Pick Your Power II - Looks like the wine is gone (SCUM WIN)

[Jack]

Official Vote Count

Players needed to lynch: 12

Bouncy.Bouncy

- 3 - RedCoyote, Devotress, Faraday - (L-9)

Cobalt

- 1 - The1fifi - (L-11)

Dramonic

- 1 - TonyMontana - (L-11)

Fate

- 3 - Bouncy.Bouncy, Dramonic, FeFiFoFum - (L-9)

FeFiFoFum

- 2 - Rayfrost, wolframnhart - (L-10)

Hoopla

- 1 - Jack - (L-11)

Pomegranate

- 3 - Socrates, Farside, Fate - (L-9)

Socrates

- 2 - Hoopla, StrangerCoug - (L-10)

The1fifi

- 4 - Ellibereth, DocPotter, Cobalt, Porkens - (L-8)

Players not voting: curiouskarmadog, Pomegranate

Jacks Lurking.
He's posting in his other 2 games on a regular basis.

Replaced into a 25 page game in one and did an ISO of everyone playing in another.

I don't agree with lynching newb town lurkers.

Hoopla is being more than stubborn.

[Jack]

Questions for reference:

1) How many people here thought about what strategy the mafia would use when picking numbers before the game?
2) How many think my math in ISO 16 is wrong?

If hardly any people do then it points very strongly to hoopla being scum. And when an experienced player is scummy you always lynch them over a scummy sounding newb player who is lurking. Newb town often lurk.

[Fate]

1) I honestly didn't think about it. I'm not really the most in-depth strategizer myself, so that could be why.
2) I didn't work either of your math's out in my head, though I'm sure they both are mathematically and logically sound, they are both

less than relevant

in my eyes.

[Fate]

slippery slop of wifom.

Are you saying my question is WIFOM? My post wasn't just a rhetorical "why would mafia do that?"; it was attached to the idea that

I believed

"doing that" would have absolutely

zero

strategic value for the mafia. As far as I know, WIFOM is only in play when both sides of the argument have some strategic value for the mafia. I genuinely saw no reason for why mafia would want to pick different numbers.

HOWEVER, I didn't think the scenarios through before I posted that. My theory was wrong and Hoopla's was right. So I'm not going to make that argument anymore.

ORL what was it that fate said that no townie would say?

"Hmmm... Any scum want to claim and tell us?"

I still think you're reason for voting me is complete and utter ***. Especially since you didn't even list this reason when you first voted me. At least have some balls and vote me just for the wagon, like Porkens or Cobalt.

I don't think Pom will claim anything but vanilla anyway, but I still don't want to lynch someone that isn't here. Bouncy is on a few people's "lists" as lynch-able too. (Hey Hoopla! Look he's #3 on the list!)

Unvote: Pom

Vote: Bouncy.bouncy

[Socrates]

V/LA until Monday.

[Devotress]

Questions for reference:

1) How many people here thought about what strategy the mafia would use when picking numbers before the game?
2) How many think my math in ISO 16 is wrong?

If hardly any people do then it points very strongly to hoopla being scum.

Hoopla is just being stubborn because he thought he stumbled across a brilliant strategy, and is ignoring your actual math post because he doesn't want to admit he was wrong. I don't think that makes him scummy. I think this whole numbers discussion, if he flips scum, is when we should analyze what he was trying to achieve with his numbers.
I'm feeling stubborn townie right now though.

[wolframnhart]

1)I did not think about strategy the mafia would use, heck i didn't even know there was strat to it i just sent in numbers.
2)I don't think anything of the numbers math thing at this time.

[DocPotter]

You're trying to show the chance that scum picked a singular number based on the results. You should look at the chance of scum picking singular numbers based on what they had to choose from.

To analyse the results we need to purely look at the results.

We know there are 5 scum. They could have picked either:
1 'x' number. Didn't happen, as the largest group has 4 people in it.
2 'x' numbers. 4:1 unlikely as bad tactics, but 3:2 is just possible but still unlikely.
3 'x' numbers 3:1:1 or 2:2:1
4 'x' numbers.
5 'x' numbers.

We kow that there were 9 'x' numbers picked, so taking the best of the likely options for town which is 3.

We get 3 of the 9 groups have at least 1 scum in it. Or 33% chance of any group containing scum.
4 'x' numbers runs to 44% with probably 1 scum per group.

5 'x' numbers gives 55% chance for 1 scum in each group.

If we assume the best scum option, then lynching the singles gives us the best chance of lynching scum but a large chance of wiping out a good town PR. If we choose the worst likely scum option, then hitting the groups of three becomes a better option. (Chance of multiple scum for a better hit chance, lower chance of PR's)


However, I still believe that we need to look for scummy players first, and once we get a flip or two, or some definate scum/town feels, then we should start looking at the numbers.

Purely going by numbers, NUFFL, or Murphy, will bite us in the arse.

ATM, I'm flippin between Jack and Hoopla appearing slightly scummy for pushing optomised math and not looking at the their assumptions properly. Could be scum trying to push theories affected by known scum choices.

[Jack]

ATM, I'm flippin between Jack and Hoopla appearing slightly scummy for pushing optomised math and not looking at the their assumptions properly. Could be scum trying to push theories affected by known scum choices.

Well, that's what I'm accusing hoopla of (gunning for the top drafters with bad math).

My assumptions are fine (maybe calculating it out of 9 instead of 15 is not accurate, but it's really just a nod towards the fact that we weren't picking randomly). You can't analyze the results by looking purely at the results, that's not how probability works:

1, 1, 1, 2, 1

That's the result from random.org I just got. So is the probability of getting "1" 80%? "No, it was 50% and that was just variance" you say. Wrong, it was 33%, I told it to pick 5 numbers between 1 and 3.

You're trying to show the chance that scum picked a singular number based on the results. You should look at the chance of scum picking singular numbers based on what they had to choose from.

I'm confused because the first part is what you and hoopla are doing, and the second part is what I did.

Here's the problem with your analysis:


Let's say that by chance 18 people picked one number, and 4 people picked single numbers. Now let's assume that the scum didn't double up on any of the numbers. That would mean that the 4 singular numbers have to be mafia. This is the essence of your guys analysis.

Problem is, the fact that the numbers turned out like that in that scenario makes it extremely probably that the mafia double/triple/quadrupled up. Your assumption that they didn't would mislead you terribly in this scenario.

((1/15)^14)/((1/15)^17) = 3375

14 townies picking the same number is 3375 times more likely than 17 townies picking the same number.

****

I can actually see now how you could reasonably get confused though.

unvote

I can buy hoop coming into the game thinking about strategy as a townie, that's not damning if it doesn't come attached to bad number analysis.

[Jack]

Pom is still a bit scummy, but less so without the hoopla connection. Not willing to vote there, much better to wait until we get a clearer picture.

I don't see any scummy fifi posts.

FeFi should be lynched. He pretends to be new in his first post, when he gets under pressure he pulls out the "I had an account on here before, bandwagons on me are pointless".

These seem to be the popular choices so far.

I would add:

dramonic--every post is hiding behind the number analysis.

wolfram--have a gut feel that his posting is fake. I feel that I demonstrated this earlier. He kept pushing me on the "answer the question" thing and later showed that he had no real reason for it. When I went after him he acted like he found me really suspicious and "really liked his vote on me". That all disappeared and he hasn't mentioned me since.

And again, this is just a scummy answer:

1)I did not think about strategy the mafia would use, heck i didn't even know there was strat to it i just sent in numbers.
2)I don't think anything of the numbers math thing at this time.

1) is worded to persuade us he's town, 2) is avoiding commenting on the issue but leaving it open with a press secretary "no comment at this time" deal.

vote:wolfram

[Cobalt]

itt more fif pressure is needed

[Jack]

itt more fif pressure is needed

You're being anti-town.

[Cobalt]

cool story
now vote fifi

[The1fifi]

cool story
now vote fifi

cool post
now vote cobalt please

[The1fifi]

V/LA

for 2 days

[DocPotter]

Just quickly Jack, because the number stuff is taking too long.

Hoopla's, and my, little spiel is looking at the distribution of scum through a known set of 'X' numbers and taking into account probably distrabution paterns. (ie: number of 'X' selected by scum over the 9 'X' numbers.

Hoopla only really talks about the set-up that most favours what she wants to do.

You however Jack, and more actively mis-representing things.

Now, to get to the math you are misrepresenting specifically.

1, 1, 1, 2, 1

That's the result from random.org I just got. So is the probability of getting "1" 80%?

With five results, and four of them known to be a '1', the probability of any of those results being a '1' is 80%, not 33%. If we didn't know the results then the chance of any result being a '1' is 33%.
(Actually with the draft we can deduce the results because we know there are five scum in the game, we know how many result groups there are, and there are limited distrabution/spreads available to the scum)

This is exactly what myself and Hoopla are getting at.
With nine result groups, and either three, four, or five* of them containing scum the chance of any group containing at least one scum is 33%, 44%, or 55% respectively.

Let's say that by chance 18 people picked one number, and 4 people picked single numbers. Now let's assume that the scum didn't double up on any of the numbers. That would mean that the 4 singular numbers have to be mafia. This is the essence of your guys analysis.

The essence of my math discussion is a chance that gets expressed as aprox 6e^-20% chance. Wow, I mean wow.

*There is a chance of 2 groups, which would be 22% but that kid of spread seems too unlikely to bother with.

ATM, I'm flippin between Jack and Hoopla appearing slightly scummy for pushing optomised math and not looking at the their assumptions properly.

I'll let Elli worry about his 'catch' himself now. (Yes I knew out what he was doing, and why)

Unvote

Jack, you seem to be attempting to do anythiing to discredit analysis of the draft result, analysis that not many people seem to be following at the moment. I might argue about when it should be applied, or the specifics of the assumptions, but not about the theory itself. I can't think of a town reason for anyone to argue against the analysis itself.

Vote Jack

Ok, it wasn't so quick

[SpyreX]

A Friendly Mod Request:

When you unvote, please say who you are unvoting (it helps me with my VC's yea I'm lazy wanna fight about it)

[Ellibereth]

aaaaaaaaaaargz.
I'm feeling The1 and bouncy as both town now...

Unvote The1

(Spyre...)
Need to rethink this.

[Ellibereth]

I'll think this out tomorrow. Don't feel like doing anything right now.

[Jack]

With nine result groups, and either three, four, or five* of them containing scum the chance of any group containing at least one scum is 33%, 44%, or 55% respectively.

Why would each group have the same chance? Scum have 17 people who could pick their number, town have 21.

Let's do an example. 2 mafia, 98 townies. After picks we have 5 groups.

1) 1 person
2) 1 person
3) 8 people
4) 10 people
5) 80 people

So with 5 result groups, and 2 of them containing scum, is the chance of group 5 containing scum 20%? And the chance of group 1 containing scum 20%?

I argued against the number analysis because it was being used to wagon fate unreasonably. I bring it up again now to gage how likely it is that hoop believed her numbers. And if you remember my initial post I didn't say it was worthless, just that it was a much smaller increase than hoopla was suggesting.

[DocPotter]

Hoopla is right with her math though*. Your math is not. And where it is, you're trying to push it as unreasonable.

Why use hypotheticals? We have the exact situation we need. Same percentage chance for both btw. 32% off of the top of my head.





*But she forgets to take Power Roles into account. Or doesn't care about them. There's also some simplification/rounding involved. But do people really want to see it all?

[Jack]

It's a direct comparison. If you really don't believe me you could play with a random number generator in excel for a while, I don't know how to prove it to you.

[Devotress]

Unvote bouncy

Vote the1fifi

Btw I was scum in the last game when I made a post where I pretended I was cobalt as I voted him.

[DocPotter]

I'm glad it's a direct comparison Jack. Now what are you trying to get at. WHat actually are you trying to say.

Plain english please.

[Fate]

Too many V/LAs+Wagons falling apart+More Mathtalk=

More and more frustrated by this game.

Ellie what changed your minds?

Is it the fact that wagons formed on them, therefore scum are on those wagons? If that's your theory no one should ever be lynched...

Glad to see Jack sees my point on dramonic.