In post 9, Yosarian2 wrote:I realize this is going to sound dodgy because of the position I'm in, but I'm going to share it anyway. I ran some numbers before the game started, and if my math is right, lynching the most suspicious person every day isn't a good strategy for town; if town just mechanically does that, scum wins about 61% of the time. (34% chance town lynches 3 town in a row and loses straight out after day 3, another 40% chance town loses after that, multiples out to only a 39% chance of a town win).
So yeah, we're going to have to actually play mafia.
1
(All this math was done before I got my role PM; now that I know my alignment is town, and I'm in the 30% slot, I know that from my perspective town's chances of winning are actually a lot lower then that lynching just using the odds, but obviously none of you know that, so...)
2
(1) You don't say! ^_^
(2) And
that
part makes this whole post strange. I can see not a single town motivation for pointing out things like "the math was done before getting the role PM".
In post 12, Yosarian2 wrote:I random vote sometimes because I have no information at the start of the game, and a random lynch is theoretically better then a no-lynch; also, assuming you're town, voting for someone who's not yourself is theoretically slightly better then voting for no one.In this game, I actually do have information; if I was going to vote for someone right now, it would basically have to be Lady Lambdadelta, because right now she's more likely to be scum then anyone else and I have no other information. But if everyone did that, then we would quicklynch LLD, and then me, and then town would probably lose.
Empty speculation.
In post 13, Murgatroyd and Miss Eliza wrote:RVS wagons can be fun, but I think they're more effective on newer players. I don't think RVS voting a good scum player is going to make them freak out and scumslip on page 1.
RVS is still the best thing to start a game, in my opinion. Not only because of possible wagons; there may be strange early posts too which are standing out.
In post 15, itlepip wrote:While the stat analysis was cool, I think the main argument is on d1. From the newbie games about 20% of the time scum gets lynched d1, so just auto lynching the top scum number would double our odds of lynching scum.
1 My suggestion is to play to day normally, but at hammer time if there isn't a consensus or nobody really scum slipped we lynch the scummiest of the top %s.
2
(1) That's flawed.
(2) On the other hand, this might work.
In post 19, petroleumjelly wrote:1.)
Yosarian2, do your numbers factor in the chunk of games where a nightkill is missed (due to Jailkeeping or No-Killing)? Those situations give Town an additional lynch. And furthermore, the odds of lynching correctly each day after the first missed nightkill would increase because the game would convert from an even number of players (e.g., 2 in 8 odds, then 2 in 6 odds) to an odd number of players (e.g., 2 in 7 odds, then 2 in 5 odds). And even those hard mathematical stats would not be able to encompass the information gained from a missing nightkill.
2.)
What do players think about having each player give hypothetical Jailkeep targets before each night?
Some pros include making it more difficult for scum to pinpoint the Jailkeeper early in the game (because at that point the Jailkeeper would not have to go out of their way to get this information across) and giving the Town more information to work with in the event the Jailkeeper is nightkilled, as there will be a paper trail. This gets
especially
powerful once the first scum is lynched.
Some cons include allowing scum to use the proposed actions to plan out the safest / most surefire kill (which would in turn make it easier to determine who the Jailkeeper is later in the game by process of elimination), and the fact that we would be forcing the Jailkeeper to decide their action before seeing the result of the lynch.
I seriously dislike this post.
Not only because he goes along with the speculation, not only because he even
encourages
Yosarian to make more calculations, but he also starts speculating about the Jailkeeper and possible Nightkills. This is very anti-town.
________
As for Setup Speculation: I don't think it helps a lot. Mathematically, only the first player could be considered a kind of semi-Innocent Child. (And obviously, their percentage won't clear them for the whole game. However, it was them who explained it why, which is something that makes me think that they're town.)
Take this example: I, who's in the middle of the list, am 4× more likely scum than them, Lambdadelta, who's at the end of the list, is 7× more likely. But if we compare Lambdadelta's chance with mine, she's less than 2× likely scum than me. Can you see my point?
In short: we should scumhunt instead of speculating about those numbers. So far Yosarian and Petroleumjelly stands out for me and I don't think the Murgatroyd-Hydra won the Role-Lottery by getting the 5% and drawing scum at the same time.