Alright. Here comes my little project I have been working on this afternoon. So I am a mechanical engineering student, with an interest in computer science. So I like patterns and concrete things. So data. this bickering between everyone kind grinds on me after a while. So what I have done is boil the game down to the votes and start looking for pattens. I'm not looking at any arguments people have made at the moment, I'm just looking at the hard data that I have: Who people have voted for. I'm not sure how useful any of this will be, but I figure it's a good thing to look at the game from a different perspective. so here goes. I apologize in advance for
Attachments and other options
shortcuts: hit alt+s to submit/post or alt+p to preview taking up all of your screen space.
Here is the list of votes from the beginning of the game (actually from pre-game) with a few notes thrown in for the important events. This should be pretty self explanatory, but I have shortned everyone's names. Also, votes work like this: [voter] -> [voted]. So if i voted for mail-mi, it looks like this: ah -> mm. Hope that makes sense. I have a few charts. The first: The whole game.
Disclaimer: I have done my best to collect the votes correctly, but I am human and am prone to error. If any of these data looks incorrect, please correct me. It is not my intention to misrepresent information.
ash -> mc
mm -> nk
lio -> everyone not playing
-------Game Start-------
mm -> nk
ash -> lio
sb -> mm
nk -> ev
mm -> ev
ah -> mm
s -> mm
lio -> RVS
-------End RVS?-------
y -> rob
sb -> ash
s -> rob
mgp -> rob
ash -> sb
mm -> ash
ev -> ash
mc -> sb
mc -> unvote
ah -> unvote
ash -> unvote
sb -> unvote
rob -> sb
sb -> rob
mgp -> sb
mc-> mgp
y -> mgp
y -> unvote
mm -> rob
mm -> ev
sb -> ev
ash -> ah
nk -> ah
ah -> mm
ash -> unvote
s -> ev
rob -> ev
mgp -> unvote
nk -> mgp
ash -> mgp
mc -> ev
mc -> unvote
sb -> mgp
mm -> unvote
ev -> mm
ah -> ev
ah -> unvote
y -> rob
rob -> mm
mm -> rob
s -> mm
lio -> rob
mc -> lio
sb -> rob
nk -> rob
s -> mm
ash -> mm
nk -> unvote
s -> mm
nk -> rob
y -> s
rob -> s
rob -> ah
lio -> rob
ash -> rob
s -> rob
lio -> s
ev -> s
y -> rob
ev -> mm
ah -> mm
s -> mm
lio -> rob
--------IC Revealed-------
ash -> unvote
rob -> y
ash -> mm
nk -> unvote
s -> mm
x -> y
lio -> s
y -> unvote
y -> s
sb -> unvote
mm -> unvote
mm -> s
x -> s
rob -> s
ash -> s ------- L-1
s -> ash
nk -> s ------- hammer
-----------D2----------
sb -> mc
mm -> mc
mc -> lio
x -> ev
y -> x
ash -> mc
x -> lio
mm -> unvote
mm -> lio
lio -> x
ev -> mm
sb -> unvote
mc -> lio
sb -> x
mm -> mc
So that's a lot to digest and it looks kinda difficult to read. So my next thought is: alright, let's boil this down into who each individual player voted for, and who in turn voted against them.
Votes by player:
ash:
ash -> mc
ash -> lio
ash -> unvote
ash -> ah
ash -> unvote
ash -> mgp
ash -> mm
ash -> rob
ash -> unvote
ash -> mm
ash -> s ------- L-1
11
votes against:
sb -> ash
mm -> ash
ev -> ash
s -> ash
4
mm:
mm -> nk
mm -> nk
mm -> ev
mm -> ash
mm -> rob
mm -> ev
mm -> unvote
mm -> rob
mm -> unvote
mm -> s
mm -> mc
mm -> unvote
mm -> lio
mm -> mc
14
Votes against:
ah -> mm
s -> mm
ah -> mm
ev -> mm
rob -> mm
s -> mm
s -> mm
ash -> mm
s -> mm
ev -> mm
ah -> mm
s -> mm
ash -> mm
s -> mm
ev -> mm
15
mc:
mc -> sb
mc -> unvote
mc-> mgp
mc -> ev
mc -> unvote
mc -> lio
mc -> lio
mc -> lio
8
Votes against:
ash -> mc
sb -> mc
mm -> mc
ash -> mc
mm -> mc
5
nk:
nk -> ev
nk -> ah
nk -> mgp
nk -> rob
nk -> unvote
nk -> rob
nk -> unvote
nk -> s ------- hammer
8
Votes against:
mm -> nk
mm -> nk
2
lio:
lio -> everyone not playing
lio -> RVS
lio -> rob
lio -> rob
lio -> s
lio -> rob
lio -> s
lio -> x
8
Votes against:
ash -> lio
mc -> lio
mc -> lio
x -> lio
mm -> lio
mc -> lio
6
sb:
sb -> mm
sb -> ash
sb -> unvote
sb -> rob
sb -> ev
sb -> mgp
sb -> rob
sb -> unvote
sb -> mc
sb -> unvote
sb -> x
11
Votes against:
ash -> sb
mc -> sb
rob -> sb
mgp -> sb
4
ah:
ah -> mm
ah -> unvote
ah -> mm
ah -> ev
ah -> unvote
ah -> mm
6
Votes against:
ash -> ah
nk -> ah
rob -> ah
3
y:
y -> rob
y -> mgp
y -> unvote
y -> rob
y -> s
y -> rob
6
Votes against:
rob -> y
x -> y
2
mgp/x:
mgp -> rob
mgp -> sb
mgp -> unvote
x -> y
x -> s
x -> ev
x -> lio
7
Votes against:
mc-> mgp
y -> mgp
nk -> mgp
ash -> mgp
sb -> mgp
y -> x
lio -> x
sb -> x
8
rob:
rob -> sb
rob -> ev
rob -> mm
rob -> s
rob -> ah
rob -> y
rob -> s
7
Votes Against:
y -> rob
s -> rob
mgp -> rob
sb -> rob
mm -> rob
y -> rob
mm -> rob
lio -> rob
sb -> rob
nk -> rob
nk -> rob
lio -> rob
ash -> rob
s -> rob
y -> rob
lio -> rob
16
ev:
ev -> ash
ev -> mm
ev -> s
ev -> mm
ev -> mm
5
Votes against:
nk -> ev
mm -> ev
mm -> ev
sb -> ev
s -> ev
rob -> ev
mc -> ev
ah -> ev
x -> ev
9
s:
s -> mm
s -> rob
s -> ev
s -> mm
s -> mm
s -> mm
s -> rob
s -> mm
s -> mm
s -> ash
10
Votes against:
y -> s
rob -> s
lio -> s
ev -> s
lio -> s
y -> s
mm -> s
x -> s
rob -> s
ash -> s ------- L-1
nk -> s ------- hammer
The numbers after each section is just a count of the votes. So now I then broke this down further (I have this on paper, not digital, so no copy-paste here). I took each person and just looked at who else that person had voted for. How do I use this information though? well, what do I know? I know 3 of these people are townies. So, let's look at who voted for the townies
key: [name] - [number of townies voted for]/[total number of different people voted for]
ash - 2/7
-> mm- 3/7
mc - 0/4
-> nk - 2/5
lio - 2/3
sb - 1/6
ah - 0/2
y - 2/3
mgp/x - 2/6
rob - 1/6
ev - 1/3
s - 1/4
The reason for the arrows on nk and mm are that they are the only two people who voted for all the currently known townies (nk of course knowing that he himself is a townie). Now, my thought is that a townie would move their vote around a lot because they are trying to get the most information from the most people. They are the ones worried about scum hunting. Although, a tricky scum could also do the same thing, so high variability does not necessarily mean town, but for this analysis I'm going to give variability town-cred. Now, the other thing to look at are the wagons: robz, sudgy, and mm. so let's list how many of the people at some point voted for someone who had a wagon (from here on out, I'm analyzing this on the fly and just kind of getting my thoughts down. Some may be useful, some may be dead ends. Let's go on an adventure!):
Red indicates that player was on max wagons (-1 if they had a wagon themselves)ash - 3mm- 2mc - 0
nk - 2lio - 2sb - 2
ah - 1
y - 2
mgp/x - 2
rob -2
ev - 2
s - 2
alright. Looking at this this actually doesn't look terribly useful. 2 of the confirmed town were on max wagons, so that doesn't really seem to tell us much about those others who were on the max wagons. I would like to note that yuma and lio have the highest on-wagon/number of players voted ratio (2 on-wagon votes/3 total different votes). However, I would lean towards scum jumping from wagon to wagon, pushing those that look ready to tip. Possibly even jumping out at the last second to not get caught up in those who lynched the D1 lynch. Let's look into that now.
so the D1 lynch was sudgy, were there any bailers last minute? Nope. looks like a dead end. Alright, where to next? let's go back to the variability thing:
least variable: ah, y, ev, lio. (var<=3)
mid: mc, s, nk (3<var<=5)
high: mm, ash, mgp/x, rob (var>5)
So my initial thought was that low variability is scummy and high variability is towny. so if I look at our known townies... I don't really get anything. Robz has high variability, and nk is on the upper end of the mid spectrum, but s is not. So I'm going to say variability is a good place to start, but definitely not the end all.
let's try looking at total number of vote changes:
ash - 11mm- 14mc - 8
nk - 8
lio - 8
sb - 11
ah - 6
y - 6
mgp/x - 7
rob -7
ev - 5
s - 10
well, it looks like this is just as confusing as the variability data. But maybe if we cross-reference the two...
low (<=6) ev, ah, y
mid(>6 && <=10) s, rob, mgp/x, lio, nk, mc
high(>10) sb, mm, ash
So where does this all leave us? I'm not sure. The of the people with low variability, only lio is missing in the low vote changes section. at the high end mm and ash are the only common two.
Alright, so overall, looks like this data thing confused me more than it helped. from this, to me, me, eevee, and yuma all look slightly scummy.
So in conclusion: the three scummy things I searched for: on-wagon/total ratio, variability in votes and the total vote count, yuma was the one constant in all 3 groups.
I'm interested what everyone else might do with this data. I think especially the votes by player section will come in handy in the future when looking for scumpairs (so not for another day or two.) I may look into relationships there more tomorrow (IRL), and even more deeply tommorrow (game time). Happy analysis and goodnight.