TWG XL: Mall Madness

[aperson]

Re: TWG XL: Mall Madness

Additionally, in case the SK is clueless: You want to keep this game as close to equilibrium as possible while giving humans just a small edge since it is better to be against humans in the end than wolves (2 humans is not an instant defeat, 2 wolves is). Since humans are now already boned in the ass you should be playing for the humans, and should continue to do so until at least two wolves are dead to restore a small advantage to the human team. If all three blues are gone you should play down to one wolf. In fact, if the wolfsbane came out you should at least give them a days breadth because killing them tonight would just be killing yourself: The humans would be completely dead and you would have to have a marvelously uncanny knack for hitting wolves to stand a chance in hell.

Kilga is right. Talisman is wrong as hell in his posts. Seriously, it sounds like he isn't even using his brain when he's trying to come up with his reasoning. Chromer is an obvious kill for reasons posted above, and Kilga already corrected him for his wolfsbane reasoning.

[talisman]

Re: TWG XL: Mall Madness

I acknowledged that kilga was right and I'll acknowledge that your theory makes a hell of a lot of sense too. That said, I still think I'm right that the bane coming out would be a bad play, as it would essentially remove his power from play and have only a 50% chance of doing any good (if the seer were alive and could work through the bane).

If your theory is right, it opens up good possibilities if the seer is still alive (either the bane or the guard would be alive and able to protect the seer). Though there would still be that 50% chance that the sk could take advantage of later in the game, if he felt the seer had outlived his usefulness.

[aperson]

Re: TWG XL: Mall Madness

First off, I'd like to point out that some of my calculations were wrong up there. I was going by Observer's post where he said 3 wolves(1 MW) in the hosting game, which made me think there were 3 wolves not 4 (the SK hits a wolf a bit over 25% of the time, not 20%). I just noticed the first post where there are... haha oh man this is class... four wolves. Good job observer you are a fucking idiot for thinking this game is anywhere near fair.

But anyway.... At this point, it is absolutely crucial that humans have as much information as possible. If we want any kind of chance of winning this game we have to trim the pool of possible wolves / SK down as much as possible. Right now, there are 14 players left, of these players, 5 of them are optimal kills. By blind luck (which is what day 1 usually is without any substantial help), we have a 35% chance of lynching a wolf or the SK today.

Let me be frank here, if we don't hit one of those players tonight, we're fucked. There is no way around it, the only way we still have any breath of a chance is if the SK hits a wolf the next night, which is still < 50%.

So what about having the seer come forward. If the seer comes forward and reveals who they saw, we now have two more players removed from the candidate pool. Additionally, the other living blue will be protected by the seer from being lynched tonight. That raises our odds of pegging a wolf to 5/11. Additionally, the SK does not know whether it is the guardian or the wolfsbane protecting the seer. Now, the SK does not kill the seer tomorrow no matter what. Even if the seer was the only blue left right now, the SK doesn't touch him. The SK is not an idiot, the SK realizes that leaving the seer alive is a dominant strategy. However, the problem is that the next night, it might be optimal for the SK to risk going for the seer if at least one wolf is dead. Why? Because if the SK kills the blue he can now start shooting for wolves, which due to their high concentration will be really easy to hit. If the kill misses, he really doesn't care because the humans are still hurting from all the kills being stacked on them anyway... Additionally, it let's the wolves know that they have no fear of dying. Both outcomes here could very well be advantageous to the SK, even MISSING HIS KILL. This is why revealing the seer could be dangerous. However, we get a 10% gain in pegging a wolf/SK if the seer comes out, and at this point, that is significant as hell.

The guardian... should not come out under any circumstances. This is obvious as hell. You are toast on night2 or night3.

If the wolfsbane comes out, then we are in a dangerous quandary: If the seer is alive, the wolfsbane cannot report who the extra human is, as doing so immediately tells the SK that he is fair game. This causes the wolfsbane to have to use some verbal trickery to try to guide us away from the seen player if we happen to try to lynch them. This is almost always a transparent move and both players would be outed.

Therefore, I think the best strategy, if we were to want anyone to come out, is to have the seer come out. If the seer comes out, we operate under what I described above. However, if after a set amount of time (24 hours or so), the seer does not respond, we assume wilkin was the seer. Now, the wolfsbane comes out. This means that the wolfsbane is most certainly protected by the guardian and thus cannot be touched. This option kicks ass as it removes some names from the lynching pool and gives the humans more information (which we need horribly right now, no matter that it gives the wolves/SK the same information as well).


Discuss.

[jwcgator]

Re: TWG XL: Mall Madness

One question:

Can the Guardian guard himself?

[aperson]

Re: TWG XL: Mall Madness

One question:

Can the Guardian guard himself?

Have you ever played this game before?

Think really hard about what happens if the answer is yes.

[talisman]

Re: TWG XL: Mall Madness

having the seer report who they saw makes me uneasy. It would seem better to me that this be done only on an as needed basis. (and on a technical note, if the seer seered a green it doesn't mean that a human has been seered, which probably factors into the odds somewhat. and the seer could have seered wilkin or chromer, of course.)

and again, I would question having the bane come out at all, unless it's absolutely needed (ie the bane is about to be lynched or whatever). if the seer is indeed dead, having the bane out removes his effectiveness (and that of the guardian since the guardian would have to be guarding him) and thus pratically guarantees the two kills a night. Whereas if he's private, both the guardian and the bane have a chance to foil the wolves/sk.

[nforcer06164]

Re: TWG XL: Mall Madness

You do realize, since your first post, you have completely contradicted yourself?

[aperson]

Re: TWG XL: Mall Madness

having the seer report who they saw makes me uneasy. It would seem better to me that this be done only on an as needed basis. (and on a technical note, if the seer seered a green it doesn't mean that a human has been seered, which probably factors into the odds somewhat. and the seer could have seered wilkin or chromer, of course.)

and again, I would question having the bane come out at all, unless it's absolutely needed (ie the bane is about to be lynched or whatever). if the seer is indeed dead, having the bane out removes his effectiveness (and that of the guardian since the guardian would have to be guarding him) and thus pratically guarantees the two kills a night. Whereas if he's private, both the guardian and the bane have a chance to foil the wolves/sk.

Right, but the problem is right now that we're behind as hell. If we don't make right moves from here on out we're completely toast. We can sit on our ass and wait for a a 1/8 wolf death or a 1/13 guarding (that's a 20% chance of a defense or kill tonight assuming that we lynch a green AND that wilkin was the seer. This fantasyland combination only happens 1/3 * 8/13 = 20.5% of the time. Any other combination lowers the odds below 20%) or we can increase our odds of hitting a wolf by about 10% right now which carries through all future days as well.

Look, this is a game where we are going to have to physically assign and approximate odds to make the best EV moves. Sitting on your ass and chirping about your worries without providing any numbers for the decisions isn't going to help. We all know that the alternate strategy is to keep them hidden and using their abilities to try to retaliate... But HOW MUCH do you value keeping a human alive compared to killing a wolf and HOW MUCH do you value killing a wolf at night and a human during the day rather than killing a human at day and locking away two important people from getting killed at night. What you can't seem to get through your thick skull is that there is a 2/3 chance that one of our 'retaliators' is dead, and at this point, maybe the proper information to let us determine the best odds is worth more than the danger they might be put in. However, until you start quantifying your posts you are going to get absolutely nowhere and are just going to stall the game and continue to look retarded.

[talisman]

Re: TWG XL: Mall Madness

if the seer is alive and comes out, the seer is already responsible for keeping one player (the other blue) alive. This is problematic for your plan in and of itself, since, like you said, if the seer tries to steer votes away from that person, it looks obvious, and (going with your plan) it would thus be obvious that that person was the guardian/bane, and that blue would be lost to the sk. If the seer doesn't reveal who they saw (if they saw a human, obviously they would if they saw a wolf), they're in the same situation, except with two people.

Either way (the seer revealing the person or not), the solution, the only way to keep everyone alive that the seer knows about, would be to put the seer essentially in charge of the game. Doing something like "ok I'm the seer today we pick from among these three names" or something like that. (which would be better than the seer saying "everyone hump my vote"). of course, the bane would be in this position as well should the bane be forced to be the alliance leader.

so I'm fine with either coming out to lead an alliance, but it would mean that they would have to dictate/limit where the votes go, and if they're doing that anyway, then there's no need (in the case of the seer) for the seer to reveal who they saw.

[aperson]

Re: TWG XL: Mall Madness

To further expand on some of my points:

There are 18 players originally, 16 are alive now. 4 are wolves, 1 is the SK (because there is only a 1/13 chance that the SK is dead, I am ignoring this chance because our strategy will change significantly tomorrow when only one death report comes)

Let's say that the value of a human being guarded is 1, and the value of a wolf dying is 1.25. Therefore, the value of a wolfsbane is 1+1.25 = 2.25. However, some of the value of a guarding is knowing that a player is human, since the wolfsbane is a special role which can protect itself, the full guarding value should be reduced to around an even 2.0. The value of a human dying can be around... let's say -0.75. Therefore the total value of a human being guarded is 1.75

If the guardian does not come out:
If we lynch a green tonight (11/16):
If the seer is dead, we have a 9.1% chance of a wolf dying and a 7.14% chance of a human being guarded.
If the guardian is dead, we have a 9.1% chance of a wolf dying tonight
If the wolfsbane is dead, we have a 7.14% chance of someone being guarded tonight
(.091*2 + .0714*1.75)*(1/3) + (.091*2)*(1/3) + (.0714*1.75)*(1/3) = 0.2046

If we lynch a wolf or the SK tonight (5/16):
If the seer is dead, we have a 8.33% chance of a wolf dying and a 7.14% chance of a human being guarded.
If the guardian is dead, we have a 8.33% chance of a wolf dying tonight
If the wolfsbane is dead, we have a 7.14% chance of someone being guarded tonight
(.0833*2 + .0714*1.75)*(1/3) + (.0833*2)*(1/3) + (.0714*1.75)*(1/3) = 0.1944

(0.2046)*(11/16) + (0.1944)*(5/16) = 0.2014

Additionally, during the daytime a human dies 11/16 times: -0.75 * 11/16 = -0.5156
a wolf dies 5/16 times: 1.25 * 5/16 = 0.3906
-0.5156 + 0.3906 = -0.1250
-0.1250 + 0.2014 = 0.0764
0.0407 is the average payoff for the blues not coming out

If a blue player comes out and 3 possible options are eliminated from the voting pool (0.66 * 14/16) = 0.5833:
Chance of lynching a wolf or SK: 5/13 * 1.25 = 0.4808
Chance of lynching a green: 8/13 * -0.75 = -0.4615
0.4808 - 0.4615 = 0.0193

If a blue player comes out and 2 possible options are eliminated from the voting pool: 1-0.5833 = 0.4167
Chance of lynching a wolf or SK: 5/14 * 1.25 = 0.4464
Chance of lynching a green: 9/14 * 1.25 = -0.4821
0.4464 - 0.4821 = -0.0359

(0.0193 * 0.5833) + (-0.0359 * 0.4167) = -0.0037

0.0764 - -0.0037 = 0.0803
So we've weighted killing a wolf at 1.25, and the total difference between the 2 strategies is only a meager 0.0803. It's only worth the lives of 0.0642 wolves to keep the seer or wolfsbane hidden. This really isn't much, in fact we would have to warp the values of each event a lot before the difference became significant. When we factor in the value of information, I'd say that the value of information is easily worth 0.0642 wolves. In fact, I didn't even weight in the intensely high value of killing the SK, which can only occur through lynching. The more we can narrow the lynching field down the better.

Really, I don't see why my strategy I posted above is a bad idea?

Edit: In case it isn't clear. I weighted the night payoffs for the no-reveal strategy such that the payoff for the reveal strategy is 0.

[DiscoBobbyPARANOiA]

Re: TWG XL: Mall Madness

Chromer could have been the SK, but it's HIGHLY unlikely (I don't know what happens if the wolves kill the SK and the SK sends a kill too).

SK can't be wolfed until the number of humans is less than or equal to the number of wolves, which currently (or when the kill took place) makes the equation look something like 18 <= 4, which is not true; the SK is still alive.

so I'm fine with either coming out to lead an alliance, but it would mean that they would have to dictate/limit where the votes go, and if they're doing that anyway, then there's no need (in the case of the seer) for the seer to reveal who they saw.

I completely agree with this part. Not much more to say about it.

AP: Being a dick isn't part of the game. Instead of insulting people, try to objectively determine if his behavior/comments make him suspicious.

Sert: You should have learned last game that going nuts when people make expected comments doesn't help.

@both: I'm not trying to be an ass or anything, just keep the game from being derailed.

I think our best course of action right now would be to have the seer reveal (assuming (s)he is alive) and if that doesn't happen within the allotted amount of time have the bane reveal.

I recall reading ap say that there was a 1/3 chance that each of the blues were killed, but that's assuming that the wolves killed chromer, if they killed wilkin then there's a smaller chance that he was the bane, but I find it unlikely that he would have used his power on anyone else night one if he were the bane. (nice run-on sentence.) Also, if I misattributed any of this, or if you added what I just said as a disclaimer with your statement, I do apologize.

Not really much more to say than... everyone needs to comment on the pros and cons of the seer/bane coming out. (even though it's almost a no-brainer)

[DiscoBobbyPARANOiA]

Re: TWG XL: Mall Madness

Really, I don't see why my strategy I posted above is a bad idea?

It's not and I don't think talisman was saying that it was.

[talisman]

Re: TWG XL: Mall Madness

ap I have no real problem with either the seer or bane coming out, although I would greatly prefer the seer for the obvious benefits. All I'm saying is that it doesn't make sense for the seer to reveal who they saw (if they saw a human) since they're going to have to be in some form of control (ie limiting the voting pool) over the game anyway to stop people from voting for the other blue.

and I'll gladly take back my doubts about the advantages of the bane being public if all that math indeed works out and leaving the bane and guardian to their own devices is an insignificant advantage.

[aperson]

Re: TWG XL: Mall Madness

okay well at this point I agree that it's questionable whether the bane should come out if the seer doesn't. However, we are at least for certain that the seer should step up if they are alive.

therefucenfore:

IF YOU GOT A PM SAYING SOMETHING TO THE EFFECT OF 'SMILE YOU ARE THE SEER' YOU SHOULD PROBABLY SAY SO WITHIN THE NEXT 24 HOURS OR ELSE YOU ARE GOING TO RUIN THIS SHINDIG FOR EVERYONE

[Kagome]

Re: TWG XL: Mall Madness

Hey sup I'm the seer.