BtPotEC GF: Post here for strategic / tactical discussions

[4sigma]

I am taking the liberty to start a thread for discussion of strategy and tactics. This include such things as whether to "pass" on an opportunity to perform a lynching. This thread is not intended for the discussion of the guilt or innocence of particular individuals.

For example, if the slayer is dead and it is the students' turn to lynch with a total of 4 players remaining, the students do better to pass and let EC kill someone. That way they have a 1/3 chance of killing EC rather than 1/4.

Before my death, I was engaged in a fairly animated discussion about strategy with via PM Gandalf and BC. For example, if EC kills the slayer while there are still the original 2 EC alive, it is often a good strategy for the students to pass that round in order to be guaranteed to get a new slayer. I'll post more details about this later this weekend.

[4sigma]

A separate strategic point -- once we successfully hit an EC, it is possible that a sleeper may activate. However, I think our best strategy at that point is to not worry too much about the sleeper, but to concentrate on nailing the second original EC. This is because:

1) We have already got some pretty good evidence, I think, of some people who can't be the original EC. So we have some ability to target the original EC. There will be less evidence of who is the sleeper.

2) There may not be a sleeper.

3) If there is a sleeper, then once we kill off both the original EC, we will be able to better detect evidence of who the sleeper is.

It is true that we won't be able to trust that anyone is innocent any more. So we will have to consider that in the PMs we receive from "innocent" people. However, I would suggest that we continue to protect them as "innocent" until we have hit both the original EC, unless some convincing evidence against them subsequently arises.

[4sigma]

Before I died, I sent an elaborate code system to Leela and to EC_Slayer, so that we can communicate to the Slayer who we are suspicious of, who we think is innocent, etc. This is the "groceries code" which EC_Slayer refers to here (http://actuary.ca/phpBB/viewtopic.php?t=31637&start=50).

One of the features in that code is the ability for the ghosts to request certain living students to send some PMs about things that we find odd, and report back to us. For example, if we think CubedBee makes an odd post about something and we want to have someone send him some PMs to get him to talk more about it, the code is set up so that we could send the message "Please have some living students send PMs to CubedBee about his post at the top of page 42" (or whatever).

There are two ways to do this. At the moment if we send such a message, the slayer will be able to read this, and will then send PMs to the students we think are most likely innocent, asking them to send PMs about the stuff we are curious about, and have them report back to us.

The other way is to choose a couple of students who we think are innocent and let them know how to read this code. So when we post the coded message "Please send some PMs to CubedBee asking him about his post at the top of page 42" these students will be able to do so directly, without waiting for the slayer to tell them. This has the advantage that if we lose our slayer, we still retain the ability for certain students to send PMs on our behalf. The disadvantage is that if we should include an EC in error, or if a student becomes the sleeper EC, the EC will become aware of who we are curious about.

At the moment I'm just throwing this out there. For the moment if we are going to use this, (and I think it makes sense to, since it looks like the slayer is not quite on the same wavelength that we are, and besides it will be an awful lot of fun) I think we should just leave the code in the hands of the slayer. However, once we hit our first EC, there is a risk of losing our slayer, and it may be worth considering at that time whether the slayer should pass the code along to one or two students who we presume are innocent.

[ahow]

I had mentioned this in one of the other threads, but thought I would lay it out in the tactical thread. I was thinking of changing my stapler picks, because I really am not very confident of them at this point (considering what I know from the ghost forum). Is there anyway we could influence the EC or other students by me changing my stapler pick? The ways I could see this working (and I am sure there are more):

1) Distract the students from lynching someone who is not our pick (ex: suddenly someone we know is innocent says the wrong thing and the mob suddenly is one vote away from a lynching, at which point I change my stapler vote and say something obvious to put the spotlight on another player of our choice).

2) When we get closer to the ECs, if the combo is not picked (which I don't see happening) I could put the pressure on them by picking that combo.

If anyone has suggestions, let me know.

[Leela]

SSH has used his stapler votes to try to protect players we believe are innocent. The other question I have is how morally correct is it for us to cast votes that we feel are correct since we have more information than the rest of the innocent students.

[Super Silver Haze]

I don't have a specific suggestion, but I think you'd definitely have to say something interesting/inflammatory to make people take notice of the post where you change your vote (which you touched on with your "say something obvious" comment). Another way to make people take notice would be to have several ghosts all change their stapler vote to a certain person right around the same time.

I actually kind of tried something along these lines already, by changing my stapler vote to the two people I least wanted to see killed at that time (4sigma and Gandalf). The idea was for the EC to think that the ghosts were considering lynching those people anyway and hold off on killing them themselves. I thought it was believable, especially since Gandalf had been a lynching candidate in the previous round. However, the EC either didn't notice it or saw right through it, as they killed 4sigma at their next opportunity.

[4sigma]

I think we should mix it up with our stapler votes. I think sometimes we should vote for our most guilty people, and sometimes for anyone else at random, with perhaps a bit more emphasis on our most innocents.

If you want to rationalize the morality of it, you can think of it that your votes based on the advantage ofbetter information is offset by your votes on people who are probably innocent. The frequency with which you alternate these is up to you and your own personal ethical standards. :)

[4sigma]

Regarding how to make this game more fun for the students, I think it will be OK if we post 2 names to choose from. Next round there will be 25 living students. If we post 2 names, that means the slayer is one of 23 students. One of these 23 probably gets slain by the slayer, leaving 22 possible "slayers" for EC to choose from.. So EC's chance to hit them is 1/22. If we had posted only one name, it would be 1/23.

The following round there will be 22 living students. Suppose we providing 2 new names (i.e. whoever, the students chose not to lynch last time, we don't nominate them again). Now we've listed 3 living students that aren't the slayer. So the slayer is one of 19 people as far as EC knows. Slayer slays one of them. So EC's chance to hit the slayer is 1/18. If we had posted only 1 name, it would have been 1/20.

Similar argument for the 3rd round. EC's chance to hit the slayer is 1/14. If we had posted only 1 name, it would have been 1/17.

Cumulative odds of slayer surviving 3 rounds if we post 1 name/round: 22/23 * 19/20 * 16/17 = 85.5%

Cumulative odds of slayer surviving 3 rounds if we post 2 names/round: 21/22 * 17/18 * 13/14 = 83.7%

Extrapolating one more round, Probability of slayer surviving 4 rounds is 79.4% vs. 75.3%. Note that the chances may be a bit better than 75.3%, since the slayer may choose to kill off some of the nominees that the students didn't lynch, or we may recycle some names for nomination before then.

I think that the fun for the students is worth the marginal risk to our slayer. Thoughts?

[ahow]

I agree with this strategy 4sig. Plus we still (at this point) have the possibility of a new slayer since both EC are still surviving...

[Leela]

I agree with this too. And us not knowing the Slayer's real ID actually helps us not give out information, I think. Since we're coming up with names that don't include the Slayer, the Slayer isn't someone who we're too suspicious of yet.

[snafu]

What are the consequences of three names but each new round does not necessarily have to be all new people. i.e. the next round's list could be (but doesn't have to be) the two remaining from the prior round plus one new name?

[Leela]

Also, if we implement this plan, when do we need to announce the names so that the students will be able to get a majority vote by the deadline? I think we should give them at least 36 hrs to reach a consensus.

[snafu]

The other potential fun for the students strategy would be to let the students start a bandwagon and see who the top candidates are. If the former slayer or current slayer is on the list then we post our suggestion of two candidates in order to protect the slayer.

If the main player candidates are OK then we post "Both choice 1 and choice 2 look good to the ghosts. Choice 3 (and maybe choice 4) are also high on the ghost's list"

[snafu]

Also, a potential mitigating strategy would be to post three names each round, BUT secretly include the slayer and/or former slayer in one of the early rounds. Since we get a replacement slayer at this point, there would be a 1/3 chance of losing a slayer turn versus a pretty-close-to-100% chance of camouflaging the slayer.

4Sigma, can your code with the slayer handle "trust us we know what we are doing?"

How badly would it hurt the ghost credibility if this gamble resulted in the slayer getting lynched?

[4sigma]

Revised to reflect the fact that EC of course will never hit themselves: :duh:

Regarding how to make this game more fun for the students, I think it will be OK if we post 2 names to choose from. Next round there will be 25 living students. If we post 2 names, that means the slayer is one of 23 students. 2 of these are EC. One of these 23 probably gets slain by the slayer, leaving 20 possible "slayers" for EC to choose from. . So EC's chance to hit them is 1/20. If we had posted only one name, it would be 1/21.

The following round there will be 22 living students. Suppose we providing 2 new names (i.e. whoever, the students chose not to lynch last time, we don't nominate them again). Now we've listed 3 living students that aren't the slayer. So the slayer is one of 17 people as far as EC knows. Slayer slays one of them. So EC's chance to hit the slayer is 1/16. If we had posted only 1 name, it would have been 1/18.

Similar argument for the 3rd round. EC's chance to hit the slayer is 1/12. If we had posted only 1 name, it would have been 1/15.

Cumulative odds of slayer surviving 3 rounds if we post 1 name/round: 20/21 * 17/18 * 14/15 = 83.9%

Cumulative odds of slayer surviving 3 rounds if we post 2 names/round: 19/20 * 15/16 * 11/12 = 81.6%

Extrapolating one more round, Probability of slayer surviving 4 rounds is 76.9% (1 name per round) vs. 71.4% (2 names per round). Note that the chances with 2 names may be a bit better than 71.4%, since the slayer may choose to kill off some of the nominees that the students didn't lynch, or we may recycle some names for nomination before then.

I think that the fun for the students is worth the marginal risk to our slayer. Thoughts?

[Leela]

How badly would it hurt the ghost credibility if this gamble resulted in the slayer getting lynched?

I think it would hurt our credibility a lot. The Slayer told me during one of the previous rounds that his name had not come up in the voting. This was before we started the suggestions in red. So, it might be possible to include the Slayer and be fairly certain the students wouldn't choose him, but that will only work if we find out who the Slayer is.

[Leela]

So, now the students (mainly Gandalf) don't like the idea of getting to choose between 2 or 3 names? There is just no pleasing them.

[Rocky]

I wanted to add something about the EC's changes in killing pattern.

It does not appear to me that 4sigma and myself were random hits. Maybe one of the two, but not both.

Why would they want us dead?

4sigma, they may have received PM's from him that made them realize that he would not be able to send the same information to students when he was dead. In the game, he did not pose a great threat with his posts, IMO.

Me, I'm not sure about. I did take jabs at EC1 every chance that I got. Also, I was not getting a lot of votes from the students. I don't think that they considered me the slayer since I was so vocal, but I'm not sure.

[Super Silver Haze]

Rocky, we got Karma Police to suggest your name to the EC to be killed, because you were pretty high up on our list of EC candidates. As it turns out, of course, you weren't EC, but at least they saved us a lynching.

As for 4sigma - this is a PM from you to the slayer...

Just to be safe, you should be careful about using code. You don't want to give too much to the EC.

I don't know a lot of the players in this game, but when I saw the code correspondence, one player came to mind immediately as a slayer candidate. If you are that player, I hope that the EC doesn't think the same way that I do!

I know even less about the players in this game than you do, but I thought you were referring to 4sigma here (he was the first one to suggest using the code). That may have been why the EC went after him - they saw us using coded messages and thought he was the slayer.

[Rocky]

SSH, I didn't exactly get that from reading through all the posts here. But, I'm glad that it worked since the students and slayer don't need to waste a turn.

A question for all. Is there any objection to my posting on the public thread that PM's of guilt/innocence should go to the EC_Slayer and a ghost?

I sent all of my guilt/innocence PM's to EC_Slayer to pass on to the ghosts since I think that he/she should be able to consider the same information for his/her kills. Strategy and general thoughts (I didn't send many) were sent to 4sigma directly for the ghost forum.

[Leela]

That's fine. Please add that they should state in the pm who it is copied to so that the Slayer doesn't forward us pms that we have already received.

[4sigma]

Rocky, I believe I already posted that, but it does bear repeating. Heck, maybe we should all go post it again.

Regarding EC's killing pattern, I think they started out the game targeting low post-counters, in the hopes that it would delay the establishment of this forum. At the end of the 2nd round there were now 6 ghosts and Leela and SSH started being active ghosts, and I think EC shifted strategy at that point to start hitting people who they thought the students considered most innocent. For example, I think I showed up on most people's innocent lists, as did Rocky.

[Rocky]

FYI - I just changed my stapler vote to Cho Da & Werewolf. In part to see if some votes would swing to Cho Da so that we could suggest him.

[Leela]

In reponse to EK's current rant, here's an idea.


What if the ghosts make their suggestions very early in the lynching process? We could suggest 2 or 3 names. This could be a way to get create discussion about certain names, like werewolf and thing, that haven't been mentioned very much yet. We would have to start our list earlier in order to run it by the Slayer in time to allow the students time to discuss.

[Rocky]

EK can smooch my tookus... :moon:

I do think that we could/should give a suggestion earlier. The drawbacks of suggesting multiple names early on, at least that I see are:

- Suggestions would obviously not be based on the most current information from the students (via PM and the public forum).

- I also suspect that this would slow down the # of PM's that we get (but that seems to be happening already).

- If providing multiple names, the votes would let the EC know who the students think is the most innocent of the names that we're listing. That is valuable information to them, IMO.

[Rocky]

I want to address the "students voting too quickly" issue.

I have a couple of ideas on this.

1) Let EC_Slayer recommend to the students (based on his/her thoughts and our coded suggestions). We'll still need to ask the students to return to suggesting (but request only one suggestion per player).

- adv's
EC slayer is certainly an innocent student (pleases us and EK since he/she is still living).

Still allows us to give input

protects the slayer 100%

could slow down voting

- disadv's
sub-optimal strategy if slayer does not follow our suggestions (which may be based on more information)

sub-optimal if students do not send guilt/innocence suspicions to slayer

may lead to many more PM's for the slayer to wade through

2) push the idea that each player makes 1 suggestion rather than voting early on

- adv's
allows us to utilize all of the information that we have (but honestly, the flow of PM's has dwindled)

keeps students fully involved

minimizes the chances of premature lynching

protects the slayer (although not 100%)

- disadv's
some players may not like the idea after having the voting frenzy of the last round or 2

will allow EK to fully blame us for misses (which he tends to do anyway)

There's also the earlier idea that someone had of posting our suggestions very early for student discussion.

Thoughts?

[snafu]

Alternatively the ghosts could start the bidding by posting 4 or so names that we would like to see pressured. We could provide rational such as these four have remained suspiciously quiet and we would like to see some voting on them. If a bandwagon forms at any time, whichever ghost happens to be around can step in and ask for some additional heat on the lowest vote getter. The idea would be to get a frenzy of votes without lynching anyone.

[Leela]

The disadvantage of this is if we try to mix up the four names each round, the EC could keep track and considerably narrow down the pool of potential slayers.

[snafu]

If we establish a precedence for stepping in anytime the voting gets too hot, then we could easily drop that slayer into one of the groups (most likely a group with at least one non-slayer leading vote getter to help keep the heaat down).

[Gandalf]

If the EC hits Avi, then we better immediately point out that the EC seems to be taking advantage of who we are protecting, and that therefore we may not suggest either of their top two. It is far better to make such an announcement before there is any pressure on anyone.

Those uncooperative students! If only they would suggest someone KP has told EC might be the slayer.

[Gandalf]

Any thoughts about how to get more PMs from players? Point out that it was Slayer who nailed cubedbee even though B^3 had little attention from students in public threads? That PMed info helps?

EK will then say the ghost forum is not secure. We could respond that EC hit Gandalf when it would be apparent from the ghost forum that Gandalf had not been the slayer.

[Gandalf]

The early evidence I saw against Butters, I easily proved false. I’m hoping the misfits in the GC didn’t muck up this slaying. To the clueless, keep an eye at the bottom of the Rebel Forum Index.
Should we respond? Would it help get us more PMs? Something along the lines of "We expect our slayer took your PMs into account when deciding who to stake. The more information the students provide, the more likely the EC will be killed."

[Leela]

The early evidence I saw against Butters, I easily proved false. I’m hoping the misfits in the GC didn’t muck up this slaying. To the clueless, keep an eye at the bottom of the Rebel Forum Index.
Should we respond? Would it help get us more PMs? Something along the lines of "We expect our slayer took your PMs into account when deciding who to stake. The more information the students provide, the more likely the EC will be killed."

I think that's a good idea. The Slayer told me last night in the pm that s/he hasn't received a pm regarding innocence and guilt or strategy in at least 5 days.

Would now be a good time to ask the remaining student to pm us a list of players arranged from most likely to be EC to least likely? I know in the last game, this helped get cubedbee. It it too early? Does the sleeper element throw this off a little?

[Gandalf]

Good idea. Ask them to focus on EC2. Most likely to least likely to be EC2. If they want to add ideas about the sleeper, that's fine. Reasons are highly encouraged, even if they don't have reasons for everyone. Who are they most likely to respond to? Leela?

[Leela]

I can post something.

[Rocky]

We don't need to respond directly to EK. He doesn't warrant that much attention, IMO.

However, the request for more PM's is a good idea.


To the clueless, keep an eye at the bottom of the Rebel Forum Index.

Not sure that I completely understand this. Just general taunting?

[Gandalf]

Good. I will probably post something too, slightly different, afterwards.

I presume you're inviting Butters and Sunny, using a special id temporarily if necessary. They could share it as long as they indicate with each post who's making it. You'll remind Sunny of the need for extreme confidentiality, I hope.

[ahow]

We don't need to respond directly to EK. He doesn't warrant that much attention, IMO.

However, the request for more PM's is a good idea.


To the clueless, keep an eye at the bottom of the Rebel Forum Index.

Not sure that I completely understand this. Just general taunting?Well, that is where the ghost forum is (for me anyway). I would assume he knows this from last time, but it worries me that he said it...

[Gandalf]

To the clueless, keep an eye at the bottom of the Rebel Forum Index.

Not sure that I completely understand this. Just general taunting?
He's probably referring to "Who's On-line", and that at some point he saw EC2 and Butters on-line together. It proves nothing, as cubedbee could have been logged on as EC2, or one person could be logged on simultaneously in two different internet explorer sessions (I just proved it for myself, as Gandalf and the mole).

[Rocky]

I'm with you Ahow...

EK seems to be referring to the location of the private ghost forum, but it almost sounds like he knows how to make it public. I'm certainly reading too much into it (but I'm guessing that he's getting at something with that post)...

Edited after Gandalf's post. I personally never look at who's online/in chat.

[Leela]

We don't need to respond directly to EK. He doesn't warrant that much attention, IMO.

However, the request for more PM's is a good idea.


To the clueless, keep an eye at the bottom of the Rebel Forum Index.

Not sure that I completely understand this. Just general taunting?Well, that is where the ghost forum is (for me anyway). I would assume he knows this from last time, but it worries me that he said it...

Could he mean the list of users that are signed in. I know that not everyone shows up there, but I do.

[Rocky]

When do we discuss KP's re-entry?

[Leela]

Would now be a good time to ask the remaining student to pm us a list of players arranged from most likely to be EC to least likely? I know in the last game, this helped get cubedbee. It it too early? Does the sleeper element throw this off a little?

So, where should I put the lists I get from the students? I have already received one from Mulan. Should I start a new thread?

[Gandalf]

(referring to letting KP back in)

That traitor? :wink: I can't see how anything is gained by letting him back in, unless we decide we definitely do not want the EC to hit Anonymouse and Will Durant. There's no way we would let him in without a convincing explanation, and EC will know he couldn't provide it. As long as he's out, they've got to wonder.

While I think Anonymouse and Will are both innocent students, I don't mind if the EC takes them out as then EC is definitely missing the slayer.

[Rocky]

Would now be a good time to ask the remaining student to pm us a list of players arranged from most likely to be EC to least likely? I know in the last game, this helped get cubedbee. It it too early? Does the sleeper element throw this off a little?

So, where should I put the lists I get from the students? I have already received one from Mulan. Should I start a new thread?

It's best to keep these in a separate thread.

[ahow]

Suddenly the sunny hit seems much more interesting since I had forgotten about the proxy vote that she had. Do you think the EC remembered that?

Sunny, did anyone mention anything about your proxy vote recently? Did you mention it to anyone???

[Sunny]

yes, I do believe the EC hit me because all the students seem to trust that I'm innocent and give me the proxy. Once anyone's become a central for students to trust and communicate freely, the EC would and should hit him.

And they did. I have no doubt that's the reason why.

[Sunny]

Suddenly the sunny hit seems much more interesting since I had forgotten about the proxy vote that she had. Do you think the EC remembered that?

Sunny, did anyone mention anything about your proxy vote recently? Did you mention it to anyone???

whoa!!!

one more thing, just remembered, don't know if that'd do anything but, EK ASKED ME WHAT THE PROXY VOTE IS FOR, AND I JUST REPLIED WITH ONE WORD, WHY, THEN HE SAID JUST TO SEE WHO TO RALLY AND LYNCH NEXT, THEN I TOLD HIM AVI!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! AND THEN HE SAID AVI IS NOT TO BE TOUCHED THIS ROUND! :o after he got my info! :evil:

ALSO, ANOTHER THING IS, WHILE EVERYONE ELSE SEEMS TO TRUST THAT I'M INNOCENT, I'M THE ONLY ONE NOT PM'D BY HIM AT ALL THESE DAYS!!!

I wonder if EK and Avi are the ECs!!!!!!!!!!!!!!!!!!!!!!!!!!!!

[Sunny]

Now who's got that stapler pick?! :tfh:

[Rocky]

Now who's got that stapler pick?! :tfh:

Sunny,

Your stapler picks are for EC1 and EC2. Since we already know that Cubedbee was EC1, EK and AVI aren't a possible winning combination.

I was wondering, what sort of behind the scenes work was EK really doing? He talked about hundreds of PM's, etc. That just seems like too much work for a game (and at most, a stapler).

[Sunny]

I believe I've forwarded some of his pm's to the ghosts. He did pm me quite a bit. My mailbox has been 100% full for a while now, I've given up deleting all of them, so many of his pm's could have been lost (when the box is 100% full one loses the oldest ones as the new ones come in).

I was trying to give hints. Don't very much care about winning anything at the moment. ;)

One thing I must point out, though, I just saw a pretty long pm by werewolf but before I posted mine his was lost. Did any of you guys see it? I didn't read the whole thing carefully but just skimmed through before I posted, but I did see that he talked in detail and then concluded to suggest or lynch Avi.

Now I just happened to read Snafu's post on werewolf, he has now become a pretty suspicious one to me now.

I'll get my summary up soon. It'll be succint, I might not be able to give you every single reason that I can think of for each, but my conclusion is mostly the combination what's deduced from everything I know and what my gut tells me.

Another thing: I am gonna watch Macroman because he has asked me, even after I was killed by the EC, what I thought of EK. He also volunteered to be my proxy at one point--I did not ask him. In general, if I were EC, I would try to get sunny to slip up information from the ghost forum now. So, I've decided only to post and edit my summary post and stop posting out there in general.

However, if any of you feel led, you can just say Sunny isn't posting out there anymore but she is active in the ghost forum and all of our decisions are also based on her input.

Mostly just for the ECs to rest in peace (or not so much peace) that they ain't gonna get anything outta me, and for the innocent students to know that what the ghost forum comes down to I've got an input too...

Working on my summary, but right now, crudely, I would recommend that we suggest Werewolf and Macroman.

[4sigma]

(referring to letting KP back in)

That traitor? :wink: I can't see how anything is gained by letting him back in, unless we decide we definitely do not want the EC to hit Anonymouse and Will Durant. There's no way we would let him in without a convincing explanation, and EC will know he couldn't provide it. As long as he's out, they've got to wonder.

While I think Anonymouse and Will are both innocent students, I don't mind if the EC takes them out as then EC is definitely missing the slayer.

I think it is right to continue the KP charade. The longer he stays out there, the more credible EC will consider the information he gave them earlier.

[Rocky]

I would think that they would act on it if they considered it credible. But, who knows...maybe they like the strategy of kill whoever the students don't lynch.

It looks like the fuss over KP (somewhat spearheaded by Sunny) has died down, so it's probably fine just to leave him out there. What's the worst that could happen to him? He's a ghost already...

[ahow]

KARMA POLICE WILL BE ORGANIZING HIS OWN SET OF DATA. PLEASE PM ME PIECE OF RELEVANT DATA. PROVIDE A CYPHER IF YOU WOULD LIKE MY ADVICE ON HOW TO USE YOUR VOTES.
BEAUTIFUL!!!

[Gandalf]

Oh, great. EK's keeping info to himself. KP's keeping info to himself. How are we supposed to know what's happening?

[snafu]

Oh, great. EK's keeping info to himself. KP's keeping info to himself. How are we supposed to know what's happening?

Maybe instead of the ghost forum this should be the mushroom cave. You know, keep them in the dark and feed them shit :D

[Gandalf]

Also, what rules have I broken? Mr. P ruled that Gandalf was full of crap in his accusation.Am I supposed to put up with his lies? Just because I wouldn't let Mr. P look in my inbox. :swear:

Seriously, does anyone think a response is a good idea?

[Rocky]

Also, what rules have I broken? Mr. P ruled that Gandalf was full of crap in his accusation.Am I supposed to put up with his lies? Just because I wouldn't let Mr. P look in my inbox. :swear:

Seriously, does anyone think a response is a good idea?

I'd reply with something like:

"That's a very loose (and back asswards) interpretation of what Mr. P. decided." :wink:

[Leela]

I agree with Rocky.

[Gandalf]

Mr. P ruled that Gandalf was full of crap in his accusation.
That's a very loose interpretation of what Mr. P decided. :roll:

Says the person ruled against.
Now what? Just :sleep:

Or ":sleep: I understand EK better than I understand you."

Or nothing at all.

I'm pretty sure 4Sigma wouldn't want a fight. Since it's KP's birthday, I've refrained from telling him where to put any info he collects. :argue:

[Leela]

You could give him a :blah: :blah: :blah:

[ahow]

Since it's KP's birthday, I'll refrained from telling him where to put any info he collects. :moon2:
This actually sounds good to me...

[4sigma]

Mr. P ruled that Gandalf was full of crap in his accusation.
That's a very loose interpretation of what Mr. P decided. :roll:

Says the person ruled against.
Now what? Just :sleep:

Or ":sleep: I understand EK better than I understand you."

Or nothing at all.

I'm pretty sure 4Sigma wouldn't want a fight. Since it's KP's birthday, I've refrained from telling him where to put any info he collects. :argue: I agree that we don't want to start a fight, at least not between KP and Gandalf. Need to get the attn back on KP. I think it is fine that we talk a bit of smack to him, but the wizard may not be the appropriate representative. I nominate snafu. :D

[Rocky]

I sent a PM to KP and asked him to spread the contempt and not focus so much on Gandalf's accusation. He's fine with that.

[Butters]

I was looking through some of the stuff in the PM thread (I couldn't read it all because it was just too long). I noticed that Cubedbee had been on 4sigma's and Sunny's innocent list for a while and they had corresponded with him. From those PMs, it appears that Mulan was being mentioned as innocent. Why hasn't Cubedbee directed the EC to hit her yet? Obviously, 4sigma would have told the ghost forum his reasons for not suspecting Mulan, and Cubedbee would have know this as well. I don't think Mulan should be hit, but why hasn't she been killed by the EC?

[Butters]

I just noticed that Anonymouse was the deciding vote to lynch Avi. If Avi is EC, does this make Anonymouse more or less likely to be the sleeper? I can see it playing out either way. If Avi is EC, he probably knew he was going to be lynched sooner or later, so he told his partner to finish him off to end the discussion and also to make Anonymouse look innocent. Since Anonymouse had already received some votes, being the deciding vote against his partner might have shifted some suspicion off of him. Of course, if Avi is innocent, this is all moot.

[Avi]

all moot :burn:

[Sunny]

I don't think Mulan should be hit

Why?

[Sunny]

I sent a PM to KP and asked him to spread the contempt and not focus so much on Gandalf's accusation. He's fine with that.

Wait, so karma has access to this forum, or not (I know it's all a hoax but does he actually have access or no)?

[ahow]

I sent a PM to KP and asked him to spread the contempt and not focus so much on Gandalf's accusation. He's fine with that.

Wait, so karma has access to this forum, or not (I know it's all a hoax but does he actually have access or no)?
Yes, he can read it, but he is not posting because a clever EC could see his post count go up, but viewable posts stay level...

[Sunny]

I sent a PM to KP and asked him to spread the contempt and not focus so much on Gandalf's accusation. He's fine with that.

Wait, so karma has access to this forum, or not (I know it's all a hoax but does he actually have access or no)?
Yes, he can read it, but he is not posting because a clever EC could see his post count go up, but viewable posts stay level...

gotcha!

So Karma, Happy Birthday to You here! :tup:

[Sunny]

Oh by the way, Karma, and everyone else, maybe it's time for you to gear up and post some serious intent to also propose the names to lynch, so to make it credible, that you really are soliciting ideas and that you will also be proposing alternatives, instead of just talk?

But I could also be not on top of all that is here (haven't caught up yet on this forum), you could also be just corresponding with whoever's pm'ing you privately? But isn't it that ghosts can't respond? So maybe Karma needs to find a way to really mean business and be on the offensive in terms of proposing who to lynch? Otherwise, if I were EC, why would I risk pm'ing Karma?

:-?

Just to let you know, I've been told by some students that they suspect the whole thing is a show and a play on the part of the ghost forum... Let me see if I can find who told me that... this could be an important lead..

[4sigma]

I will say that I think the way we win this game is by figuring out who is NOT EC and then lynching whoever is left. For example, there are 3 students I am utterly convinced are not EC2 right now -- the current slayer, the former slayer, and Jadzia. The first two also cannot be the sleeper EC.

There are several other students I'm pretty sure are not EC2. As long as we keep hitting someone else, we will eventually hit EC2.

[Avi]

For example, there are 3 students I am utterly convinced are not EC2 right now -- the current slayer, the former slayer, and Jadzia.Why?!

[ahow]

For example, there are 3 students I am utterly convinced are not EC2 right now -- the current slayer, the former slayer, and Jadzia.Why?!Jadzia is French-Canadian and we figured out that she has a grammatical error every few posts. Neither EC made any blunders of that sort, therefore we pretty much ruled her out 100%. Perhaps 4sig can point you in the right direction as to where we discussed it...

[4sigma]

For example, there are 3 students I am utterly convinced are not EC2 right now -- the current slayer, the former slayer, and Jadzia.Why?!Jadzia is French-Canadian and we figured out that she has a grammatical error every few posts. Neither EC made any blunders of that sort, therefore we pretty much ruled her out 100%. Perhaps 4sig can point you in the right direction as to where we discussed it...

Easiest is to just search for all posts by jadzia. Here are a few key quotes of hers from her early posts in March:

I've teached pretty much all high school programs in mathematics. I'm from Montreal, and I've never heard of putting GPA on a resume. In fact, in formations on employment, I was told the contrary. Another trouble is that all universities in Quebec (i don't know if it is the case elsewhere) don't all have the same grading system : And if we are in Quebec, we can be audited by two agency... I showed him the reef, and he now thinks all actuaries are crazy !!!

So... is craziness part of actuary ?

I do not believe this person is capable of making the posts that EC2 has made.

The death of Cubed Bee does bring additional evidence of certain people as likely innocent. For example, I think it reinforces that Asynchronous is likely innocent, due to his help in constructing the EC1 profile. Not that the ghosts used it successfully, mind you, but CubedBee did fit the profile that he voiced. This seems improbable for EC2 to have done.

But enough about the EC2 profile -- we have a separate thread for that. The point is the longer we can keep known innocents in the game, the better our odds of hitting EC with each of our strikes. I think we have at least 4 at the moment -- Current slayer, former slayer, Asynch, and jadzia. Perhaps we can identify a couple of others. The more we identify, the better our odds at hitting EC2, and then beginning to smoke out the sleeper, if there is one.

[Avi]

Does EK know about the Ignorance/Fear Paranoia connection?!?!?!

http://actuary.ca/phpBB/viewtopic.php?t=30397&start=2744

[4sigma]

Back to the subject of strategy / tactics, it is worth noting that it may be a good strategy to pass as we near the engame. In particular this may apply starting in round 10, if our slayer is still alive.

Note that if we can reach the scenario where it is our turn to lynch and only 3 players remain, including the slayer, then the students are guaranteed to win.

Because of this, if we are in the scenario where it is our turn to lynch and only 5 players remain, including the slayer, then we may pass and let the slayer strike. The odds work as follows:

Option 1: We lynch someone, and the slayer hits someone. These combined hit EC 1/2 the time. 3 players now remain and it is EC's turn. If they hit the slayer (1/2 the time) they win, otherwise the slayer slays them and wins. Overall odds of EC victory are 1/4.

Option 2: We pass. The Slayer hits EC 1/4 of the time. EC now strikes. 2/3 of the time they miss the slayer and the 3-player scenario above now exists, i.e.students win. If EC hits the slayer (probability 1/3), the students still have a 1/3 chance of hitting EC with their last lynching. Overall odds of EC victory are 3/4 X (2/3 X 0 + 1/3 X 2/3) = 1/6.

Since EC wins 1/4 of the time in Option 1 and 1/6 of the time in Option 2, the students should prefer Option 2 and pass.

Backing up one more round, suppose there are 7 players, including the slayer.

Option 1: We lynch someone, and the Slayer hits someone. These combined get EC 1/3 of the time. If we miss, EC now strikes and hits the slayer 1/4 of the time. If EC hit the slayer, there are 4 players remaining (including EC) and the students clearly pass. Odds of missing the EC the next round are 2/3. a priori odds of EC victory is 2/3 * 2/3 = 4/9 when they hit the slayer.

If EC misses the slayer, the students and the slayer get 2 shots at EC out of 3 non-slayers, so hit EC 2/3 of the time. a priori odds of EC winning are 2/3 * 1/3 = 2/9.

Composite odds of EC victory are (1/4 X 4/9) + (3/4 X 2/9) = 5/18 = 27.8%, since they hit the slayer 1/4 of the time.

Option 2: We pass when there are 7 players remaining, including 1 slayer and 1 EC. Slayer now hits EC 1/6 of the time. EC strikes. If they miss the slayer (4/5 of the time) we are back to the above scenario, where the students pass and win 5/6 of the time. If EC hits the slayer, (1/5 of the time) there are now 5 players remaining. EC survives the next 2 lynchings with probability 4/5 X 2/3 = 8/15.

Composite odds of EC victory in Option 2 are 5/6 X [(4/5 X 1/6) + (1/5 X 8/15)] = 1/5 = 20.0%.

As you can see 20% < 27.8%. So the EC's odds of winning are less if the students pass in the scenario named, i.e. 7 remaining players including 1 EC and the slayer.


Note that this assumes that all parties select randomly. To the extent that we think we have some good information of who is EC, this would improve our odds of lynching compared to passing.

Grateful acknowledgement is made to BC for his work in modelling the game, which contributed significantly to my awareness that this scenario could occur.

Edited to reflect the fact that 1 <> 5/6

[Gandalf]

Option 2: We pass when there are 7 players remaining, including 1 slayer and 1 EC. Slayer now hits EC 1/6 of the time. EC strikes. If they miss the slayer (4/5 of the time) we are back to the[/color] above scenario, where the students pass and win 5/6 of the time. If EC hits the slayer, (1/5 of the time) there are now 5 players remaining. EC survives the next 2 lynchings with probability 4/5 X 2/3 = 8/15.

Composite odds of EC victory in Option 2 are (4/5 X 1/6) + (1/5 X 8/15) 18/75 = 24.0%.

As you can see 24% < 27.8% (EC chances if students don't pass). So the EC's odds of winning are less if the students pass in the scenario named, i.e. 7 remaining players including 1 EC and the slayer.

The odds in favor of passing at 7 if both EC and slayer are alive are better than the above calculation. The 24% is EC's chances if the slayer misses at 7. Since EC must survive that staking, EC's chances are only 5/6 * 24% = 20%, and 20% is significantly less than 27.8%.

That's different enough that I think we'll want to pass at 7, then again at 5, assuming in both cases that EC and slayer are alive.

As 4Sigma noted, the probabilities become different since we have some idea about who is more likely and less likely to be EC, and EC may have some idea about who is more and less likely to be the slayer. I doubt these would change the odds enough to justify hitting at 7 or 5 when we have a slayer.

In a different post I suggested that we might want to claim that the original slayer is alive and we know who it is. If we do that, we will want to redo these calculations to take into account that in our and slayer hits we can eliminate original slayer and the current slayer targets as long as both survive. [Not for our considerations as to whether to pass, but in presenting it to the students, if we want the claim about the original slayer to be believable.]

[Gandalf]

As to the KP charade, I just realized: cubedbee is a ghost, so KP is allowed to PM him, isn't he? Would we be better off having KP converse with cubedbee in PM, where cubedbee could grill him in greater detail but where we hope KP could allay his suspicions?

We would have to rely on KP being able to think quickly, not seeking our advice before replying. I'm willing to count on KP to pull it off. He should keep us aware of anything that happens, so we don't say/do anything contradictory.

One detail that we would need to have a single story about is the original slayer. KP got the EC to hit Rocky, didn't he? Is it best to have the EC think Rocky was the original slayer; or that the two current KP suggestions are the current slayer and the original slayer? Rocky = original slayer is probably the better lie, since EC can't know that one is false, while if one of the current KP suggestions is EC, then they'll know those two are not original and current slayer.

[Casper]

Macroman here:

As to the KP charade, I just realized: cubedbee is a ghost, so KP is allowed to PM him, isn't he? Would we be better off having KP converse with cubedbee in PM, where cubedbee could grill him in greater detail but where we hope KP could allay his suspicions?

We would have to rely on KP being able to think quickly, not seeking our advice before replying. I'm willing to count on KP to pull it off. He should keep us aware of anything that happens, so we don't say/do anything contradictory.

One detail that we would need to have a single story about is the original slayer. KP got the EC to hit Rocky, didn't he? Is it best to have the EC think Rocky was the original slayer; or that the two current KP suggestions are the current slayer and the original slayer? Rocky = original slayer is probably the better lie, since EC can't know that one is false, while if one of the current KP suggestions is EC, then they'll know those two are not original and current slayer.

Won't this look suspicious now? I'd be very wary of such an approach. Of course we wouldn't have much to lose as long as KP is careful. One thing we could lose is the possibility that EC would act on the earlier info. That seems remote as the EC has definitely been trying to hit the slayer, which was the supposed original bait.

I think the EC has written off KP as a double agent already or they would have acted on the info by now. I think we have effectively told the EC that these people are not the slayer.

[Gandalf]

If we're convinced that they aren't going to hit KP's suggestions, then we need to start hitting those targets ourselves.

Is there anything that would make a PMed contact from KP to cubedbee more believable?

Suppose, for example, that snafu or someone was to taunt the EC with the claim that the original slayer is alive, well, and being protected by the ghosts. After which KP PMs cubedbee "Snafu is lying. Rocky was the original slayer." KP could also make a public post saying the GC is just trying to look good, with a claim that they know no one can disprove.

[4sigma]

KP did not cause EC to hit Rocky. When the KP opportunity arose, the thread moved extremely fast. We had just about decided to have KP tell EC to hit Rocky, when they did it anyway. So we quickly had to reverse course and give them someone else. That is when we came up with Will Durant and Anonymouse. They were the only ones we were willing to hit at the time that we knew were not the slayer.

I agree with the observation that EC has apparently decided that this is a hoax, since they haven't hit either Will Durant or Anonymouse yet. I think this is a fairly strong argument for hitting Will Durant and Anonymouse. If one of them is EC, it is obvious that the hoax would fail. If EC is anyone else, the hoax might well have succeeded. The failure therefore tends to suggest Will Durant or Anonymouse as likely EC.

Even if Will and Anon are not EC, EC has apparently decided that they aren't the slayer. I say this because their last 3 hits appear to have been attempts to hit the slayer, and none of them were Will Durant or Anonymouse. Therefore it will help protect the slayer if we hit Will and/or Anon, since this leaves EC a larger field of people that could possibly be the slayer, hence they are more likely to miss.

At this point, I believe we continue the KP hoax until Will and Anon are dead. And I recommend that we now kill them off as soon as possible. If it turns out that one of them is EC, KP gets full credit for creating the opportunity that led to identifying them.

I am updating my summary list to raise Will Durant and Anonymouse as possible EC, due to the above reasoning.

[Rocky]

I prefer Anon to WD right now. We should hit Anon and suggest strongly WD (as others have mentioned). Nothing new or insightful here (5 Samuel Adams bars much insight), just agreeing with earlier analyses from other ghosts. As the poker punks would say...I'm "all in" for hitting Anon this round (unless he drops from the top 3, in which case I'd go for WD/EK/Jadzia in that order)...

Happy 4th of July to all that are celebrating it!

[4sigma]

Following is my strategy/tactics post from above, updated to include various assumptions about our ability to target EC, and EC’s ability to target the slayer.

In summary, the more we think we know about who is really innocent, the more the odds favor us lynching someone in round 10 rather than passing. If we are confident of two innocents besides the slayer, it is pretty clearly right to lynch. If we are confident of one innocent besides the slayer, it is close to a toss-up whether we should lynch or pass, but the odds appear to slightly favor passing.

I invite anyone who is as much of a numbers nerd as myself to check my math or challenge my assumptions. :D

Note that if we can reach the scenario where it is our turn to lynch and only 3 players remain, including the slayer, then the students are guaranteed to win.

Because of this, if we are in the scenario where it is our turn to lynch and only 5 players remain, including the slayer, then we should pass and let the slayer strike. The odds work as follows:

Option 1: "Don't pass." We lynch someone, and the slayer hits someone. We hit EC 50% of the time if we know of nobody that's innocent. Or if we are sure of one player that is innocent, our odds of hitting EC in this round are 2 out of 3.

If we miss, it is EC's turn. 3 students remain, including EC. If EC hits the slayer they win, otherwise the slayer slays them and wins. EC’s odds of randomly hitting the slayer are 1/2, but they likely will have some better information and perhaps hit the slayer 2/3 of the time in reality. Overall odds of the students winning under this option: 2/3 if we pick randomly, 7/9 if we know of one innocent.

Option 2: We pass. The Slayer hits EC 1/4 of the time if he hits randomly. If we know of 1 innocent, the slayer hits EC 1/3 of the time. EC now strikes with 4 players remaining, including EC. Odds of EC hitting the slayer are 1/3 if they pick randomly, but are more likely to be closer to 1/2 in reality.

If EC misses the slayer the 3-player scenario above now exists, i.e. students win. If EC hits the slayer the students still have a chance of hitting EC with their last lynching. Odds if choosing randomly are 1/3 if there are no innocents known, or 1/2 if we know of one innocent.

Odds of the students winning under Option 2 are 3/4 if we pick randomly, or 5/6 if we know of one innocent besides the slayer among the initial group of 5.


Since our odds of winning are always better under Option 2 than under Option 1, we should prefer it always if we are in this situation.

Note that if we are certain of 2 innocents in the initial group of 5 in addition to the slayer, our odds of winning are 100% under either option.

Backing up one more round, suppose there are 7 players, including the slayer. For the moment, I’ll assume that we know two students other than the slayer are innocent, in the initial group of seven.

Option 1: We lynch someone, and the Slayer hits someone. These combined get EC 50% the time. Game over. If we miss, EC now strikes.

Case 1a) If EC hits the slayer, there are now 4 players remaining (including EC) and we know 2 of them are innocent. The students clearly pass. Assume EC takes out someone we know is innocent, then we choose one of the 2 non-innocents for lynching, and our odds of hitting EC with our final lynching are 50%. In reality, EC may occasionally hit someone that we did not know was innocent, so our odds here may be closer to 55%, but probably not much better.

Case 1b) If EC misses the slayer, there are 4 players remaining, one of which is the slayer and at least one of which we know is innocent. Between our lynching and the slayer, we are guaranteed to hit EC 100%.

In order for EC to win, they have to survive our initial 50% chance of hitting them in Round 10, as well as survive either Case 1a/1b, depending on whether they hit the slayer.

EC's chances of hitting the slayer if they choose randomly is 1/4. In reality, they probably have better information and hit the slayer perhaps 40% of the time, which makes case 1a/1b composite to a 82% chance of us hitting EC even if we miss in round 10. So overall odds for the students to win is 91%.

Testing the sensitivity of how often EC hits the slayer -- if they hit the slayer 100% of the time, our odds drop to 77.5%. If they hit the slayer only 25% of the time, our odds improve to 94.875%

Option 2: We pass when there are 7 players remaining, including 1 slayer and 1 EC. Slayer now hits EC 25% of the time. EC strikes.

Case 2a) If EC hits the slayer, there are now 5 players remaining, but we know 2 of them are innocent. We hit EC 2/3 of the time, assuming that EC hits someone we know is innocent when there are 4 players remaining. Since there is perhaps a small chance that they don’t, we can say that overall the students hit EC 70% of the time in this case, assuming that EC survived the slayer’s initial strike.

Case 2b) If EC misses the slayer we are back to the above scenario, where the students pass and win 83.3% of the time if we know of one innocent, or 100% if we know of 2 innocents. Let’s assume EC usually hits someone that appears to be innocent if they miss the slayer, so the students win perhaps 85% of the time in this case, even if they missed in Round 10.

In order for EC to win, they have to survive the slayer’s 25% chance of hitting them in Round 10, as well as survive either case 2a/2b, depending on whether they hit the slayer.

EC's chances of hitting the slayer if they choose randomly are 1/5 in this scenario. Since they likely have better information, assume their odds of success are as good as 1/3. Then case 2a/2b composite to an 80% chance of hitting EC even if we miss in round 10, so the overall odds for students to win is 85%.

Testing the sensitivity of how often EC hits the slayer -- if they hit the slayer 100% of the time, our odds drop to 77.5%. If EC picks randomly, our odds improve to 86.5%.

Note that under these assumptions, our odds of winning are 77.5% / 91% / 94.875% under Option 1, compared to 77.5% / 85% / 86.5% under Option 2. Therefore if we are confident that 2 students other than the slayer are innocent, we should prefer to lynch in Round 10 rather than pass.

This is what the analysis looks like if there are 7 players remaining including the slayer, and if we know of only one innocent other than the slayer in the initial group of 7.

Option 1: We lynch someone, and the Slayer hits someone. These combined get EC 40% the time. Game over. If we miss, EC now strikes.

Case 1a) If EC hits the slayer, there are now 4 players remaining (including EC) and we know 1 of them is innocent. The students clearly pass. Our odds of hitting EC are 1/3 if they take our our known innocent, or 1/2 if they do not. Assume on average that we correctly hit EC 37.5% of the time with our last lynching in this case.

Case 1b) If EC misses the slayer, there are 4 players remaining, one of which is the slayer. If we also know one is innocent, we are guaranteed to hit EC 100%. If not, our chances of hitting EC are 2/3. Assume on average that we correctly hit EC 70% of the time in this case.

In order for EC to win, they have to survive our initial 40% chance of hitting them in Round 10, as well as survive either Case 1a/1b, depending on whether they hit the slayer.

EC's chances of hitting the slayer if they choose randomly is 1/4. In reality, they probably have better information and hit the slayer perhaps 40% of the time, which makes case 1a/1b composite to a 57% chance of us hitting EC even if we miss in round 10, so overall odds for the students to win is 74.2%.

Testing the sensitivity of how often EC hits the slayer -- if they hit the slayer 100% of the time, our odds drop to 62.5%. If EC picks randomly, our odds improve to 77.125%

Option 2: We pass when there are 7 players remaining, including 1 slayer and 1 EC. Slayer now hits EC 1/5 of the time. EC strikes.

Case 2a) If EC hits the slayer, there are now 5 players remaining, but we know 1 of them is innocent. We hit EC 1/2 of the time, assuming that EC hits someone we know is innocent when there are 4 players remaining. Since there is a chance that they don’t, let's say that overall the students hit EC 55% of the time in this case, assuming that EC survived the slayer’s initial strike.

Case 2b) If EC misses the slayer we are back to the above scenario, where the students pass and win 83.3% of the time if we know of one innocent, or 75% if we know of no innocents. EC has a 1/4 chance of hitting our last innocent if they pick randomly, but if they pick a bit better than randomly, our overall odds of winning in this case are perhaps 80%, even if we missed EC in Round 10.

In order for EC to win, they have to survive the slayer’s 20% chance of hitting them in Round 10, as well as survive either case 2a/2b, depending on whether they hit the slayer.

EC's chances of hitting the slayer if they choose randomly are 1/5 in this scenario. Since they likely have better information, assume their odds of success are as good as 1/3. Then case 2a/2b composite to a 71.67% chance of hitting EC even if we miss in round 10, so the overall odds for students to win is 77.33%.

Testing the sensitivity of how often EC hits the slayer -- if they hit the slayer 100% of the time, our odds drop to 64%. If EC chooses randomly, our odds rise to 80%.

If I've done the math right, the odds of success under Option 1 are 62.5% / 74.2% / 77.125% compared to 64%/ 77.33% / 80% under Option 2, depending on how well we think EC has the ability to target the slayer. Option 2 is therefore a slight favorite if we only have one student we are convinced is innocent.


Grateful acknowledgement is made to BC for his work in modelling the game, which contributed significantly to my awareness that this scenario could occur.

[4sigma]

kp has these thoughts regarding strategy:
From the time when we have 6 players left, I have worked out an end-game strategy that maximizes victory, assuming the Slayer still lives.

6 left: Slayer takes 2 shots - 1 by directing students.

5 left: Slayer takes 1 shot

Reasoning: We want the EC to have four players left on his turn. He will have a 1/3 chance at hitting the slayer. If he does, we have a 1/3 chance of hitting him back. This is a 1/9 chance of victory. If he misses the slayer (2/3), we automatically win as the Slayer calls the next two shots. So, if it is the ECs turn and there are 4 players left, we have 7/9 chance of winning (assuming the EC doesn't pass). If he does pass, at least we have earned ourselves an extra two shots.
I don't think your situations are possible. At the moment, before staking, there are 9. After staking and EC hit, there will be 7, not 6 or 5.

The tough decision is at 7, with slayer alive. Students can pass at 7, then (if slayer survives) at 5. Or students can hit at 7, and then (if slayer remains alive) it will be student turn at 4, where students should hit at 4 and slayer at 3.

Well, first of all, this is probably moot as I think WD is the EC. However, if he is not, and assuming the Slayer does not get hit, this is the scenario.

First of all, in order to maximize probability, I think the slayer should take all shots after this one. That leaves us with:

A) We take 2 shots: 1/3 chance of victory. Assuming we miss (2/3)
A1) EC hits slayer (1/3 chance). We must pass. EC kills another. Then we have a 1/3 chance of winning. Victory chance by this path - 2/27
A2) EC misses slayer. We take two shots. (2/3 Chance). Chance of winning by this path - 6/27.

Total chance of victory by path A = 17/27 or .6296.

B) We take 1 shot: 1/6 chance of victory. Assuming we miss (5/6)
B1) EC hits slayer (1/4 chance). We get 2 more shots. Chance of winning down this path is (1/5 + 4/5 * 1/3) = (7/15). Total win probabilty with this path = (7/64)
B2) EC misses slayer (3/4 chance). Slayer takes one shot. (1/5 chance)
If he misses, we have a 7/9 chance of victory. Victory chance assuming this path is 37/45. Total win probability by this path is 37/72.

Total chance of victory by path B = .7899.

I think that we should clearly only take one shot when there are 7 players left. edited to add my comment, below I think kp has made a typo in scenario A, and the probability should be 19/27 instead of 17/27, which comds out to 70.4%. Still, I think he is correct that if everyone is picking randomly, the students should prefer to pass rather than lynch this round. The question, however, is whether or not we should pass or lynch, based on what we think is our ability to correctly target EC, and what we think is EC's ability to target the slayer.

[Klaymen]

Dude, get a life.

[Gandalf]

Anonymouse, ghosts don't have lives. You'll catch on sooner or later.

[4sigma]

Dude, get a life. Anon, I would venture to guess that you've spent similar amounts of thought analyzing the best move in a chess position, or the best line of play on a bridge hand. If you start throwing stones, you may find that your glass house is in jeopardy. ;)

[Gandalf]

I'm still trying to evaluate whether passing at 7 is right, leaning toward yes. To reemphasize what I said in another thread, even if we present the "random" result, I think we should present results that say we are sure of one innocent student besides the slayer, which if asked we claim is the original slayer.

[Gandalf]

Items in red are changes by me; in blue are clarifications
First of all, in order to maximize probability, I think the slayer should take all shots after this one. That leaves us with:

A) We take 2 shots when 7 are left: 1/3 chance of victory. Assuming we miss (2/3)
A1) EC hits slayer (1/4 chance). We must pass. EC kills another. Then we have a 1/3 chance of winning. Victory chance by this path - 1/18
A2) EC misses slayer. We take two shots. (2/3 Chance). Chance of winning by this path - 1/9.

Total chance of victory by path A = .7222.

B) We take 1 shot when 7 are left: 1/6 chance of victory. Assuming we miss (5/6)
B1) EC hits slayer (1/5 chance). We get 2 more shots. Chance of winning down this path is (1/5 + 4/5 * 1/3) = (7/15). Total win probabilty with this path = (7/90)
B2) EC misses slayer (4/5 chance). Slayer takes one shot. (1/4 chance)
If he misses, we have a 7/9 chance of victory. Total win probability by this path is .5555.

Total chance of victory by path B = .80.

[ahow]

What if the EC passes as well after the slayer is gone? Is that plausible? Just so they can force our bad hand???

[Gandalf]

The rules say the EC will notify Mr. P of a victim each round, so they cannot pass.

The rules allow the students to pass, by failing to reach a majority in favor of lynching someone before the deadline.

The slayer cannot pass.

[Avi]

According to my preliminary calculations, we cannot reach the stae where only three players remain including the slayer at a lynching point IF we choose each round until there are 3.

The order is Student-->Slayer-->EC

We now have 7 left on a student turn.

Assuming that the students heed our advice and we can protect the slayer.

We will have 6 left entering the slayer turn, 5 for the EC turn, 4 for the student turn and 3 for the slayer turn.

At that point there is a 50/50 chance of winning as if the slayer hits a student, the game is over and the EC wins as there is no majority.

We will have to pass now AND next student turn to get 3 on a student turn.

The format would be

7 left entering the slayer turn, 6 for the EC turn, 5 for the student turn, 5 for the slayer turn, 4 for the EC turn, and then 3 for the students.

I have not completed my probabilistic analysis of whether or not the increase from 50% win to 100% win at that point is worth giving up the two attacks that may hit EC earlier.

I'll post my findings if I ever calculate any :dsmile:

[Gandalf]

KP asked me to post my calculations. Here's an attempt.

In the "Our Team" column, that represents either one or two tries, depending on whether we have a slayer (and would take two tries if we do). Each row represents a path leading to a student win. First is the description of the path, then the probability for each step. If the path involves fewer than 5 steps, steps to the right are probabilty 1.

Column "Product" is the product of the probabilities for the steps on that path. "Total" is the sum of all the paths.

The variations here assume: we hit at random, except we never hit the slayer and never hit the one or two other students we consider surely innocent (and assume those students are indeed innocent.) These assume (and this is certainly not necessarily valid) that the EC has a decent idea who are slayer and innocent. In particular, these that I am posting assume that giving the EC a smaller number of choices does not increase the chances that the EC hits the slayer.

In particular, what I am posting for the slayer + 2 innocent case assumes that the EC will hit one of those three at its first turn (each equally likely) and another at its second turn (now 50-50). That first turn assumption applies whether there are 6 or 5 total choices then. If the EC does have an accurate short list, 6 or 5 does not matter. If the EC does not, making them hit with 6 hurts their chances.

What I am posting for the slayer + 1 innocent case assumes that on its first turn the EC will hit one of those each with probability 1/3 at its first turn (again whether there are 6 or 5 choices), and will hit with probability ½ at its second. (EC hitting a student not known innocent is labeled “random” in the EC column.)

Example: one way we win with one innocent student, not passing, is (a) we and slayer miss; (b) EC hits the slayer; (c) we pass; (d) EC hits a known innocent; (e) we lynch EC. This is the the third row of numbers below. The associated probabilities are (a) = 2/5, since the 7 living students = 1 slayer + 1 known innocent + 5 possibles. (b) 1/3, the assumption discussed above. (c) 1; we know we pass. (d) ½, the assumption discussed above; (e) 1/3. There are three living students, none known to be innocent. Probability each of (a) thru (e) = product = 0.05

Under these particular assumptions, if we start with one known innocent plus the slayer, our chances are 82% if we try at 7; 77% if we pass. If we start with two known innocent plus the slayer, our chances are 92% if we try; 83% if we pass.

Starting with Slayer + 1 innocent

If we don’t pass

Our team EC Our team EC Student Product
hit
0.4000 1.0000 1.0000 1.0000 1.0000 0.4000
both miss slayer pass innocent hit
0.6000 0.3333 1.0000 0.5000 0.3333 0.0333
both miss slayer pass random hit
0.6000 0.3333 1.0000 0.5000 0.5000 0.0500
both miss innocent hit
0.6000 0.3333 0.6667 1.0000 1.0000 0.1333
both miss random hit
0.6000 0.3333 1.0000 1.0000 1.0000 0.2000

Total 0.8167

If we pass

Slayer EC Our team EC Our team Product
hit
0.2000 1.0000 1.0000 1.0000 1.0000 0.2000
miss slayer hit
0.8000 0.3333 0.2500 1.0000 1.0000 0.0667
miss innocent hit
0.8000 0.3333 0.2500 1.0000 1.0000 0.0667
miss random hit
0.8000 0.3333 0.3333 1.0000 1.0000 0.0889
miss slayer miss innocent hit
0.8000 0.3333 0.7500 0.5000 0.3333 0.0333
miss innocent miss slayer hit
0.8000 0.3333 0.7500 0.5000 0.3333 0.0333
miss random miss slayer hit
0.8000 0.3333 0.6667 0.5000 0.5000 0.0444
miss slayer miss random hit
0.8000 0.3333 0.7500 0.5000 0.5000 0.0500
miss innocent miss innocent hit
0.8000 0.3333 0.7500 0.0000 1.0000 0.0000
miss innocent miss random hit
0.8000 0.3333 0.7500 0.5000 1.0000 0.1000
miss random miss innocent hit
0.8000 0.3333 0.6667 0.5000 1.0000 0.0889
miss random miss random hit
0.8000 0.3333 0.6667 0.0000 1.0000 0.0000

Total 0.7722

Starting with Slayer + 2 innocent

If we don’t pass

Our team EC Our team EC Student Product
hit
0.5000 1.0000 1.0000 1.0000 1.0000 0.5000
both miss slayer pass innocent hit
0.5000 0.3333 1.0000 1.0000 0.5000 0.0833
both miss slayer pass random hit
0.5000 0.3333 1.0000 0.0000 1.0000 0.0000
both miss innocent hit
0.5000 0.6667 1.0000 1.0000 1.0000 0.3333
both miss random hit
0.5000 0.0000 1.0000 1.0000 1.0000 0.0000

Total 0.9167

If we pass

Slayer EC Our team EC Our team Product
hit
0.2500 1.0000 1.0000 1.0000 1.0000 0.2500
miss slayer hit
0.7500 0.3333 0.3333 1.0000 1.0000 0.0833
miss innocent hit
0.7500 0.6667 0.3333 1.0000 1.0000 0.1667
miss random hit
0.7500 0.0000 0.5000 1.0000 1.0000 0.0000
miss slayer miss innocent hit
0.7500 0.3333 0.6667 1.0000 0.5000 0.0833
miss innocent miss slayer hit
0.7500 0.6667 0.6667 0.5000 0.5000 0.0833
miss random miss slayer hit
0.7500 0.0000 0.5000 0.3333 1.0000 0.0000
miss slayer miss random hit
0.7500 0.3333 0.6667 0.0000 1.0000 0.0000
miss innocent miss innocent hit
0.7500 0.6667 0.6667 0.5000 1.0000 0.1667
miss innocent miss random hit
0.7500 0.6667 0.6667 0.0000 1.0000 0.0000
miss random miss innocent hit
0.7500 0.0000 0.5000 0.6667 1.0000 0.0000
miss random miss random hit
0.7500 0.0000 0.5000 0.0000 1.0000 0.0000

Total 0.8333
If we keep the assumption about our decisions (i.e., that one or two students beside the slayer is absolutely innocent and will never be hit by us), and have the EC make all its decisions randomly, the odds barely favor hitting with two known innocent but not one: 96.9% (not 97.5%) if we hit; 95.8% if we pass when there are 2 innocent plus the slayer. 86.9% (not 89.5%) if we hit; 87.5% (not 88.5%) if we pass when there is 1 innocent plus the slayer.

(edited to correct odds in the last paragraph)

[Avi]

Note that if we can reach the scenario where it is our turn to lynch and only 3 players remain, including the slayer, then the students are guaranteed to win.

Explain once again?

I can see this only if a) the slayer goes public at this point and b) the remaining student volunteers to be lynched.

[Gandalf]

Anyone so inclined is welcome to check the calcs. In addition to the possibility that some numeric entries in the tables are wrong, you should consider that I may have missed some paths that lead to a student win.

[Avi]

Anyone so inclined is welcome to check the calcs. In addition to the possibility that some numeric entries in the tables are wrong, you should consider that I may have missed some paths that lead to a student win.

I am calcualting a probabilty tree with PnP. Gimme a sec :D

[Gandalf]

Note that if we can reach the scenario where it is our turn to lynch and only 3 players remain, including the slayer, then the students are guaranteed to win.

Explain once again?

I can see this only if a) the slayer goes public at this point and b) the remaining student volunteers to be lynched.
Players left A,B,slayer. Ghosts say lynch A. At a minimum, B and slayer so vote. A is dead.

If A is not EC, it is slayer's turn. He slays B.

[Gandalf]

How sure are we that one or more particular students are innocent?
That is indeed a key question. It must be less than 100%. I suppose it tends to favor passing, because one important passing route to success is the EC missing the slayer twice. If that happens, we win even if we were totally wrong about who is innocent.

OTOH, despite the EC's apparent cluelessness so far, I'm not sure we want to put our fate so much in their hands. If we hit, then there is only one chance for them to affect our destiny (their try with 5 left). If we pass, they can hurt us with 6 left or with 4 left.

[Avi]

OK. Here is my Pen and paper analysis of the following 2 scenarios.

The first set has the assumptions that we do not skip any turns, and that we succesfully protect the slayer at each round of the student voting.

Let S = Slayer, T = Student, E = EC. The game obviously ends on a string ending with E. I wish

Round 1, Students turn. 1 EC, 1 Slayer, 5 Students

1/6 we hit EC, 5/6 the game continues. Prob Student win at 1 = 1/6

Round 2, Slayer turn. 1 EC, 1 Slayer, 4 Students

1/5 hit EC, 4/5 hit Student. Prob Student win at 2 = 5/6 * 1/5 = 1/6

Round 3, EC turn. 1 EC, 1 Slayer, 3 Students.

1/4 hit slayer, 3/4 hit student. (I wish I could post a tree diagram).

Assuming student hit.

Round 4, Student turn. 1 EC, 1 slayer, 2 Students.

1/3 we hit EC, 2/3 we hit student. Prob of hitting EC = 5/6 * 4/5 * 3/4 * 1/3 = 1/6

Round 5, Slayer turn. 1 EC 1 Slayer 1 student

1/2 we hit EC. 1/2 we hit student In which case EC wins.

Prob of Student winning = 5/6 * 4/5 * 3/4* 2/3 *1/2 = 1/6

Assuming Slayer hit in Round 3

Round 4, Student turn. 1 EC, 3 Students.

1/4 we hit EC, 3/4 we hit students. Prob (win) = 5/6 * 4/5 * 1/4 * 1/4 = 1/24

Round 5, EC turn. 1 EC, 2 Students.

100% hit student Game over as no majority left.

Total Prob of student win = 1/6 + 1/6 + 1/6 + 1/6 + 1/24 = 17/24

The second set of assumptions is that the students pass BOTH times in order to get to the 3 people left including slayer opportunity. This is even worse, so please bear with me.

Round 1, Students turn. 1 EC, 1 Slayer, 5 Students. SKIP

Round 2, Slayer turn. 1 EC, 1 Slayer, 5 Students.

1/6 hit EC, 5/6 student. Prob (win) = 1/6

Round 3, EC turn. 1 EC, 1 Slayer, 4 students.

1/5 hit slayer, 4/5 hit students.

Assuming slayer survives round 3.

Round 4, Student turn. 1 EC, 1 Slayer, 3 Students. SKIP

Round 5, Slayer turn. 1 EC, 1 Slayer 3 Students.

1/4 hit EC 3/4 hit Student.

Prob (win) = 5/6 * 4/5 * 1/4 = 1/6

Round 6, EC turn. 1 EC, 1 Slayer, 2 students.

1/3 hit slayer 2/3 hit student.

Assuming slayer survives round 3 AND round 6.

Round 7, Student turn. 1 EC, 1 Slayer, 1 Student.

STUDENTS WIN. Prob = 5/6 * 4/5 * 3/4 * 2/3 = 1/3

Assuming Slayer survives round 3 and dies round 6.

Round 7, Student turn. 1 EC, 2 students.

Prob hit EC = 1/3 Prob student = 2/3 <-- EC WINS.

Prob (win) = 5/6 * 4/5 * 3/4 * 1/3 * 1/3 = 1/18

Assuming slayer dies round 3.

Round 4, Student turn. 1 EC, 4 students.

Prob hit EC = 1/5 Prob student = 4/5

Prob (win) = 5/6 * 1/5 * 1/5 = 1/30

Round 5 EC turn. 1 EC, 3 students.
100% hit student.

Round 6, Student turn. 1 EC, 2 students.
Prob hit EC = 1/3. Prob student = 2/3 <-- EC WINS (no majority)

Prob (win) = 5/6 * 1/5 * 4/5 * 1/3 = 2/45

Total Prob of student win = 1/6 + 1/6 + 1/3 + 1/18 + 1/30 + 2/45 = 4/5

4/5 > 17/24 (80% vs. 70.83%)

I propose we have the students skip both turns, unless there is a flaw in my math.

[snafu]

Not following, please correct my reasoning.

Nevermind, I'll leave the mess below, which forgets to properly account for the EC hits. Forgive me as I'm on some good meds today.

OK, I worked through some probabilities under the assumption that the only party we know is NOT EC is the slayer and the only party EC knows is NOT Slayer is the EC. Here are the conditional probabilities that I come up with.

Without Passing
Students Round 10 = 1/6 Prob Student Win
Slayer Round 10 = 1/6 Prob Student Win = 5/6 Prob Student Miss x 1/5 Prob of Slayer Hit
EC Round 11 = 1/6 Prob Student Loss = 5/6 Prob Student miss x 4/5 Prob Slayer Miss x 1/4 Prob EC Hit
Student Round 11 = 1/6 Prob Student Win = 5/6 x 4/5 x 3/4 x 1/3
Slayer Round 11 = 1/6 Prob Student Win = 5/6 x 4/5 x 3/4 x 2/3 x 1/2
EC Round 12 = 1/6 Prob Student Loss = 5/6 x 4/5 x 3/4 x 2/3 x 1/2 x 1

Total Prob of Student Win = 4/6

With Passing
Student Round 10 = Pass
Slayer Round 10 = 1/6 Prob Student Win
EC Round 11 = 1/6 Prob Student Loss = 5/6 x 1/5
Student Round 11 = Pass
Slayer Round 11 = 1/6 Prob Student Win = 5/6 x 4/5 x 1/4
EC Round 12 = 1/6 Prob Student Loss = 5/6 x 4/5 x 3/4 x 1/3
Student/Slayer Round 12 = 2/6 Prob Student Win = 5/6 x 4/5 x 3/4 x 2/3 x 100%

Total Prob of Student Win = 4/6

If this is correct then the odds are the same under either scenario. Furthermore, the no passing scenario would seem to be preferred as EC might have better than random information and thus decrease the odds of ever getting to the Student/Slayer Round 12 scenario.

Please let me know what I'm missing. I agree that the students win 100% under the 3 player slayer still alive scenario but disagree that this is the best strategy.

[Gandalf]

Avi: In the first scenario, you have the students lynching at 4 with no slayer. They would always pass at 4 with no slayer. Thus the total probability for scenario 1 rises to 72.22%

There is no doubt whatsover that it is right to pass at 7 if no hits by us or them is better than random. The complication arises because we at least hope our hits are better than random, and fear (with no justification so far) that the EC hits may find the slayer more often than random.

[Mulan]

If we lynch someone this round, could we pass next round and get the same result?

[Gandalf]

Snafu: you are overlooking that the students may win without a slayer.

E.g., you assign p=1/6 to a student loss if the EC hits the slayer at first opportunity in both scenarios. In fact, students still have hope as we can lynch.

In the second scenario, you also assign p=1/6 to a student loss if the EC lynches the slayer at its second opportunity. The students would still have a lynch opportunity left.

[4sigma]

I'm not able to respond to everyone indivdually on this, as I have some actual work to do this afternoon.

snafu, if the slayer is hit, the students don't lose. They get to try and lynch EC still.

Avi, I think with Gandalf's adjustment, your math is correct. If we (and EC) are picking randomly, the correct action this round is to pass.

The question is how many students, if any, can we rule out as not EC. And how many students can EC rule out as the slayer.

The more likely EC is to hit the slayer, the more important it is for us to act this round and give them fewer shots.

The more likely we are to hit EC, the better are odds are by lynching rather than passing.

[Avi]

Avi: In the first scenario, you have the students lynching at 4 with no slayer. They would always pass at 4 with no slayer. Thus the total probability for scenario 1 rises to 72.22%

There is no doubt whatsover that it is right to pass at 7 if no hits by us or them is better than random. The complication arises because we at least hope our hits are better than random, and fear (with no justification so far) that the EC hits may find the slayer more often than random.

Correct. Still less than 80%.

According to further calculations, using the first student turn and skipping the second gives rise to only a 45.27% win chance.

Skipping the first and using all remaining (could be more than 1 if slayer is killed in round 3) gives 74.44% BTW, since in this scenario, there are 1 EC and 5 students, having both chances comes to 7/15 > 5/15 (1/3) of skipping both and waiting until there are only 3. This also depends on the slayer being allowed to skip the "student" turn when only 2 are left and act as a slayer. If the students lose at only 2 left even if one of them is the slayer, Prob(win) falls even further.

All in all, I think that skipping both student turns will maximize the chance of the EC dying.

[Gandalf]

If we lynch someone this round, could we pass next round and get the same result?
No. Passing with 4 left is a bad idea if EC will hit the slayer at random or anything more often than random.

Likewise passing now and hitting at 5 is not a good idea (assuming everything at random). It is better than hitting now, but not much.

[Avi]

If we lynch someone this round, could we pass next round and get the same result?According to my calculations, the chances drop precipitously to about 45%, becuase we end up with more scenarios where the EC gets to win by default.

Edit

Even with Mr. P's ruling, I get 67/90 for skipping the first and taking the second.

I suggest we try and convince the students that the best possibility of winning lies with skipping their turns.

These probabilities are only valid if we have no data about anyone other than the slayer, and that the EC has no data about anyone.

[4sigma]

Avi, take a look at Gandalf's and my calcs where we try and quantify the fact that neither we nor the EC is picking randomly. The less random things are, the more the odds favor us lynching this round. The more random things are, the more of the 7.8% advantage (80% - 72.2%) we retain by passing.

[Gandalf]

KP's PM to me (which I posted) is really a good way of seeing the main issue. One item in favor of the pass twice strategy is that it gives us a reasonable chance of reaching the 3 left, slayer alive, students turn, which guarantees us a win even if we are wrong about who is innocent.

The hit immediately strategy gives us a chance (better than the chance above) of reaching 4 alive, including slayer and at least one known innocent student, a guaranteed win if we are really sure that the innocent student is innocent.

The belief that certain students are innocent affects a lot of other paths, too, but that particular one is fairly critical to the odds.

[4sigma]

Gandalf's calculations assume we know with certainty that someone is innocent. We cannot be that sure. If we are wrong, we definitely lose. At this point, I am becoming more in favor of random choice. Certainly, I don't think we should abandon the mathematical choice in favor of our past assumptions.

I agree it's not possible to be certain whether someone is innocent. This is admittedly an assumption for the purposes of modeling. It is perhaps a better wording to say that some people are more likely to be EC than others, so our odds of hitting EC if we lynch someone now is better than 1/6. How much better is an assumption. 1/5 or 1/4 is perhaps not an unreasonable assumption.

I'll post your PM and my reply in the strategy thread. Thanks for all your contributions to this.

Instead of thinking of it as "one student we are sure is innocent" perhaps a better way to think is, "If we lynch someone, what are our chances of hitting EC?" Picking randomly, our chances are 16.67%. If we think we can do better than random, our chances maybe perhaps 20% or 25%. These are the actual probabilities that the model requires, not the "number of innocent students."

I am beginning to incline towards passing this round, though it is still in question.

[Gandalf]

My earlier odds calculation in the case that everything was random except that we knew one or two students were innocent was slightly off. That post has been updated. As before, the decision is extremely close: barely favoring hitting if we know 2 plus the slayer are innocent; barely favoring passing if we know 1 plus the slayer is innocent.

Just to emphasize that it depends on the EC's knowledge as well as ours: assume we will never hit our slayer, but are otherwise clueless. If the EC will hit our slayer right 50% of the time with 5 left (its only relevant decision if we lynch), we win 66.7%.

If we pass and the EC will hit the slayer 50% of the time at 6 and 50% (of the times we reach 4) at 4, we win 67.3% by passing. If we pass and the EC will hit the slayer 50% at 6 and 75% at 4, then we win only 62.15% by passing. Those are high hitting % for the EC, but we don't know how many names they may have eliminated.

[Macroman]

ursyjohn's post in the public forum:
:oops: I knew better than to get into this, but there is no more entertainment left other than to shoot the bull with urysohn. :shake:
:rofl:
subtle. Very subtle. But I don't think they're (he's? she's? depends on who you were taunting, I guess) going to pick up on it, EK.

I have (I hope) been authorized to post this PM:
I am the slayer.
I give my solemn word that this quote is not taken out of context. Now, there's always the chance I've been duped, but I believe it.
Why would the slayer out themselves, you ask? Good question. I'd say it's right there in front of you in 4sigma's post -- the students think it might be in their best interest to skip a turn anyway. Getting the slayer knocked off would pretty much do that, wouldn't it?
Or it could be a bunch of hooey and EK is using me to draw a hit away from the true slayer. Got me - I haven't been putting enough time into this game to figure it out for sure (I have it narrowed down to two - and EK has been at the top of the list for awhile).

Clearly if this post is believed and true we cannot reasonably pass. I think that only EK would play the slayer this way forsaking ghost forum protection as the slayer has.

[ahow]

ursyjohn's post in the public forum:
:oops: I knew better than to get into this, but there is no more entertainment left other than to shoot the bull with urysohn. :shake:
:rofl:
subtle. Very subtle. But I don't think they're (he's? she's? depends on who you were taunting, I guess) going to pick up on it, EK.

I have (I hope) been authorized to post this PM:
I am the slayer.
I give my solemn word that this quote is not taken out of context. Now, there's always the chance I've been duped, but I believe it.
Why would the slayer out themselves, you ask? Good question. I'd say it's right there in front of you in 4sigma's post -- the students think it might be in their best interest to skip a turn anyway. Getting the slayer knocked off would pretty much do that, wouldn't it?
Or it could be a bunch of hooey and EK is using me to draw a hit away from the true slayer. Got me - I haven't been putting enough time into this game to figure it out for sure (I have it narrowed down to two - and EK has been at the top of the list for awhile).

Clearly if this post is believed and true we cannot reasonably pass. I think that only EK would play the slayer this way forsaking ghost forum protection as the slayer has.Although maybe he can draw some fire from the Slayer this round if the EC buys it and they hit him...

[Macroman]

ahow,

What makes you think EK is not the slayer? And what makes you think that ursyjohn would do something to help us out? To me this seems to be EK's final power play to force our hand once more.

[ahow]

ahow,

What makes you think EK is not the slayer? And what makes you think that ursyjohn would do something to help us out? To me this seems to be EK's final power play to force our hand once more.I don't think that even EK would play this risky as the slayer. Be vocal? Yes. Be a thorn in the sides of every single player in the game? No. Do things just short of collude with the EC? So far, very close. Of course most of this is baseless but it has a good feel to it, in my opinion...

[Avi]

Uh folks, read this: http://actuary.ca/phpBB/viewtopic.php?p=577250&highlight=ek+hit+slayer#577250

[Macroman]

EC 101- If someone claims to be the slayer and is not hit by the EC at the first opportunity, they are almost certainly EC themselves.

Since EK had just recently posted the fact that he was conversing with ursyjohn the post is most probably made with EK's blessing. If it is otherwise EK will deny the assertion in short order.

[ahow]

Uh folks, read this: http://actuary.ca/phpBB/viewtopic.php?p=577250&highlight=ek+hit+slayer#577250Ok, so what if (to further protect his identity) the slayer, aka EK, decided to throw his name into a list of 5 people because that actually reduces the chance that we give the go ahead there. Note that all four of those others have become ghosts, while EK has not...

EDIT: Note that if he were the Slayer, he would have access to info such as who our preferences are via the code...

[Jables]

Is it true that we really don't know who the slayer is, and have only been getting the slayer's OK as to who from our pool of nominees we are safe to hit? Or was the slayer's identity revealed at some time and we're just talking about the slayer like this now?

If we don't know the slayer's identity, which of the remaining students have been nominated for lynching and cleared by the slayer?

If the slayer is keeping his/her identity a secret, might I ask why? Is it just a personal challenge to try and make it through the game without revealing themselves? I didn't read through all of the posts from the original ghost forum last game, but did JT make herself known to be the slayer last time?

[Macroman]

....If the slayer is keeping his/her identity a secret, might I ask why? Is it just a personal challenge to try and make it through the game without revealing themselves? I didn't read through all of the posts from the original ghost forum last game, but did JT make herself known to be the slayer last time?

That's how I view it. Of course if the slayer believes the GF to be insecure as EK has stated, any slayer might choose to remain unknown. Occaissonally suggesting oneself amounst a large list might not be a bad misdirection tactic especially if you have final approval of the ultimate nominee via the coded messages.

[Werewolf]

At this stage of the game, with the ghosts posting that the students should delay while strategy is discussed, is it suspicious that some people voted early anyway?

[Mulan]

I don't think EK is the slayer or EC - and if you think about it, neither do you all. Besides we could ask the slayer to reveal him/she self, or at least ask him/her if he/she is EK.

Let's not get distracted folks.

It would be good if we acted all upset that EK has revealed his slayer-hood this way, so the EC will be sure to hit him and not the slayer.

[Gandalf]

I think it is best if we ignore the statement that EK is the slayer. Let him deal with it as he sees fit. It would not surprise me in the least if he asked urysohn to post the PM.

If we react, I think EC will be even more sure he is not the slayer. (Like Mulan, I think he is not.)

[Gandalf]

At this stage of the game, with the ghosts posting that the students should delay while strategy is discussed, is it suspicious that some people voted early anyway?
Not at all suspicious. 4Sigma and I each said it was fine for them to vote, so long as they don't get more than two votes for anyone. It's good to have any info whatever they want to provide.

[Rocky]

We should ignore EK's assertion to urysohn. If he wants to die, it's his prerogative. But, I do not think that the EC will believe him for one minute.

[4sigma]

Those of you who have noted that the slayer's ID is not posted on the forum, you are correct. The slayer has left it to my discretion whether or not to post their identity, recommending that I do not. For the moment, I have chosen to honor that recommendation.

The end of the game is near enough now that all will be revealed fairly shortly. If the slayer is still alive when there are 4 or fewer players remaining, I will make the slayer's identity public in that round, since when it is our turn, all cards go onto the table.


In the meantime, here is a mystery that I can help clear up for everyone. The posts between myself and EK in the public thread 16 hours ago were regarding an alliance that EK formed with several other students near the start of the game. It appears that the alliance may have had some small effect in swinging votes, but not much.

All the members of that alliance but two are now dead and were proven to be innocent students. I will leave it up to them whether they wish to discuss their involvement, since their death has cleared them of any possible guilt. At this point, the alliance has dwindled to an alliance between EK and RSF.

The alliance stirkes me quite honestly as an effort by a bored EK to liven up a game where the living have less power than the dead. It is still my opinion that EK is less likely than average to be EC. (Less likely than the mean -- I have him at the median in my current list.) This is not based particularly on the alliance data, except that it confirms his profile as someone who is trying to play the game to maximize his personal entertainment.

My opinion of RSF has not changed much except that it is now apparent that he has been more active than I had been aware, which lowers his suspiciousness very marginally in my book.

[Macroman]

....
In the meantime, here is a mystery that I can help clear up for everyone. The posts between myself and EK in the public thread 16 hours ago were regarding an alliance that EK formed with several other students near the start of the game. It appears that the alliance may have had some small effect in swinging votes, but not much.

All the members of that alliance but two are now dead and were proven to be innocent students. I will leave it up to them whether they wish to discuss their involvement, since their death has cleared them of any possible guilt. At this point, the alliance has dwindled to an alliance between EK and RSF

....

My opinion of RSF has not changed much except that it is now apparent that he has been more active than I had been aware, which lowers his suspiciousness very marginally in my book.

To me this revelation makes RSF look all the more suspicious. EK's suspicious actions are explained as an attempt to aid an ally. EK's recurring attacks against All Clear deflect attention that should go to RSF. He also makes himself look suspicious to deflect attention.

I think if we're going after the EC we should take both shots RSF and Traina rather than pass. If the objective is still to protect the slayer I think we should take just the single shot at WD.

edited to add... I forgot the obvious, how did you know about this alliance, except that you were a member. You should be able to tell us more, I think.

[Casper]

This is J.T. everyone.... :wave:

Sorry to be joining you :cry: , but here is my input. I was a member of the alliance, and EK gave 4sigma the info to see what we had put together. It was an opportunity to spice up the game more. There are a few of us (I won't say who, since I don't have the authority to out other alliance members) that suspected RSF. He has always been the most suspicious alliance member in my opinion.

The main purpose of this alliance was not only to protect each other (which we really didn't have to do, since the EC or slayer killed most of us), but to just spice up the game. I'm not sure that it accomplished that, but at least it kept a little extra variety in the game.

[Leela]

Breaking news from Mr. P .....

The Slayer is allowed to pass. So, we can get to the 3 people left scenario by having the Slayer pass this round and either the Slayer or students pass next round.

The Slayer has said s/he will pass if we want him/her to. We'll need to come to a decision fairly soon though.

[Leela]

Now, if the Slayer can pass, that means the EC can also. Just something to think about...

[Tim><]

I don't think we need to come to any decision soon. WD hasn't posted a death scene yet.

[Leela]

The Slayer only has 24 hours after Will posts his death scene to give Mr. P a name or pass. Considering how long it took us to reach a conclusion about whether the students should pass this round, we should at least start discussing.

[Mulan]

I think the slayer should slay while she/he is still alive. Or leave it up to the slayer - do we even have a code that says "skip this turn"??

I thought we'd already settled on RSF, thinking that the EC will execute Traina.

[Leela]

I think the slayer should slay while she/he is still alive. Or leave it up to the slayer - do we even have a code that says "skip this turn"??

The Slayer is the one who actually brought this up and asked Mr. P about it. The Slayer has told us to just say we want him/her to pass in the publice thread without coding it if we decide that's the best strategy.


I thought we'd already settled on RSF, thinking that the EC will execute Traina.

That was when we thought the Slayer and EC couldn't pass. Now, that we know they can, it opens up new possibilities.

[ahow]

I don't like the idea of the slayer passing. It would be especially bad if the slayer passed and then was slain by the EC...

[Leela]

I tend to agree with ahow. Plus, if the Slayer passes, the EC will know they can pass. I don't know if they would think about it if the Slayer doesn't pass.

[Rocky]

The slayer should strike, IMO.

I thought that RSF would be a good choice based on our discussions in this forum.

[Gandalf]

I wonder if the EC can pass? Probably not, otherwise the game could end in a total stalemate.

[The following are under the assumption that WD was not EC. If he was, all is irrelevant. All assume we never kill slayer.]

If I did the calculations correctly, the odds in favor of passing if all chances of hits (us hitting EC; them hitting slayer) are even more strongly in favor of passing now: 76% if we pass, 66.7% if slayer stakes. Someone should try to verify this.

What is clear is that IF the EC is definitely going to hit the slayer next, it doesn't matter who goes first.

If you assume we are only random, but EC has a 50% chance at its first opportunity and a 50% chance at its second (thus 25% that both EC chances would fail), then it is still barely in favor of pass (60% hit; 60.8% pass). EC 50% at first shot, but random thereafter gives 60% hit; 65% pass.

Let the EC hit at 50% then 50%, but assume we have one non-slayer we are sure as innocent, and pass is better by 70.9 to 67.5%, I think.

Here are the paths for that last one, if someone want to try to check (there are also assumptions about whether if the EC misses the slayer it will hit the person we know is innocent; assumes they do so very slightly better than random).
DON'T PASS

Our team EC Our team EC Student Product
hit
0.2500 1.0000 1.0000 1.0000 1.0000 0.2500
both miss slayer pass innocent hit
0.7500 0.5000 1.0000 0.2500 0.3333 0.0313
both miss slayer pass random hit
0.7500 0.5000 1.0000 0.7500 0.3333 0.0938
both miss innocent hit
0.7500 0.2000 1.0000 1.0000 1.0000 0.1500
both miss random hit
0.7500 0.3000 0.6667 1.0000 1.0000 0.1500

Total 0.6750

PASS

Slayer EC Our team EC Our team Product
hit
0.0000 1.0000 1.0000 1.0000 1.0000 0.0000
miss slayer hit
1.0000 0.5000 0.2500 1.0000 1.0000 0.1250
miss innocent hit
1.0000 0.2000 0.2500 1.0000 1.0000 0.0500
miss random hit
1.0000 0.3000 0.5000 1.0000 1.0000 0.1500
miss slayer miss innocent hit
1.0000 0.5000 0.7500 0.2500 0.3333 0.0313
miss innocent miss slayer hit
1.0000 0.2000 0.7500 0.5000 0.3333 0.0250
miss random miss slayer hit
1.0000 0.3000 0.5000 0.5000 0.5000 0.0375
miss slayer miss random hit
1.0000 0.5000 0.7500 0.7500 0.5000 0.1406
miss innocent miss innocent hit
1.0000 0.2000 0.7500 0.0000 1.0000 0.0000
miss innocent miss random hit
1.0000 0.2000 0.7500 0.5000 1.0000 0.0750
miss random miss innocent hit
1.0000 0.3000 0.5000 0.2500 1.0000 0.0375
miss random miss random hit
1.0000 0.3000 0.5000 0.2500 1.0000 0.0375

Total 0.7094
I'm for passing. Note however, that this may increase the chances EK gets the stapler if we have the students rather than slayer hit RSF next round. If RSF is indeed guilty, I think EK has earned the stapler.

[Leela]

This is the message I received from Mr. P

As I told the slayer:

Yes. Since there is, more or less, a 24 hour deadline to slay, failure to do so would mean you forfeit your turn. By the same token, rather than waste time, I will allow the EC_Slayer ID to post "PASS" in the game thread.

I have asked him if this means the EC can pass also. I believe there is a 24 deadline for them also. So, I think they would be able to pass, but they won't know that unless they ask Mr. P or if we let the Slayer pass.

[Leela]

Here's the response from Mr. P. (Boy, he's quick.)

As I told the slayer:

Yes. Since there is, more or less, a 24 hour deadline to slay, failure to do so would mean you forfeit your turn. By the same token, rather than waste time, I will allow the EC_Slayer ID to post "PASS" in the game thread.

Does this mean the EC can pass also?

Yes, it does.

But if everybody just keeps passing, I may have to kill somebody.

:D

[Gandalf]

We would love to have the slayer pass and then the EC pass. Having 6 left and it be the student lynch turn is a great situation for us.

[Mr. Penguin]

Here's the response from Mr. P. (Boy, he's quick.)


I'm refreshing constantly today, waiting on Part 4 results.

Damn EC!!!!

:swear:

[Gandalf]

Good luck on Course 4, Mr. P.

If you get good news, will you be so happy that you tell us who the EC is?

[Leela]

So, what do we think EC's take on EK's recent "confession" will be? Will the believe him or rule him out as the Slayer?

I think this would cause them to rule him out as the Slayer. So, I am leaning towards passing or getting EK.

[Gandalf]

I suspect the EC will rule him out, but am not sure they will. I think it would be terrible move on our part to hit him, especially without passing first. We may have only 2 shots left. I don't want to risk using one on someone who is not in the top two to be EC.

I like passing, then lynch RSF, then slayer passes again. Or passing, then lynch Traina, then slayer passes again.

[Rocky]

We got HIM!

I think...

[4sigma]

A bit ambiguous even in death, isn't he?

Still, it looks pretty good. I'm ready to party! :party:

[Macroman]

Macroman here:

As to the KP charade, I just realized: cubedbee is a ghost, so KP is allowed to PM him, isn't he? Would we be better off having KP converse with cubedbee in PM, where cubedbee could grill him in greater detail but where we hope KP could allay his suspicions?

We would have to rely on KP being able to think quickly, not seeking our advice before replying. I'm willing to count on KP to pull it off. He should keep us aware of anything that happens, so we don't say/do anything contradictory.

One detail that we would need to have a single story about is the original slayer. KP got the EC to hit Rocky, didn't he? Is it best to have the EC think Rocky was the original slayer; or that the two current KP suggestions are the current slayer and the original slayer? Rocky = original slayer is probably the better lie, since EC can't know that one is false, while if one of the current KP suggestions is EC, then they'll know those two are not original and current slayer.

Won't this look suspicious now? I'd be very wary of such an approach. Of course we wouldn't have much to lose as long as KP is careful. One thing we could lose is the possibility that EC would act on the earlier info. That seems remote as the EC has definitely been trying to hit the slayer, which was the supposed original bait.

I think the EC has written off KP as a double agent already or they would have acted on the info by now. I think we have effectively told the EC that these people are not the slayer.

I'd just like to point out that it was I that initially pricked the KP hoax balloon which ultimately led to the EC's demise. I'm sure the other ghosts would soon have come to the same conclusion, but I got it started shortly after arriving.

[Tim><]

Macroman here:

As to the KP charade, I just realized: cubedbee is a ghost, so KP is allowed to PM him, isn't he? Would we be better off having KP converse with cubedbee in PM, where cubedbee could grill him in greater detail but where we hope KP could allay his suspicions?

We would have to rely on KP being able to think quickly, not seeking our advice before replying. I'm willing to count on KP to pull it off. He should keep us aware of anything that happens, so we don't say/do anything contradictory.

One detail that we would need to have a single story about is the original slayer. KP got the EC to hit Rocky, didn't he? Is it best to have the EC think Rocky was the original slayer; or that the two current KP suggestions are the current slayer and the original slayer? Rocky = original slayer is probably the better lie, since EC can't know that one is false, while if one of the current KP suggestions is EC, then they'll know those two are not original and current slayer.

Won't this look suspicious now? I'd be very wary of such an approach. Of course we wouldn't have much to lose as long as KP is careful. One thing we could lose is the possibility that EC would act on the earlier info. That seems remote as the EC has definitely been trying to hit the slayer, which was the supposed original bait.

I think the EC has written off KP as a double agent already or they would have acted on the info by now. I think we have effectively told the EC that these people are not the slayer.

I'd just like to point out that it was I that initially pricked the KP hoax balloon which ultimately led to the EC's demise. I'm sure the other ghosts would soon have come to the same conclusion, but I got it started shortly after arriving.

I don't get any credit?

[Macroman]

....
I don't get any credit?

You know I would think so, you ran a masterful snojob....did you come out and say that the hoax had failed?

[4sigma]

Macroman here:

As to the KP charade, I just realized: cubedbee is a ghost, so KP is allowed to PM him, isn't he? Would we be better off having KP converse with cubedbee in PM, where cubedbee could grill him in greater detail but where we hope KP could allay his suspicions?

We would have to rely on KP being able to think quickly, not seeking our advice before replying. I'm willing to count on KP to pull it off. He should keep us aware of anything that happens, so we don't say/do anything contradictory.

One detail that we would need to have a single story about is the original slayer. KP got the EC to hit Rocky, didn't he? Is it best to have the EC think Rocky was the original slayer; or that the two current KP suggestions are the current slayer and the original slayer? Rocky = original slayer is probably the better lie, since EC can't know that one is false, while if one of the current KP suggestions is EC, then they'll know those two are not original and current slayer.

Won't this look suspicious now? I'd be very wary of such an approach. Of course we wouldn't have much to lose as long as KP is careful. One thing we could lose is the possibility that EC would act on the earlier info. That seems remote as the EC has definitely been trying to hit the slayer, which was the supposed original bait.

I think the EC has written off KP as a double agent already or they would have acted on the info by now. I think we have effectively told the EC that these people are not the slayer.

I'd just like to point out that it was I that initially pricked the KP hoax balloon which ultimately led to the EC's demise. I'm sure the other ghosts would soon have come to the same conclusion, but I got it started shortly after arriving.

GMTA. I did mention this before your demise, Macroman, though perhaps you had not read it. This is my comment in the Round 8 thread, when we ultimately chose to lynch BC instead of you.

EC is clearly trying to hit the slayer with their last 2 picks. They've been hitting the quiet folks (Maine-iac, Mulan).

I find it interesting that EC hasn't targeted the 2 "slayer" names we gave them in the KP hoax. I believe this means that they have seen through the hoax. It is possible that we didn't execute it well enough, but it is also quite possible that the hoax failed because either Anonymouse or Will Durant is EC2 or the sleeper. Because of this, Anonymouse and Will Durant are now moving up my target list, particularly as the sleeper.

I am warming to getting Will Durant nominated soon, since he does have several indications as EC2 and would also explain why EC has not bitten on the KP hoax. Gandalf, how compelling do you think is the evidence for his innocence?

KP also clearly gets full credit for the excellent execution.

[Tim><]

Macroman here:

As to the KP charade, I just realized: cubedbee is a ghost, so KP is allowed to PM him, isn't he? Would we be better off having KP converse with cubedbee in PM, where cubedbee could grill him in greater detail but where we hope KP could allay his suspicions?

We would have to rely on KP being able to think quickly, not seeking our advice before replying. I'm willing to count on KP to pull it off. He should keep us aware of anything that happens, so we don't say/do anything contradictory.

One detail that we would need to have a single story about is the original slayer. KP got the EC to hit Rocky, didn't he? Is it best to have the EC think Rocky was the original slayer; or that the two current KP suggestions are the current slayer and the original slayer? Rocky = original slayer is probably the better lie, since EC can't know that one is false, while if one of the current KP suggestions is EC, then they'll know those two are not original and current slayer.

Won't this look suspicious now? I'd be very wary of such an approach. Of course we wouldn't have much to lose as long as KP is careful. One thing we could lose is the possibility that EC would act on the earlier info. That seems remote as the EC has definitely been trying to hit the slayer, which was the supposed original bait.

I think the EC has written off KP as a double agent already or they would have acted on the info by now. I think we have effectively told the EC that these people are not the slayer.

I'd just like to point out that it was I that initially pricked the KP hoax balloon which ultimately led to the EC's demise. I'm sure the other ghosts would soon have come to the same conclusion, but I got it started shortly after arriving.

GMTA. I did mention this before your demise, Macroman, though perhaps you had not read it. This is my comment in the Round 8 thread, when we ultimately chose to lynch BC instead of you.

EC is clearly trying to hit the slayer with their last 2 picks. They've been hitting the quiet folks (Maine-iac, Mulan).

I find it interesting that EC hasn't targeted the 2 "slayer" names we gave them in the KP hoax. I believe this means that they have seen through the hoax. It is possible that we didn't execute it well enough, but it is also quite possible that the hoax failed because either Anonymouse or Will Durant is EC2 or the sleeper. Because of this, Anonymouse and Will Durant are now moving up my target list, particularly as the sleeper.

I am warming to getting Will Durant nominated soon, since he does have several indications as EC2 and would also explain why EC has not bitten on the KP hoax. Gandalf, how compelling do you think is the evidence for his innocence?

KP also clearly gets full credit for the excellent execution.

Well, I can't take full credit. I had excellent guidance. :notworth: