Quote Originally Posted by Totally not Taffy View Post
It doesn't matter whether he's faking the reads. His initial read on Ender was premature and almost everything he has brought up is NAI for Ender. But it's early game. Ime Asks juggles people in and out of POE/towncore continuously, just to see how they (and the other players) respond. Rinse, repeat until all the pieces fit.

It's a process very similar to coldreading; you start with something that's just statistically likely and the reactions you get from your audience lead you to the correct answer.

I am not sussing him at this point, but even if I were I think this is a terrible idea, it doesn't matter what order we yeet the wolves in, so on the off chance you're wrong and he's town you should try to find the others first.
You've never played with me, so that's not what I am doing but that's probably what it looks like from the outside.

The reads are in flux but compare it to a chess analysis engine which is very, very old and very very slow, so it's behaving like a human grandmaster instead of rapidly coming to the maximum correct move.

It sees a position it thinks is strong, suggests moves that lead to it... and then when it sees refutations by the opponent, decides that position is no good, tries a different position to see if that one is stronger, and continues doing that at the speed of a snail trying to make its way through molasses, until it decides this position also sucks, but on the third or fourth attempt at finding a good move, it sees a strong position that more likely finds wolves.

However, it still doesn't see the mate. It usually doesn't see the mate until near LYLO.

Chessmaster 3000 era chess engine, on the oldest possible processor it can run on, and that's me analyzing games.

Takes me literal days or weeks to decide the endgame wolf is a wolf, and to find certain townies as town also takes days sometimes.

When I post my reads and it looks wacko because the towncore is moving down and the scum suspects are moving up in a manner no one else understands because no one follows a dozen solving processes one after the other on every player slowly until the game is solved.

The analysis engine gets more and more accurate (usually) the more it cogitates but the first few attempts at analysis almost always suck and need to be checked carefully for any part of it that was actually true, before moving on to a better move and solve attempt.

Don't want to say, accuse 4 names, hit 2 correctly, then abandon those 2 names because you decided to townread the other two names.

Any accurate reads at all during even the shitty party of analysis, if valid, should be carried over to the next solve attempt, a piece at a time.

Sometimes it takes three, four, or five solve attempts to get to one where, after exhaustive analysis, it can be determined that someone who was townread in the past four analyses actually shouldn't be towncore because I've found too many other townies elsewhere I believe in more, and I found wolves that are likely to be paired with them, and I can see the wolf plan more clearly.

Thats why the names bounce all around yet it looks like I was confident the whole time and then increasingly so, while throwing out tons of reads.

I'm confident even on the first pass I'll find a good solid four townies or so, easy.

Making a true masonry of four is already a massive upset to the game balance. Even if the rest of my reads suck, if those townreads are good, it doesn't matter that the first pass, first analysis, first suspicion scattershot, first attempt at building a wolfcore, was terribad.

Doesn't matter, because if those townreads carry over to the second scattershot and analysis and they were right, then it makes the second attempt more accurate.

That's why you see lots of names bounce around at first, then settle into positions, then more settle, then very few bounce around, then I'm moving one name around the list at a time, not more than one.

One person I suspect, who then gets killed or townread and put at the top of the pile, pushing everyone down.

That's the process.

The more I analyze, the fewer names are in motion or the reads dependent on how other people flip, and fewer universes remain, and fewer possible scum teams.

That's why it's like the lotto balls. The thing is just a jumble of nonsense at first, then a number comes out and stays put. One by one, numbers come out and stay put, until you get the daily lotto numbers.

It's an analogy but I'm not making the reads for the reactions, the reactions are making the reads, and the questioning and pushing and interactions create the reactions.