EPTTL Mafia LXX: Guardians of the Galaxy - Page 344
| Forum Index > TL Mafia |
|
Artanis[Xp]
Netherlands12968 Posts
| ||
|
Artanis[Xp]
Netherlands12968 Posts
| ||
|
Trfel
7015 Posts
Thoughts? | ||
Fecalfeast
Canada11355 Posts
On March 26 2015 10:54 Artanis[Xp] wrote: Actually with all my townreads it has to be exactly Slam/Superbia/Fecal by PoE. Interesting. So your poe is at least 33% wrong | ||
|
Artanis[Xp]
Netherlands12968 Posts
On March 26 2015 10:56 Trfel wrote: Artanis, I've heard from a very experienced and capable mafia player that process of elimination is not advisable when there are lots of players still alive in the game. Thoughts? It makes sense. With 13 people there are 10 that are town and 3 mafia, meaning any slot has 77% chance of being town. It's much harder to go from 77% town to 95%+ town on 10 slots than it is to go to ~50% mafia on one slot. However, with the way the game has gone and all the blueclaims and orbs, I feel it's landed us much closer to a situation where PoE is possible. SL and you are confirmed through orb usage, Toad/Vivax through how the claim went down, and Eden through general gameplay. Taking out myself, that knocks it down to 7 slots in which 3 are mafia. Suddenly, PoE becomes a lot stronger. | ||
|
Trfel
7015 Posts
Artanis, why are you townreading raynpelikoneet/ritoky again? | ||
|
Half the Sky
Germany9029 Posts
Fecalfeast (3): Eden1892, Vivax, Alakaslam Onegu (2): Artanis[Xp] (1): Superbia Eden1892 (0): Alakaslam (0): Not Voting (7): sicklucker, Onegu, Breshke, ritoky, Damdred, Trfel, Artanis[Xp] Currently, Fecalfeast is set to be executed. Day 3 ends in at 23:00 GMT (+00:00). Reminder to make sure to unvote before voting, if you have already voted someone. Voting is mandatory and in this thread. You may NOT abstain. Posting after the deadline and prior to flip is subject to a modkill. | ||
|
Artanis[Xp]
Netherlands12968 Posts
On March 26 2015 11:01 Trfel wrote: Hm, I disagree with the Onegu town read. Maybe there's something you're seeing in his tone that I'm not. Artanis, why are you townreading raynpelikoneet/ritoky again? Palmar went after him quite hard throughout D1 and Palmar generally doesn't bus. Ritoky in general has played pretty well since subbing in too. He's kinda falling off again though. | ||
|
Trfel
7015 Posts
| ||
|
Artanis[Xp]
Netherlands12968 Posts
On March 26 2015 11:13 Trfel wrote: Does Palmar really never bus? I find that hard to believe. And unreliable, since eventually he's going to bus someone. I'm sure he busses sometimes, but I find it doubtful he'd start out with a shitfight on Rayn of all people as scum on scum. | ||
|
Eden1892
United States5866 Posts
Posting Rate Analysis: An Advanced Look I've talked about this before, but never actually bothered to try to do this analysis for an actual game. See, I think the way a lot of people use filter lengths to analyze players is too simplistic. It's lazy, frankly. They just quote a number from a game, guesstimate whether it's relatively active or inactive given game size, and then say "oh well his filter is X pgs long so he's <alignment>." Naw bitches we steppin this game up. I've created a formula to generate values for posting rates that correct for the sizes of games and the duration of that player's life in the game. Observe: PostRt_This_Game = Filter_Length (pgs) / Player_Count (players) / Time_Alive (days) * 1000 PostRt_Alignment = AVG(PostRt_1, PostRt_2, ...) So for example, if I'm playing in a 20 person game, my filter length is 20 pages and I die n1, my rating will be 1000 and I'll promptly commit seppuku because seriously that's absurd even for me. (20 / 20 / 1 * 1000 = 1 / 1 * 1000 = 1 * 1000 = 1000) The determinations per alignment is just taking the average of the value for each game registered for each alignment. I used the database and in order to get a more accurate sampling, I defaulted to using games from only 2015. Where I didn't feel I had enough games to do it (i.e. had 1-2 games), I went as far back as I could to find enough samples to run it. I went ahead and did the calculations for 3 of the 4 players I would lynch today. (Left out FF because he only had like 1-2 games in the database as mafia.) Breshke postrt_this_game: 0.8721 postrt_town: 5.0262 postrt_scum: 3.4863 Superbia postrt_this_game: 2.3256 postrt_town: 8.5083 postrt_scum: 3.8854 ritoky postrt_this_game: 4.7438 postrt_town: 3.7115 postrt_scum: 1.9685 So what did we find?
Unfortunately I couldn't do FF's, since he has one valid scum game (the 4-hr mini is probably too different to be factored here). But I think this is a decent start. Superbia likely mafia, ritoky likely town, Breshke maybe mafia but really his inactivity is too extreme to know what to do with it. | ||
|
Eden1892
United States5866 Posts
Sorry, instead of time alive, it's number of pages at time of death. So amend my example to say that there were 100 pages of filter. 20 pgs of filter for me, I die n1, 20 players in game. 20 / 20 / 100 * 1000 = 10.0, which is still extreme lol | ||
|
Eden1892
United States5866 Posts
| ||
|
Eden1892
United States5866 Posts
BUT "HE TALKS 2ME LIKE IM CONFIRMD TOWN SO I MUST B TOWN" - tlmafia.net | ||
|
Alakaslam
United States17336 Posts
More like 67% I am the 34% sson | ||
|
Damdred
15669 Posts
Anyway art I see what you mean on the tone here. I hate to do this but I'm starting to doubt two vig again for some reason... I've been pretty lazy tommorow I am going to do my favorite thing. | ||
|
Damdred
15669 Posts
I keep going back and forth on balance and looked at fanfic its the same role claims basically besides tracker for alignment cop and no masons. Which the orb could take care of bah. I hate this. Reading super, and ff in the morning I think ff is still town though | ||
|
Damdred
15669 Posts
| ||
|
Toadesstern
Germany16350 Posts
On March 26 2015 04:58 Trfel wrote: I think that Onegu's mason claiming doesn't say he is town. Imagine if Onegu claimed mason with Holyflare on Day 1, and then Holyflare responded "What? I'm not a mason." That would say nothing at all about Onegu's alignment. So, Onegu's mason claim with Holyflare has no downside. And Onegu has two potential positives for doing this. If Holyflare accepts, Onegu gets a free pass through Day 1 and Night 1. And regardless of what Holyflare does, Onegu shows that he is willing to fake-claim mason, thereby giving him an out from all mason claims he makes for the rest of the game. Then Onegu claimed mason with raynpelikoneet. Regardless of raynpelikoneet's alignment, I think this is a poor move for town here. Rereading Onegu's filter in this section, he seems very determined in this statement. And he provided a false breadcrumb. He is much more determined about this mason claim than he was with his previous mason claim with Holyflare. His explanation for this is that he was certain of raynpelikoneet being town, and was defending him in this way. If he is this certain of raynpelikoneet being town, then he ought to explain this read, instead of just claiming mason. And Onegu I don't think Onegu ever explains this read. Basically, this is a bad play for town, as it provides false information. I also note that Onegu said he hadn't read over half the game, so how is he sure enough in his townread of raynpelikoneet that he claimed in such a convincing manner, while most people were very suspicious of raynpelikoneet at this time? Since then, Onegu rescinded his mason claim. I don't recall Onegu doing much else of note. My conclusion is that Onegu is scum here. I don't see why his play has any benefit for town, and claiming mason is a very low-risk play for scum. Ignoring his claims, Onegu hasn't done very much. I'm scumreading Onegu right now regardless of raynpelikoneet/ritoky's alignment, but I think there's a good chance that they are scum together. this one needs more attention because it's damn right. That being said 5:20 am and I'm slowly getting my brainpower started. Someone want to explain to me why we're not lynching between Onegu/Slam anymore and it's moved on to FF? @the above science: Except for some single cases where it's just too extreme to make sense for them to be mafia I don't really like post-count analysis for this game. It's a big game already thus by the very nature of that people WILL have more than average postcounts or otherwise it would not have become such a big game... so unless it's Vivax / Artanis I honestly couldn't care less about that. | ||
|
Eden1892
United States5866 Posts
| ||
| ||
