Mini 436 - Game over - Mafia wins with no casualties!

[ryan]

Vollkan: What part of my statement isn't correct? If you aren't helping the town you are hurting them right? Miz had a plan (that he thought would help) and look what it did, caused a ton of suspicion on somebody even I thought was pro town earlier and also has been the main topic for what, 10-15 pages? Instead of getting some of our lurkers talking (ie: TrustGossip) we went round and round right up to deadline, had 3 people not even vote or help us find the scum and than we lost a townie. I place some of that blame on a "plan" that didn't work (not all but some) Why wouldn't Miztef be sitting right with me? He tried a little trickery, knew he would get caught by the mod or the town and than tried to drop a "whoops I'm sorry, bad mistake, but I'm still town" bunch of comments. He did pull an anti town move (IMO) and that's why I placed a confident vote on him. It's not like I voted Deathsauce and than two pages later unvoted and voted Miztef. And actually you and Miztef had votes on Deathsauce (who ended up town) why is it so unlikely that you two aren't working together?

Votecount:
Not voting : everybody

With 7 alive it takes 4 votes to lynch.

[Paradoxombie]

I am not saying that Ryan cannot be scum. What I am saying is, that we can't consider the Night Kill as the evidence to find scum

Well I am willing to vote ryan based on VDs case, anyway. If anything, his proven innocence shows that his arguments were genuine.

Vote: Ryan

[vollkan]

Now it's just a matter of waiting until everyone has checked in. Given the recent lull and the complete absence of Trust (absence of trust in a game of mafia), that may take a while.

I think maybe the vote is a little premature; I was planning on voting Ryan too, but after we had heard from everyone and discussed a bit, anyway the evidence is certainly all there even if we didn't get to hear all of what VD had to say.

[Miztef]

Voting is severely risky to me right now. We have (most likely) 3 scum, that means that 1 vote on a town by a town could mean all the scum jump on and win the game (this is likely LY/LO).

That jump vote gives me some incentive to believe you are scum paradox, it is way too premature. I don't have nearly enough of a read on SirTornado, Sweenytoad, and trustgossip to dare send a vote out in this situation, I don't see how anyone is willing this early in the day. (sorry for any name typos in advance)

[Paradoxombie]

That's an interesting conclusion, Miztef.

[ryan]

I'm not sure how anyone has a read on TrustGossip.

[vollkan]

Voting is severely risky to me right now. We have (most likely) 3 scum, that means that 1 vote on a town by a town could mean all the scum jump on and win the game (this is likely LY/LO).

On that note, I want to get the numbers flowing again; I figure that could be a good way of helping to assess the situation. I had errors in my numbers before, please correct me if I have any problems here.

I have bolded each possible outcome of D3 and N3 combined.

Assuming it is 4:3:

Townie Lynched D3 = 3:3

Scum NK of N3 =

2:3 (Loss. This is the outcome if we really screw up and lynch the vig)

If we have a vig and the vig survives the D3 lynch, then the vig will also get an NK. If the vig chooses not to NK, we lose since it is 2:3 (am I correct here?)

If the Vig NKs a town =

1:3

(Loss again)
If the Vig NKs a scum =

2:2

(I will get to this in a minute, since it is the focus of my question)

Scum Lynched D3 = 4:2

Scum NK of N3 = 3:2
Since vig will obviously have survived to get a NK in this case,
Vig NK of Townie =

2:2

(This is also the focus of my question)
Vig NK of Scum =

3:1

(This is, obviously, the best outcome we could get)

Now, the two ones which are confusing me are where it is 2:2. Here, I am ignoring any information we actually learn and am focusing on just the numbers.

In #625, DS told me that in an even split scum wins in "most games". My question is, does that apply here if the vig survives for the reasons I shall give now?

Obviously, the scum won't vote each other which would mean that D4 could only end in either a No Lynch or a Townie being lynched.

If No lynch D4

= 2:2
Scum NK N4 = 1:2
Vig NK of scum N4 = 1:1

This is, of course, providing that the vig does not also choose the scum's target. If the vig made that error, we would lose since D5 would open at 1:2.

If the vig were to target the other townie and the scum were to target the vig, it would result in 0:2.

Anyway the point of that is that if it is 1:1 there can be no lynch on D5.
Hence, on N5 wouldn't the NKs cancel each other if the vig was not NKed by the scum and cause...a tie??

If town lynched D4

= 1:2
Scum NK N4 = 0:2
assuming vig is not the lynchee,
Vig NK N4 = 0:1 (We lose)

Hence, today is basically LYLO/D (lynch or lose/draw). If a town is lynched today, the best we can hope for is a tie (that is to say, 1:1 with a scum and vig and assuming such an outcome is actually a tie which I may be wrong about).

This next bit is

very important

! Providing the lynchee is not the vig, scum will not hammer today. If they did, the vig would NK and it would end up 2:2. Since a hammer would most likely give away the scum's identities, it seems most definite that D4 would be a No Lynch and hence, D5 will open 1:1. If the town is vig, then it seems like a tie (again, I think), if the town is a townie, then it will be a loss.

If the scum were to hammer today, it would dash all hopes of us winning. However, it would also give the scum a 50:50 chance of a draw based on whether they are able to NK the vig. Their chances of winning, hence, are enhanced by not hammering today but lynching a townie as normal because that would make the vig more likely to misdirect his NK and, thus, decreasing the likelihood of a 2:2 arising where it pretty much comes down to a 50:50.

That's all for now and should give a bit of fuel for discussion.

[Sir Tornado]

If the scum were to hammer today, it would dash all hopes of us winning. However, it would also give the scum a 50:50 chance of a draw based on whether they are able to NK the vig. Their chances of winning, hence, are enhanced by not hammering today but lynching a townie as normal because that would make the vig more likely to misdirect his NK and, thus, decreasing the likelihood of a 2:2 arising where it pretty much comes down to a 50:50.

Actually, it is more than a 50-50 chance of winning for the scum. If we lynch a townie today, we have 3-3 split. At this point, the scum will have 2 Nights to NK the Vig. If they choose randomly, they get 33% chance of hitting the scum first Night, and then if they fail, 50% the second night, which gives them roughly a 66% chance of winning the game without the Vig hitting his target incorrectly.

The Vig on the other hand has to hit the scum on 2 consecutive nights. That is he has something like 16% chance (if done randomly). So, 66% of 16% of chance has the scum winning even if Vig gets his targets right (comes to somewhere around 10.33%) and 84% chance of scum winning by the Vig messing up a NK and hits a townie. It boils down to 94% over all chance of a scum victory if we mis lynch. That is almost a certainty.

PS: It has been a couple of years since I studied probability... might have got some numbers wrong. Please correct me if you spot any mistake there.

[vollkan]

Actually, it is more than a 50-50 chance of winning for the scum. If we lynch a townie today, we have 3-3 split. At this point, the scum will have 2 Nights to NK the Vig. If they choose randomly, they get 33% chance of hitting the scum first Night, and then if they fail, 50% the second night, which gives them roughly a 66% chance of winning the game without the Vig hitting his target incorrectly.

Yes, true. Let me clarify where the 50-50 comes into it. The 50-50 arises if the scum were to hammer in rapid succession or something, giving the vig a perfect idea as to the scum's identities.

See, if the scum hammer it goes to 3:3.
The vig should have a good idea who the scum are bringing it to 3:2.
Now, the scum's NK brings that to 2:2. Here, there is a 1/3 chance of the vig being NKed.

If the vig survives,
D4 is 2:2
No Lynch
Vig NK = 2:1
Scum N4 NK = 1:1

This is where the 50-50 comes in. If the scum N4 NK is the vig, we lose. If the scum N4 NK is not the non-vig, it is a tie. Hence, there is a 50-50 choice for the scum at that particular point. I did not for a second mean the probability of a scum win is 50%.

Hence, if we mislynch then it is very dangerous because the vig is at a disadvantaged position and, hence, the 94% of Sir Tornado raises its ugly head. However, if the scum were to hammer they are virtually staking their win on being able to pick the vig on N4 since the vig would most likely be able to NK the scum on N3 and N4 consecutively were the scum to hammer.

[ryan]

Anyone have anything new they'd like to bring to the game? (awfully quiet so consider this a post to get discussion moving again)

[Paradoxombie]

I think Prods are in order

[ryan]

SweenyTodd and Trustgossip could use a nudging

[Lawrencelot]

I will be on vacation this friday. I think starting D4 before I'm gone would be a bit too tight, so just take your time for D3. I will be back before August I hope. I'll prod these 2 if they didn't post here before I'm gone.

[vollkan]

It's been over a week since anybody has posted. If the mod is going to prod the two lurkers then we at least need to have some discussion going on if this is going to progress forward at all.

To continue what was being said re: Para's vote being premature. I ask Miz, do you stand by your suspicion of Para in light of the statistics I have given? (This could lead nowhere but it is better than nothing given the level of activity here)

[ryan]

SweenyTodd and Trustgossip are two that I have ZERO read on. I have a few suspicions on Vollkan and Miz, but without some more content from the two players I mentioned, I'm not confident in a vote at this time.

[Lawrencelot]

Alright, I'm back. I'm not sure how much time I have now, but if someone could tell me who needs to be prodded most I would appreciate it. Then hopefully the game will continue at normal speed. I will also set a deadline.

Deadline: 10 days from now (August 7).

[ryan]

SweenyTodd and Trustgossip need prods, Miztef might need one as well

[vollkan]

Prod Sweeney and Trust and then we'll see who else needs to be prodded once discussion picks up again.

[Lawrencelot]

Prodded SweenyTodd and TrustGossip.

[ryan]

Prodded SweenyTodd and TrustGossip.

What is the latest on the prod?

[Lawrencelot]

Alright, a deadline isn't really fair when 2 people didn't pick up their prods. There will be no deadline now, but once SweenyTodd and TrustGossip picked up their prods I might set one again, unless there is enough discussion.

Btw, I'll put this in the rules now: at a deadline, the one with the most votes is lynched, and in case of a tie I'll check who got the amount of votes first.

[ryan]

Any update on Sweenytodd or TrustGossip? I'd like to get some discussion going but it seems like most of the game isn't checking in anymore. Sir Tornado? Miztef? Vollkan? Para? You guys still checking in?

[Lawrencelot]

Still no word from Sweenytodd and TrustGossip. I'll try to find a replacement for them.

[Sir Tornado]

I am reading this thread, yes.

[Lawrencelot]

=Confused= will replace Sweenytodd. Thank you.