For all of hoopla's wanting to run up a wagon to l1 to get a claim by the midway point, she votes for no people on it. I'ma gonna go through the posts in detail tomorrow when my head is less clear so I can get an accurate read.
Also, excellent vote gobbles.Excellent. If you do a meta analysis of all of my games I correctly call scum 100% of the time.
by your logic you should self vote at this point so we can get information. Hell, your arguments would indicate that mass claiming would be a good idea.
Posted: Sat Jun 06, 2020 1:17 pm
by CooLDoG
hmmmm... I should stop tving hoopla for a bit. Get some more juicy scum reads.
Posted: Sat Jun 06, 2020 1:20 pm
by Hoopla
In post 403, CooLDoG wrote:by your logic you should self vote at this point so we can get information. Hell, your arguments would indicate that mass claiming would be a good idea.
i have 0% chance of flipping scum, and have good value to town's on later days. that is, to town's that put faith in me.
one to two claims for day 1 is optimal.
Posted: Sat Jun 06, 2020 1:35 pm
by gobbledygook
In post 403, CooLDoG wrote:by your logic you should self vote at this point so we can get information. Hell, your arguments would indicate that mass claiming would be a good idea.
I actually think that mass claims are usually positive for town on day 1 since scum are locked into a claim and usually it makes PoE so much easier.
agar? despite misreading you, he is becoming more and more town by the post. he belongs in the town-pile. i'm thinking we should send him some flowers or chocolates, and hope he joins this hot new wagon that is forming.
Posted: Sat Jun 06, 2020 3:24 pm
by Emperor flippyNips
ahhh there's my first of two prods i shallnt be making it three
Posted: Sat Jun 06, 2020 3:25 pm
by Emperor flippyNips
whats the haps party people. i'll be reading later im in a smite game rn
Posted: Sat Jun 06, 2020 3:53 pm
by Hiraki
hoopla -
do you have any evidence to backup whatever this stupid "HAHA" system is? those numbers have no value unless there's something behind it. i didn't see anything after a quick search. not to mention, your model has a lot of assumptions that i don't know who would get behind, without further analysis (which is why I am skeptical of it in the first place, as neat as I think game theory in mafia is) - for example, your model goes into the fact that 39/50 games have "scum" as being under 23%. How is the 23% calculated? Could there potentially be a point where 2 scum are slightly above that threshold and one scum drags them down? Then I don't think 23% is a fair number, no? And if 23% goes up, I'm sure that 39/50 is not solid either. Which means that your underlying "towntelling" theory isn't very valid and doesn't make sense.
i did a little digging too - no one in this game has posted over 23% when taking a sum of the average and the average of all in the thread as of your data set was 30.615. that would mean that your circle conveniently misses the one person that everyone has put up weird looks for in your town block -
renaissance
. what is even
more
convenient is that the number you went with to highlight your data was 23 while renaissance has
24
posts.
lastly, if we get rid of the two outliers in this thread (i think this helps the dataset become more uniform but i understand the underlying problem here) - your average goes from that 30.615 above to 32.991 which gets rid of every player but Gypyx, gooble & CooLDoG
i don't think this really makes a lot of sense and i'm starting to just feel weird about the clearly soft defenses of renaissance here that don't consider anything that renaissance has posted
well then, this post looks really weird to me, might be another one of Flippy's joke posts, but it really stands out in his ISO
It’s not my fault them buddying up on page 2 or 3 looked bad to me
Sorry I think you don't get my point, what i'm saying is that this post really feels like you're a mafia traitor trying to reveal yourself to your team, and i may be overthinking this, but the punctuation is really weird too, like, each dot represting a mafia member and the dot next to a question mark represting the unknown mafia member (traitor therefore) makes a lot of sense to me
(I hope this answers the questions other guys had about that)
Hoo fuckin boi I don’t like this.
Why is that?
Spoiler:
Traitor discussion out of nowhere is a biiiig red flag. Without trying to delve too far into theorycrafting optimal traitor play, there is no town reason to bring this up in an accusation, but there’s plenty of reason to do so if you’re a traitor or if you know one is out there.
i see, so is gyphx sus to you then?
Posted: Sat Jun 06, 2020 4:08 pm
by LicketyQuickety
VOTE: Espe
Since they seem to be trying to dodge the wagon on them.
In post 403, CooLDoG wrote:by your logic you should self vote at this point so we can get information. Hell, your arguments would indicate that mass claiming would be a good idea.
I actually think that mass claims are usually positive for town on day 1 since scum are locked into a claim and usually it makes PoE so much easier.
Half the reason I made that comment was to bait people like the above. So now I have 4 scum reads. jeez. mass claims on d1 are great for scum. Just absolutely great.
do you have any evidence to backup whatever this stupid "HAHA" system is? those numbers have no value unless there's something behind it. i didn't see anything after a quick search. not to mention, your model has a lot of assumptions that i don't know who would get behind, without further analysis (which is why I am skeptical of it in the first place, as neat as I think game theory in mafia is) - for example, your model goes into the fact that 39/50 games have "scum" as being under 23%. How is the 23% calculated? Could there potentially be a point where 2 scum are slightly above that threshold and one scum drags them down? Then I don't think 23% is a fair number, no? And if 23% goes up, I'm sure that 39/50 is not solid either. Which means that your underlying "towntelling" theory isn't very valid and doesn't make sense.
i did a little digging too - no one in this game has posted over 23% when taking a sum of the average and the average of all in the thread as of your data set was 30.615. that would mean that your circle conveniently misses the one person that everyone has put up weird looks for in your town block -
renaissance
. what is even
more
convenient is that the number you went with to highlight your data was 23 while renaissance has
24
posts.
lastly, if we get rid of the two outliers in this thread (i think this helps the dataset become more uniform but i understand the underlying problem here) - your average goes from that 30.615 above to 32.991 which gets rid of every player but Gypyx, gooble & CooLDoG
i don't think this really makes a lot of sense and i'm starting to just feel weird about the clearly soft defenses of renaissance here that don't consider anything that renaissance has posted
don't get baited into a technical debate my guy. Look at the actual behavior in the game thread and how obscuring behind a system masks a lack of reasons for reads and votes which contradict the claims being made.
Man, I love being right. I'm typically right...
Posted: Sat Jun 06, 2020 4:39 pm
by CooLDoG
dude, I had a flash of brilliance as to why I now have 4 scum reads. We gobbles is the traitor.
do you have any evidence to backup whatever this stupid "HAHA" system is? those numbers have no value unless there's something behind it. i didn't see anything after a quick search. not to mention, your model has a lot of assumptions that i don't know who would get behind, without further analysis (which is why I am skeptical of it in the first place, as neat as I think game theory in mafia is) - for example, your model goes into the fact that 39/50 games have "scum" as being under 23%. How is the 23% calculated? Could there potentially be a point where 2 scum are slightly above that threshold and one scum drags them down? Then I don't think 23% is a fair number, no? And if 23% goes up, I'm sure that 39/50 is not solid either. Which means that your underlying "towntelling" theory isn't very valid and doesn't make sense.
i did a little digging too - no one in this game has posted over 23% when taking a sum of the average and the average of all in the thread as of your data set was 30.615. that would mean that your circle conveniently misses the one person that everyone has put up weird looks for in your town block -
renaissance
. what is even
more
convenient is that the number you went with to highlight your data was 23 while renaissance has
24
posts.
lastly, if we get rid of the two outliers in this thread (i think this helps the dataset become more uniform but i understand the underlying problem here) - your average goes from that 30.615 above to 32.991 which gets rid of every player but Gypyx, gooble & CooLDoG
i don't think this really makes a lot of sense and i'm starting to just feel weird about the clearly soft defenses of renaissance here that don't consider anything that renaissance has posted
oh yes, a magical spreadsheet filled with numbers beyond your wildest dreams. perhaps you'd like to know my collection method for this set of data?
1) record the d1 posts of 50 different 3:10 games. all the games selected were 13 player 3:10 games, so i have a stable ratio of scum to town across each game.
2) iso the mod, subtract their posts from total
3) iso the scum (including any replacements)
4) left over amount = town posts, and we can now observe the frequency of posting by both alignments relative to each other
~~
these numbers are a way to visualise the activity of scumteam as a whole, not so much as individuals. so, for this game for example, we have 398 posts made. 23% of the game should be scum, which means scum should have posted ~91 times collectively to be average. according to my data, if the trend i've observed over time has continued here, we'd expect scum's
actual
output by now to be ~74 posts so far. combinatorially, this starts to make scumteams comprising of multiple active players less likely, and we should expect more scumteam combos containing the likes of our prana's, our lickety's, our espeonage's etc.
admittedly, i'd have preferred to reveal this data and other findings later in the day when we've had more posting, but it seems there are some paranoid townsfolk around doubting the veracity of my prestigious H.A.H.A system.
and again, this is one metric of many i base day 1 lynch choice on. i am strongly in favour of utility lynching. lurkers being a good starting place to build wagons.
Posted: Sat Jun 06, 2020 5:00 pm
by Hoopla
i realised in my haste, there were a couple of things i skimmed over in Hiraki's post and didn't respond to properly. please forgive me:
In post 411, Hiraki wrote:Could there potentially be a point where 2 scum are slightly above that threshold and one scum drags them down?
yes, that is definitely possible. but like the composition of scum on lynch wagons (we can observe the density of scum on/off lynch wagons, or even positionally), we're more observing frequencies. and simply put, there is a much denser saturation of scum in the lurkier slots of towns across time, and probably in this game too.
In post 411, Hiraki wrote:
i don't think this really makes a lot of sense and i'm starting to just feel weird about the clearly soft defenses of renaissance here that don't consider anything that renaissance has posted
why are you hung up on this one town read i've made?
Posted: Sat Jun 06, 2020 5:19 pm
by gobbledygook
Hoopla doesn't Espeonage better fit the profile of scum you're saying? Less posts and almost at prod range.
VOTE: Espeonage
Plus this gives me the ammo I need to kill Agar
Posted: Sat Jun 06, 2020 5:29 pm
by Hoopla
yes, espeonage fits the profile well. he scores higher on the metric of "
could be useful later if he gets his act together
" than players like Flippy, who i view as a lost cause.
generally speaking, i give some amount of grace to lurkers with dormant town potential not yet actualised, but in the interest of cohesion...
UNVOTE: VOTE: espeonage
Posted: Sat Jun 06, 2020 6:01 pm
by Emperor flippyNips
so page 10 i like gnar, una, cooldog, uhh uhh thats it. but im only on page 10 things may change
Posted: Sat Jun 06, 2020 6:14 pm
by Emperor flippyNips
In post 253, Espeonage wrote:It's not like I went and checked, I just sort of posted, but CD was always my vote after the post he made.
