

And lastly, the only other one who has been brought to my attention is Pleonast by way of storyteller. But granded, he's the lowest of them all.

Now that CJ was subbed out, I'd really like to see a lot of posts from Kyrie.

sachertorte

06-27-2007, 03:50 PM

You seemed obsesseed with the role almost or at least learning all about it.
Obsessed is a good word; I don't deny it. Although I'm obsessed with analyzing game mechanics in general, only so far it's been fairly investigation oriented. I posted my thoughts on the Avatar which is admittedly radical, but does present a pretty cool option for what we could do if investigators found the Avatar but we had no nightkillers to dispatch him at night. I don't have any good ideas on the Prophet. Obviously we'd like to dunk him ASAP, or at the very least get a recruitment to happen sooner rather than later, but I have no ideas on how to enourage this to happen.

Lately my thoughts have been drifting to non-believers. Someone upthread said something about thinking the number of non-believers could equal the number of Believers. That got me thinking. I don't think that would be the case, because that's quite a lot of millers. The fact that the game is designed so that non-believers can win with believers makes me think the number of non-believers is fairly low, but not tiny. I've also been thinking about what to do about non-believer investigation results or role claims.

Take a look at the forbidden thread for M3. I think that pretty much establishes my mindset, approach to game analysis, and general obsessive demeanor.

Malacandra

06-27-2007, 03:52 PM

See I think this is where you misinterpreted my earlier post. There IS a risk. If the Apprentice is NOT in the group that the Oracle is investigating, he has a 0% percent chance of improving his accuracy. The fact that the Apprentice CAN improve is an important nuance to his role. Even though the chance of that happening are 50%, I think that is too big of a risk to lose what could end up being two Oracles (if the Apprentice reaches 80% prior to the Oracle's death).

I see what you're saying, but I don't think it matters. The Oracle will investigate a certain number* of randomly chosen people during the game. It doesn't matter what criteria he uses to split the pool - either his investigations will include the Apprentice or they won't. I'm assuming no evidence arises during the game to suggest a possible Apprentice to the Oracle; if it does, then he should follow his hunch, of course. But otherwise the only difference is that an external observer with full knowledge could see that the Oracle was predestined to fail. He could see that anyway, if the Oracle published the entire list of investigatees at the start of the game, whether the Oracle had split the player list in half by number or by any other means.

Summarising: If the Oracle and Apprentice are investigating opposite halves of the player list, there is no objective difference in the chance that the Oracle will find the Apprentice; but there is the small improvement in that the two are guaranteed not to investigate the same person. (Not that this is a huge chance anyway.)

* About 7 each given the death count per day and the probability of the Oracle or Apprentice getting whacked.

sachertorte

06-27-2007, 04:03 PM

Queuing

06-27-2007, 04:08 PM

Summarising: If the Oracle and Apprentice are investigating opposite halves of the player list, there is no objective difference in the chance that the Oracle will find the Apprentice; but there is the small improvement in that the two are guaranteed not to investigate the same person. (Not that this is a huge chance anyway.)

Thankyouthankyouthankyouthankyouthankyou.

Just to clarify my position;

We knew this already. We all get your point sachertorte. The benefit gained by this idea of yours is so infinitesimal that ANY risk it adds to the oracle is soooo not worth it. Enough of this however, at least from me.

HazelNutCoffee

06-27-2007, 04:18 PM

sacher, the way you cling to the Oracle/Apprentice discussion is amazing. For the record, we understand your plan perfectly, as Queuing has already stated. We just don't think it's a good idea, as has been explained so many times already I have lost track. I would vote for you if I thought a cultist would be stupid enough to paint such a wide target on his own back. I am more inclined to think you may be just an overzealous believer, but we'll see.

Idle Thoughts, re-reading your post, I see you did not defend MadTheSwine per se. And I have already stated that I found Clockwork Jackal somewhat suspicious as well. I am STILL pointing an FOS at MtS. He originally pegged both CJ and Hal as suspicious. When CJ defended her(him?)self and responded to MtS by saying he himself looked scummy, MtS accused her of a revenge vote and voted for her. It was pointed out that CJ did not, in fact, cast any kind of vote. MtS acknowledged this but kept his vote. If his vote had been prompted by the supposed revenge vote, why wasn't it retracted after he found out his mistake? If it was the original "scum tell" that prompted the vote, why CJ and not Hal?

MtS also accused CJ of twisting his words, but to me CJ's interpretation seemed like a valid one. Saying that discussion of X is a scum tell does sound like an attempt to stop people from discussing X. (X = the distribution of the roles amongst the players.)

But like I said, I'm waiting for MtS to come and defend himself before I cast a vote.

Mtgman

06-27-2007, 04:20 PM

Um, ok...so why bring it up in the first place? Are you at least backing away from the idea of random.org, because I gotta say, you still haven't addressed my points as to why I think that's a phenomenally bad idea that can only, only benefit scum.The anser to your second question, if I'm no longer advocating a random Day 1 vote for the town. Not because I'm convinced it's a bad idea, or unimplementable, but because with so many people opposed to the idea I'm not interested in shoveling sand against the tide. Plus the approach is only mathematically sound if employed by the entire player base, which it obviously won't be.

Ok, going back to post 294 (http://boards.straightdope.com/sdmb/showpost.php?p=8719665&postcount=294) we find the following list of objections, specifically.1) There are still going to be those who delay even a minute or two that could throw the vote a certain way by casting a non-random vote and saying it's random. And even if that doesn't happen, if the votes truly are random then suspicion will fall upon those people who voted for the dunk unfairly.We could also do something like post a random number generator based on an anagram or hash of your username, so the randomness is transparent and in-thread. So we would know the votes were not cooked. Probably too much work though. Say everyone went to a website and ran an MD5 of their username and then ran all the numbers in it together to make one big number, then ran that through a modulus operation for the number of players. That would be random, but verifiable, so no one could cook their votes. No one could "respond" to a vote by throwing their weight behind a townie. Multiple passes would be necessary to achieve a majority(making a new player list from the ones who had the most votes and using a lower mod number the next time), but it would be impossible to cook the votes this way, so the time factor goes away.2) The chances a pro-town power role could be revealed are small, but possible; in fact, probably at a greater percentage than a normal vote. If a power role is revealed, I damn well want a trail to who forced that to happen. Plus, will there even be time for a role-claim? Will that role-claim be more or less believable?Ok, I gotta ask. How in the heck did you figure out the odds of hitting a power role with a "normal" vote? I guess you could say the town rarely hangs power roles because those roles roleclaim at the 11th hour and avert disaster, but then they're as good as dead anyway because the scum will kill them for us(probably after we speed lynch some other poor townie).

With purely random votes we're looking at a number of scum divided by number of players probability of lynching scum. We're looking at the same formula for pro-town power roles. We know we have at least 7 pro-town power roles, and, based on NAF's info, probably 7 or 8 scum. So the odds of hitting a scum are probably equal to the the odds of hitting a power townie and both are less than hitting some other role. Roughly 23% chance of hitting scum, and 77% chance of hitting pro-town, a 23% chance of hitting a pro-town power role.

I'll cover your comment about traceability of the reasons behind a vote after the next bit.3) This negates any information that could be gleaned from today's conversations. There will be no trail to follow. Not true, the conversations still contain nuggets of info(although probably not useful until we have a larger body of data including how people react to night kills and the revelation of who we dunked) it's just the first day's votes which provide no useful information. Which, let's face it, the first day's votes aren't that useful anyway. You can't establish a pattern from a single data point, and scum can pretty much all vote for scum on the first day and they'll be lost in the noise. Aside from the case of them all voting as a bloc(which would be suicide), we probably won't be able to glean any useful information from Day 1 votes, ever. So why not make the scum vote randomly?

Let's look at the pros of a, verifiable, random voting process. Firstly, it's verifiable. ;) Secondly, it does what so many people here are trying to do with their FOS and early votes, except with teeth. It straps someone into the dunking chair and says "tell me why I shouldn't let the fishes nibble your tasty bits." A honest townie can either take one for the team, or roleclaim. If they're a power role they get a reprive, if they're a Monk, they get off. If we aim the vote process at scum they can either claim power role, in which case we expect them to be sacrificed during the next night or two or swing the dunking chair back their way, or they can take one for the team. Scum claiming Monk is a Bad Idea™ because all it takes is one real Monk counter-claiming and then we dunk them both and have a nice trade of 1/23rd of our citizens for 1/7th of the scum.

Thirdly, it's an objective, transparent, process. If we use subjective processes, such as our own guts, then they can be manipulated by those with more information than us, i.e. the scum. If we start a bandwagon for someone the scum know is the Prophet, they'll work hard, subtly, but hard, to derail it or start another for someone else. How could they do this with an objective, transparent, process? Say, "please don't kill him/her, he's/she's the prophet" or support his/her false roleclaim(and give themselves away in the process)?

Ultimately it's down to math and game theory. The study I cited shows the random vote to be an optimal strategy for the case with no detectives. I say this applies to Day 1 because the Oracle, and his/her Apprentice, have no information, so they may as well not count. The math works out, and even though I've said it before, the standard day 1 voting tactic on Mafiascum.net is random voting. It hits a townie more often than not, but it negates the scum's greatest advantage by forcing them to vote without including the hidden knowledge they have in how to vote.

More to come, including a meta-discussion about the value, or lack thereof, in discussion pro-town strategies and some particular comments.

Enjoy,
Steven

Last edited by Mtgman : Today at 4:21 PM. Reason: Added fake edit tag to mess with people's heads. :)

Mtgman

06-27-2007, 04:23 PM

Now I wish I really could edit. :( That first sentence is mangled. Suffice to say I'm no longer advocating a random voting process.

Enjoy,
Steven

fluiddruid

06-27-2007, 04:32 PM

This is one of my pet subjects, so I'd like to respond (anything to get off the subject of the Apprentice. Yeesh. I'd FoS the whole lot of you, but disagreeing with me on strategy doesn't equal scum tell, something I will struggle to remember as the game goes on but am trying to keep firmly in mind for the moment).

There is no best way for scum to succeed, or rather, no best one way. Pick any given pattern of behavior, and chances are at least some scum will act that way, and some will not. Early in M4, Gadarene and NAF were filling page after page with posts, Autolycus was as silent as he's been in this thread so far, and tirial was relatively quiet and inoffensive as you suggest. If the scum are all acting in the same way, as your last paragraph suggests, then they are handing us the keys to the kingdom if we catch even one of them. Well, of course, there are no absolute tells. Whatever people identify as "mafia tells" are avoided wisely by the mafia, and presenting such tells also manipulates the townsfolk into acting defensively and not exuding such "tells" so they can continue play. That was in essence the point of my point - defending other posters in Day One isn't a true tell, as there are many reasons why not to do it as Mafia and reasons to do it as Town, contrary to FCOM's assertions.

Nonetheless, some strategies are more viable than others and I maintain that it is best to stay off the radar for the first few days for Mafia. Loud people, or, contrarily, people that lurk too much are accused, but seldom are inoffensive or unaccusing souls. These tend to be accused later, but not in the early game, and when it comes down to it, these threads are so long that mild patterns (unlike voting/FOS) are difficult to track and digest real life days and weeks later.

In any case, my ultimate point is that NAF should not be suspected based on the strength alone of an argument that doesn't hold true, which is, defending someone is scum behavior. I don't believe that it is based on my experience in the games; it is very likely to be one of the following:

- An honest Citizen explaining their thoughts innocently
- A scum trying to hide their trail
- A power role trying to breadcrumb (unlikely now)

It is accordingly less likely the earlier it is posted to be done by a pro-town detective-type power role as it is very dangerous to post anything in such a role that you are not sure about, as it is very likely to be taken as God's truth if you die (though this is arguably not the case for Day One as long as town is on the ball enough to point out that no investigation could have been done yet, but, it is also the case that if the person ends up actually posting suspicions of them due to an investigation result, and dies, it dilutes the effect somewhat if the trail goes back to Day One as well, so caution is advised).

Hockey Monkey

06-27-2007, 04:32 PM

What would be your threshold then? 1/13? 1/20?

If the necessity should arise by then (that is: both players are still alive and well and have not outed themselves), the discussion should held then.

Of course, if they have a comprehensive list of living players and their investigation results at such a time (4, 5 Days?), they should out themselves and publish the investigation and that would render the scenario moot, anyhow.

Huh? Wha? Did you just say that they should out themselves after 4 or 5 Days? Why would they need to do that?

FOS DiggitCamara

Mtgman

06-27-2007, 04:48 PM

In general I agree the pro-town power roles should not take marching orders from the hoi polloi. This doesn't mean the group should refrain from discussing them. It's not like we're pointing to a particular person's post and saying, "Oracle tell!" or "That's a very Priestly thing to say." The Cultists are still on their own to find the individuals who have the pro-town power roles. Discussing pro-town strategy does two important things. It helps educate the newer players on how those roles work. Also, it provides chatter for the holders of those roles to hide amongst if they wish to leave breadcrumbs or get feedback on a particular strategy they're considering. I disagree with the idea that talk about pro-town strategies should be muzzled. Consider the reverse, would you think it wise to forbid discussion of Cultist strategies in case we come across a good strategy they hadn't considered? There are probably three times as many townies as there are Cultists, invariably we'll have a broader spectrum of ideas on how the Cultists should approach the game than the smaller group would. Strategies they may not have come up with in their Night-based discussions may well come up here. Should we refrain from discussing them because we don't want to help them beat us? I think not. I think it's important to have frank and open discussions about each and every topic anyone feels is strategically important. By the same token, I don't want the game bogged down with theoretical discussions about extremely unlikely events.

This may seem weird to anyone who interpreted my post on page six(number 298 in the thread) (http://boards.straightdope.com/sdmb/showpost.php?p=8719692&postcount=298) as saying we should shut up about the Apprentice and Oracle. That's not what I meant, I meant we should consider the Apprentice as a confounding factor whose information is so unreliable as to be useless, and therefore not worth much in the way of brain power to analyze the role. A large number of low probability events would need to converge in order to make the Apprentice(pre-death of the Oracle) a useful asset for the town. Sorry about this, whoever got the Apprentice, it's nothing personal. If you could try to do a prescient "I'm about to die, here's who I investigated, the results and when I did it" post right before you get sacrificed/crusaded/psycopathed/avatared/dunked I'd still appreciate it. Thanks ;)

Enjoy,
Steven

SnakesCatLady

06-27-2007, 04:48 PM

Mtgman, in your discussion about random voting above, I think you said that power roles should roleclaim on Day 2 - the day following the random vote. I hope I misunderstood you, as I cannot imagine why you would want to expose the town's secrets so soon, and for no good reason. Actually I can imagine why - if you are Cult.

If I misunderstood you I am sorry - some of those long posts make my eyes cross.

DiggitCamara

06-27-2007, 04:48 PM

Huh? Wha? Did you just say that they should out themselves after 4 or 5 Days? Why would they need to do that?

FOS DiggitCamara
Because, hopefully, they'll have enough information by then to reduce the number of unsuspected/suspected townies by a bucketful.

Think about it: after 5 Days we'll have at least 10 players less (it'll tend to be around 12, unless the Psychopath is activated and the math gets fuzzy from there).

Let's assume 10 players less:

Therefore 20 players left. Deduct 2 "confirmed" town folk (Oracle + Apprentice). Imagine they only inspected living players until then (5 inspections each). Which pretty much gives the town an overview of about 50% of the Town--who should be suspected, who shouldn't, who should be re-inspected, etc. etc. It forces the Cultist's hand, since they'd have to attack unsuspected Townies.

Sounds pretty good to me.

Which is why it most certainly ain't gonna happen.

Idle Thoughts

06-27-2007, 04:51 PM

Idle Thoughts, re-reading your post, I see you did not defend MadTheSwine per se. And I have already stated that I found Clockwork Jackal somewhat suspicious as well. I am STILL pointing an FOS at MtS. He originally pegged both CJ and Hal as suspicious. When CJ defended her(him?)self and responded to MtS by saying he himself looked scummy, MtS accused her of a revenge vote and voted for her. It was pointed out that CJ did not, in fact, cast any kind of vote. MtS acknowledged this but kept his vote. If his vote had been prompted by the supposed revenge vote, why wasn't it retracted after he found out his mistake?

I have no idea...but the way I see/saw it (which is why I am where I am now) is because of this post (http://boards.straightdope.com/sdmb/showpost.php?p=8719002&postcount=232) from MtS (the original one that started it). Reading that, it just seems it's already apparent Mad had suspicions about CJ. Not only that, but it was something I notced myself, before I ever even seen Mad's post there. So keeping his vote, even though what finally stoked it was misinfo about a revenge vote, still doesn't strike me as all, off the wall, odd or shady. I'm not saying that I trust Mad any farther than I can throw him, I'm just saying, based on what I've observed and felt myself, that his kept vote isn't all that off base, that's all.

And about Hal, that's something else entirely. It must have tripped his (Mad's) meter but it didn't mine. YMMV.

Blaster Master

06-27-2007, 04:54 PM

sachertorte (2) - SnakesCatLady, Pleonast
Clockwork Jackal / Kyrie Eleison (1) - MadTheSwine
MadTheSwine (1) - storyteller0910
NAF1138 (1) - FlyingCowOfDoom

Voter - Action - Votee - Post
Pleonast - Vote - USCDiver - 221
DiggitCamara - Vote - ComeToTheDarkSideWeHaveCookies - 225
MadTheSwine - Vote - Clockwork Jackal - 244
Pleonast - Unvote - USCDiver - 248
DiggitCamara - Unvote - ComeToTheDarkSideWeHaveCookies - 251
storyteller0910 - Vote - MadTheSwine - 281
NAF1138 - Vote - Clockwork Jackal - 344
FlyingCowOfDoom - Vote - NAF1138 - 383
Pleonast - Vote - MadTheSwine - 387
sachertorte - Vote - FlyingCowOfDoom - 388
NAF1138 - Unvote - Clockwork Jackal - 394
SnakesCatLady - Vote - sachertorte - 403
Pleonast - Unvote - MadTheSwine - 405
Pleonast - Vote - sachertorte - 405
sachertorte - Unvote - FlyingCowOfDoom - 416

HazelNutCoffee

06-27-2007, 05:00 PM

DiggitCamara

06-27-2007, 05:15 PM

A LOT can happen in 4 or 5 days, and there's no telling what the town will look like by then. We'd be better off discussing what to do now, rather than worrying about an uncertain future. I am tempted to say that power-roles should not role-claim until they absolutely must, but then again I am reminded of sturmhauke's play in the first Werewolf game, where his initial role-claim seemed much too early to be sensible but ended up being the key to the town's victory. So it depends on the circumstances; but again, what good does it do to discuss these things now, when we're not even halfway through the first day?

DiggitCamera, you never answered my question about your post 342 (http://boards.straightdope.com/sdmb/showthread.php?p=8720002#post8720002), when you said that a "random vote" might be an excellent strategy. Care to elaborate?
HockeyMonkey: I forgot to mention one other thing--I was referring to Days, not days, by which I meant to say that an outing of the Oracle and his/her Apprentice would take place in about 5 gameDays.

As to your question, HazelNutCoffee... heh... well, I didn't want to wake up that pack of dogs unless someone else understood my implicit idea (I've done that before, remember, Idle Thoughts?)... on the other hand there is a REAL advantage in having two separate "detectives" around... but here goes nothing:

If everyone (and by everyone I mean at least 80% of the current player population) were to publish a "random vote" (with unvoting, of course), the Apprentice could hide his/her "random vote" among the noise.

With a truly random vote (by everyone else) you'd have votes flying every which way. However, the Apprentice with his/her "random" vote could actually purposefully be voting for the Oracle. The Oracle would only have to cast his/her real night-time investigation in the Apprentice's direction...

Since everyone would be adding their own really random vote to the mix, the noise would make the hidden information all but inaudible to the Cultists.

Anyway, it's just an idea.

Mtgman

06-27-2007, 05:23 PM

Huh? Wha? Did you just say that they should out themselves after 4 or 5 Days? Why would they need to do that?

FOS DiggitCamara
It's actually a good Oracle strategy to hide for a few days and then reveal what you know(almost certainly at the cost of your own life). The key to understanding why is to know what the real value of the Oracle role is. The role doesn't sniff out Cultists, it validates townies. After four or five nights you've got probably one to two Cultists and three to four townies you can verify with a tell-all post. The town dunks the Cultists, and the verified townies join the Monks as a trusted block of players who can see the game through. You can't get a majority without either convincing this block(and it narrows the field of people you can convince them to dunk if you're a Cultist trying to misdirect the town). The life expectancy of an Oracle is usually about the number of residents divided by four, so we can expect them to live, at max, eight Days. Witness the crushing blow dealt by the Oracle in the original Werewolf game when the "wait a bit, then tell all, in a clear and unambiguous way" strategy was employed. I think, but I haven't run the numbers, that if we have five Day/Night cycles worth of data from an Oracle, that would be enough to break the Cult's back. It depends on how fast we lose Monks really.

One thing I haven't seen mentioned, and I think we should, would be the night kill math. How do we interpret a night with just one kill? No kills? Three kills? How do we detect the recruitment? These are going to be key questions. I'll think about them a bit and post my thoughts later.

Enjoy,
Steven