Mini Normal 2153: Post-game

[bob3141]

And I STILL absolutely need Flippy to get in here.
I'm afraid they are scum here, and they want to remove someone who knows them well enough N1 before actually starting to participate.
And the replacement for Sujimichuka has done nothing of value either.

That's two potential lurkers starting to slide from the get go, and I've lost enough games to lurkers already.
I think D1 should be reserved for them unless we find someone who is blatantly scummy.

I'm still not sure Hoopla is that.
The same way I don't think Dunn or Bella are.
But I do think the wagon on Hoopla is the towniest one so far, so I guess I'll play sheep for today and see where that lands me.

Wouldnt be suprised if una is scum that the rest of his hoopla wagon is town

[bob3141]

i should also mention;

ok, should i get on this bob wagon?

idk i think bob is mislynch bait, im not feeling it

this sequence is good. i think town is more likely to show restraint/concern here given a bob-vote could have easily been justified by popo-scum (assuming bob is town).

i find with low-effort players like bob, the best way to read them is from the collective interaction others have with them, rather than anything they do/say themself. if they start receiving a disproportionate level of suspicion relative to what their behaviour ought to realistically warrant, it's an indication there is scum taking advantage of an easy target.

its become apparent to me that bob's wagon seemed likely scum-pushed in the end.

im not a low effort player

I'm a concise player as I spend too much time actually analysing the thread.

You've only played with me when I was pretty sure I had pegged two of the three scum. Day one with the third being caught by agar. Was a good thing as I had been town reading him.

Shame i never got the flash wagon on gypx day one. The game would have been even quicker lol.

[bob3141]

Was fun game. Two games for two. in where i was town roleblocker that i blocked one kill.


One loss and one win. First one town decided to give a free pass to obvous scum. Got nigthkilled so couldnt do anything about it.

[notscience]

Can you do a tiered list bob

[Hoopla]

oh hoh hoho! looks like while the cat's away, the mice will play.

allow me a quick catchup and to defend myself against this fevered mob.

[Hoopla]

I think I've only seen Hoopla's scumgame once, and it was years ago and I replaced in as her buddy. Last game (*takes shot*) Hoopla said she follows a similar formula every time D1,

i haven't been scum in years, so i genuinely don't know how i'd play it these days. historically, i have had a fierce devotion to mimicking my town game as much as possible which won me plaudits for over-the-top levels of deception, and has fuelled deep-seated paranoia about me that still echoes through towns to this very day.

my town game has been developing in a more nuanced direction in recent times. i view the point of the day 1 as a day of content creation and will favour utility-executing as a way to manually steer the town into a more balanced/cohesive gamestate rich with information. towns that are mindful of this and have a good balance of voices tend to win. for the most part, the execution rate of scum D1 is barely above random, and when towns do execute scum D1, it doesn't improve town's overall winrate - perhaps because the town has less associative tells (but i degress). my point is, the outcome of the D1 execution is largely irrelevant; it's more about the process, as a day rich with information (one that is parse-able for analysis on future days) is a big factor in towns winning.

observers of my play have noted a more deliberate/calculated tone to my posts. this is a fair assessment, because i have been. i dislike going into my processes too much, because revealing why i do some of the things i do defeats the purpose of them, but for the sake of transparency; the role i tend to play on D1 depends on how the rest of the town is acting, and i alter my behaviour in a way to influence a town operating in a gamestate conducive to town success. for example;

- in games with many big egos and town leaders, i tend to lay lower, reward good cases with votes, sheep accurate scumhunters, and offer left-field observations that haven't been seen.
- in games rich with spammers/lurkers, i try to shame the spammers into quieting down and bring lurkers into the discussion to balance the flow of the game

this game specifically was a largely chatty affair with little to no early game information to work off. my attempts to stimulate a more chaotic gamestate and wagon with more ferocity was a deliberate adjustment in response to the early-game malaise that was lingering in town hearts. i should also note, i do have another reason for why i have played in this way, but i would rather hang onto it for now, and intend to switch focus to a more analytic bent in my next post and give some more detailed thoughts on what i am seeing.

as a quick observation: the rapid acceleration of my wagon (when the possibility of a hoopla-wagon was wafting through the air) should give the town a clue as to my alignment.

[Isis]

as a quick observation: the rapid acceleration of my wagon (when the possibility of a hoopla-wagon was wafting through the air) should give the town a clue as to my alignment.

I love clues!
A scummier rear end to the wagon means the scumteam doesn't think you're seriously likely to get exiled...
You put up only a token defense here, continuing to "keep secrets" and hold back trump cards, which indicates you don't think you're seriously likely to get exiled...
What circles do you run in, Hoopla?

[Dunnstral]

Dunnstral did you win a scumgame or something

https://forum.mafiascum.net/viewtopic.php?f=23&t=83142

Dunn putting it at L-1 is sketchy but not inherently bad.

Why do you keep taking potshots at me?

[citation needed]

Here are my sources:

I hadn't quite realized that Bob had only posted the 'second' lol. That does make Dunn's vote pretty weird.

VOTE: Bellaphant

Your favorite show is on downstairs Bella :P

I dunno, there's not much to work with. Dunn could stand to try a lot harder, because if everyone had his approach to the early-game then we'd never reach Page 80. I'm surprised you're engaging as much as you are, because you usually say exactly what Dunn's saying right now about the early game. What's different about this game?

People I'd be interested in lynching today is starting to look something like this:

bob3141
Sujimichi●
Bellaphant
Emperor flippyNips
Hoopla
popopopopopopo

With like meh maybes on Galron and Dunn.

Don't relieve pressure now, Hoopla hasn't even seen what we've done with the place :P
I remember there being a few fast wagons last game? I guess the difference is that they weren't on Hoopla.

I think the core of this wagon has been town-driven. Isis feels town, Bella's approach to it has been good, you and Una are town. Dunn putting it at L-1 is sketchy but not inherently bad.

This is amidst you calling me town sporadically so it's pretty weird

[Dunnstral]

A scummier rear end to the wagon means the scumteam doesn't think you're seriously likely to get exiled...

As a reminder I'm town and it doesn't matter how "scummy" I am

[notscience]

I think I took a lot of steam out of it being a nervous nelly, Isis so that’s prob why she doesn’t

I’m curious what she’s garnished before I vote her again I think, if I don’t decide someone else is better

[Bellaphant]

See, hoop, that doesn't massively address my two concerns - either you thought the Gera wagon was great and Towny or you didn't, plus like...your post to me felt really Pockety when I feel like id been clear that I wasn't interested in Gera today. I dunno, maybe a reads list from you would help

Dun, I'm not sure what point you are making?

[Bellaphant]

....oh sorry, I can't read right now dun, ignore me

[Hoopla]

to pick up from my last post, i'm aware that i am an opaque player that is largely unreadable early in games. over time, i have since come to accept that fact about myself and lean into it. i do as i please, and hope the town can see the game is going in the right direction. that usually quells paranoia about my play for a while, but i acknowledge in games where we are frequently mis-executing or things are going off the rails i should definitely be a suspect. this puts pressure on me to be correct early in games, or to do enough d1/d2 to scare scum into NK'ing me.

on to some analysis:

fans of my work may have seen in recent times i have been honing my system of "Holistic Abstracts and Heuristic Analytics" (or H.A.H.A for short). this is a special blend of large-scale data analysis (mostly of d1 play, as this is the cleanest data to work with) and with a healthy sprinkling of intuitive magic to guide it all. i am still weighting how valuable each metric is (this is where intuition plays a role), but here are a couple of observations i've noted that are currently influencing my reads;

a useful metric with enough statistical significance i've seen is that of early game wagons. for this data set, i've collected the alignment of every player that has received a wagon of four players within the first 150 posts of games. these numbers were chosen as they symbolically represent the tipping point of a wagon to "something serious", and by ~150 posts or so, we should be out of the RVS stage of the game. these are fluid concepts that vary from game to game (which is why i use the largest sample size i can), but it's a useful data-backed illumination of a tell i often see early in games.

over the last 5 years of 3:10 mini normals:

town

/

scum

ratio of players receiving four votes before 150 game-posts:

86.3%

/

13.7%

.

scum make up ~23% of the town's population D1, so this is a very significant observation and why i think bob is likely town. to explain why this is the case, i hypothesise that the earlier the game is, the more any vote can be justified. and when any vote can be justified, townies will more or less randomly vote/wagon any player, whereas there are 3 scum-votes that are more likely to land on town by virtue of ignoring teammates more often than random. scum are much willing to place the third or fourth vote on a town player for the sake of wagoning than on a partner because any vote can be justified - it's not damning to ignore a teammate's early wagon when it is currently "random". this discrepancy expresses itself in the numbers. townies are much more likely to be the target of an early wagon.

this is also why i pivoted to a dunnstral vote in 52. he was the fourth voter on the wagon citing "scummy posts", when bob had posted just once. that was the vote that made me think that this wagon had landed on town, and this belief grew throughout the game as more players jumped on bob.

~~

folks who played with me in mini 2146 would also remember i revealed some large scale D1 lurking data. i'm relying on this less this game now that is has been exposed, as i suspect players like klick, una and notscience would act in a deliberate way to subvert this, and definitely inform scumbuddies to stay active in any of them are indeed scum. so, i am not considering lurking data in any real way, but a useful corollary i didn't reveal last game is voting frequency to post-count frequency. town vote-hop on d1 at much higher rates than scum, and though i am still in the process of completing this data-set, it appears significant.

this was an influential reason in why i powered and sat on the geraintm wagon as much as i did. he was a classic candidate of scum finding a vote acceptable enough to be unchallenged, then sitting on it without much reevaluation. it takes effort to be constantly reevaluating the gamestate as scum and to find convincing reasons for your opinions to change. this is why we see scum sit on hivemind-approved votes longer than town, as switching votes incurs the risk of receiving additional attention/critique of your play.

players like klick and notscience seem vote-hoppy-happy anyhow, so i don't extend the tell to capable, aware players like them, but i suspect this tell is meaningful for a player like geraintm and why i still suspect him.

we have had a lot of useful wagons so far; one on galron that i haven't spoken about yet comes to mind. realistically, i'd prefer to speak about the d1 wagons as a whole in conjunction with alignment flips (if i make it that far), and will next switch to some behavioural things and PoE work i have been doing behind the scenes.

i will also say i am flustered and annoyed that i am being forced into collating all my views so early in the game. i prefer to do that after churning up the game-thread for information, but it appears this is the way the dice have fallen this game, and i will give the town as much as i can to work with, come what may.

[Isis]

A scummier rear end to the wagon means the scumteam doesn't think you're seriously likely to get exiled...

As a reminder I'm town and it doesn't matter how "scummy" I am

I've said I townread you, so why did you assume you were the part of the rear end of the wagon I was indicating? I've been liking you but that has a sound like "out, damned spot" to it

[Isis]

These all sounds like delightful knickknacks I will be happy to deadsheep but can't really view as town-indicative in their own right from Hoopla.

[Hoopla]

These all sounds like delightful knickknacks I will be happy to deadsheep but can't really view as town-indicative in their own right from Hoopla.

totally understandable.

i humbly accept my role as blueprint developer for mid-game sheep to follow.

i have more to say about others, but a question to you while i get on with things;

are you not at least a little suspicious of the rapid nature of my wagon? when do you ever see those land on scum? or do you suspect significant bussing in a hoopla-scum universe?

[UnaBombaH]

High Fluff High Content > Medium Fluff Low Content

I don't know why people pretend fluff subtracts from how much content someone has posted, game after game.

Don't be a flareon

Wait, wtf?
I didn't call fluff bad.
Just them calling fluff from others bad.
At least by implication.

[UnaBombaH]

Yea, lot of quality shit from Hoopla.
Not in the AI-sense necessarily, but exactly the kind of thing people can expect from her.
That being said, I really dislike the earlier unvote from someone.
And the fact that bob seemed now to be implying scum!Una because I townread them? Or did I misunderstand something?

Anyway, I've been drinking a little, and need to get some sleep now.
I'll try and wrap my head around my reads sometime tomorrow.

[notscience]

I for one am shocked you disliked something I did

[UnaBombaH]

Like, fuck me for being negative, but I feel like I've already compromised by joining the Hoopla-wagon and now it seems to be overwhelming to push it over the finish line.

[UnaBombaH]

I for one am shocked you disliked something I did

Hey hey, don't be like that.
It's just that I don't think it was a good time to unvote for anyone.
Don't necessariöy thik it was AI.

[notscience]

It wasn’t tbh I just think lolhammers ruin the game so I try to make sure I Avoid them

[Dunnstral]

A scummier rear end to the wagon means the scumteam doesn't think you're seriously likely to get exiled...

As a reminder I'm town and it doesn't matter how "scummy" I am

I've said I townread you, so why did you assume you were the part of the rear end of the wagon I was indicating? I've been liking you but that has a sound like "out, damned spot" to it

Because I was the last vote?

[Dunnstral]

to pick up from my last post, i'm aware that i am an opaque player that is largely unreadable early in games. over time, i have since come to accept that fact about myself and lean into it. i do as i please, and hope the town can see the game is going in the right direction. that usually quells paranoia about my play for a while, but i acknowledge in games where we are frequently mis-executing or things are going off the rails i should definitely be a suspect. this puts pressure on me to be correct early in games, or to do enough d1/d2 to scare scum into NK'ing me.

on to some analysis:

fans of my work may have seen in recent times i have been honing my system of "Holistic Abstracts and Heuristic Analytics" (or H.A.H.A for short). this is a special blend of large-scale data analysis (mostly of d1 play, as this is the cleanest data to work with) and with a healthy sprinkling of intuitive magic to guide it all. i am still weighting how valuable each metric is (this is where intuition plays a role), but here are a couple of observations i've noted that are currently influencing my reads;

a useful metric with enough statistical significance i've seen is that of early game wagons. for this data set, i've collected the alignment of every player that has received a wagon of four players within the first 150 posts of games. these numbers were chosen as they symbolically represent the tipping point of a wagon to "something serious", and by ~150 posts or so, we should be out of the RVS stage of the game. these are fluid concepts that vary from game to game (which is why i use the largest sample size i can), but it's a useful data-backed illumination of a tell i often see early in games.

over the last 5 years of 3:10 mini normals:

town

/

scum

ratio of players receiving four votes before 150 game-posts:

86.3%

/

13.7%

.

scum make up ~23% of the town's population D1, so this is a very significant observation and why i think bob is likely town. to explain why this is the case, i hypothesise that the earlier the game is, the more any vote can be justified. and when any vote can be justified, townies will more or less randomly vote/wagon any player, whereas there are 3 scum-votes that are more likely to land on town by virtue of ignoring teammates more often than random. scum are much willing to place the third or fourth vote on a town player for the sake of wagoning than on a partner because any vote can be justified - it's not damning to ignore a teammate's early wagon when it is currently "random". this discrepancy expresses itself in the numbers. townies are much more likely to be the target of an early wagon.

this is also why i pivoted to a dunnstral vote in 52. he was the fourth voter on the wagon citing "scummy posts", when bob had posted just once. that was the vote that made me think that this wagon had landed on town, and this belief grew throughout the game as more players jumped on bob.

~~

folks who played with me in mini 2146 would also remember i revealed some large scale D1 lurking data. i'm relying on this less this game now that is has been exposed, as i suspect players like klick, una and notscience would act in a deliberate way to subvert this, and definitely inform scumbuddies to stay active in any of them are indeed scum. so, i am not considering lurking data in any real way, but a useful corollary i didn't reveal last game is voting frequency to post-count frequency. town vote-hop on d1 at much higher rates than scum, and though i am still in the process of completing this data-set, it appears significant.

this was an influential reason in why i powered and sat on the geraintm wagon as much as i did. he was a classic candidate of scum finding a vote acceptable enough to be unchallenged, then sitting on it without much reevaluation. it takes effort to be constantly reevaluating the gamestate as scum and to find convincing reasons for your opinions to change. this is why we see scum sit on hivemind-approved votes longer than town, as switching votes incurs the risk of receiving additional attention/critique of your play.

players like klick and notscience seem vote-hoppy-happy anyhow, so i don't extend the tell to capable, aware players like them, but i suspect this tell is meaningful for a player like geraintm and why i still suspect him.

we have had a lot of useful wagons so far; one on galron that i haven't spoken about yet comes to mind. realistically, i'd prefer to speak about the d1 wagons as a whole in conjunction with alignment flips (if i make it that far), and will next switch to some behavioural things and PoE work i have been doing behind the scenes.

i will also say i am flustered and annoyed that i am being forced into collating all my views so early in the game. i prefer to do that after churning up the game-thread for information, but it appears this is the way the dice have fallen this game, and i will give the town as much as i can to work with, come what may.

tl;dr; you're scumreading me because of 'statistics' rather than my one play, and you never reevaluated

[Dunnstral]

Isis, would you support a wagon on popopopo?