TWG XL: Mall Madness

[iggymatrixcounter]

Re: TWG XL: Mall Madness

I haven't been around long enough to really pick up on the dominant strategy.

This is your 5th game and I know you've been a wolf at least once and a vigilante another time. You know how the game is played and ignorance is no longer an excuse for you. It's a simple concept once you've played a few games.

[FoJaR]

Re: TWG XL: Mall Madness

No I think kagome should make bad picks while we lynch off humans....
Besides, define a 'good' seer pick. They're all based off gut feelings and minute suspicions.

Also:

Not sure if this is the same rule but last game where there was a SK, the game still went on if the SK was alive and wolves > humans. So in theory that would buy us another day of lynching and SKing if that were the case. Having the SK alive with a bunch of wolves helps us out. (7 players: 4 wolves + 1 SK + 2 humans = GAME ON!) I'm not sure if the humans could win at that point but it would be a game between the wolves and SK, but still XD.

what i mean is, make a pool of wolf suspects, but dont try to lynch them. get kag to seer from the pool. we need to try to lynch the SK as soon as possible.

[sertman]

Re: TWG XL: Mall Madness

This is your 5th game and I know you've been a wolf at least once and a vigilante another time. You know how the game is played and ignorance is no longer an excuse for you. It's a simple concept once you've played a few games.

Considering one of the games I was killed night 1, another game i was gone night 3, my first game (when I was a wolf) was played so badly by the humans nothing good came out of it, and the game where I was a vigilante was a completely ****ed up game with weird rules and no real logic attributed it, I don't want to say that I have a legit "ignorance" excuse, but at this point it's the best I can give you.

This is the first game I've played in where the humans are ****ed right off the bat, and theres a 3rd party thrown into the mix. I'm not asking for sanctuary, but what's better: opening my mouth and being wrong or keeping my mouth shut and not adding to the game?

[Afrobean]

Re: TWG XL: Mall Madness

This is the first game I've played in where the humans are ****ed right off the bat...

wait a second...

my first game (when I was a wolf) was played so badly by the humans nothing good came out of it

So... your first game you were a wolf, and in that first game the humans were easily defeated? I'd say we're on a straight track down to a human loss thanks partly to broken role numbers and partly to ****ty luck of the night 1 kill.

And you statistics people...

ap: When I read your post where you assigned numerical value to certain events, I was like "is he seriously being so subjective and trying to use that in statistics to support his claim?" Is that what you meant by "uncomprehendable"?

rioting: Don't throw out random statistics that help nothing, and in many instances are quite wrong. As for showing things from the wolf's perspective: As someone else has already said-- bad idea. It's one thing to think from their perspective to deduce their actions thus far, or to get into their heads as a blue trying to find them using your power, but giving random stats about chances of reds hitting a blue... what the ****. Anyway, most of ANYONE's statistics are nearly worthless because they're not drawing a random lot. They're using logic to choose their kills, just like we're using logic for our kills.

Hell, if no one else had already called you on it, I'd probably be campaigning for your lynch right now. I'll leave that up to whoever it was that already called you though.

Actually, **** it. I need to get to sleep soon and after that I have class. After class, I'm going to work. I doubt I'll be around during end day, so I'll put out my vote right now. rioting.

ps who the **** is rioting rofl

pps I have 4 classes per week and I work ~3 days every week. Don't expect me to be nearly as active as I usually am. Then again, I'll probably still be more active than a lot of you ****s.

[nforcer06164]

Re: TWG XL: Mall Madness

Afrobean, I see rioting as a new player who made an honest mistake simply because he's trying to be helpful but is a bit confused on the inner workings of the game. Take that into consideration when you vote for him. I'd be agreeing with you if it was a vet, but there's a bit of a difference when it's a newbie.

Aperson, I have the capacity to check your numbers, mind you. I actually thought about it at one point. I didn't for two reasons:

1) Why would you go to all the trouble to procure those numbers anyway, to push forth a strategy that (seemingly) makes sense?
2) I'm lazy.

But now, my curiosity has been pushed a bit far, so I'm going to check them anyway.

[DiscoBobbyPARANOiA]

Re: TWG XL: Mall Madness

Rioting is a friend of Nija, and he didn't understand the whole concept of an alliance, and that's why his numbers were off (I talked to him about it on aim shortly after his post). I'd say freaking out when we get a noob that actually thinks and tries is a bad idea. We could have gotten another jwc. (No offense jwc, you've been trying a lot harder lately, and it's shown)

I talked to rioting on aim for a while last night, and he seems to have his stuff straightened out. I explained the whole 'voting pool' vs 'free voting' (with the seer directing us away from the other blue) and the working of an alliance and he understands where he went wrong now.

Not trying to defend him from your vote, just the unwelcoming feeling that noobs get when they come here.

[nforcer06164]

Re: TWG XL: Mall Madness

Okay, aperson, whether it was an honest mistake (which I doubt), or if you're trying to trick us with the overdose of numbers, or if I'm the one making a mistake here, your post, when reanalyzed, doesn't sit well with me.

First off, to start, we are required to assume certain numbers hold true before your equations are carried out. In logic, making the second party accept certain terms as fact before carrying out the logic process is a strong sign that something within the logical process, or the soundness of the argument, may be false. That aside, let's look further.

We really shouldn't accept your numbers right then and there, because we all cannot possibly agree on the value of a human life, a guarding, or whatever else in a numerical form. It's all subjective. The numbers will never be the same for every person, and therefore a sound argument for consistency is debunked.

It doesn't end there, though. Because your numbers no longer have merit, I don't feel the need to analyze each equation. The whole "payoff" strategy that your post touts has nothing to compare itself to; it's easy to take your numbers and see them as percentages, rather than see that you're comparing the numbers you set forth in the beginning to themselves. We cannot possibly accept the "payoff" because it just comes back to itself to prove its point. I guess you could say the numbers talk in circles.

Your end point is your ultimate goal, and the conclusion is what we need to look at. I don't think there's fault in your numbers; I think they are skewed. You've placed the difference of revealing blues versus not revealing blues as worth the lives of 0.0642 wolves. That value was found by dividing the "payoff" value between the two strategies (0.0803) by the subjective value of a wolf death (1.25). I'm sorry, but it is very clear that the end result is still in the units of the payoff value, and not a wolf itself.

[Payoff strategy difference(p)] / [Payoff value of killed wolf(p)] = [Difference of payoff value of killed wolf between strategies(p)]

Because your conclusion does not match the argument through a simple switch of words, it clearly does not have any logical merit. It's really a simple fallacy that I'm amazed that nobody caught.

Anyone else want to check my work on this, or wish to add any thoughts? Remember, aperson, I'm not trying to say that the strategy we have chosen to roll with is not the best one. I just want to make it very clear that your numbers are complex enough to be just plainly accepted without being checked. Heck, you got me the first time.

One more thing I need to add: If, in reality, you meant that we gain 6% in the chance of killing 100% wolf, you are right. The way you presented your argument, you claim we are 100% closer to killing 6% wolf, which is silly.

[iggymatrixcounter]

Re: TWG XL: Mall Madness

Because your conclusion does not match the argument through a simple switch of words, it clearly does not have any logical merit. It's really a simple fallacy that I'm amazed that nobody caught.

Yea because we all understand math XD.

My question is that even if ap's numbers are off but the end result is still optimal, then why even bring it up at all?

Side point: I know some may be caught up in some college ball tonight (GO BUCKS!) but remember day ends at 11:59pm TONIGHT. (Eastern I think)

[iggymatrixcounter]

Re: TWG XL: Mall Madness

what i mean is, make a pool of wolf suspects, but dont try to lynch them. get kag to seer from the pool. we need to try to lynch the SK as soon as possible.

Read this over and over and tell me the flaw to this plan...

You see if we carried out this plan we would be lynching who we basically thought were humans. (I mean if we save wolf suspects for seerings, then logically only human suspects are left right?)

What kag does is independent of the thread's actions IMO. He will do his own thing when it comes to who to seer. The thread is responsible for lynching the best wolf using the best subjective evidence that any given player comes up with.

[aperson]

Re: TWG XL: Mall Madness

Okay, aperson, whether it was an honest mistake (which I doubt), or if you're trying to trick us with the overdose of numbers, or if I'm the one making a mistake here, your post, when reanalyzed, doesn't sit well with me.

First off, to start, we are required to assume certain numbers hold true before your equations are carried out. In logic, making the second party accept certain terms as fact before carrying out the logic process is a strong sign that something within the logical process, or the soundness of the argument, may be false. That aside, let's look further.

Welcome to game theory where everything rests on the arbitrary value of UTILITY. You've made one of the fundamental arguments that every idiot who first steps into a game theory class takes: How can you quantify some of the things we do? I realize they are estimates of utility, however I think the values are accurate in terms of human payout. If you disagree, then create your own values and run the math. Yes, the numbers are fuzzy. Not every application branch of math is black & white like you wish it was.

We really shouldn't accept your numbers right then and there, because we all cannot possibly agree on the value of a human life, a guarding, or whatever else in a numerical form. It's all subjective. The numbers will never be the same for every person, and therefore a sound argument for consistency is debunked.

Okay then let's let everyone create their own numbers for the system given that a wolf death is 1.25 and we'll take the mean and run it.

It doesn't end there, though. Because your numbers no longer have merit, I don't feel the need to analyze each equation. The whole "payoff" strategy that your post touts has nothing to compare itself to; it's easy to take your numbers and see them as percentages, rather than see that you're comparing the numbers you set forth in the beginning to themselves. We cannot possibly accept the "payoff" because it just comes back to itself to prove its point. I guess you could say the numbers talk in circles.

Uhh, that's because the original utility values are arbitrarily defined, the payoffs to be scaled back in terms of the original values or they have no meaning. As much as I'd like to say you're talking in circles, you're just talking in a straight line out to nowhere. This also explains your statement below:

Your end point is your ultimate goal, and the conclusion is what we need to look at. I don't think there's fault in your numbers; I think they are skewed. You've placed the difference of revealing blues versus not revealing blues as worth the lives of 0.0642 wolves. That value was found by dividing the "payoff" value between the two strategies (0.0803) by the subjective value of a wolf death (1.25). I'm sorry, but it is very clear that the end result is still in the units of the payoff value, and not a wolf itself.

[Payoff strategy difference(p)] / [Payoff value of killed wolf(p)] = [Difference of payoff value of killed wolf between strategies(p)]

Because your conclusion does not match the argument through a simple switch of words, it clearly does not have any logical merit. It's really a simple fallacy that I'm amazed that nobody caught.

Or you're an idiot. Occam's Razor.

Anyone else want to check my work on this, or wish to add any thoughts? Remember, aperson, I'm not trying to say that the strategy we have chosen to roll with is not the best one. I just want to make it very clear that your numbers are complex enough to be just plainly accepted without being checked. Heck, you got me the first time.

One more thing I need to add: If, in reality, you meant that we gain 6% in the chance of killing 100% wolf, you are right. The way you presented your argument, you claim we are 100% closer to killing 6% wolf, which is silly.

They're the same, actually.


Edit: On the arbitrary value issue. Here is the prisoner's dilemma game: http://plato.stanford.edu/entries/prisoner-dilemma/
Now you tell me how the fuck they give any numerical value to the different payoffs without being arbitrary.

[aperson]

Re: TWG XL: Mall Madness

rioting: Don't throw out random statistics that help nothing, and in many instances are quite wrong. As for showing things from the wolf's perspective: As someone else has already said-- bad idea. It's one thing to think from their perspective to deduce their actions thus far, or to get into their heads as a blue trying to find them using your power, but giving random stats about chances of reds hitting a blue... what the ****. Anyway, most of ANYONE's statistics are nearly worthless because they're not drawing a random lot. They're using logic to choose their kills, just like we're using logic for our kills.

To the contrary, given all the snares I posted above, I think the wolves might actually find it best to use random selection for kills. Hell, that could even be a slip on his end.

[nforcer06164]

Re: TWG XL: Mall Madness

aperson, I'm not trying to say you're wrong, or that I'm correct. I have a beef with everyone else who immediately assumes that you may be the smartest person in the game and therefore follows you blindly. I was ready to dismiss you as human right away after your "payoff" post. I did the same thing. That post really doesn't matter, because it isn't necessary to qualify the strategy you set forth.

Honestly, I'm just sick of people takeing the attitude of "omg look guyz ap is so smart lets hump his strat lol". I'm not upset at all that you took charge, but not one person (except maybe DBP) has challenged your methods. They let it fly by every time and just assume that you're right. I'm guilty of that on several occasions. We can't really verify the validity of any strategy, especially this game, without plenty of counter-arguments that can be shot down to verify how solid the original strategy is. And that's what I'm trying to encourage here.

[iggymatrixcounter]

Re: TWG XL: Mall Madness

Honestly, I'm just sick of people takeing the attitude of "omg look guyz ap is so smart lets hump his strat lol". I'm not upset at all that you took charge, but not one person (except maybe DBP) has challenged your methods. They let it fly by every time and just assume that you're right. I'm guilty of that on several occasions. We can't really verify the validity of any strategy, especially this game, without plenty of counter-arguments that can be shot down to verify how solid the original strategy is. And that's what I'm trying to encourage here.

1) A bunch of people posted about ap's "non challenged" strategy before the seer actually came out.

2) I don't know much about ap, but I know when it comes to numbers, he's the best bet. (I mean he's got basic calculus in his sig, he should know what he's talking about to that degree.)

3) This thread is way above average when it comes to TWG experience and outside intelligence (compared to the last few games at least). Ideas will be shot down, fear of being flamed, and extra thinking will/should go on this game. I don't think there's any worry in a dumb idea getting past EVERYONE. Just because people aren't saying anything to dispute ideas, doesn't mean that everyone is blindly accepting it as fact. (In fact, unchallenged ideas denote good ones since no one is thinking up something to decredit it IMO..... unless you're a wolf that is XD.)

[rioting]

Re: TWG XL: Mall Madness

rioting: Don't throw out random statistics that help nothing, and in many instances are quite wrong. As for showing things from the wolf's perspective: As someone else has already said-- bad idea. It's one thing to think from their perspective to deduce their actions thus far, or to get into their heads as a blue trying to find them using your power, but giving random stats about chances of reds hitting a blue... what the ****. Anyway, most of ANYONE's statistics are nearly worthless because they're not drawing a random lot. They're using logic to choose their kills, just like we're using logic for our kills.

The problem with having so little information is that it makes nearly everyone suspect. I’m not surprised to be under attack, one for being a newbie and likewise for being an unknown. My “random statistics” were conceived in naiveté and a will to participate. Afrobean, your suggestion that they “help nothing” is an ignorant hyperbole. True, they may be put to better use among a more receptive audience or perhaps better kept to myself, but to say they help nothing is unjustifiable. Certainly they helped you reach your brash decision to have me lynched. As for whether they are “quite wrong”: I’ve already acknowledged this point when iggy responded to them. In learning more about the game, I know that I made a few faulty assumptions and perhaps some bad calculations. At the very least though, I tried to be a participant whereas we still have yet to hear from MiniNeo, FFR4EVA_00, or HansSky who, for all we know, could in fact be wolves or the SK, using their silence to remain unsuspected because they’re known. Myself being a newbie and admittedly unknown, I’d rather not be a suspect because of my silence but instead be a part of the game.

It’s interesting, Afrobean, that you know the wolves are “using logic to choose their kills.” I wonder how it is you know that. aperson argued that “To the contrary, given all the snares I posted above, I think the wolves might actually find it best to use random selection for kills.” We have no way of knowing which one of you is correct, for all that we know is that two have died but not which by the wolves and which by the serial killer. You though, claim to know that the kills were logical when in fact we have no evidence of that either way, and when they could be as aperson said, “random selection.” However it seems ironic that you continue claiming “we're using logic for our kills.” What logic are we using? Or should I ask, what logic are you using? When we decide tonight whom to lynch, I should hope that we do so logically, based on sound reasoning. You, however, have already made up your mind with subjective and angered finger-pointing. You may be suspicious of me for my naïve comment, but you, having played before, are all the more suspicious to me for yours. If you can’t see that my comment came from inexperience then your quickness to accuse me just further adds more suspicion to yourself.

[nijatwo]

Re: TWG XL: Mall Madness

Ok, I am back.

My “random statistics” were conceived in naiveté and a will to participate. Afrobean, your suggestion that they “help nothing” is an ignorant hyperbole. True, they may be put to better use among a more receptive audience or perhaps better kept to myself, but to say they help nothing is unjustifiable. Certainly they helped you reach your brash decision to have me lynched.

This is a quite an over the top defense, similar to what happened with sertman last game, where people were throwing around harshly worded offenses before there was enough evidence to suggest either direction. It may just be the fact that this is your first game, but it's really wolfy. The goal of afro's orginial post may have been just to get a reaction out of someone. Additionally, one vote doesn't get you lynched. You could be 4 votes up at 11 pm and by 12 be in a totally different position.

Last game, two days in a row afro's mostly-safety votes ended up costing us two days of progress because I took them a little too seriously and the people that reacted to them pointed out how wolfish it looks (ie. sertman and I day 1).

AP- one thing that really bugs me about your posts is that initially you dealt in some absolutes which always worry me day 1. The only way to know some of the information is fact is to either be the source, or be in direct contact with the source. The only merit I saw in the assumption that the wolves went for chromer was potentially flushing out a response from an unsuspecting player to confirm one way or another, but to me it looks more like you did that to cover yourself. I've never played with you before, but from what I hear you're alright. Your game theory is dead on, the issue is unless other people have studied it it's one of those things that can't really make sense. I thought I understood it for the longest time and then after your post I took a gander at my number theory text for next semester, and it's a lot more complicated than first glance. Though it is a good way to assess the chances of the random lynch on day 1 (psuedo-random really.), chances are still just chances. To go along with the game theory, some basic stats will tell you that 13% and 87% are not generally considered significant (those are just arbitrary numbers, not taken from anything specific).

What might help, is the next time the game theory is aspect is analyzed, we should work with an alpha level of 10% maximum, probably even 5% (for those without stats, the alpha level is the highest possible value that shows significance. So for alpha=10, if probability<10%, 'no' significance is assumed with x amount of confidence; conversly 90%>P, 'yes' significance is assumed.)

I'm excited that this has come up though, I've been telling dbp i've wanted to analyze a game that way since the election one that Fojar ran.

The rest of you: I've been trying to get a hold of everyone on aim, but there are still some people I haven't been able to talk to. If I'm set to 'away' I'm probably still online, so don't hesitate to declare your twg-talk-ability.