Pre-flip Associations

[Yukari]

We have been guilty of this. Just wondering why is this bad?

[wgeurts]

As a pre-flip associative is basically using evidence based on presumptions.
We already can't be 100% sure of someone's alignment outside of PR's so basing someone elses alignment on an all ready shaky foundation makes the strength of the evidence gained through a pre-flip associative a lot shakier.
It's fine as a theoretical thought which you could use to figure out how the game state will go, however it shouldn't ever be used as evidence unless the first player has been copped or something.

Being wrong with a pre-flip associative can also be a lot more destructive.

[sthar8]

Huge waste of time, also. You build a house of cards on sand, then you have to redo all the work when it collapses.

[implosion]

I've argued about this in games a lot. Another good way to look at it is a probabilistic argument.

Take a couple of common setups - 2:7 and 3:10 scum:town ratios. Imagine you have a read that a particular player is scum. Since you are town, in the first case there are 8 other players with 2 scum, and in the second case there are 12 other players with 3 scum - so in each case, your read has a 25% chance of being correct a priori (that is, before considering the evidence).

Now imagine that you see something that makes you think two players are scum together. In the first case (2:7), there are 8 other players and (8 choose 2) = 28 possible scumteams among them. So your pair of people that you suspect have an a priori 1/28 chance of being scum together. In the latter case (3:10), there are 12 other players and (12 choose 3) = 220 possible scumteams among them. The pair of people you suspect could be scum with any of the other 10 players, so there are 10 possible scumteams that would make you correct out of 220. So your a priori odds of being right are 1/22. In both cases it's less than 5%.

This basically means that if you want to suspect a particular pair of people being scum, even through perceived associative tells, you should have a much higher burden of evidence before you are convinced. It's the kind of thing that may be worth noting, either for the sake of remembering it if one of those people flips scum in the future or for future bragging rights if you're right - but IMO, it isn't worth basing decisions on.

Note that this is mitigated as the game goes on, as there are fewer and fewer pairs of people alive that could be scum together. In particular pre-flip association hunting can be very useful in, say, 5-person LYLO if you want to eliminate certain team possibilities.

[Yukari]

Hmm, ty for the responses. Will have to think on this a bit.

Just seemed that since their is a scum "team" that one thing to look for would be scum on scum interactions such as the buddy/bus. Thinking like a programmer, a buncha if/else type statements start automatically popping up in our head.

If(X == S)
Y = T
Else(Y = S)
//Or
If(X == S)
Y = S
Else(Y = T)

Based almost entirely on situation/gut. Most of this info rarely makes it into thread, just something that goes on in our head for some reason. We think, can x and y possibly be scum together? Then go from there.

Although, we can also tend to get struck by fits of wifom induced madness where there is no tell and no real way to determine alignment outside of PR. When that happens sometimes it seems like the only way out is to reactionfish and judge based on others perceptions since thankfully town outnumber scum.

Never played as scum, but seems like it would be easier than having to actually scumhunt.

@_@

[DeathRowKitty]

It's not hard to find associative tells between town-town or town-scum. Using associative tells pre-flip might or might not increase your odds of finding a second scum if your first is correct, but it doesn't make you any more likely to be right about your first scum, while probably conf-biasing you towards two players. It's safer to just find individual scum and to make a note of possible connections to look at later on.

[Ankamius]

I ignore pre-flip associative tells unless I have a reason to believe they're worth pursuing. I don't think they're inherently bad to use but it's like meta in that in your table of a case, it's a fairly brittle leg.

[SilverWolf]

I have found when myself or others use them too early in the game, without a scum flip, it's a huge waste of time. Sometimes I end up getting neighbor tells or mason tells or it's just town interacting with town or even town interacting with scum. I either need the flip to look for associative tells or it has to be late enough in the game where there aren't a lot of people left and I know how many scum are still alive.

[sthar8]

Hmm, ty for the responses. Will have to think on this a bit.

Just seemed that since their is a scum "team" that one thing to look for would be scum on scum interactions such as the buddy/bus. Thinking like a programmer, a buncha if/else type statements start automatically popping up in our head.

This is absolutely true, but scum interactions are mostly unusable until you KNOW that one of them is scum. You have to test the premise before you can rely upon the conclusion.

[kuribo]

It also poisons the well for future wagons

Like this: you feel both me and DGB are scum and push that association (I've seen it happen exactly like this)

DGB flips town and you've just made it harder to lynch me because you've given me an excuse to discredit you completely

And swerve: I'm scum but I'm looking like gold because of your shitty case

[kuribo]

There's exactly one interaction that I see frequently enough to give me pause when I see it. And even then it's not a smoking gun until one of them flips

(Hint: it's the aborted bus. When one buddy is the inevitable lynch and his buddy wants to be seen on the wagon, but doesn't want that to be the lynch, it can be easy to spot when you see it)

[Flubbernugget]

It seems to be okay to stay off a wagon because too many scum reads are on it

[Flubbernugget]

Even though this tends to stay early game sheeping someone for being a town read happens too

Either that or I'm misremembering

[Accountant]

It also poisons the well for future wagons

Like this: you feel both me and DGB are scum and push that association (I've seen it happen exactly like this)

DGB flips town and you've just made it harder to lynch me because you've given me an excuse to discredit you completely

And swerve: I'm scum but I'm looking like gold because of your shitty case

If I were you in that case, I would kill DGB that night specifically to trigger this. Plus he's not going to have docs on him because someone just made a compelling case that he's scum.

[kuribo]

kill DGB that night specifically to trigger this

DGB would be insulted if I didn't try to kill her N1

[Katsuki]

You're in a game with DGB and spyspy. Who do you try to kill first?

[kuribo]

You're in a game with DGB and spyspy. Who do you try to kill first?

DGB, she's more unpredictable and more likely to turn on me at any given point, even if it's for the wrong reason. I have to keep calling her town as long as she's alive because she knows I can read her--- I can never push her mislynch because she'll see it as a rock-solid scumtell.

SpySpy I can treat with guarded skepticism because he's been scum against me so many times, and anyway he typically takes a day or two to gather up steam. He's my N2 kill though, no repeats of Gay Mafia for me.

That would be my strategy:

Day 1, call DGB town, fake skepticism that SpySpy is town but warm up to the idea as the day wears on
Night 1, Kill DGB
Day 2, Buddy SpySpy even if I have to bus my own buddy.
Night 2, Kill SpySpy
Day 3, Fake a guilty on mastin lol