|
Day One
Apr 17, 2020 13:46:08 GMT -8
Post by stinkypete on Apr 17, 2020 13:46:08 GMT -8
Do whatever with regards to out of game communication, just don't post in HQ if you're dead. Except for the first 24 hours, right?
|
|
|
Day One
Apr 17, 2020 14:14:07 GMT -8
Post by Ron Wednesday on Apr 17, 2020 14:14:07 GMT -8
Yeah, I saw that after the fact and edited my post. Thank you for confirming we do indeed have zero non-participants. If we're going by forum access, this might count people who posted in the 'i'll play' thread. Just by a cursory glance at this thread past the initial hello, it doesn't seem like 18 active people. Maybe we can get a quick check in and our first real datalist of people choosing something. Who would *not* like to be on the list of people we choose to random sacrifice?
|
|
|
Day One
Apr 17, 2020 15:26:00 GMT -8
via mobile
Post by merlotbrougham on Apr 17, 2020 15:26:00 GMT -8
Maybe we can get a quick check in and our first real datalist of people choosing something. Who would *not* like to be on the list of people we choose to random sacrifice? Checking in. I would be much obliged if you left me off the sacrifice list.
|
|
Super Sarah
Super Person
Dancer
Mildtropolis: safe from Coronavirus but the crowd will kill you!
Posts: 381
|
Day One
Apr 17, 2020 16:33:28 GMT -8
via mobile
Post by Super Sarah on Apr 17, 2020 16:33:28 GMT -8
So how do we pick and who do we pick? Will our weekend be crazy enough for most of us to gang up against one person? (I’m thinking this in an old-timey tv announcer voice.)
|
|
Mac Murdock
Super Person
Jazz Singer
To alcohol! The cause of -- and solution to -- all of life's problems.
Posts: 125
|
Day One
Apr 17, 2020 16:47:16 GMT -8
via mobile
Post by Mac Murdock on Apr 17, 2020 16:47:16 GMT -8
How many people are resigned to not vote? We need 50% of the vote to go to one person to send them home and judging from the other thread it seems like a handful of people were gonna abstain.
|
|
|
Day One
Apr 17, 2020 16:49:03 GMT -8
via mobile
Post by bsketchy on Apr 17, 2020 16:49:03 GMT -8
I'm willing to vote. I'd like to stick around.
|
|
|
Day One
Apr 17, 2020 16:55:05 GMT -8
Post by stinkypete on Apr 17, 2020 16:55:05 GMT -8
I'll admit there is a solid distinction to be made with checking in once in 2 days, but the whole "they'll be ahead by getting an elimination at night" is the sunk cost fallacy. It's throwing bad money after worse. Let me explain. Our relative voting advantage over villains is expressible as A = H/V where H is the number of Heroes and V is the number of Villains. I'm not sure if the true ratio is known from the start, but historically it's been around 1/3. That would mean with 18 people at the start, A = 12/6 = 2 When making a decision, we should consider the expected impact on A. We want to keep A large for as many cycles as possible, so we can use our powers and work things out through discourse and figure out how to establish trust (we may have to sacrifice ourselves to the villains by outing ourselves, be proven right after the fact and our posts mined for honest observation, etc). If our expected gain in relative advantage (E ) is positive or zero, it's probably worth doing if it yields information. If our expected gain is negative, there would have to be a compelling, high-stakes reason to go through with it. That is, someone would have to make themselves the prime target for the next vote if they're wrong.
When we know nothing but participation volume, it's still basically random, so 66% chance hero, 33% chance villain. An example of something that changes those chances is when several observations clearly pin a murder on one person, so the choice of that person is more like 90% villain. Anyway, if we do nothing, then the villains kill one hero tonight and E[H] = 11 and E[V] = 6, so E = 11 / 6 = 1.83
If we kill a random person (and then the villains kill one hero), E[H] = 10.33 and E[V] = 5.66, so E = 10.33 / 5.66 = 1.82 < 1.83
House always wins, as it were. Everyone is still adjusting and finding
We will have to message people that we know who haven't posted much so that they actually use their powers tonight
So I don't see us gaining .01 advantage worth of info by killing on the first day. Most information from how people talk in the first week is junk, but info from the powers gives us some bedrock
|
|
annie2
Super Person
I sense fiendish activity afoot.
Posts: 84
|
Post by annie2 on Apr 17, 2020 17:00:27 GMT -8
Stinky, bro, TLDR.
Seems like you seem minimal gain from killing during week one but lots of gain from seeing how powers get used. Correct me if I'm wrong.
|
|
|
Day One
Apr 17, 2020 17:02:37 GMT -8
via mobile
Post by merlotbrougham on Apr 17, 2020 17:02:37 GMT -8
Indeed. Seems unclear to me too. Maybe Evan can address some of that in the phone booth to clear up some of the gray area about out of game communication. Do whatever, just don't post in the Headquarters when you're dead. Whoops I accidentally edited this when I meant to quote it. So sorry, I will watch out so I dont make that mistake in the future. Do whatever with regards to out of game communication, just don't post in HQ if you're dead. As a matter of disclosure, here is what Evan edited off the bottom of my original post: "Back to the subject at hand - There are posts in the voting thread on the matter below, but I was going to post here and try to leave that thread for actual voting. I agree with what some others have said more articulately than I could about needing to have a vote for a kill on day one. In order to figure out who the villains are, we need information. By doing nothing, we get zero information. At least by getting people on the record, we can see where people stand, and what positions they are pushing, and hopefully we catch out a villain while we are at it, but if not, that's part of the game. I don't see any good reason for a hero to advocate a no-vote scenario"
|
|
|
Day One
Apr 17, 2020 17:06:25 GMT -8
Post by stinkypete on Apr 17, 2020 17:06:25 GMT -8
Stinky, bro, TLDR. Seems like you seem minimal gain from killing during week one but lots of gain from seeing how powers get used. Correct me if I'm wrong. We get a small amount of expected loss by killing someone in the first week
|
|
|
Day One
Apr 17, 2020 17:15:02 GMT -8
via mobile
Post by merlotbrougham on Apr 17, 2020 17:15:02 GMT -8
We will have to message people that we know who haven't posted much so that they actually use their powers tonight Since it appears that outside collusion and communication is also not only allowed, but encouraged for the heroes, yes, this should happen regardless.
|
|
|
Day One
Apr 17, 2020 17:25:21 GMT -8
Post by Ron Wednesday on Apr 17, 2020 17:25:21 GMT -8
Stinky, bro, TLDR. Seems like you seem minimal gain from killing during week one but lots of gain from seeing how powers get used. Correct me if I'm wrong. We get a small amount of expected loss by killing someone in the first week Unless I'm mistaken, our powers results arn't publicly posted, so it's not like we get any extra magic from 'using our powers'. The only true measure is from who you vote for and who gets killed. It's not just about the mathmatical results, which I admit are fairly equal. It's about losing a very limited set of useful data. Who someone voted for, who actually logged in and said 'dont kill me', and then who dies and their side. Each of us individually will get some info from powers, but that doesn't mean we get an out in the open result. Not like votes where we have to actually opt in and then kill someone.
|
|
|
Post by stinkypete on Apr 17, 2020 21:40:06 GMT -8
When people die we find out if they were heroes, and they can explain what their power was and all the information they got by using it. Look at the list of powers, some of them provide objective information that can be pieced together, but we don't know who has those yet, so we risk screwing ourselves more by killing a power that we desperately need for rooting out villains.
Every living hero is a growing investment that cashes out at death. Further, by not killing a hero we get more cycles before the villains can reach a point where they can vote as a mask-off bloc and rush us
|
|
Amer
Super Person
My super power is tremendous wealth.
Posts: 372
|
Post by Amer on Apr 17, 2020 22:25:24 GMT -8
There's probably ~75% chance that we'll randomly vote to kill a hero. But will the villains realistically let us randomly vote to kill a villain? Not likely. So it's almost a 100% chance we'll kill a hero.
• The heroes might see some questionable gain by voting to kill someone this week.
• The villains will absolutely see a positive gain by getting us to kill someone this week. And when it turns out to be a hero, they can start casting accusations around at other heroes who voted to kill that hero. It's a double win for the villains.
It seems pretty clear that villains will prosper by us voting this week.
|
|
|
Day One
Apr 18, 2020 2:36:59 GMT -8
via mobile
Post by merlotbrougham on Apr 18, 2020 2:36:59 GMT -8
I've been consulting with an expert in statistics. Perhaps you've heard of Professor Scott Steiner?
He says if we sit idle, there's a one hundred forty one and two thirds chance the first person dead is a hero.
|
|
|
Day One
Apr 18, 2020 7:46:19 GMT -8
Post by stinkypete on Apr 18, 2020 7:46:19 GMT -8
We're choosing between two dead heroes and one dead hero
|
|
annie2
Super Person
I sense fiendish activity afoot.
Posts: 84
|
Day One
Apr 18, 2020 7:48:31 GMT -8
Post by annie2 on Apr 18, 2020 7:48:31 GMT -8
I suspect that without even trying, the villains have set about a feeding frenzy between the heroes.
|
|
|
Day One
Apr 18, 2020 12:07:51 GMT -8
via mobile
Post by bsketchy on Apr 18, 2020 12:07:51 GMT -8
I don't know if there's a feeding frenzy, but there is some drama brewing. I see both sides for sure. On one hand, it would be terrible to blindly kill a hero. When in all likelihood, another hero will die overnight, giving the villains a significant leg up going into day two. On the other hand, there's a non-zero chance that we vote for a villain, thus keeping the playing field weighted in the heros' favor for at least one more day. Damned if you do, damned if you don't.
|
|
Mac Murdock
Super Person
Jazz Singer
To alcohol! The cause of -- and solution to -- all of life's problems.
Posts: 125
|
Day One
Apr 18, 2020 12:47:01 GMT -8
via mobile
Post by Mac Murdock on Apr 18, 2020 12:47:01 GMT -8
I'll admit there is a solid distinction to be made with checking in once in 2 days, but the whole "they'll be ahead by getting an elimination at night" is the sunk cost fallacy. It's throwing bad money after worse. Let me explain. Our relative voting advantage over villains is expressible as A = H/V where H is the number of Heroes and V is the number of Villains. I'm not sure if the true ratio is known from the start, but historically it's been around 1/3. That would mean with 18 people at the start, A = 12/6 = 2 When making a decision, we should consider the expected impact on A. We want to keep A large for as many cycles as possible, so we can use our powers and work things out through discourse and figure out how to establish trust (we may have to sacrifice ourselves to the villains by outing ourselves, be proven right after the fact and our posts mined for honest observation, etc). If our expected gain in relative advantage (E ) is positive or zero, it's probably worth doing if it yields information. If our expected gain is negative, there would have to be a compelling, high-stakes reason to go through with it. That is, someone would have to make themselves the prime target for the next vote if they're wrong.
When we know nothing but participation volume, it's still basically random, so 66% chance hero, 33% chance villain. An example of something that changes those chances is when several observations clearly pin a murder on one person, so the choice of that person is more like 90% villain. Anyway, if we do nothing, then the villains kill one hero tonight and E[H] = 11 and E[V] = 6, so E = 11 / 6 = 1.83
If we kill a random person (and then the villains kill one hero), E[H] = 10.33 and E[V] = 5.66, so E = 10.33 / 5.66 = 1.82 < 1.83
House always wins, as it were. Everyone is still adjusting and finding
We will have to message people that we know who haven't posted much so that they actually use their powers tonight
So I don't see us gaining .01 advantage worth of info by killing on the first day. Most information from how people talk in the first week is junk, but info from the powers gives us some bedrockRN I’m gonna vote for stinky because I hate math and this post gave me flashbacks.
|
|
Farrell
Super Person
Bureau of Investigation Agent
Posts: 398
|
Post by Farrell on Apr 18, 2020 13:16:34 GMT -8
People are arguing that we don’t get any information by not voting. That’s not true, I’m learning a lot from this discussion.
|
|