[REVIEW] Angels Vs Demons

[Isis]

I think it's imbalanced then.

I think 6:3 nightless is scumsided, even in the current meta. Last 8:4 nightless was a townstomp and some comparable setups recently (Crossfire, FakeGod Dance, the hearsay on coalitions and nominations if I'm recalling them properly it's harder to gauge games I didn't play in) have been good for town but I think the condensed size of 6:3 helps scum in so many ways that something between 6:3 nightless and 6:3 mountainous isn't what you want to do. Actually, it's probably just a good time in site meta to try 6:3 nightless and see what happens.

[MURDERCAT]

I have to add all this up later, but basically how well town plays D1/N1 determines what nightless setup you enter

elim town D1 (6/9; 66%)

elim town D2 (5/8; 62.5%)

0 NK (6%) - Town win 14%
1+ NK (94%) Town win 0%

elim scum D2 (3/8; 37.5%)

0 NK (6%) - Town win 43%
1 NK (35%) - Town win 33%
2 NK (46%) - Town win 20%
3 NK (13%) Town win 0%

elim scum D1 (3/9; 33%)

elim town D2 (6/8; 75%)

0 NK (18%) - Town win 43%
1 NK (54%) - Town win 35%
2 NK (28%) - Town win 20%

elim scum D2 (2/8; 25%)

0 NK (18%) - Town win 66%
1 NK (54%) - Town win 60%
2 NK (28%) - Town win 50%

[MURDERCAT]

Also, even if EV is scum sided the friendly neighbors should balance it out at least somewhat

[Gypyx]

Technically there is a breaking strategy

D1 no elim

N1 - everyone targets the same person

D2 - the targeted person says who's name he received, we exe him to make sure he's not mafia and lying, if he isn't we have all scum, if he is do it again

I belive scum can't beat this

[MURDERCAT]

This kinda still works if town wants to prevent night kills, but I did make it so a single player can only gets a single message.

[Gypyx]

Ah okay my bad

[MURDERCAT]

No worries that is exactly why I put that in and it took me a while to realize it

[Gypyx]

Also, i'm pretty sure that no Exe d1 has a way higher EV strat potential to this setup, which might be a bit boring

Well that strat still works to an extent, if like, you define the 2/3 top scummiest persons to visit someone you've chosen as D2's exe

Either you gain a clear on someone who was suspicious, or you exe a scum or you now know 2 scum, which is pretty cool

Or lemme see, breaking strategy of targeting in circles

Scum will have to 1v1 with the person below them, so 3 1v1's in total

If the town is exed, 5 v 3, scum kills the FN targeting the now outed scum 4 v 3, we send a 1v1 to clear themselves to the now confirmed townie (only scenario where he isn't confirmed is if the fn'd person he sent to is also in a 1v1 and there aren't any other 1v1, which actually confirms him as scum)

So we got 2 outed scum, meanwhile the last 1v1 pair will probably be able to sort themsleves by both going to whoever ends up being most likely town


Tl;dr : complex as fuck, but i think town wins this 100%

[MURDERCAT]

I'm not sure that no elim is a good approach because the difference between 6:3 and 5:3 on D2 doesn't really matter because the odds of scum being able to just kill you N2 are pretty high. So you are basically just giving up one of your shots for free. Maybe I missed something though.

I think that first targeting strategy is reasonable gameplay. If you haven't elimed scum on D2 you are at risk of just losing outright but town should gain an advantage if you do hit scum.

I'm not sure I follow the 1v1 scenario but I think it's not taking into account the possibility that 2 scum are paired together and clear each other? I'm a bit lost but does it also require multiple FN shots? Because those are 1 shot and fire mandatorily on N1 now.

[MURDERCAT]

Also, I summed up the EV and it's laughably bad, around 20%

Looking at the branching above I think that the most likely outcome is that scum manages to get a D1 town elim.
Town then gets 5 FN shots, so they probably hit scum on D2? But maybe I am overestimating the power of the 5 FN shots to help town coordinate.
Town probably also uses a strategy to limit scum NKs or least isn't random bad (easier to find town than scum) so I think the 33% and 43% branches are then more common than the EV calculation says.
From there you are at a slightly more reasonable EV plus the knowledge gained from the FN shots.

So anyway I'm not convinced it is so scum sided as to be unfun to play, but maybe it's just bad

I like this idea of town being able to choose how much information they gain at the risk of helping scum, but maybe there is a better way to explore that idea.

[northsidegal]

Angels vs Demons

-snip-

this is a really interesting setup

even if i don't really have an intuitive sense right now for how balanced it is

. the dual friendly neighbor / arsonist thing is really cool.

yeah, just realized the "everyone targets one player" strategy which kind of breaks the setup. even if it is true that that player can only receive one name, it basically eliminates the scum nightkill and reduces the game to 3:6 nightless (plus one basically conftown in the person who gets targeted – i assume town collectively can find one locktown), which i would consider (and i imagine you would also consider) something of a setup failure in that it sort of sidesteps the intended mechanics.

i think i might've formulated it slightly differently. even assuming people don't go with the breaking strategy, i think it kind of misses out on that aspect of scum being able to interfere with strong players, because we imagine that a strong townie can probably find at least one lock townread, and thus they won't be able to get killed by scum. only people who end up townreading scum and thus send them their name will be able to be killed (which now that i think about it, is a sort of EV-invisible factor that plays against scum – they're only able to get rid of people who trusted them, and can't get rid of people who didn't).

i also just think that in general when you have some unique mechanic that works with the nightkill that it makes more sense to play it at a higher playercount, 9 people seems a little cramped for this setup.

I think 6:3 nightless is scumsided, even in the current meta. Last 8:4 nightless was a townstomp and some comparable setups recently (Crossfire, FakeGod Dance, the hearsay on coalitions and nominations if I'm recalling them properly it's harder to gauge games I didn't play in) have been good for town but I think the condensed size of 6:3 helps scum in so many ways that something between 6:3 nightless and 6:3 mountainous isn't what you want to do. Actually, it's probably just a good time in site meta to try 6:3 nightless and see what happens.

Half Mast Nightless, something which I think is strictly more townsided than nightless (at least at the same numbers) was a scum victory. So I do think there's probably some element of EV-breakdown as you approach smaller and smaller setups

[MURDERCAT]

Yeah I think that is all very true. I will keep thinking of similar mechanics that play with the night kill, I think that's a fun idea.

Maybe I will run 6:3 nightless as Isis suggested, I'm curious how that goes.

[Isis]

So anyway I'm not convinced it is so scum sided as to be unfun to play,

Yeah, t's fine to run setups that are somewhat scumsided or townsided as long as they're within certain bounds