Re: Pandas vs Trash Pandas [Day 1]
Posted: Tue Feb 08, 2022 10:31 pm
probably we want constant detail on nanook and then allow townies to use their best judgement to protect other townies?
Murder, Mayhem, and Mafia
https://www.mafiathesyndicate.com/
Dude, why is everything i say insulting to everyone here.NANOOKTHEGREATANDFEARSOME wrote: βTue Feb 08, 2022 10:20 pmK, insulting me isn't the way to go hereGuillotine wrote: βTue Feb 08, 2022 10:18 pmIt's not ego. Just because you are confirmed town it does not mean that whatever you come up with is the perfect plan or solve to win the game, pull your head out of your buttocks and work with me. We have had our differences but im not a thrower.NANOOKTHEGREATANDFEARSOME wrote: βTue Feb 08, 2022 10:13 pm Like I don't really want to policy shoot but I also don't want to lose because someone can't follow a plan cause they have too much ego, y'know
It's not you specifically, it's anyone that won't agree to follow the mech plan. Like...that's just how this game is, it requires cooperation, so the choices are either work with the group or get shot
You're welcome to give input on the plan obviously, but if you're just gonna say "no fuck that im not following any plans!" That's just...not something we can afford in this setup, lol
Oh, Creature suckin up to Nook, eh?Creature wrote: βTue Feb 08, 2022 10:09 pmNot opposed to this. Waiting 48 hours for a flip feels unbearable.NANOOKTHEGREATANDFEARSOME wrote: βTue Feb 08, 2022 10:05 pm I'm gonna shoot at 24 hour mark every day, first person to complain about it or call it anti-town gets shot today
what i mean is that we need to ensure the risk of losing by leaving nanook unprotected is less than the risk of scum losing by missing out on a kill. for example, if we assume scum have a WR of 50% going into n1, and if we assume that scum WR drops to 45% if the kill fails, then as long as we keep the risk of nanook being unprotected under 5% we're good.NotAnAxehole wrote: βTue Feb 08, 2022 10:22 pmIt's scum's probability of winning if they target Nanook exclusively.Urist wrote: βTue Feb 08, 2022 10:15 pmthe acceptable risk is roughly:NotAnAxehole wrote: βTue Feb 08, 2022 9:55 pmYeah, I'll calculate how much we can allot to that after... And once we all agree on the risk factor, we lock it in and stop playing the mechanics, and just play mafia.
the amount of "winrate" scum loses by missing a kill... minus 1%
A vanilla nightless town EV is calculated through [NUMBER OF TOWN PLAYERS - NUMBER OF MAFIA PLAYERS]/[NUMBER OF TOTAL PLAYERS] so an 11:4 nightless town EV would be: [11 - 4]/[15] = 46.67% or wolves EV = 53.33%.
"i believe nanook can kill mafia quickly"NotAnAxehole wrote: βTue Feb 08, 2022 9:45 pm Ok, so basic risk calculation indicates that if we all roll 3, and we only save on a 3 (1/3 of the time)... There's a 1.7% probability of mafia winning instantly with a shot on Nanook -> This isn't a viable shot for them.
If we maintain a ~1.5-2% chance each day (just need to calculate the numbers for each amount of pandas), every scenario, should be trivial...
That gives mafia about a 7% chance of winning in by day 4 (exclusive targeting Nanook), 13% chance with 8 days of targeting...
This is a very good number for town, and I think Nanook can kill mafia faster than this.. I have not calculated the probability of us running out of saves, but I would probably do that privately...
Nobody ever saves anyone but Nanook, and nobody ever claims saves.
CorrectionGuillotine wrote: βWed Feb 09, 2022 12:43 am Im not good at plans but i think i got the plan to keep wolves at bay with our saves.
1) We establish the PoE
2) Assign one player in PoE to protect Nanook and we chop them
3) If for any reason Nanook dies, that means the PoE player didnt protect Nanook.
4) Repeat.
That way we make our PoE protect our red panda and the non-poe choose to protect whoever they townread! Genius!
What you think?
i can't even begin to understand how you thought this would workGuillotine wrote: βWed Feb 09, 2022 12:43 amCorrectionGuillotine wrote: βWed Feb 09, 2022 12:43 am Im not good at plans but i think i got the plan to keep wolves at bay with our saves.
1) We establish the PoE
2) Assign one player in PoE to protect Nanook and we chop them
3) If for any reason Nanook dies, that means the PoE player didnt protect Nanook.
4) Repeat.
That way we make our PoE protect our red panda and the non-poe choose to protect whoever they townread! Genius!
What you think?
I need some sleep Β―\\_(γ)\_/Β―Urist wrote: βWed Feb 09, 2022 12:56 ami can't even begin to understand how you thought this would workGuillotine wrote: βWed Feb 09, 2022 12:43 amCorrectionGuillotine wrote: βWed Feb 09, 2022 12:43 am Im not good at plans but i think i got the plan to keep wolves at bay with our saves.
1) We establish the PoE
2) Assign one player in PoE to protect Nanook and we chop them
3) If for any reason Nanook dies, that means the PoE player didnt protect Nanook.
4) Repeat.
That way we make our PoE protect our red panda and the non-poe choose to protect whoever they townread! Genius!
What you think?
Urist wrote: βWed Feb 09, 2022 12:33 am s = scum remaining
t = town remaining
d = doctor shots remaining
Pn = odds of protecting nanook
Po = odds of protecting someone else (the target is left to the individual panda's discretion)
Pn = (d / s) * (1 - (t / (s + t)))
Po = Pn / 3
In general, the odds of nanook being unprotected by anyone are (1 - Pn) ^ t
So on night one, regardless of who's eliminated, the probabilities work out to...
Pn = (4 / 4) * (1 - (10 / (4 + 10))) = ~28.6%
Po = Pn / 3 = ~9.5%
The chance of nanook being unprotected is only ~3.5% which is close to ideal imo. Anything under 2% is too low because we'll likely run out of protects too quickly.
I think this formula achieves the goals of:
1) making a protect on nanook more likely if scum have fewer players left alive.
2) making a protect less likely if you've already protected before.
3) allowing a small chance of protecting a townie who isn't nanook, to give the opportunity for saves in other places.
Odd NightsUrist wrote: βWed Feb 09, 2022 12:33 am s = scum remaining
t = town remaining
d = doctor shots remaining
Pn = odds of protecting nanook
Po = odds of protecting someone else (the target is left to the individual panda's discretion)
Pn = (d / s) * (1 - (t / (s + t)))
Po = Pn / 3
In general, the odds of nanook being unprotected by anyone are (1 - Pn) ^ t
So on night one, regardless of who's eliminated, the probabilities work out to...
Pn = (4 / 4) * (1 - (10 / (4 + 10))) = ~28.6%
Po = Pn / 3 = ~9.5%
The chance of nanook being unprotected is only ~3.5% which is close to ideal imo. Anything under 2% is too low because we'll likely run out of protects too quickly.
I think this formula achieves the goals of:
1) making a protect on nanook more likely if scum have fewer players left alive.
2) making a protect less likely if you've already protected before.
3) allowing a small chance of protecting a townie who isn't nanook, to give the opportunity for saves in other places.
Are you saying he should only read and react to pings? If so,weird take. He controls the elimination, I think he should probably be up to speed on what is happening in the thread.hollowkatt wrote: βTue Feb 08, 2022 10:16 pm that's probably ideal from nanook tbh to just read and maybe react to pings or something like that
Oh, maybe hk was saying that Nook not posting much is a good plan. Not sure either, if this makes a difference, but I can understand Nook wanting to observe more than interact.NANOOKTHEGREATANDFEARSOME wrote: βTue Feb 08, 2022 10:14 pm Anyways I'm not gonna post much, you guys do your thing
No, we don't have any way to ensure Nanook is protected, and to have sufficient saves to play out the game. Coordinating saves on Nanook also gives scum an exact count.Spacedaisy wrote: βWed Feb 09, 2022 2:23 am Oh god, more math.
I'm thinking about it, and I like the dice roll idea, but it's statistics, and just because it's low doesn't mean it can't happen. If somehow no one protects Nook and he gets hit, we're done. We can't risk that.
What if we mix the ideas of stacking protects and dice roll. Something like, stack protection on Nook, two on one kill nights, and three on two kill nights. And then everyone else rolls to possibly protect someone else (either one of the ones protecting Nook or someone else they think might be at risk), or themselves, or not protect at all? Because if they mafia know who is unprotected, it makes it easy for them to pick us off. But we can't risk completely random because we might accidentally end up with nook unprotected, right?
Someone want to do math on that? Is it sustainable for the length of the game or are we going to run out of kills too quickly if we stack too methodically on Nook?
I don't think we should talk night strategy past this - On N2 if you still have 4 saves, and you miss Nanook, you can try to make a play...NotAnAxehole wrote: βWed Feb 09, 2022 2:09 amOdd NightsUrist wrote: βWed Feb 09, 2022 12:33 am s = scum remaining
t = town remaining
d = doctor shots remaining
Pn = odds of protecting nanook
Po = odds of protecting someone else (the target is left to the individual panda's discretion)
Pn = (d / s) * (1 - (t / (s + t)))
Po = Pn / 3
In general, the odds of nanook being unprotected by anyone are (1 - Pn) ^ t
So on night one, regardless of who's eliminated, the probabilities work out to...
Pn = (4 / 4) * (1 - (10 / (4 + 10))) = ~28.6%
Po = Pn / 3 = ~9.5%
The chance of nanook being unprotected is only ~3.5% which is close to ideal imo. Anything under 2% is too low because we'll likely run out of protects too quickly.
I think this formula achieves the goals of:
1) making a protect on nanook more likely if scum have fewer players left alive.
2) making a protect less likely if you've already protected before.
3) allowing a small chance of protecting a townie who isn't nanook, to give the opportunity for saves in other places.
Panda % No Save % Save
10 0.666666667 0.333333333
9 0.633333333 0.366666667
8 0.6 0.4
7 0.55 0.45
6 0.5 0.5
5 0.45 0.55
4 0.35 0.65
3 0.25 0.75
2 0.15 0.85
Even Nights
Panda % No Save % Save
9 0.6 0.4
8 0.55 0.45
7 0.5 0.5
6 0.45 0.55
5 0.35 0.65
4 0.25 0.75
3 0.1 0.9
2 0 1
Realistically, @NANOOKTHEGREATANDFEARSOME has max 2 miss-chops. If we stick to this, only using saves on Nanook, then by N5 (which is the longest we have to hold out in this scenario) we should on average have 1.8 saves per player on average... This is basically a worst-case winning scenario calculated based on the misschops being D1 & D2. Misschops later in the game offer more flexibility than early.
I donβt understand what you are saying.NotAnAxehole wrote: βWed Feb 09, 2022 2:47 amI don't think we should talk night strategy past this - On N2 if you still have 4 saves, and you miss Nanook, you can try to make a play...NotAnAxehole wrote: βWed Feb 09, 2022 2:09 amOdd NightsUrist wrote: βWed Feb 09, 2022 12:33 am s = scum remaining
t = town remaining
d = doctor shots remaining
Pn = odds of protecting nanook
Po = odds of protecting someone else (the target is left to the individual panda's discretion)
Pn = (d / s) * (1 - (t / (s + t)))
Po = Pn / 3
In general, the odds of nanook being unprotected by anyone are (1 - Pn) ^ t
So on night one, regardless of who's eliminated, the probabilities work out to...
Pn = (4 / 4) * (1 - (10 / (4 + 10))) = ~28.6%
Po = Pn / 3 = ~9.5%
The chance of nanook being unprotected is only ~3.5% which is close to ideal imo. Anything under 2% is too low because we'll likely run out of protects too quickly.
I think this formula achieves the goals of:
1) making a protect on nanook more likely if scum have fewer players left alive.
2) making a protect less likely if you've already protected before.
3) allowing a small chance of protecting a townie who isn't nanook, to give the opportunity for saves in other places.
Panda % No Save % Save
10 0.666666667 0.333333333
9 0.633333333 0.366666667
8 0.6 0.4
7 0.55 0.45
6 0.5 0.5
5 0.45 0.55
4 0.35 0.65
3 0.25 0.75
2 0.15 0.85
Even Nights
Panda % No Save % Save
9 0.6 0.4
8 0.55 0.45
7 0.5 0.5
6 0.45 0.55
5 0.35 0.65
4 0.25 0.75
3 0.1 0.9
2 0 1
Realistically, @NANOOKTHEGREATANDFEARSOME has max 2 miss-chops. If we stick to this, only using saves on Nanook, then by N5 (which is the longest we have to hold out in this scenario) we should on average have 1.8 saves per player on average... This is basically a worst-case winning scenario calculated based on the misschops being D1 & D2. Misschops later in the game offer more flexibility than early.![]()
Even though I don't know how mechanically sound that is.
Are people gonna have to spend another 2 rl weeks begging you to read the rules postGuillotine wrote: βTue Feb 08, 2022 9:10 pm Im telling you right now, i aint attempting to save anyone in this game but myself, i believes in flips over saves.
Falcon doesn't have a truth tell.Dragon D. Luffy wrote: βTue Feb 08, 2022 8:42 pm Hmm
Not sure how hardcore your truth meta is in this case tbh, like theres no work to fake claim the only available town role. But some players have truth metas that are like a religion and they'd rather self-modkill than break them.
No, it's also a mafia game. We should all also solve alignments obviously, and that will also make it easier for Nanook as well as others to solve the game. That is how mafia works. For example, if my solving is good, I'm likely town. If my solving is bad, I'm likely mafia.NotAnAxehole wrote: βTue Feb 08, 2022 9:29 pm Also, for anyone who's confused, this isn't a mafia game, it's a statistics problem, with a side of mafia for Nanook to play.
Look at this obvious scum slip.
No, it's up to everyone to solve. It's up to you to make the kills.NANOOKTHEGREATANDFEARSOME wrote: βTue Feb 08, 2022 10:21 pm Anyways this game is literally up to me to solve so like
Antagonizing me and refusing to cooperate isn't gonna help things
Literally just asking you to be a team player
Why are you so adamant about not cooperating and doing what you want, when you don't seem to understand the basics of the setup in the first place?Guillotine wrote: βTue Feb 08, 2022 10:32 pmDude, why is everything i say insulting to everyone here.This is getting tiring.
Im not opposed to plans, just take it from someone whom is great at turning plans made by town against them as a wolf, you gotta have a wild card, someone capable of doing an 180 on the wolves without risking a throw. The moment you commit to a plan in mafia, that is the moment you lose because scum can manipulate that.
Im not saying i dont want yo cooperate, i do, im saying⦠we dont want to lay it all out for scum.
Why Urist?
I agree with a lot of this, but not at all on the last sentence. We should definitely open up for other saves than on Nanook.NotAnAxehole wrote: βTue Feb 08, 2022 9:45 pm Ok, so basic risk calculation indicates that if we all roll 3, and we only save on a 3 (1/3 of the time)... There's a 1.7% probability of mafia winning instantly with a shot on Nanook -> This isn't a viable shot for them.
If we maintain a ~1.5-2% chance each day (just need to calculate the numbers for each amount of pandas), every scenario, should be trivial...
That gives mafia about a 7% chance of winning in by day 4 (exclusive targeting Nanook), 13% chance with 8 days of targeting...
This is a very good number for town, and I think Nanook can kill mafia faster than this.. I have not calculated the probability of us running out of saves, but I would probably do that privately...
Nobody ever saves anyone but Nanook, and nobody ever claims saves.
Falcon says he's town when he's mafia as well.Dragon D. Luffy wrote: βWed Feb 09, 2022 6:37 am Falcon told me last game he never lies about his role claims.
I strongly recommend you don't attempt to solve it by yourself and just follow whatever the majority is saying unless you have REALLY strong reasons not to.NANOOKTHEGREATANDFEARSOME wrote: βTue Feb 08, 2022 10:21 pm Anyways this game is literally up to me to solve so like
Antagonizing me and refusing to cooperate isn't gonna help things
Literally just asking you to be a team player
I don't care if you don't care.Dyslexicon wrote: βWed Feb 09, 2022 6:42 amFalcon says he's town when he's mafia as well.Dragon D. Luffy wrote: βWed Feb 09, 2022 6:37 am Falcon told me last game he never lies about his role claims.
I don't know or care about the role claims.
I don't really think he's suspicious, but he doesn't have a "I'm town" truth tell.
Sure.Dragon D. Luffy wrote: βWed Feb 09, 2022 6:58 amI don't care if you don't care.Dyslexicon wrote: βWed Feb 09, 2022 6:42 amFalcon says he's town when he's mafia as well.Dragon D. Luffy wrote: βWed Feb 09, 2022 6:37 am Falcon told me last game he never lies about his role claims.
I don't know or care about the role claims.
I don't really think he's suspicious, but he doesn't have a "I'm town" truth tell.
This game has no vanilla town roles. Falcon said he doesn't fake claim roles. He claimed a town role. So unless we believe his previous statement about having that meta was a lie, falcon should be treated as confirmed or nearly confirmed town.
I'm not saying falcon has a "no lie" meta. I'm saying he has a "no fake claim" meta. Those metas exist, I've met people with them in other sites before.