Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Uncleeurope]

That’s a lot of faith in the forum to assume that is a strong that doesn’t occur often.

Honestly, it is a compliment more than anything.

[Uncleeurope]

That’s a lot of faith in the forum to assume that is a string that doesn’t occur often.

Honestly, it is a compliment more than anything.

EBWOP

[mail-mi]

everyone should join me on joth

or, well, 2 more people.

[Glooble]

everyone should join me on joth

or, well, 2 more people.

No you come over to sudgy.

[faust]

Glooble is totally the next one to get their head chopped off once joth flips scum.

[mail-mi]

Glooble is totally the next one to get their head chopped off once joth flips scum.

sounds good

[pubby]

I will gladly hammer joth

[Galzria]

I will gladly hammer joth

Need one more vote

[LaLight]

vote: joth

Here you go!

[pubby]

vote: joth
bed man bad

[2.71828.....]

Hey, I was going to hammer. Need to improve me hammer stats

[mail-mi]

Woot

[Galzria]

Hey, we did a thing.

[Joseph2302]

Hey, I was going to hammer. Need to improve me hammer stats

I wanted to hammer

[Brightgalrs]

Here are some numbers from a simulation I ran:

71474 rejected games out of 100000 trials.

Expected values (roles):
   Town: 12.8580593143
   Mafia: 3.46585571058
   Werewolf: 3.58904157611
   Alien: 1.78654560752
   Cult: 0.466030989273
   SerialKiller: 0.502629180397
   Lyncher: 0.127217275468
   Survivor: 0.204620346351

Expected values (discard):
   Town: 16.3984084695
   Mafia: 2.31988361495
   Werewolf: 1.18113300147
   Alien: 0.610180186497
   Cult: 0.053670335834
   SerialKiller: 0.630792960808
   Lyncher: 0.489903947276
   Survivor: 1.03940265021
   WildCard: 0.276624833485


Our actual discarded values are:
   Town:
   Mafia: 2
   Werewolf: 2
   Alien: 1
   Cult: 0
   Serial Killer: 0
   Lyncher 1
   Survivor 1
WildCard: 0

[Brightgalrs]

Accidentally submitted before adding any commentary:

Simulated players would pick between two roles, always picking better role, and if it was a tie, they'd pick randomly.

The main input into the simulation is that role hierarchy I mentioned in my first post. But that's extremely debatable, so these numbers only represent what I think it should be (or what I want you to think I think it should be ).

I did 100000 trials and around 71% of them were invalid setups - almost always too few townies. So that might bring in some doubt about the accuracy of the simulation. But use rolling that 30% chance for a valid setup, first try, seems alright to me.

The expected vs actual values of the discarded roles is pretty spot on. So that part is a bit encouraging.

[Brightgalrs]

Our actual discarded values are:
   Town:

Should be Town: 16

But yeah, you should expect a narrow majority, 13 town vs 10 scum . Well, 12 town with Awaclus's death today...

And tonight will have 3-4 kills.

[MiX]

We did it!

I've either not woken up fully or Bright's analysis of the setup is xyz% wrong. It's one of the two, I'm sure.

Did you sort ALL PRs?

[SpaceAnemone]

Space Count

jotheonah (11): faust, mail-mi, sudgy, Joseph2302, DatSwan, SpaceAnemone, pingpongsam, MiX, Galzria, LaLight, pubby
sudgy (6): 2.71828....., Glooble, jotheonah, WestCoastDidds, Eevee, EFHW
faust (1): Uncleeurope
LaLight (1): raerae
Not Voting (3): Shraeye, Archetype, Brightgalrs

[MiX]

Space Count

jotheonah (11): faust, mail-mi, sudgy, Joseph2302, DatSwan, SpaceAnemone, pingpongsam, MiX, Galzria, LaLight, pubby
sudgy (6): 2.71828....., Glooble, jotheonah, WestCoastDidds, Eevee, EFHW
faust (1): Uncleeurope
LaLight (1): raerae
Not Voting (3): Shraeye, Archetype, Brightgalrs

Uuuuuuuuuh does that mean it wasn't hammer yet?

[MiX]

Space Count

jotheonah (11): faust, mail-mi, sudgy, Joseph2302, DatSwan, SpaceAnemone, pingpongsam, MiX, Galzria, LaLight, pubby
sudgy (6): 2.71828....., Glooble, jotheonah, WestCoastDidds, Eevee, EFHW
faust (1): Uncleeurope
LaLight (1): raerae
Not Voting (3): Shraeye, Archetype, Brightgalrs

Uuuuuuuuuh does that mean it wasn't hammer yet?

No wait ADK counted Bright as voting joth so it WAS hammer.

[Brightgalrs]

We did it!

I've either not woken up fully or Bright's analysis of the setup is xyz% wrong. It's one of the two, I'm sure.

Did you sort ALL PRs?

My narrow definition of power role included doctors and investigative roles.
I only sorted the roles into 6 blocks, so there's lots of simulated random picks when it's obvious which is the better one. I have a more granular list on my desktop that I would've liked to use, but I doubt I'll get to it before twilight ends.

[SpaceAnemone]

I'm enjoying Bright's commitment to analysis... so many decimal places ;-)

I'll probably go ahead and replicate it for myself for the fun of it, so if there are any hidden assumptions now is the time to state them..

Also, I think the model needs a nonzero chance of picking the lower-expected-utility role (as dictated by the heuristic), at least for some fraction of the players.

[SpaceAnemone]

No wait ADK counted Bright as voting joth so it WAS hammer.

Oh, sorry, you're right! There's a vote for "jothenoah" that my script threw a warning on, but I didn't notice... it's clearly meant to be a Joth vote, so I assume ADK counted it that way.

[SpaceAnemone]

Updated Space Count

jotheonah (12): faust, mail-mi, sudgy, Joseph2302, DatSwan, SpaceAnemone, Brightgalrs, pingpongsam, MiX, Galzria, LaLight, pubby
sudgy (6): 2.71828....., Glooble, jotheonah, WestCoastDidds, Eevee, EFHW
faust (1): Uncleeurope
LaLight (1): raerae
Not Voting (2): Shraeye, Archetype