Good luck everyone. Be wary.
Guys, I think that we’re getting played here. Here’s the post count in this thread:
Blaster Master 214 - Confirmed town(dead)
Queuing 197
NAF1138 171
zuma 151 - Confirmed town(dead)
Gadarene 118 - Confirmed town(dead)
Fern Forest 112 - Confirmed town(dead)
pimaspinner 108
Rysto 98
glee 90
Autolycus 86 - Confirmed town(dead)
Kat 85
MadTheSwine 85
JSexton 81 - Confirmed town(dead)
FlyingCowOfDoom 75 - Confirmed scum(dead)
Pygmy Rugger 67
Malacandra 67
MonkeyMensch 66
Lakai 63
storyteller0910 56
Menocchio 55
ArizonaTeach 55
Pleonast 54 - Confirmed town(Mason)
Winston Smith 50 - Confirmed town(dead)
Suburban Plankton 48 - Confirmed scum(dead)
Aguecheek 47
brewha 47
Fretful Porpentine 44
kivvik 43 - Confirmed scum(dead)
StarvingButStrong 42
hocow 39
chrisk 34
sturmhauke 29 - Confirmed town(Mason)
nesta 27 - Confirmed scum(dead)
CaerieD 18 - Confirmed town(dead)
cowgirl 17
Smitty 13
We keep killing off the prolific posters and they keep coming up Town. Meanwhile, our SK kept killing the low post-count players, and they kept coming up Mafia. I think that we’ve reached the point where we have to demand that people participate more, or we’ll lynch them.
Pima dies tonight. So, really, it’s not worth arguing about what we should do regarding future readings.
I make no apologies for yesterday’s decisions. It all made perfect sense. Autolycus wasn’t the playing the game. His behavior being a dodge for the GF was simply the more charitable of the options.
Fern Forest was at least a quarter way to a lynching before the cop claim. And that was a 50% probablity of being correct. In the absence of a better idea, it was a decent choice.
What does suck is that the SK and Mafia both were able to outmanuever JSexton.
So, I guess the moral of today’s story is, kill the least suspicious folks?
Shit. Shit. I want a button on my keyboard thay types “shit” over and over.
While I completely understand if you guys want to lynch me today, I am a beat cop. And I did preface my investigation of Fern Forest that I could be completely wrong. Then at the end asked you to look at the evidence and make a determination yourselves. The reason he was being investigated in the first place is he was almost lynched in the sted of Suburban Plankton. I should have known that everyone being in agreement on **Autolycus ** was going to be a bad bad thing. I did investigate **Lakai ** last night. The reading was town. 50/50 is not good enough for me for anymore readings.
I feel really bad about the way things played out, and the fact that I was a direct contributor. By the way guys, I’m a girl. So now you know I feel that much worse about it. 
Yes, I can explain. Yesterday, there were 28 players left alive (I think – anyway, the exact numbers don’t matter since this is just an explanation for illustrative purposes). Pimaspinner decided to investigate one of them. There is a one in twenty-eight chance that the person pimaspinner investigated was actually the serial killer. Since a Beat Cop has a 50% chance of getting a correct read on a given investigation, pimaspinner actually had one chance in fifty-six of correctly identifying the Serial Killer.
An investigation where the cop is wrong has a one in seven chance of identifying the investigatee as the Serial Killer, since the result is random and all of the roles have an equal opportunity of coming up. Pimaspinner thus has one chance in fourteen of getting a WRONG read of Serial Killer (since half of the Beat Cop’s investigations are wrong, and one-seventh of those will come up as Serial Killer).
Therefore, the odds of Pimaspinner’s read on Fern Forest being correct were 4:1 against, even though Beat Cops are right half the time. (This is possible because if the investigation had resulted in a reading of Citizen, there would be a much higher than 50% probability that Pimaspinner was right – a false reading has a one in seven chance of identifying the player as a Citizen, while a true reading has, for example, a 70% chance if 70% of the players are citizens.)
Clear as mud? (I think this is right, anyway – if it turns out to be an illustration of why English majors shouldn’t do math, I apologize.)
You see, the only thing remotely suspicious about any of that was that he was defended by mafia members. Plus, with a detective dead, why would anyone try to attract an investigation by a beat cop. Attracting suspicion is not in the mafia’s favor. You have to stop defending obviously faulty logic which leads to townsmen getting killed. Only mafia would do that.
I also would like to here answers from certain people on certain subjects. I won’t post links for everything because I don’t have enough time before I go to work to find them. However, if you want links later, I will find them for you.
Malacandra, why were you suspicious of hocow for half your career as a townsman.
Rysto, I suspect you’ll tell us soon, but why do you suspect **Storyteller **as scum.
Cowgirl, about this:
I’d like to know why you are suspicious of MadtheSwine. You implicated two people who I think are town in that same post (me and Pygmy Rugger) I want your reasons for the third.
You also did not answer my defense of your post which can be found here .
I’m still wondering why **MadtheSwine **and **StarvingbutStrong **voted for **JSexton **on day 3 and why **Storyteller **voted for **Pimaspinner **on day 3. They were the only ones that ended the day voting for someone other than Suburban Plankon or **Fern Forest. **
NAF1138, can we post the Beat Cop probability question in the GQ forum?
That will end the confusion right there.
Suspected. Not sure what to think anymore.
Ok, so really, what’s the point of being the beat cop…
Dammit. Well, we got screwed that night, that’s for sure. The only thing that’s in our favor now is that the mafia (quit laughing, you mafia) absolutely has to try to target the SK right now, so we may have a little breathing room tonight. Very little.
[quote=glee]
Autolycus posted almost exclusively Mafia-speak content-free posts.
He was defended on several occasions by what turned out to be known Mafia.
He posted illegally at night and edited illegally.
He didn’t seem to understand the rules.
Clearly he was either a Townie playing for himself, or the Godfather trying to attract a Cop to investigate him and reveal a ‘Townie’.
Finally his lynch defence was not convincing.
(Also you really wouldn’t have wanted him to be one of the last Townies, would you?)
I think your logic is faulty!
Firstly it is perfectly reasonable for the Godfather to want to be investigated by the Detective. The result will be ‘Citizen’.
Autolycus began his Mafia-speak immediately the game started, and could easily have been the first person investigated. There wasn’t exactly much information at that point.
Next, having started in that vein, it would look suspicious for Autolycus to change tack once the Detective was killed.
Even though Autolycus was a Town, I repeat my question. Would you want him to be one of the last Citizens?
His lack of knowledge of the game and his selfish approach could easily lead to him failing to lynch the last Mafia.
Finally there is no ‘equivalent’ of Autolycus left. Therefore whether my logic in his particular case was faulty or not, it won’t affect future challenges.
Further to my logic, have a look at my pursuit of Suburban Plankton. I followed Blaster Master’s lead and eventually there was a successful lynch. (Note that two other suspects wwere well in the lead at the time.)
Perhaps you would like to suggest how we find further Mafia.
Well your argument is quite interesting! 
And of course probability is always a tricky topic.
My first response is that PimaSpinner did not try to find the Serial Killer. If he was trying to do that, I think your maths has some basis. The chance that Pima both first selects a rare case and then finds precisely that is unlikely. (Unless Pima is very good at the game!)
(There’s also not exactly a 1 in 28 chance if you accept that three Masons had been identified.)
But all we know is that once Pima selected a suspect, he would get some response. And that would be 50% to be right and, if wrong, equally likely to be one of about 9 choices.
It may seem strange that the probablilities alter if Pima decides beforehand who he wants to find, but I’m confident it’s true!
Lakai:
Only because of this:
referring to CaerieD and Menocchio. I thought MTS’s response above was awfully aggressive towards what is now a known and a suspected townie. That still doesn’t make sense to me.
I had further niggling suspicions but couldn’t confirm them so I dropped them, but I would still like to know why he would say something so definitive in the very early days.
Re: your other post - I was satisfied with your reasoning. I am still suspicious of folks whose posts are more about theorizing than about finger-pointing, but I am going to be re-evaluating everything with the additional knowledge I have now, so I will take an extra close look.
Malacandra, I would also like an answer to this:
Sure, go for it. Nothing else though.
Fuckity fuckity fuck. Sorry people.
I’ll have to review everything and see what I can come up with. For now, here’s yesterday’s voting records.
Everyone voted for Autolycus in the end, and for most people that was their only vote. brewha, Pleonast, and Rysto unvoted temporarily in order to allow time for Auto to raise a defense. Menocchio first voted for JSexton, while Queuing first voted for Fern Forest.
I’ve uploaded my spreadsheet here. I’m not sure if that link works for anyone but me; let me know if there are any problems.
True. More like a 1 in 23 chance, actually, since I miscounted the number of remaining players anyway. But the general principle, I think, still stands, and I’m not convinced that it makes any difference whether pimaspinner sets out to find the SK. (I’m sure this will get answered in GQ by wiser heads than mine, though.)
I can see it. Nice spreadsheet – kinda Mondrian-esque with all the color coding
That makes sense to me even though I am on the low end of post #'s. I guess I’m just gonna have to come in here and babble to increase my posts. The problem for me is that after I catch up on the 200 posts a day, I don’t have the time and energy to say what my make of the whole thing is.
If I post without content, I’ll probably be lynched like Auto.
So far, my suspicions have all been wrong and I’m going to look like scum if I keep fingering town.
I really don’t know what to do at this point. Our major canidates for scum have been proven otherwise.
I’m also not completely sure we should lynch Pima. A beat cop has a pretty good chance of getting the wrong reading.
Well… felch.
My brother actually called me on the road to tell me the outcome of the night’s victims and I can tell you it looks no better in photons than it did in sound waves.
I take back saying that reading is more enjoyable with facts. I think facts are teh suxxor.
Well that sucked. We suffered a big loss in that last day and night.
I don’t think we did much wrong though. So FF wasn’t the SK, a beat cop said he was, and had a decent theory to back it up as well. I don’t see much point in worrying about the math and probability of the correctness of the beat cops reading. It boils down to 50/50 folks. Regardless of what reading he gets, the chances are the same. The cop is either correct or incorrect. Really whether or not the reading is a power role is irrelevant. Its 50/50 right or wrong.
We all knew that. We all choose to believe the cop. I see no reason to stop this. Right now we have another confirmed townie in my mind, Lakai. Sure there is a chance that the reading was wrong, but for right now we must not worry about that.
The GF theory was a good one. It made sense, and things seemed to have fallen in place. We were wrong, and that sucks, but I make no apologies for being wrong.
The SK is now a danger to both factions, and I imagine that the scum will be targeting the SK as well.
I really have no idea whom to vote for at this moment, nor do I have time to read through this thread again. We should take as long as possible for this day as well.
I will not be posting again today, and probably not tomorrow or only 1 post tops. Just FYI.
OK. I thought Aguecheek’s response: “ho-ly cow” was a disguised shout-out as there was no need to hyphenate the word “holy”. I’m now convinced that it wasn’t a shout-out, that there’s no particular reason to suppose that Aguecheek is our other beat cop, and that even if he were there is no reason to suppose that he would have picked up **hocow ** correctly as a killer.
And now to post what I had ready three hours ago when my internet connection dropped out (thank you so much, BT Broadband).
Oh carp. Four pro-town players gone in one day. Well done all, especially pimaspinner. We’re scrod.
Now in a spirit of closing the stable door after the entire herd is nothing more than a settling cloud of dust on the horizon, here’s some mathematics. Suppose the Beat Cop has investigated someone and turned up role X. He can have done this either by finding someone who is role X and correctly identifying them; or by finding someone who is not, incorrectly identifying them, and randomly picking role X. Let there be:
[ul]p players in the game (other than the cop)
[li]s instances of role X in the game[/li][li]*r * distinct roles in the game[/ul][/li]
The probability that the Cop did indeed find a player who is role X is s/p. The probability that he then identifies that player’s role correctly is ½. So the probability of picking a role-X player and identifying him correctly is s/2p.
The probability that the Cop found someone who is not role X is (p-s)/p. The probability that he then identifies that player’s role *incorrectly * is ½. The probability that he then picks Role X out of all possible roles is 1/r. Multiplying these together the probability of picking a non-X player and identifying him as role X is (p-s)/2pr.
Multiplying through by 2pr * we find that the ratio of correct to incorrect identifications as Role X is rs:(p-s). * In the case of a unique role *s=*1 and so this simplifies to r:(p-1). * In other words there will be r * correct identifications to (p-1) incorrect ones (which is not good, as we know that r is appreciably less than p). In the case of a non-unique role there will be rs correct identifications to (p-s) incorrect ones – rather better, if still not good.
Meanwhile… Well, the scum are now looking nearly as likely as we are to want to off the SK, ‘cos they can’t win while the SK is still in the game. For what that’s worth.