Civilization Mafia - Town Wins!

[Puzzle]

Ok, I got the error. I assumed for some weird, totally stupid reason that investigations only happen if the previous one fails. Obviously that is wrong. EmpTy has the correct numbers. I think the best results could be achieved if we would agree on who is checking who, then it could not happen that a scum gets no investigation at all, increasing our chances. So, who is gonna pick that?

I fear that may be optimistic : there are chances for scums to be asked to check other scums and it seems hard to avoid.
For the other, I'd say either the subscription order, or alphabetical order, or anything that doesn't really involve personal choices.

[VisMaior]

ugh, so you mean like this

alphabetical order investigates 1. post order
Astronaut -> inHimshallibe
armlx -> VisMaior
Bamboomancer -> Astronaut
corporateclaw -> dybeck
d_rouge -> d_rouge
dybeck -> Puzzle
EmpTyger -> rolandofthewhite
HezLucky -> swinkee
Iammars -> HezLucky
inHimshallibe -> rajrhcpfreak
Mastermind of Sin -> Bamboomancer
Mr. Flay -> armlx
mlaker -> Mastermind of Sin
Puzzle -> Mr. Flay
rajrhcpfreak -> EmpTyger
rolandofthewhite -> corporateclaw
swinkee -> mlaker
Thok -> Iammars
VisMaior -> Thok



?

[dybeck]

That doesn't really make much sense - especially in the d_rouge area of things...

Surely alphabetical order OR post order would be smarter? Where everyone checks the person following them?

Or even just pick the people we think are scummy - it makes more sense to have multiple investigations on scummy people - since there's a high probability that an individual investigation will fail.

[Puzzle]

i have the impression that quite a few haven't posted yet.

Prods ?

[VisMaior]

Or even just pick the people we think are scummy

Any suggestions on these? I had the impression earlyer that EmpTyger might be one, but he pretty much cleared himself by now.

[Mr. Flay]

If we go alphabetical-investigates-alphabetical, it might be difficult to tell the difference between a scum whose investigation 'fails' and a player who actually doesn't get any information. And multiple investigations on a single target will make it obvious very quickly if we've got sanity issues to contend with.

But I'm with Puzzle, we've got waaaay too many people lurking quiet, probably as a result of the number crunch going on. And I *like* numbers...

[corporateclaw]

I've read the thread over once, I'm probably going to try and reread it -- i'm getting a little lost in the numbers -- but I have one fear. I would think that the barbs have some sort of unit-building capacity, and if they can build units that can vig-kill, then going no lynch for a long time seems very dangerous.

Basically, since we will need to take several days of building investigators and investigating, and THEN swith over to vig-kill units I feel like that is time that the barbs can be upgrading or building more vigging units. At some point, since I assume they know who all the barbs are and don't need to build investigative units, I think that could turn out really bad.

If somebody can address this issue, I would feel safer going no lynch.

[Bamboomancer]

My units aren't so great at investigation and I sorta like my unit so I dunno about all that. I'd rather wait till night three to start investigation.

[Puzzle]

Message from Armlx : since the move, he's having massive problems logging in.

His posting may be sporadic.

[armlx]

Ok. Thanks for posting that Puzzle. I got on this time, but I don't know about later. I had a good post on page 3, but the summary was

- Inhim stops complaining about game breaking
- We lynch people since no-lynch gives them as much tempo.
- We go with the plan.
- I

Vote Inhim

for not wanting to game break.

[Iammars]

I'm here, checking math. But before that post I would like to say a few things.

-Some of us may have extra units, any may only be able to investigate with a 1 tonight.
-Never try to utguess the mod about the scum's civilizations. No one know what SO did, so there's no point in guessing.

[EmpTyger]

The only reason I see to discourage lurking today is to ensure that everyone is on the same page about what we are doing. And I think the simplest way to do that is to wait for those who haven’t been participating as much to cast no-lynch votes.

But let me repeat that there is *no* point in discussing suspicions today. Write them down or type them up and save them for another day. The only thing to be discussed today is the optimal plan for tonight. (if not the next few nights)

As for that, I see 4 possibilities:
1) Make assignments of Player A -> Player B (as in [76]). I’m nervous about the idea of specific assignments, because I’m instinctively against creating the opportunity for the mafia to interfere.
2) Make assignments of Civ A -> Player A. But that still has the problem of discerning the setup of civilizations.
3) Divide the set of possible civilizations, i.e. so that some focus on protection while the rest investigate, but leaving the assignment to the player. This has some of the disadvantages of the unknown civilization setup, and also some of the disadvantages of leaving things unorganized.
4) No specific nighttime instructions, letting everyone determine their own nightaction as they will. This makes it hardest for the mafia to interfere, and also allows what happened Night 1 to be best taken advantage of. But this leaves the town so disorganized that it cannot take best advantage of its abilities.

I’m still not confident enough to recommend a specific course of nightactions as optimal (or dissuade one as suboptimal). And there might be other possibilities I am overlooking- permutations of doing one thing tonight and something else another night. I am particularly wondering about delaying mass investigations until Night 3 or even 4, to give everyone a chance to build an investigative unit. (Of course, this might leave the town vulnerable to mafia blocking.) Perhaps something like “by night X, Y should investigate Z”?

Anyone else have any ideas about this, analyzed or not?

[VisMaior]

I think the general consensu is that we lynch today too. Are you against this, Emp?

I also see these 4 possibilities. I personally think the 3. one is best, as it combines the advantages of the others without giving too much info for the mafia.
Im pretty much lost with the extended versions, I have to think about them some more. The mass investigations of corse would need everyone to have an investigative unit, so if there are people not having them right now, then we cannot really do that, so delaying till night 3 sounds like a plan.

[Puzzle]

1) Interference by the Mafia can take 2 forms, as far as I can see :
- kill someone who checks one of them : suspicion falls on who was investigated.
- kill someone who doesn't check one of them : higher risk of getting caught.
A lose / lose situation for them, imo.
And scums don't know anyway who built cops and not, so as long as we keep them in the blur, they may target people who didn't and that plays in our court.

2) Gives away who is who. Gets people to check themselves. I'm against.

3) Loop of investigation incomplete, potentially leaving scums through the net. Also requires to tell the scums who the docs and cops are. The advantage is that we can start this plan earlier than others in case not everyone chose to build cop on night 1. I'd prefer 1 or 4 though.

4) A bit the same as 3 but with the bonus of giving away less info to scums.


So for me, it's rather 1 or 4, with a preference on 1. The mass investigation scheme has no reason to be delayed, as long as scums don't know who built cops on night 1 and not. Those who didn't can just stay silent, like all who don't find scum.
Whoever finds scum should obviously come forward tomorrow.

And I think we should change the order of who checks who every day.

[d_rouge]

I agree with Puzzle here that number 1 is probably the best option, since there's not much scum can do about that without casting suspicion upon themselves.
The only problem might be if two scum wind up in a row, so that one is supposed to check the other, but I guess that changing the order everyday is a good way to solve that. It may waste us a few investigations, however, because if A succed in investigating B and finds him innocent, of course tells nothing about that the day after, and whoever checks B the night after is wasting his investigation.
I don't see a solution to this...

[Mr. Flay]

1) Interference by the Mafia can take 2 forms, as far as I can see :
- kill someone who checks one of them : suspicion falls on who was investigated.
- kill someone who doesn't check one of them : higher risk of getting caught.
A lose / lose situation for them, imo.
And scums don't know anyway who built cops and not, so as long as we keep them in the blur, they may target people who didn't and that plays in our court.

You're forgetting:
- kill someone who is targeting someone suspicious : suspicion falls on the target. I have a feeling people did not tend toward building cops N1, so there's a good chance scum will slip through tonight anyway.

I wonder if a variant of 3 is to capitalize on the 'tendencies' of the civs? Militaristic build Vig units, Pacifistic build docs, Expansionist builds cops, etc...? Then we don't have to know who is who right away.

Whoever finds scum should obviously come forward tomorrow.

And I think we should change the order of who checks who every day.

Agreed on both.

[Puzzle]

1) Interference by the Mafia can take 2 forms, as far as I can see :
- kill someone who checks one of them : suspicion falls on who was investigated.
- kill someone who doesn't check one of them : higher risk of getting caught.
A lose / lose situation for them, imo.
And scums don't know anyway who built cops and not, so as long as we keep them in the blur, they may target people who didn't and that plays in our court.

You're forgetting:
- kill someone who is targeting someone suspicious : suspicion falls on the target. I have a feeling people did not tend toward building cops N1, so there's a good chance scum will slip through tonight anyway.

I wonder if a variant of 3 is to capitalize on the 'tendencies' of the civs? Militaristic build Vig units, Pacifistic build docs, Expansionist builds cops, etc...? Then we don't have to know who is who right away.

Thing is : the scums don't have many means to know who made what (apart from our behaviours).
Even if we may not have a complete inspection circle, it's worth trying to get as many inspections as possible, as the scums are quite blind to who is effective on night 1 or not.
And I doubt people built vigs (no comment pls). So, if some built docs (idem), they should now build cops : Docs delay scums, cops burn them. This way, on next nights, we get the highest odds to catch rats.

Even if night 1 doesn't bring scums, the repetition and high number of cops will make it more likely to catch scums in the following nights. Additionally, the more cops we have (all of us if possible), the harder it is for scums to get rid of them.

[armlx]

The only problem I see is after a few nights the scum get a bunch of RB's and block our units. Still, they will randomly miss with the RB and we will hit them.

[Puzzle]

The only problem I see is after a few nights the scum get a bunch of RB's and block our units. Still, they will randomly miss with the RB and we will hit them.

1. I'm not sure at all that scums get units (imagine each with a vig).
2. The problem is the same even if we build docs or vigs, except that cops will be more efficient in purging the town.
3. Even if scums can build blockers, this goes in the way that we shouldn't go No Lynch anyway.

[armlx]

No, I agree with the plan. I'm just stating something we need to think about for later.

[EmpTyger]

d_rouge:

I agree with Puzzle here that number 1 is probably the best option, since there's not much scum can do about that without casting suspicion upon themselves.
The only problem might be if two scum wind up in a row, so that one is supposed to check the other, but I guess that changing the order everyday is a good way to solve that. It may waste us a few investigations, however, because if A succed in investigating B and finds him innocent, of course tells nothing about that the day after, and whoever checks B the night after is wasting his investigation.
I don't see a solution to this...

Has any analysis been done about whether there’s any advantage to announcing innocent results? I’ve been assuming there isn’t based on conventional strategy, but with this setup that doesn't necessarily mean anything. I suppose it depends on what protective roles we might be having.

I’m not even sure about guilty results. They should definitely be announced immediately, not only to prevent accidental reinvestigation, but more importantly, it prevents valuable results being lost to a future nightkill. But I’m not certain what the optimal is for the town to do after a single guilty result, say, A announcing B is guilty. I’m not sure whether it is best to lynch or no-lynch; [at least] 1 of {A, B} is guilty, but I don’t see any advantage in guessing. If a third party has an innocent results on either A or B (or a corroborating guilty result) then a mislynch would trade 1 innocent for 2 guilty, so it might be worth it then. Alternatively, having B claim their civilization to see if there’s a counterclaim from a third party might achieve the scenario (and if not, perhaps having A likewise claim). On the other hand, no-lynching with the results announced guarantees that the townsperson gets another nightaction, maintaining our numerical advantage while leaving all options for a future day.


armlx:

The only problem I see is after a few nights the scum get a bunch of RB's and block our units. Still, they will randomly miss with the RB and we will hit them.

This isn’t a problem, because the town outnumbers the antitown. So even if scum get a bunch of troops at night, the town will get more. This is another reason why it is better to no-lynch to ensure that our numerical advantage is preserved.

I’m still not completely convinced about option 1, but it seems that its potential for antitown interference is more than outweighed by its advantages.

[armlx]

-Some units have a Bombard value - these can be used as roleblockers. The player can choose to have the Bombard-capable unit target the player's cities or units. Should they target the cities, then that player's unit production or era-advancement for the night will be nullified. Should they target the units, then all that player's units will be unable to perform any actions.

1 mafia with 2 roleblockers shuts down (potentially) 2 townies.

[EmpTyger]

-Some units have a Bombard value - these can be used as roleblockers. The player can choose to have the Bombard-capable unit target the player's cities or units. Should they target the cities, then that player's unit production or era-advancement for the night will be nullified. Should they target the units, then all that player's units will be unable to perform any actions.

1 mafia with 2 roleblockers shuts down (potentially) 2 townies.

Hm. This is a good point.

Unvote: No Lynch

since I can't see any way around this.

[Thok]

EmpTyger-one way around bombard is to build docs first and have people doc protect themselves once and also doc protect others semirandomly. That forces scum to waste bombards on their kill targets (to get through the doc protection) and also to try to guess in the dark on who else they'd need to bombard.

Heck, the threat of doc protection should be enough to force scum to waste bombards.

[SinisterOverlord]

Sire, the votes stand as follows:

No Lynch: 3 (rajrhcpfreak, Thok, dybeck)
dybeck: 2 (VisMaior, Puzzle)
inHimshallibe: 2 (rolandofthewhite, armlx)
Astronaut: 1 (Mr. Flay)
Mr. Flay: 1 (HezLucky)
HezLucky: 1 (d_rouge)

10 to lynch.