Open 680.1 C9++ | Endgame

[ThinkBig]

VC 1.7

Umlaut

(3): Green Crayons, FireScreamer, Narna

FireScreamer

(2): FrankJaeger, CommKnight

Necta

(2): davesaz, Alchemist21

Narna

(1): Umlaut

Alchemist21

(1): RoryMK

Titus

(1): Brian Skies

shannon

(1): Titus

FrankJaeger

(1): shannon

Not Voting

(1): Necta

With 13 alive, it takes 7 to lynch.

Deadline

: (expired on 2017-05-03 20:06:48)

[Umlaut]

(off-topic: I wonder what the most improbable C9++ setup ever rolled is.)

[Narna]

What do you have out of rvs that's stronger than my shit? "Double mislynch" my ass1

[Green Crayons]

I'm certainly not lynching anyone today that was scum last game, because that is counter-intuitive. It is a gambler's bet. But it's a bet even you know is likely to be correct. Rolling something twice in a row is highly unlikely. Especially if we end up being in a 2 mafia vs 10 VT & 1 1-shot PR game.

I ignored this kind of bad, self-serving logic this last go around--to my detriment. It's literally a 4/12 chance both times, and while I get there's a youtube video about how the probability of rolling scum both times is somewhat less likely than rolling town after rolling scum the first go around, the fact that you're staking out the position on

page 8

that you are going to be a slave to that probability is. very. stupid.

Mafia is a game of personalities. Probabilities divorced from personalities is a scum tool.

UNVOTE: Umlaut
VOTE: Comm

[Umlaut]

GC, does Narna/Comm ping for you? I know it's way early to be thinking about this but I felt like Narna jumped down my throat the second I started questioning Comm.

[Narna]

Buttfuck pre-flip associations. Especially pre-flips from page 9!

[Green Crayons]

Anyone suspecting anyone for their suspicions on a third player before page 20/30 when we stop stretching our legs is probably not great fodder for finding scum (on the basis of defending the third player) because town are looking for any bad holes in arguments, which necessarily requires that sort of situation.

[Green Crayons]

If defense continues after game is not just right out of the box, feel free to make wild speculation into more concrete arguments.

For now, I'd evaluate arguments based on their own validity, rather than who they do and do not "defend" or whatever.

Also you asking me this triggers my buddy spider-sense which I laid to rest last game but now I'm super paranoid.

[CommKnight]

This is C9++ meaning that the set-up is actually found on the site's wiki.

C9++ Wiki

So in this set-up the mod rolls SEVEN numbers of 0-100

1-50 = T (Townie. This actually directly influences scum powerroles.)
51-65 = C (Cop)
66-75 = D (Doctor)
76-85 = V (Vigilante)
86-95 = M (Mason)
96-100 = B (Blocker)

As you can see each set of numbers have an assigned letter.

T = 50
C = 15
D, V, M = 10
B = 5

Meaning T has 50% chance of being rolled for each of the 7 randomly generated numbers. C has 15%, and so on.

If you check the link, they'll show you what roles there are upon rolling each one.

Then you go to the scum roles.

TTTTTTT = Goon + Godfather, Serial Killer (Investigation Immune OR 1-Shot Bulletproof)
TTTTTT = Goon + Godfather
TTTTT = Goon + Godfather, Serial Killer (Investigation Immune OR 1-Shot Bulletproof)
TTTT = 2 Goons + Roleblocker
TTT = 2 Goons + Roleblocker, Serial Killer (Investigation Immune OR 1-Shot Bulletproof)
TT = Goon + Roleblocker + Godfather
T = Goon + Roleblocker + Godfather, Serial Killer (Investigation Immune OR 1-Shot Bulletproof)
0 Ts = Goon + Roleblocker + Godfather

If all 7 are T's you go with the first set-up for scum. Work your way down the list for each C, D, V, M or B that's rolled in place of a T.

So let's say for example the mod rolls CBMMTTT.

This means the game will have 2 goons, mafia roleblocker, Serial Killer, 1-shot cop, town roleblocker, 2 Masons and 5 VTs. For the rest of the configurations you can check.

But as we flip mafia roles/SK is found/town PR's die, we can fill in the ??????? questions with letters and prevent scum from fake claiming. Or we can actually CATCH scum in fake claiming.

Fastposted a few times. LOL, I'm gonna go ahead and say I'm on the right track here. Already Umlaut is trying to make associations and GC doesn't like the fact I won't lynch the other two that were mafia last round DAY 1. (All other days are open season, but Day 1, I'm not lynching them. If you think that's a smart move, I may need to teach you on how to play with probabilities as town). You can hate me all you want, but the numbers don't lie. My preference is in the Umlaut/FS duo today.

[CommKnight]

@GC, if you're paranoid about being buddied. Look at Umlaut's question to you about a Narna and me duo. Like legit. He's trying to get on your good side. Also for being the math guy, he's not even looking at the numbers for this. Just sucking up to you to mislynch D1. I'm a betting man and I will bet in odds when they are favourable to catching scum even if they act townie. They can never hide from probability.

[Green Crayons]

lol

okay

let's kill Comm

[Narna]

!

[Umlaut]

Gambler's Fallacy (MafiaScum Wiki)
Gambler's fallacy (Internet Encyclopedia of Philosophy)

pedit

@GC, if you're paranoid about being buddied. Look at Umlaut's question to you about a Narna and me duo. Like legit. He's trying to get on your good side. Also for being the math guy, he's not even looking at the numbers for this. Just sucking up to you to mislynch D1. I'm a betting man and I will bet in odds when they are favourable to catching scum even if they act townie. They can never hide from probability.

This feels exactly like Comm from last game.

VOTE: CommKnight

[Umlaut]

Also for being the math guy,

he's not even looking at the numbers for this

. Just sucking up to you to mislynch D1. I'm a betting man and I will bet in odds when they are favourable to catching scum even if they act townie.

They can never hide from probability.

What's the probability of getting scum in a C9++ setup?

What's the probability of getting scum in a C9++ setup, given that you got scum last time?

Unless you can answer these questions, you don't actually give a shit about "looking at the numbers." You're just making things up.

[CommKnight]

Also for being the math guy,

he's not even looking at the numbers for this

. Just sucking up to you to mislynch D1. I'm a betting man and I will bet in odds when they are favourable to catching scum even if they act townie.

They can never hide from probability.

What's the probability of getting scum in a C9++ setup?

What's the probability of getting scum in a C9++ setup, given that you got scum last time?

Unless you can answer these questions, you don't actually give a shit about "looking at the numbers." You're just making things up.

Well you have to make an assumption of how many anti-town are in the game. It can range from 2-4 and SK is every second T, so let's say on average there is 3 anti-town every game.

This means that there is 3/13 chance of rolling scum in ANY given C9++ set-up. (Roughly 23%).

While it is true, both games are indeed independent, when they are the same number, same set-up, same probability created by the same randomizer. We have a probability stat to look at.

(3/13) * (3/13) = 9/169 , in other words, 5.3%

So you have to gamble on me being scum on a 5.3% margin. Higher if there are 4 scum, lower if there are only 2.

Probability. You just can't beat it.

Meanwhile. Let's look at the probability of rolling town twice on the assumption of 10 townies each game shall we?

10/13 = 77% (rounding up)

(10/13) * (10/13) = (100/169) OR 59% In other words, if I were to take a random shot at a person who was townie last game, they'd have a significantly higher chance of being scum this game than I would.

For those who don't understand. Out of the two games, there is 13 slots. We're estimating a 10:3 ratio. 13 slots per game makes 169 total combinations. Me being scum in both games only works in 9/169 combinations (or 5.35% of the combinations). While someone being townie in BOTH games works in 100/169 combinations (or 59%). Obviously the percentages seems off when you think 41% chance of every town player from last game being scum this game. But what the probability is saying, is that if we were to branch this town into every possible outcome in the game, then only 9/169 combinations of the two games would have me being scum in both.


On another note @Umlaut. Here's the lowest probability setup: BBBB(VVV/DDD/MMM). Technically BBBBBBB would be the rarest but there is NO set-up available for such a scenario. So the above would be the rarest. The probability would be as follows:

(5/100) (5/100) (5/100) (5/100) (10/100) (10/100) (10/100) = 6.25e-9 (Or in more understandable terms... 625,000/100,000,000,000,000 = 625/100,000,000,000) Yep, so if we had time to make 100 billion games randomly, 625 of them would contain the BBBB(VVV/DDD/MMM) setup!

[CommKnight]

Note on combinations: It doesn't account for what OTHER people are for every slot on am. Just the fact that there are 3 available scum slots for me each round and 10 available town slots. It measures the probability of me falling into any of the 3 scum slots both games, not what others would fall under for the overall "combination". Just where my slot would fall under. I hope that clears that up. Probability is fun. But it can really mess with your head.

[Umlaut]

Well you have to make an assumption of how many anti-town are in the game. It can range from 2-4 and SK is every second T, so let's say on average there is 3 anti-town every game.

This means that there is 3/13 chance of rolling scum in ANY given C9++ set-up. (Roughly 23%).

While it is true, both games are indeed independent, when they are the same number, same set-up, same probability created by the same randomizer. We have a probability stat to look at.

(3/13) * (3/13) = 9/169 , in other words, 5.3%

So you have to gamble on me being scum on a 5.3% margin. Higher if there are 4 scum, lower if there are only 2

This is wrong but at least it makes me think you're genuinely mistaken, because it's the genuine mistake someone would make.

Estimating the number of scum at 3 isn't perfect (the expected number is about 3.27) but close enough. The real problem here is that the (3/13)*(3/13) number isn't "the probability of being scum in a second game given that you were scum in the first," but just "the probability of being scum in both games."

As of now there is

zero

probability that you weren't scum in the first game. You just were. So the probability that you are scum in both isn't (3/13)*(3/13), it's just 1*(3/13).

[Umlaut]

If what you were saying was true it would invalidate the entire concept of re-rolling a game.

[Umlaut]

Also it would imply that the probability of your being scum in the first game and town in the second is (3/13)*(10/13) = 30/169.

Meaning your probability of being scum

or

town this game is 33/169, and the remaining 136/169 of the time the universe explodes.

[Umlaut]

Meaning your probability of being scum or town this game is

39/169

, and the remaining

130/169

of the time the universe explodes.

EBWOP

[Umlaut]

(sorry everyone, CommKnight pressed my nerd rage button)

[CommKnight]

Just because we know the outcome of the first roll, does not make the probability any less. Because in the sum we are looking at percentages. I mean I could very well be scum in both games. But the percentage overall is lower. I mean it'd be like saying if I won the lottery this week and picked the exact same numbers next week that I'd "still have the same chance". But no I wouldn't because the law of probability dictate that picking the exact same numbers as the week before would be stupid and you'd be throwing your money away.

It's like rolling dice. If I roll a six four times in a row. You can bet your ass my money would be on rolling something else the fifth time. It isn't that I could never roll a six a fifth time, but that the probability of me rolling a six five times in a row is a lot lower than just saying I got 1/6 chance of rolling a six every single time and completely ignoring the probability from the last roll.

Now usually you can't do this probability in mafia because there's different mods, time in between the rolls, different sizes, etc. But in an open game with the same number twice in a row. We can use it to our advantage D1 at least of all days. D2 I'd be more interested in getting a solid cop claim. But D1 when we all know each other and any "reads" are now more NAI than anything. I'm willing to risk taking a D1 probability lynch.

Now here's the question. Why is there so much resistance when I go up against FS? I'd say the probability of him being scum is tipped a bit higher than a random lynch by now. Because of how much protection/buddying he is currently receiving.

Fastposted by Umlaut. See, 30/169 is a bigger chance than 9/169.

[Umlaut]

Yes, and the ratio between them is 30:9, or 10:3.

[Umlaut]

It's like rolling dice. If I roll a six four times in a row. You can bet your ass my money would be on rolling something else the fifth time. It isn't that I could never roll a six a fifth time, but that the probability of me rolling a six five times in a row is a lot lower than just saying I got 1/6 chance of rolling a six every single time and completely ignoring the probability from the last roll.

Remind me to do this with you for money some time.

[CommKnight]

So then we're in agreement. I'm less likely to be scum than anyone who was town last game.

Also FS is definitely acting differently this game than last game. Also would you think I'd draw attention to myself this game as scum if I did so last game as scum? I mean fool you once, I'm doing my job. Fool you twice, and LMAO. But in all seriousness. Do you believe that after the game ending early last game and me showing up as scum playing the way I did, would it not make more sense for this time if I were scum to lay lower and not draw as much attention to myself? You really gotta ask yourself that one. Which is why I find it funny GC points it out yet continues to vote me like "Hey, he's doing the same thing twice in a row as scum, he must be scum guys! He'd NEVER change his strategy or anything if his alignment was the same!"

Fastposted. You're on Umlaut. I'll have you know, I predicted a coin flip correctly 23 times in a row one time... Netted myself $300.