TWG XXXV - Knick It In The Bud

[blahblah18]

Re: TWG XXXV - Knick It In The Bud

but my initial signup plan was to wolf cypher.. now its impossible? boooooo this game sucks

[Omeganitros]

Re: TWG XXXV - Knick It In The Bud

Woooohoohoohoooo!

Sup night.

[Tps222]

Re: TWG XXXV - Knick It In The Bud

Sup.

Tass's plan seems fine.

[aperson]

Re: TWG XXXV - Knick It In The Bud

To run some quick numbers:

Assuming that the wolves are brilliant enough to not submit the same person for a lottery guard as they did for a wolf, here are some odds:

Interesting situation #1) If the vigilante comes out and all humans guard the vigilante (this model assumes worst-case scenario where no wolves are lynched in the daytime):

Night 1: 13/16 = 0.8125
Night 2: 11/14 = 0.7857
Night 3: 9/12 = 0.75
Night 4: 7/10 = 0.7
Night 5: 5/8 = 0.625
Night 6: 3/6 = 0.5

Now, this is pretty interesting to look at. With full cooperation from the humans, even with the vigilante public the wolves have a very low chance of successfully wolfing them. In fact, I think this chance is so low that it would even be in the wolves best interest to not attempt to wolf the vigilante because it would likely result in 2 to 3 failures (and wasted kills) before their plan is successful. In fact, if the wolves tried to wolf the vigi for the first three days in a row, their chance of succeeding by the third day is 0.5212... Just barely over half!

In a sense, I think this makes the lottery guarding have the same impact as standard guarding, because the success rate of the wolves is so low that it will prevent them from trying it at all... And even here, my model implies the worst-case scenario where no wolves are found out. In actuality, the chances of a successful guarding jump up to around 80% with a single wolf removed.

Does anyone see any flaws to this logic?


Interesting situation #2: Pure randomosity.
Assuming no cooperation between the humans and wolves smart enough not to guard their selection, the odds of a successful guarding are as follows (this follows worst-case scenario model as well):
Night 1: 0.0507
Night 2: 0.0561 (0.1068 cumulative)
Night 3: 0.0625 (0.1693 cumulative)
Night 4: 0.07 (0.2393 cumulative)
Night 5: 0.078 (0.3173 cumulative)
Night 6: 0.083 (0.4003 cumulative)

So basically, with random chance we have a 40% chance to have a successful guarding by the time the day is over. (This model uses the equation: n = number of players, w = number of wolves: ((n-w)/n)*(1/n). The first part calculates the mean chance of them being selected for a guard that night, the second part calculates the chance of the first part actually being selected.)

[iggymatrixcounter]

Re: TWG XXXV - Knick It In The Bud

Well...I am the wolfsbane, so if you planned on voting me for guarding, no need to worry about that.

Come on... consecutive special roles? XD

Anyways since you can vote for yourself in the lottery, I was thinking that maybe we should post who we voted.

It may give us nothing but it would be like day voting patterns. Might give us something to work with in case someone is wolf or whatever.

Course the only bad thing I see from this is the bane (cypher) may know the vigi and maybe the wolves would try to kill whoever cypher chose.

So maybe everyone but the bane posts their pick or something to remedy that. Just an idea to get us more infromation.

[blahblah18]

Re: TWG XXXV - Knick It In The Bud

no AP, and i basically me ntioned the same thing to Tass already

[blahblah18]

Re: TWG XXXV - Knick It In The Bud

Iggy, lying about that would be the easiest thing of all time, and really gains nothing

[Tasselfoot]

Re: TWG XXXV - Knick It In The Bud

Iggy... no. the whole point about hiding our guarding picks is to make it hard for the wolves to pick who to wolf. By revealing that info, it is like Inf1... where all the humans were revealing where they were going, which made it uber easy for the wolves to kill them. Not exactly the same thing, but along the same lines.


And ap... the reason we don't have the vigi come out is because there is NO REASON for him to come out. We already have 1 public figure. Don't need 2. Cypher will make sure we don't lynch him. And above that, having the vigi public will force the human's hands on who to guard. Which leaves the wolves free to attack anyone else... which is bad for you, me, and blah.

Yes, those stats are probably right (no sense in double checking), but that isn't the only factors to consider. Plus, the wolves likely WOULDN'T still try and kill the public vigi, they'd take the freebie and kill someone else. 0% chance of failure.

BECAUSE THIS IS THE MAIN THING.... THE ONLY WAY WE CAN PROVE SOMEONE HUMAN IS THROUGH A GUARD. It isn't 100%, but it is pretty damn close. There is no seer or psychic... WE NEED GUARDS. And having a public figure that requires us to basically waste our guardings is a huge hinderence towards winning this game.

[aperson]

Re: TWG XXXV - Knick It In The Bud

Did you look at my odds for a successful guarding? Seriously, they're pretty low, and because of the nature of this game they're lower than they would be with a standard guardian.

Having the vigilante come out would be the equivalent of having one successful guard because it gives you one extra known human. There is only an ~40% chance of an equivalent state rising by the end of the game if you are to use a random guarding system as we would be without having a force-guard on the vigilante. And to make this more worthwhile than having the vigilante come out we'd need to have two people be guarded throughout the course of the game. What are the chances of that happening...? Somewhere around 13%.

Now I know this is bad for you, me, and blah... But wasn't it a long, long time ago that we stopped making it where you had to be alive at the end to be counted a victor? I'm fine taking a sacrifice for the team to develop the optimum strategy even if it means that once again I won't see day 1 for like... the third time in a row now.

You're a poker player and you know how important the actual odds are of a situation. Seriously, sit down and think this over very well to see whether it is +EV for us or not, because I certainly think it might be close either way.

[iggymatrixcounter]

Re: TWG XXXV - Knick It In The Bud

I was talking about revealing them after the night was over (if I wasn't clear on that.)

Plus no human would have a reason to lie anyways blah.

However, I do see why it could be bad because you would be giving the wolves SOME idea of where people are putting their guards. And thus act accordingly with kills. But that would only benefit higher players.

Also I was thinking, why not bank people in two guard picks? (Like say half vote talisman, half tass for example.) We would be able to cast doubt on who the wolves should go for and possibly save two people per night from being picked by the wolves.

Course doing this would mean that anyone not mentioned would be up for grabs but if there is no discussion about who to guard and everyone kind of does their own thing, chances are that there would be no guarding this game anyways.

[Tasselfoot]

Re: TWG XXXV - Knick It In The Bud

look at last game... all the smartest humans were killed off by day 3... which left the wolves able to just run over the remaining stupid humans.

we're needed alive for the team, much more than the others. not to mention, your odds are factoring an even and random probability of the guard. you know this isn't the case. clearly certain people are much more likely to be gaurd targets and probably the same people are likely to be wolf targets. and if the wolves choose to use reverse logic on that and kill off stupid people... well, i consider that a victory too. because smart humans will trump smart wolves every time.

but it really comes back to the point of the vigi won't be unknown. Cypher will know him. it will be just like any alliance where 1 person is working out in the open and another power player is hiding in the shadows. It has worked countless times in the past, and I see no reason to adjust that now, especially given the reasons above.

[talisman]

Re: TWG XXXV - Knick It In The Bud

having read through this I think I gotta go with tass on the issue of the vigi coming out... obviously the pro of knowing the extra human is nice (although double-edged), cypher should be able to keep the vigi hidden well enough, and if the vigi gets wolfed it's no real loss, as the vigi power is one that really never comes into play, and should never need to.

we could do a compromise though... like if the vigi is a "power" player then have them come out (as that increases the chances of that player not being wolfed) and if it's a "weaker" player not have them come out. and if the wolves try then to wolf weaker players because no one comes out, then more power to them, because like tass said smart humans left alive will generally win most of the time.

[iggymatrixcounter]

Re: TWG XXXV - Knick It In The Bud

we could do a compromise though... like if the vigi is a "power" player then have them come out (as that increases the chances of that player not being wolfed)

well the only problem with that talis, is that other power players would die because the vigi is a strong player, so more guarding votes would go to him and not other good players.

[talisman]

Re: TWG XXXV - Knick It In The Bud

edit scratch that compromise thing I just read through the player list and pretty much everyone is good. this game is stacked.

actual edit: this was supposed to be an edit I'm retarded...

[iggymatrixcounter]

Re: TWG XXXV - Knick It In The Bud

Booya, who can die when you've got a sweet av like I do?

Anyone who changes is just a copy of the real deal XD.