Behind. Responding as I go.
@skumpy and @Swan, if you're going to go through pairings and look at how likely they are to be the one true scum-pair left, you really have to be a lot more systematic and organised about it, or at least present your ramblings as something other than sensible elimination.
I mean, I do agree that having a set of likely third partners is nice when lynching a second scum. Ideally, we want to look at every possible trio that contains Teproc, assign a likelihood to each of his possible pairs of buddies, then marginalize to see which single player has the highest chance of being in a pairing, and go for the lynch on that one. But that only works if you're willing to consider all 21 pairs from your personal perspective, not just the 11 from Skumpy's post at #958. (He only posted e-gkrieg, then the 10 pairs not including himself, e or gkrieg).
Not sure what 'something other than sensible eliimination' entails.
I'm not going to actually finish this because
a) time. It's going to take a lot of work to present solid cases against all 10 pairs, which I'm not willing or have access to at this point
b) I got overzealous, I don't think I can get rid of all 10. I got eager because there were a bunch that seemed easy to get rid of (such as Awaclus/Andrew), but there are a bunch that it's going to be a stretch to get rid of.
My first (rather successful!) attempt was #1154 in M105. I took the 10 potential teams of 3 from a group of 5 that excluded myself, the IC, and 2 unknowns I wanted to isolate, and tried to show none made sense. What I was trying to do with that was avoid getting shutout my first time playing, and make a lynch pool of 2 that would guarantee 1 correct lynch on the next 2 days. But since I'm not going to be finishing it, I agree that the next best thing is to eliminate as many teams as possible and lynch the greatest common denominator. That is something I think I can manage at least some of. For instance, Awaclus/Andrew reminds me of Andrew's arguing semantics, and I don't think it's a bluffed argument. I don't know Awaclus' history with tunnel bussing, but I would imagine it's low. And as I alluded to earlier, I don't see why scum!Andrew stays off the Teproc wagon only to join another wagon against scum!e, and then flip to the other, so Andrew/2.7 seems unlikely as well.
@skumpy. What do you mean Teproc stayed far away from Andrew? Isn’t that what you do as partners? What do you expect partners to do in mafia?
Usually stay away, but I wouldn't call it a truth universally acknowledged. And I'd expect at least some interaction between 2 scum somewhere. But what I was saying was: if youre going to make a case against Andrew, I think you should include a case for the third partner that tries to finish the puzzle, which includes a D1 featuring 1 correct wagon that at least 1 scum did not bus. And then we'll see if it explains everything.
I see your point, but that’s not always how it goes. I don’t like to call scum teams too early because then people discount the team because of the third teammate. Also scum then knows who I think the third one is which gives them a lot of information. So I won’t name a third person.
I held back the one key reason why I wanted the third person: if Andrew is scum, then who could possibly be the partner, based on the fact that a lot of Andrew teams are unlikely? In my brief thinking, without much rereading admittedly, I felt the most likely partner for scum!Andrew....is you! (and maybe IDPTG, which sadly ruins the dramatic emphasis).
Now that I'm seeing Space vs. World, I'm teamSpace. It's D3, with a successful lynch behind us. That's more than enough information to knock out quite a few possibilities. It doesn't have to narrow all the way down to a correct pair, or even a pool of two as I once planned, but it gives probabilities to lynches, which is helpful. Will it err sometimes and a scumteam will figure out how to disguise themselves. Sure. But if it's accurate, say, 60% of the time...I'd take a 60% winrate any day of the week. I guess I'm a progressive when it comes to mafia strategy.
Does anyone want to even comment on what I am saying about talking about specific pairings???
I kind of did.. I said that I think marginalizing over all possible pairs containing a player is actually what would be good. Your example about disregarding a scum pair based on the identity of the third person kind of missed the point, because you were worrying about thinkinking player A might not be scum just because pair A-B wasn't good, but really we'd have to consider A likely to be scum in A-B, A-C, A-D etc too in order for A to be a good lynch. Enumerating the rest of those pairings is exactly what Skumpy and DatSwan were doing particularly badly, so I agree with you in that limited respect, but I don't agree that it's a bad approach in general.
What I am saying is that gives scum a lot of information over who will be lynched if someone flips scum, and it is wayyyy too early to be talking about entire scumteams at this point. It also gives scum an opportunity to pair people together to try to line up bad lynches. I think you want to save your time and just do targeted rereads and look at interactions, instead of trying to exhaustively look at every single pair of scumteams, which tells scum who they should kill.
...except there's a Vig with IC potential still alive. And maybe more (Doc/Neo).
2ish days left. Ok with voting 2.7, not ok with Andrew until someone makes a solid, all-inclusive case.
Vote: Awaclus I'm done giving this a pass. There's gain even he's town with regards to above-enumerated-interactons-list. Is that a terrible perspective on D3? Probably! But people talk and I'm gullible and I believe them. I wasn't feeling them before...now I am starting to sense some marionette strings. And I do not appreciate it. Gut still tells me a scum's under the radar.