Large Normal 92 - Game Over! Scum Win!


User avatar
RedCoyote
RedCoyote
Jack of All Trades
User avatar
User avatar
RedCoyote
Jack of All Trades
Jack of All Trades
Posts: 8036
Joined: October 19, 2008
Location: Houston, TX

Post Post #475 (ISO) » Fri May 22, 2009 4:37 am

Post by RedCoyote »

Mod-Edit Votecount 1-19

Mastin - 11
(Caboose, Devestation, zu_Faul, Empking's Alt, orangepenguin, NanooktheWolf, zoraster, King, zwetschenwasser, ryan2754, Phoebus)

OrangePenguin - 5
(cateraction, Maturin24, hewitt, Mastin, Azhrei, AceMarksman)

Zwetschenwasser - 1
(zer0ph34r)

hewitt - 1
(RedCoyote)

NanooktheWolf - 1
(Amished)


Not Voting - 6
(Everyone Else)


With 27 alive, it takes 14 to lynch.

Seeking a replacement for over_9000.




King 473 wrote:Of course. That's the risk we typically MUST take. But we DON'T have to take that risk today because we have someone who definitely is NOT A TOWNIE. How was that not obvious?
Do you typically try to avoid as many risks as possible?
User avatar
zoraster
zoraster
He/Him
Disorganized Crime
User avatar
User avatar
zoraster
He/Him
Disorganized Crime
Disorganized Crime
Posts: 21680
Joined: June 10, 2008
Pronoun: He/Him
Location: Belmont, CA

Post Post #476 (ISO) » Fri May 22, 2009 5:31 am

Post by zoraster »

RedCoyote wrote:
Nanook 433 wrote:Even though we could "do better" with a lynch as Tar has stated earlier, we also could still do worse (lynching a pro town role such as doc, cop, vig, etc.)
Yeah, the argument keeps being brought up. And?

This fear of lynching a townie is such a scummy sounding argument to me. The town cannot be afraid to lynch, period. We'll lose if we let the fear of lynching power roles keep us from taking risks.
It's not about taking risks or not taking risks. It's about taking calculated risks. Is option (a) a lower risk than option (b)? If so, the reason for choosing option (b) has to be much higher than (a).

---

[/quote]
zora 434 wrote:I'm fine with this. But what constitutes a "better candidate" will likely be contentious.
Welcome to the game of Mafia. We argue and have contentious debates over who is guilty and who isn't. :D
zora 434 wrote:Those not in favor of lynching Mastin today (e.g. RC) will have a much lower standard for what a "better candidate" is.
No, I disagree. I'm scumhunting like I normally would. Why is that a problem?[/quote]

Fluff. And disingenuous fluff. My point was that those not in favor of lynching Mastin today may try and scum hunt, but the level of certainty they have to have to lynch someone over mastin will be far lower than I will. Keep in mind my statement was in response to your quoting Tar with the following (paraphrasing): "If we have a better target, we should lynch them. If not, we should lynch Mastin."
zora 434 wrote:But if you're town, then that vested interest is far less important. Something nice (no active player wants to be lynched), but it's not of great interest.
Well, I agree and disagree. I see what point you're making. I'm coming at it from a different perspective though. I see myself as a Lyncher's target, which gives my arguments lot more creditability. Keeping Mastin alive is self-serving in that sense, no doubt, but it's not based on preservation for the sake of playing the game, it's preservation based on the sake that me and Dev are the closest things the town has right now to being clear. Having our opinions out here is going to be very likely the closest thing to a genuine, town perspective on the game that one will get.
You are not remotely clear. Yes, if Mastin is masoned and he confirms you really were clear, then you'd be clear. But that possibility requires two things: (1) that there's a mason who masons Mastin (say that five times fast); (2) that you really are Mastin's target.

I find the first only somewhat likely. I'm pretty sure the second is just not true.

However, also consider the dangers here that I've mentioned. D2 comes around and Mastin fake claims Mason. He knows he hasn't be masoned, so he suspects there aren't masons and this is his best chance to stay alive. So now he's either "cleared" himself (and you) in a fake way or masons have to come out and disown him. Now we have the masons outed.
zora 437 wrote:Let's consider, shall we? Will Mafia kill you? Very unlikely. Will SK kill you? Unlikely as well.
Why? Why are you completely discounting the possibility of Masons/Psychiatrists?

I mean, the mafia, and SK... these players don't want us to have a cleared role.
Why would mafia kill someone we're likely to lynch day 2? Why would Mafia kill someone we're likely to vig n1? The fact that he MIGHT be cleared is probably not outweighed by the fact if they get rid of Mastin, they're doing our job for us.



---
Caboose 440 wrote:I think Mastin is lying about his lynchee.
Who is it?
No one knows which is why Mastin is so freaking dangerous to keep around. Not only is it dangerous because we might lynch his real target, it's dangerous because we'll have to view every lynch possibility through the lens of "is this guy Mastin's target?"

But if you want my best guess based on game play to this point, it's Dust.

---
Nanook 445 wrote:Ace, why couldn't we gain any information from Mastin's lynch? I've heard this comment be stated by others as well I think, and I can't see how there couldn't be ANY information obtained as well as the fact that we could exterminate a threat for tomorrow.
I want to answer this too because I think this is an important question.

Any forum mafia player worth their salts will tell you that the flip after a lynch or a kill is worth so much more than all of the hot air that everyone in the game creates.

The mafia's big advantage over the town is always information. Lynching someone is the best way for the town to get more information.

I think few people on the Mastin wagon disagree with the fact he's a Lyncher, because it's likely that he is. Obviously we don't
want
to lynch a town member, but lynching someone else, even if they are town, will cause them to flip. If we know whether or not said person was scum, we can base further opinions we have tomorrow and the day after that on what that player said and did today.

If we lynch Mastin, and he comes up Lyncher, all we'll get is that Dev is basically a confirmed JOAT, and the same back and forth about who Mastin's target was.

If we lynch someone else, we not only get the information from that lynch, but we create a chaotic WIFOM situation for all anti-town roles at night by leaving Mastin alive.
All things equal, I wholeheartedly agree with this line of reasoning. It's why no lynches in mafiascum are rarely a good idea.

But things are not equal. I think most believe there will be at least two night kills tonight. Those night kills lead to some great information as well. If we decide to lynch Mastin tomorrow, we've thrown away a good opportunity.

I'll agree that lynching Mastin doesn't provide quite as much information as lynching someone else. But it's a great reason to lynch him today rather than tomorrow.

Mastin is an anti-town force. I can only say that so many times.



---
zora 453 wrote:My thought is that there's been a lot of assuming lately. We assume there's a vig. We assume there are masons.

So why don't we assume there's a doctor, one of the most basic roles? Or for that matter that Dev can't hide one of his nights.
It's a two-way street. You can't use this as an argument, it cancels out with the fact that Mastin could be converted or killed.

How are Masons and Vigilantes not "basic" roles?
you've taken me out of context. This was in direct response to the idea that Dev was in mortal danger IF we lynched Mastin because it confirms Dev. My point is that those arguing for saving Mastin have argued that Masons and psychiatrists* are probably around to turn him town. In addition to this argument, they say that we shouldn't lynch Mastin because Dev will be confirmed, endangering him. So I wanted to basically make the point you have: you can't assume one without giving thought to the other.

*By the way, does anyone have a single game where Lyncher was saved by a phsyiciatrist? http://www.mafiascum.net/wiki/index.php ... ychiatrist is the entry for psychiatrist and it's only about SKs. It's my toughht the term Psychiatrist is just added to mason each time to make it sound like there's even a higher probability that he'll be turned town... which isn't true.


---
Mastin 455 wrote:The math is here. People "conveniently" ignored it, because they didn't want to hear how they're wrong and I'm right
I don't know if that's the reason, but I agree with this logic 100%.

The sheer number of positive possibilities of leaving you alive outweigh the negatives to such a degree that I'm confident the scum/third-parties are rooting for a Mastin lynch today. zora can flash all the statistics he wants, there's no way to account for the sensibilities of players with killing powers that have to make a serious decision during the night. You can't simply say mafia has X percent chance of shooting someone, we don't know who the mafia is, we don't know if they're risky or conservative, we don't know whether or not there is a Mason, we
don't know
.
And you can dismiss statistics all you want, but they're still an important part of the discussion. No one is arguing that the statistics I listed are exactly correct by any means... we don't operate in a vacuum. But the question is whether the relative statistical difference between the two is reversed given scum hunting? My thought is that the difference is actually
larger
given scum hunting.


zora 460 wrote:If we don't lynch you today because we believe the RC claim, we believe you're in it for the town, provide a different perspective, etc., then I don't see why lynching you tomorrow would be a good idea.
No one is saying that Mastin will be for sure town tomorrow, and I'm certainly not saying Mastin should be kept alive indefinitely.

The idea that we won't be able to lynch Mastin tomorrow, or any other day, because of potential reports, is ludicrous. Assuming we have a Cop/Tracker/Watcher, assuming they get a guilty report, assuming the supposed guilty doesn't counter them... so many assumptions.
I quite agree that your statement is ludicrous. It's not what I said. I'm saying the reasons for lynching Mastin tomorrow are lower than they are today. I say nothing of the chances of reports. My argument is based on the fact we'll have night kills to go on, and the probabilities of finding scum are higher.
In any case, not getting any new information from the D1 lynch could seriously hinder this town, especially after we may have just lost our JOAT's usefulness.
Not getting rid of the lyncher can seriously hinder this town. Lynching him d2 rather than d1 will seriously hinder the town relative to the situation we would have been in had we lynched day 1.

-------------

I know lots of people don't read all these posts, so I just want to highlight a few things:

1. Mastin will likely claim Masoned tomorrow regardless of whether he actually is.
2. Dismiss statistics if you want, but do so if you have solid reasons for it. The mere fact we scum hunt does not negate a statistical advantage. .
3. Lynching Mastin today is superior to tomorrow, in no small part because of point 1 but also for a host of reasons discussed ad naseum in this post and others before.
.
User avatar
RedCoyote
RedCoyote
Jack of All Trades
User avatar
User avatar
RedCoyote
Jack of All Trades
Jack of All Trades
Posts: 8036
Joined: October 19, 2008
Location: Houston, TX

Post Post #477 (ISO) » Fri May 22, 2009 6:34 am

Post by RedCoyote »

zora 476 wrote:It's not about taking risks or not taking risks. It's about taking calculated risks. Is option (a) a lower risk than option (b)? If so, the reason for choosing option (b) has to be much higher than (a).
Okay, fine, I'll even concede to your definition, that it's calculating to the degree of said risk rather than a more black and white argument.

What is the risk of Mastin winning today, zora?
zora 476 wrote:My point was that those not in favor of lynching Mastin today may try and scum hunt, but the level of certainty they have to have to lynch someone over mastin will be far lower than I will.
Huh? Why?

My response was not removed from this, you are effectively labeling those on the Mastin wagon as supportive of lower standards in their scumhunting. I'm saying that just because I don't want Mastin lynched doesn't mean I would want our lynch to be a random person.

Moreover, I think I could just as easily make the argument that those relying on the Mastin lynch as a crutch are being
far
too lazy in their current analysis of the game. Players like OP, King, zwet, and even you zora, have seemed the abandon entirely the idea of any actual scumhunting outside of pushing Mastin. Granted, this is by far the biggest issue at hand, but, if Mastin has made any solid point today, it's that he's unarguably been interested in
scumhunting
today.

How honest it is is absolutely up for debate, his methods are absolutely up for debate, but the level of interest in
finding scum
between, say, OP and Mastin, is no contest. Mastin beats OP ten times over in that department.
zora 476 wrote:Yes, if Mastin is masoned and he confirms you really were clear, then you'd be clear. But that possibility requires two things: (1) that there's a mason who masons Mastin (say that five times fast); (2) that you really are Mastin's target.
You don't need a Mason to accomplish this.

Look, you say flat out that I'm not Mastin's target. Fine. Whoever Mastin's target is has a great propensity than anyone else (aside from Dev) to be town, right?

Don't you think said player would be the closest thing to a clear townie we have?
zora 476 wrote:D2 comes around and Mastin fake claims Mason. He knows he hasn't be masoned, so he suspects there aren't masons and this is his best chance to stay alive. So now he's either "cleared" himself (and you) in a fake way or masons have to come out and disown him. Now we have the masons outed.
This is just more fear pandering. The past handful of posts you've been saying for us to not so easily accept the fact that there are Masons, and then you come out with a hypothetical that leaving Mastin alive will out all of our Masons?
zora 476 wrote:No one knows which is why Mastin is so freaking dangerous to keep around. Not only is it dangerous because we might lynch his real target, it's dangerous because we'll have to view every lynch possibility through the lens of "is this guy Mastin's target?"
So do you think Dev is lying about the fact that Mastin can't win until tomorrow? Otherwise, why does this worry you today?
zora 476 wrote:But things are not equal. I think most believe there will be at least two night kills tonight. Those night kills lead to some great information as well. If we decide to lynch Mastin tomorrow, we've thrown away a good opportunity.
What opportunity are we throwing away by lynching Mastin tomorrow? Hypothetical guilty Cop report lynches?
zora 476 wrote:I'll agree that lynching Mastin doesn't provide quite as much information as lynching someone else. But it's a great reason to lynch him today rather than tomorrow.
What? This is just a blantant contradiction. Not getting as much information is a great reason to lynch Mastin today?
zora 476 wrote:Mastin is an anti-town force. I can only say that so many times.
And mafia are more anti-town than Mastin. And using our lynch for information rather than getting rid of a Lyncher who can't win today will give the town more to work with tomorrow. I can only say these things so many times.
zora 476 wrote:My point is that those arguing for saving Mastin have argued that Masons and psychiatrists* are probably around to turn him town.
This is just one possibility of many, zora. More than this, I contend, the WIFOM that leaving Mastin alive tonight is enough a reason to keep him breathing, let alone the possibilities of lynching a real threat today.
zora 476 wrote:And you can dismiss statistics all you want, but they're still an important part of the discussion.
I agree with that, but statistics like this ignore the bigger picture. Your basing your statistics off of a specific setup that you have no idea whether or not is close to the actual setup. Granted, you try to make the hypothetical as close to basic as you can, but, honestly, how much good are statistics that have no real basis in the reality of the game? When you don't know what roles there are, it's hard to argue that
either side
has a statistical advantage, can't you at least agree to that?
zora 476 wrote:My argument is based on the fact we'll have night kills to go on, and the probabilities of finding scum are higher.
And using our lynch on an unknown variable rather than Mastin will make the probabilities tomorrow even higher than that.
User avatar
zoraster
zoraster
He/Him
Disorganized Crime
User avatar
User avatar
zoraster
He/Him
Disorganized Crime
Disorganized Crime
Posts: 21680
Joined: June 10, 2008
Pronoun: He/Him
Location: Belmont, CA

Post Post #478 (ISO) » Fri May 22, 2009 7:00 am

Post by zoraster »

What is the risk of Mastin winning today, zora?
Low to zero, but as you make the point, the big picture is necessary. I'm not worried about Mastin winning today. I'm worried about him winning tomorrow or the next day if left alive. Even more so, though, I'm worried about him simply throwing town off in multiple ways by his mere existence.
My response was not removed from this, you are effectively labeling those on the Mastin wagon as supportive of lower standards in their scumhunting. I'm saying that just because I don't want Mastin lynched doesn't mean I would want our lynch to be a random person.
I hope I haven't done this. I'm saying that those not on the wagon presumably have a lower standard for what makes someone a better lynch than Mastin.
You don't need a Mason to accomplish this.

Look, you say flat out that I'm not Mastin's target. Fine. Whoever Mastin's target is has a great propensity than anyone else (aside from Dev) to be town, right?

Don't you think said player would be the closest thing to a clear townie we have?
And how do we know who Mastin's target is to clear them? Without a mason, it's impossible. Even if Mastin tomorrow says, "whoops. you caught me. RC wasn't my target. Person Y who just died was my target, but now I REALLY can't win" we have to question the veracity of the statement.
zora 476 wrote:
D2 comes around and Mastin fake claims Mason. He knows he hasn't be masoned, so he suspects there aren't masons and this is his best chance to stay alive. So now he's either "cleared" himself (and you) in a fake way or masons have to come out and disown him. Now we have the masons outed.

This is just more fear pandering. The past handful of posts you've been saying for us to not so easily accept the fact that there are Masons, and then you come out with a hypothetical that leaving Mastin alive will out all of our Masons?
This is not fear pandering AT ALL. It is a discussion of the risks we are at if Mastin isn't lynched. I'm pretty mystified why you'd label a complete discussion as fear pandering. This is not some sort of wild theory I have. I think anyone who reads that will agree it seems to make a great deal of sense.

Let me be clear though: the risk of outing masons less concerns me than the risk of Mastin claiming Mason and there being no other masons. Both are certainly negatives though which is why both are discussed.
zora 476 wrote:
No one knows which is why Mastin is so freaking dangerous to keep around. Not only is it dangerous because we might lynch his real target, it's dangerous because we'll have to view every lynch possibility through the lens of "is this guy Mastin's target?"

So do you think Dev is lying about the fact that Mastin can't win until tomorrow? Otherwise, why does this worry you today?
No, I don't think he's lying. And it worries me because I'm here to play the complete game. Not just live day to day.
zora 476 wrote:
But things are not equal. I think most believe there will be at least two night kills tonight. Those night kills lead to some great information as well. If we decide to lynch Mastin tomorrow, we've thrown away a good opportunity.

What opportunity are we throwing away by lynching Mastin tomorrow? Hypothetical guilty Cop report lynches?
I believe I've answered this, and I worry you're just trying to bury my reasoning:

1. If we lynch Mastin today, tomorrow we can lynch without worrying about lyncher. Moreover, we can do so with additional information: the nightkills. If there are cop reports or whatever, all the better. But this is not necessary.

2. If we lynch Mastin tomorrow, we have to lynch today. We do so without knowledge of those night kills.

Thus, we've lost the opportunity to lynch based on additional information.
I'll agree that lynching Mastin doesn't provide quite as much information as lynching someone else. But it's a great reason to lynch him today rather than tomorrow.

What? This is just a blantant contradiction. Not getting as much information is a great reason to lynch Mastin today?
It's not a contradiction. I'm arguing against this half-ass theory that we should lynch Mastin tomorrow over today. We lynch Mastin today, we miss out on the information we would have gained from the lynch that would have occurred otherwise. We lynch Mastin tomorrow, we miss out on the information THAT lynch would have given us.
zora 476 wrote:
Mastin is an anti-town force. I can only say that so many times.

And mafia are more anti-town than Mastin. And using our lynch for information rather than getting rid of a Lyncher who can't win today will give the town more to work with tomorrow. I can only say these things so many times.
The fact he can't win today is irrelevant. In fact, it seems logical that we should want to eliminate Mastin BEFORE his power activates.

And yes, I agree mafia are more anti-town than Mastin. That's why I'd rather get rid of Mastin today and focus on those more anti-town elements tomorrow.
zora 476 wrote:
And you can dismiss statistics all you want, but they're still an important part of the discussion.


I agree with that, but statistics like this ignore the bigger picture. Your basing your statistics off of a specific setup that you have no idea whether or not is close to the actual setup. Granted, you try to make the hypothetical as close to basic as you can, but, honestly, how much good are statistics that have no real basis in the reality of the game? When you don't know what roles there are, it's hard to argue that either side has a statistical advantage, can't you at least agree to that?
Give me another potential setup, and I'll do the statistics for that as best I'm able. The results will probably be similar whether you have two scum teams, 2 vigils, etc.
.
User avatar
Phoebus
Phoebus
Hall Monitor
User avatar
User avatar
Phoebus
Hall Monitor
Hall Monitor
Posts: 3743
Joined: October 19, 2003

Post Post #479 (ISO) » Fri May 22, 2009 7:17 am

Post by Phoebus »

Pray, can someone tell me where this idea of masons recruiting people came up?

if you're all being hypothetical... that's pretty unproductive. not to mention profligate on the posts which may well lead to nothing at all...
Your happiness is intertwined with your outlook on life.
User avatar
AceMarksman
AceMarksman
Mafia Scum
User avatar
User avatar
AceMarksman
Mafia Scum
Mafia Scum
Posts: 1432
Joined: January 23, 2009
Location: The OA

Post Post #480 (ISO) » Fri May 22, 2009 8:05 am

Post by AceMarksman »

Zoraster, thank you for this:
zor wrote:Scenario 1: We let Mastin live d1 and lynch him day 2.
Probability of Lynching correctly day 1: 5/26= 19%
Probability of Vigil hitting scum night 1 if we lynch wrong (81% of the time): 5/25= 20%
Probability of Vigil hitting scum night 1 if we lynched correctly (19% of the time): 4/25: 16%
Total chance of lynching or shooting at least one scum: 45%
There's also a 3% chance we could kill TWO scum.

Scenario 2: We lynch Mastin d1 and scum hunt day 2.
Probability of vigil hitting scum n1: 5/25: 20%
Probability of lynching scum d2: 5/22: 23%
Probability of lynching or shooting at least one scum: 48%
Probability of lynching or shooting at least TWO scum: 4.6%
That's enough to convince me to
Unvote, Vote: Mastin


Mastin: While I know that this sucks and is unlucky for you, it's the most pro-town course of action. It's nothing personal at all.
amished wrote: I want clarification on this. You're voting OP for pushing a lynch on somebody that you can eventually see being a threat to the town, and who's lynch you would push if only the day was different?
I realize now that a mastin lynch is a good course of action. At the time of this writing, however, I thought that mastin would not be a good lynch because of the possibility of lynching scum that are a threat right now. Based on zor's math, however, I now realize that Mastin's lynch is ideal. That explanation enough?
Show
"either he's scum and yay 1 less scum, or he's town and yay 1 less zwetschenwasser. " ~Moratorium
Words to live by.

My record: W/L/T/A
Overall:5/2/0/1
Town:5/2/0/1
Scum:0/0/0/0
3rd Party:0/0/0/0
User avatar
Caboose
Caboose
Mafia Scum
User avatar
User avatar
Caboose
Mafia Scum
Mafia Scum
Posts: 2139
Joined: July 28, 2008

Post Post #481 (ISO) » Fri May 22, 2009 8:16 am

Post by Caboose »

Caboose wrote:Mastin is lying about his lynchee, and I think his lynchee is OP.

We need to lynch Mastin today.
In fact, I'm not even fully buying the lyncher claim.
I wouldn't be surprised if Mastin was actually an SK, asking to be killed at night when he knows he's NK immune.
User avatar
Mastin
Mastin
She/Her
Unabridged
User avatar
User avatar
Mastin
She/Her
Unabridged
Unabridged
Posts: 1622
Joined: October 7, 2008
Pronoun: She/Her
Location: Scumread Inc.

Post Post #482 (ISO) » Fri May 22, 2009 8:20 am

Post by Mastin »

I wasn’t joking when I said I’d do all the math. Posting what I’ve done so far.

We have 27 players—that’s fact.

Now, assume one scum, no other killing roles. (I DID say ALL)

1/27-->3.7% chance of lynching scum day one.
One night-kill.
Two bodies, 25 alive.

Day two.
We lynch me, the outed lyncher. 0% chance of lynching scum.
Mafia kills again.

Day three. 23 alive.
1/23-->4.35% chance of lynching scum.
Lynch.
Night-kill.

Day four. 21 alive.
1/21-->4.76% chance of lynching scum.
Lynch.
Night-kill.

Day five. 19 alive.
1/19-->5.26% chance of lynching scum.

18.07 % chance of finding scum within five days.


We lynch me. The outed lyncher. 0% chance of lynching scum.
One night-kill.

Day two.
Two bodies, 25 alive.
1/25-->4% chance of lynching scum.
Mafia kills again.

Day three. 23 alive.
1/23-->4.35% chance of lynching scum.
Lynch.
Night-kill.

Day four. 21 alive.
1/21-->4.76% chance of lynching scum.
Lynch.
Night-kill.

Day five. 19 alive.
1/19-->5.26% chance of lynching scum.

It worked out to 18.37 % on my calculator.
18.37% / 5-->3.674 % chance of hitting scum per day, on average.

But this is NOT due to me being lynched d1—This is due to how the chance of lynching scum increases every day by statistics. Every day, less players alive. And the following day, that means a greater chance of hitting the scum. Lynching me produces the extra .3% due to one factor only:
The dead body n1.
One less person alive to be a suspect.
My death, a ZERO percent chance of lynching scum, in both scenarios, adds nothing.
It’s the body count that matters. The body dead n1, instead of two bodies, actually seems to decrease the chance of hitting scum.

Same probability in both scenarios, in my mind, despite what the statistics say. Logic and common sense to me say that, with more confirmed not-lyncher bodies dead, the chances of finding scum drastically increase compared to half that number of not-lyncher bodies dead, and the lyncher. I could be lynched any day and have the same outcome.

Just to be sure. Let’s run the math with 26 players instead. Because we know I’m not scum.

1/26-->3.85% chance of lynching scum day one.
One night-kill.
Two bodies, 25 alive.

Day two.
We lynch me, the outed lyncher. 0% chance of lynching scum.
Mafia kills again.

Day three. 23 alive.
1/23-->4.35% chance of lynching scum.
Lynch.
Night-kill.

Day four. 21 alive.
1/21-->4.76% chance of lynching scum.
Lynch.
Night-kill.

Day five. 19 alive.
1/19-->5.26% chance of lynching scum.

The Math, 18.22%
18.22 / 5-->3.644 % chance of hitting scum per day.


Now, that puts it even closer to even. 0.15 % difference. Per day, it’s even less, at .03%. Because it’s a far more accurate statistic, due to people KNOWING I am not scum.


Two scum. No other killing roles.

Day one.
2/26-->7.69% chance in hitting scum.
Assume scum not lynched.
Night-kill.

Day two, I’m lynched.
Night-kill.

Day three.
2/23-->8.7% of catching scum.
Night-kill.

Day four.
2/21-->9.52% chance of hitting scum.
Night-kill.

Day five.
2/19-->10.53% chance of hitting scum.

All this assumes no scum is lynched. Yet every day, the stats add up, making each day more likely to hit scum.

The math, 36.44% total of possibly hitting scum.
36.44 / 5-->7.288 % chance per day of catching scum.

Day one.
I am lynched.
Night-kill.

Day two.
2/25-->8%
Night-kill.

Day three.
2/23-->8.7% of catching scum.
Night-kill.

Day four.
2/21-->9.52% chance of hitting scum.
Night-kill.

Day five.
2/19-->10.53% chance of hitting scum.

36.75%
.31 % extra. Not much.
36.75 / 5-->7.35 % chance per day.
A difference of 0.062 %. Again, not much.





Let’s also do the math for leaving me alive, on ALL days and see how that works out.

Again, starting with the one-scum scenario.

1/26-->3.85% chance of lynching scum day one.
0% chance of lynching my target.
One night-kill.
Night kill-->1/25-->4 % chance of being my target.

Two bodies, 25 alive.

Day two.
1/24-->4.17 % chance of hitting scum.
1/24-->4.17 % chance of lynching my target.

Night kill-->1/24-->4.17 % of being my target.
Day three. 23 alive.
1/22-->4.55% chance of lynching scum.
1/22-->4.55 % chance of lynching my target.

Lynch.
Night-kill-->1/22-->4.55% chance of being my target.

Day four. 21 alive.
1/20-->5% chance of lynching scum.
1/20-->5% chance of lynching my target.

Lynch.
Night-kill-->1/20-->5% chance of being my target.

Day five. 19 alive.
1/18-->5.56% chance of lynching scum.
1/18-->5.56% chance of lynching my target.

The Math, 23.13% chance of lynching scum.
19.28 % chance of lynching my target,
17.72 % chance of my target being night-killed.

23.13 / 5-->4.63% chance of hitting scum per day.
19.28 / 5-->3.86% chance of hitting my target per day,
17.72 / 4-->4.43% chance per night of my target being shot.

THOSE look like much better odds to me. And this doesn’t even take into account who I claim to have as a target, and the fact that ANY night, I could be night-killed.



There's a LOT more to come. I've barely even scratched the surface. This will take hours to complete, and when I'm done, it'll show you why lynching me today is bad. For now, the best I've got is the last scenario.

Oh, and HUGE
-S^1 + F
(eh, can't find the link quick enough) to Ace, for bandwagoning on the Math, when I SPECIFICALLY SAID to wait for ALL the math.
I'm back! Well, kind-of.
No Access on Weekends
. :/
Advid reader/contributor to MD, as I'm far better in theory than I am in reality. :P

True to my word, I'm retiring. Totally not me. :P
User avatar
Mastin
Mastin
She/Her
Unabridged
User avatar
User avatar
Mastin
She/Her
Unabridged
Unabridged
Posts: 1622
Joined: October 7, 2008
Pronoun: She/Her
Location: Scumread Inc.

Post Post #483 (ISO) » Fri May 22, 2009 8:22 am

Post by Mastin »

Cabooscum wrote:In fact, I'm not even fully buying the lyncher claim.
Again, that would make Dev a liar, when Dev put great care into the details of what role I am.

If Dev's a role cop of some sort, wouldn't Dev just out me as a Serial Killer?

And also, why do people think that Dev would be clear?
Mafia have rolecops just as often as the town does.
I'm back! Well, kind-of.
No Access on Weekends
. :/
Advid reader/contributor to MD, as I'm far better in theory than I am in reality. :P

True to my word, I'm retiring. Totally not me. :P
User avatar
Mastin
Mastin
She/Her
Unabridged
User avatar
User avatar
Mastin
She/Her
Unabridged
Unabridged
Posts: 1622
Joined: October 7, 2008
Pronoun: She/Her
Location: Scumread Inc.

Post Post #484 (ISO) » Fri May 22, 2009 8:28 am

Post by Mastin »

Keep in mind that, to do all the possible scenarios, it will take hours.

From one to nine scum, no other killing roles.
From one to nine, with a serial killer.
From one to nine, with a serial killer and a vig.
From one to nine, with a serial killer and two vigs.
Two scum factions from two to five members each.
Two scum factions from two to five members each, and a serial killer.
Two scum factions from two to five members each, a serial killer, and a vig.
Two scum factions from two to five members each, a serial killer, and two vigs.

Crossfire between all killing factions.

Different days for me to be lynched, if at all.

It's a HUGE amount of math.

But I'll do it all. I will, to prove Zoraster's math wrong.

To prove how lynching me today is NOT the best course of action we can take.



To make it official,

HUGE
-S^1, + F
, Ace
,

For having defended me, then blindly switching to me, AFTER I HAD SAID TO WAIT FOR
ALL
THE MATH AND NOT JUST ZORASTER'S SMALL SAMPLE!
I'm back! Well, kind-of.
No Access on Weekends
. :/
Advid reader/contributor to MD, as I'm far better in theory than I am in reality. :P

True to my word, I'm retiring. Totally not me. :P
User avatar
AceMarksman
AceMarksman
Mafia Scum
User avatar
User avatar
AceMarksman
Mafia Scum
Mafia Scum
Posts: 1432
Joined: January 23, 2009
Location: The OA

Post Post #485 (ISO) » Fri May 22, 2009 8:44 am

Post by AceMarksman »

Caboose wrote:I wouldn't be surprised if Mastin was actually an SK, asking to be killed at night when he knows he's NK immune.
How do you know than an SK would have NK immunity in this game? It's not information that a townie would have. Small slip, perhaps?
IGMEO Caboose


Mastin: That Fos is OMGUS and you know it.
Show
"either he's scum and yay 1 less scum, or he's town and yay 1 less zwetschenwasser. " ~Moratorium
Words to live by.

My record: W/L/T/A
Overall:5/2/0/1
Town:5/2/0/1
Scum:0/0/0/0
3rd Party:0/0/0/0
User avatar
AceMarksman
AceMarksman
Mafia Scum
User avatar
User avatar
AceMarksman
Mafia Scum
Mafia Scum
Posts: 1432
Joined: January 23, 2009
Location: The OA

Post Post #486 (ISO) » Fri May 22, 2009 8:46 am

Post by AceMarksman »

I mean, I know it's not fair to you, but I do feel that zor's math spelled it all out, and your math with the speculation about the setup isn't as valid as it leaves too much for chance.
Show
"either he's scum and yay 1 less scum, or he's town and yay 1 less zwetschenwasser. " ~Moratorium
Words to live by.

My record: W/L/T/A
Overall:5/2/0/1
Town:5/2/0/1
Scum:0/0/0/0
3rd Party:0/0/0/0
User avatar
Mastin
Mastin
She/Her
Unabridged
User avatar
User avatar
Mastin
She/Her
Unabridged
Unabridged
Posts: 1622
Joined: October 7, 2008
Pronoun: She/Her
Location: Scumread Inc.

Post Post #487 (ISO) » Fri May 22, 2009 9:02 am

Post by Mastin »

Nanook wrote: I believe you've said it yourself that there is more than enough information in day 1, (19 pages worth, 5-6 of them belonging to you Mastin).
1: As I pointed out, no anti-town player contributes THAT much to the game early on. Anti-town role, sure, but they're definitely not going to be anything other than a pro-town player with that much content overall.

2: We gain less information from a lynch of me, whose alignment is already known. What're you going to do, analyze my reactions with others? Their reactions towards me? It's nothing you don't already know.
We gain far more information from the lynch of a player whose alignment we don't know. We get to see how others reacted to them, how they reacted to others, and also gives us better insight as to who dies during the night, why they were targeted, who likely shot 'em, etc.
I applause your effort to claim honorary townie
Roleclaim: Honorary Townie
.

I can be extremely open about my opinions without consequences. The scum most likely do not want to kill me. I scum hunt like a normal townie, I give opinions like a normal townie, and I give insight to the game like a normal townie...except for deeper. I show great logic, and back it up where a regular townie would be afraid to do so for risk of looking scummy.
:P
but I just don't feel I can trust you regardless of what you've just posted.
After all of this...after I've proved time and time again how I want to help the town...after having claimed several times, and given proof, of my target, and how he won't be lynched...you still don't trust me.

I can't help that, Nanook.

But I could've let Dev die. Bank all my hopes of people believing me day two when saying Dev was lying. I didn't. I admitted to who I was, perhaps my greatest contribution to the game. I semi-cleared a suspicious player (Dev), semi-cleared a player we hadn't heard much from (Red), and I cleared myself from being mafia. When none had been true before.

I can't do much more than what I have. Give the math, research what the setup likely is, give theories as to what the scum and what the town can do, and scum hunt. That's no more than what any pro-town player can do. And I've done exactly that.

What more can I do?
Also in regards to knowing if OP is scum .. You read that statement how you wish, but from my perspective I believe it still to make perfect sense in how it related to the question I was answering.
This seems more like an admition than a defense.
I don't know if OP is scum, if I did it would be him who I'd be voting for.
Unless you're partners, and don't want to buss him.
I believe the term tunneling was coined (this is new for me, just heard of it since getting back to the site over a year).
You seriously haven't heard of Confirmation Bias?

It was around well before I joined in October. At least a few months.
Zor wrote:You want us to out more PRs for your sake?
No, for the town's sake, to test the setup, test the power of the scum and the power of the town.
For example, tomorrow you come out and claim masoned. If we have real masons, they have to make a choice: CC you and call you out but out themselves?
Zor...If I were to lie, I wouldn't have claimed today. I would've claimed to be the cop, and watched as Devestation was lynched.

No.
I wouldn't lie day two about being masoned.
If there are no masons, no one ccs, and town has to make a choice of whether to believe you or not (which is the same thing that probably happens if there are other masons).
Again, I won't lie.

The thought had occurred to me.

That lying was the best hopes.
I could lie, say I'm masoned, claim that my target was someone else, and lynch Red for scummy comments and easily accepting that I said he was my target.

I wouldn't.
And if I lived, I won't.

I am not going to lie.

I haven't once lied since I claimed.

I've shown exactly what I would've done, shown the breadcrumbs and meta, pointing to my lyncher target STRONGLY being one of only three people...I can't be more honest that that, Zor.
But the last tells us very little. There are a plethora of reasons you wouldn't die and no one would flip mason.
If it's the last, then we gain no information.

If any other, then we procure a great amount of information from the night.
And I think we have both scum and nonscum on Mastin and off it.
Precisely why the lynching of a person whose alignment is known has a negative impact on information.
You say it's in the numbers and you say it'll take time.
There are millions of variables.
I will have to do them all, and it might take days. Doesn't matter. I have the determination, and motivation, to prove that your math is incorrect. I will do as much as necessary to prove it, I will take as long as needed to prove it without a doubt, and that those voting me off of it (Ryan, Ace) are being misguided.
But you're going to have to do better than that.
I've posted my defense several times, and proven how I've been scum hunting. Add the complete math (which I've only done a fraction of, mind you), and you get one of the best defenses ever seen on MS.net. If that isn't enough to convince you, you'd either be lying scum, or tunneling town to disbelief.

It'll just take time to finish.

Really, we have a LONG time 'til the deadline.

Players can WAIT for me to do the math.
We can argue about which assumptions to make, but the numbers will be similar either way.
I'm going to assume the worst case scenarios first, of no scum deaths in lynches and night-kills. That gives the worst case scenario for win percentages, and the best-case scenario for numbers.
And we can test the setup without using you.
Not nearly as well. I volunteered as a Guinea Pig. No other person's death can give as much information as mine, during the night.

(NOTE: Key words:
During the night
. During the day, my death is what gives the LEAST amount of info.)
I've already done numbers
For ONE VERY SPECIFIC SCENARIO, yes, you have.
BUT ONLY ONE.
and they clearly support my statement
Numbers are so easy to manipulate to support your cause. Even if the math is correct, it's the given numbers that are in question. (Five total scum, a vig, etc.)
although you have not directly addressed them.
At least in my head, the math itself works out, to me. There's nothing to address about it. All I can do is counter it, show how in most circumstances, you're incorrect about the math favoring a d1 lyncher lynch.
I don't think that changing the number of scum teams, vigils, etc. will cause a difference in the end result (i.e. lynching you today is superior to lynching you tomorrow), but I'm willing to look at the math.
Again, it'll take time. Lots of it. I stand firm by the facts that the math will support me.
but those who ignore statistics are ignoring a major tool.
People who use statistics alone are the exact same people who suggest no-lynches day one, or policy lynches of certain players like Zwet/Emp, on occasion.
But if scum hunting on day 2 give us a more statistically probable chance of catching scum, then it's a strong indication that's a route we should pursue.
Again, I disagree.

A flip day one of another player gives us their reactions, and the reactions of others to them. Plus the night-killed people, who likely killed them, who the night-killed people interacted with and who interacted with them, etc.

My flip day one gives us nothing about interactions with the lynch,
And only the night-kills.
Nothing else.
I don't think I've manipulated the numbers at all.
Giving a specific scenario instead of all of them qualifies.
certainly, but it will not change the result.
Oh, yes it will.
I'm barely at the surface, and I already gave one (albeit risky) scenario where the numbers favor keeping me alive for a greater chance at hitting scum.



Posting what I have.
I'm back! Well, kind-of.
No Access on Weekends
. :/
Advid reader/contributor to MD, as I'm far better in theory than I am in reality. :P

True to my word, I'm retiring. Totally not me. :P
User avatar
zoraster
zoraster
He/Him
Disorganized Crime
User avatar
User avatar
zoraster
He/Him
Disorganized Crime
Disorganized Crime
Posts: 21680
Joined: June 10, 2008
Pronoun: He/Him
Location: Belmont, CA

Post Post #488 (ISO) » Fri May 22, 2009 9:03 am

Post by zoraster »

First of all, your math does not in any way dispute my numbers. Your math says the difference is "small" (although you use one and two scum so the result will be small as will only having one night kill), but it doesn't come out with a different result.

Anyway, I admit that the probabilities are different because of the body count. But that's a nonstarter. My question is "so what?"

Second, My argument (and the math supporting it) with statistics is about whether or not lynching you day 1 or day 2 is a better option. This is why I have stated that it's preferable to lynch you day 1 vs. day 2.

If, on the other hand, the debate is between whether we lynch Mastin AT ALL, then I have very different responses.

So which is it? Are we having an argument about whether to lynch you day 1 vs. day 2 or are we having an argument about whether to lynch you at all? Because this is important.

---
Mastin wrote:23.13% chance of lynching scum.
19.28 % chance of lynching my target,
17.72 % chance of my target being night-killed.
Don't you see how this is a negative result for town? Of course, there are more scum around, so the likelihood of lynching scum is higher, but a 19% chance of losing the game outright over five days? That's not a risk I'm willing to take when we can easily eliminate you now. Admittedly, town can mitigate this by being aware of what you're doing, but again... do we really want to try and foil Mastin every day we want to lynch?

---

Last, consider his charge that my logic was bad because of my math. Then he uses, to prove that the differences are insignificant, an unrealistic setup -- which, by the way, reaffirms that my math was right and holds true whether you're talking 1 scum or 4+sk+vigil.

Rather than show all the math, Mastin, why don't you show us the math in the scenario that you think most helps your case (so long as the setup is within the realm of possibility)?
.
User avatar
Mastin
Mastin
She/Her
Unabridged
User avatar
User avatar
Mastin
She/Her
Unabridged
Unabridged
Posts: 1622
Joined: October 7, 2008
Pronoun: She/Her
Location: Scumread Inc.

Post Post #489 (ISO) » Fri May 22, 2009 9:07 am

Post by Mastin »

Ace wrote:Mastin: That Fos is OMGUS and you know it.
No. It is not.

You were defending me.
DEFENDING.

Now you've suddenly switched stances, over ONE
very
specific scenario, EVEN AFTER I SAID TO WAIT FOR ALL THE MATH,
IN ORDER TO PROVE ZOR WRONG
.

And THAT is INCREDIBLY scummy, Ace, to no ends.
but I do feel that zor's math spelled it all out
Of course it did. It was one very specific scenario. That's why.
and your math with the speculation about the setup isn't as valid as it leaves too much for chance.
Zor's math is ALSO speculation about the setup, and ALSO leaves JUST AS MUCH to chance!

Hypocrite, much?

Yea. Seems that way to me.



Something which I'll also have to factor into the math:

Dev, a rolecop, a semi-clear.
The possibility of the vig being lynched, instead of caught in the cross-fire.

Lots.
And
Lots.
And
LOTS
of math.

But I will do it, nonetheless. I will do whatever it takes to show you why Zoraster is wrong.
I'm back! Well, kind-of.
No Access on Weekends
. :/
Advid reader/contributor to MD, as I'm far better in theory than I am in reality. :P

True to my word, I'm retiring. Totally not me. :P
User avatar
zoraster
zoraster
He/Him
Disorganized Crime
User avatar
User avatar
zoraster
He/Him
Disorganized Crime
Disorganized Crime
Posts: 21680
Joined: June 10, 2008
Pronoun: He/Him
Location: Belmont, CA

Post Post #490 (ISO) » Fri May 22, 2009 9:13 am

Post by zoraster »

I'm barely at the surface, and I already gave one (albeit risky) scenario where the numbers favor keeping me alive for a greater chance at hitting scum.
Again, my discussion of numbers was about whether d1 or d2 makes a better lynch.

The discussion about whether to lynch you at all is a very different discussion and involves the question of "is it worth having a higher chance at getting scum if there's also an infinitely higher chance of losing by lynching lyncher target?" Getting a 5% advantage to getting a scum in exchange for a 19% chance of lynching your target is less than optimal. And I'll argue the existence of a lyncher alive only makes that chance of finding scum smaller.

---

Anyway, I didn't cherry pick my scenario. It seemed one likely to occur. To my mind, much more likely than the 1 or 2 scum scenario you've painted.

But as I've said, I'm not sure anyone wants to read through pages of numbers... or for you to have to do all the work required to do all those numbers. Simply coming up with the most helpful numbers for your case in at least a plausible setup can do wonders for the town's understanding of the probabilities. That includes my understand as well, by the way.
.
User avatar
Mastin
Mastin
She/Her
Unabridged
User avatar
User avatar
Mastin
She/Her
Unabridged
Unabridged
Posts: 1622
Joined: October 7, 2008
Pronoun: She/Her
Location: Scumread Inc.

Post Post #491 (ISO) » Fri May 22, 2009 9:23 am

Post by Mastin »

Zor wrote:First of all, your math does not in any way dispute my numbers. Your math says the difference is "small" (although you use one and two scum so the result will be small as will only having one night kill), but it doesn't come out with a different result.
YET. (And it DOES come out with a different result in the last--a far higher chance of hitting scum if I'm left alive, with an insignificant risk of hitting my target, especially when I claimed my target, and gave proof as to why he is truly my target)

I said I'd do ALL the math.

That isn't even CLOSE to all the math.

That's probably only 1/1000th of the math that can be done.
I'm working my way up.
And will do dozens of different calculations for the same setup.

I will show you why your math is wrong, even if I haven't done so thusfar.
My question is "so what?"
So, the body count makes a HUGE different in scum hunting, ESPECIALLY when considering WHO the dead bodies are, and what roles they have.
with statistics is about whether or not lynching you day 1 or day 2 is a better option. This is why I have stated that it's preferable to lynch you day 1 vs. day 2.
And I will give multiple options.

Day one lynching.
Day two lynching.
Day three,
Day four,
Day five,
And never being lynched.

I'm 99% sure, a confidence level I don't even claim to have when cop with a guilty (that's 95%. ;)), that I can prove that it is best not to lynch me day one. I already gave one scenario which proved why leaving me alive gives a far greater chance of nailing scum, with minimal risk, where to lynch me gives far less.
If, on the other hand, the debate is between whether we lynch Mastin AT ALL, then I have very different responses.
The argument is when it is best to lynch me, if at all.
I said that the safest bet was day two.

But I will give statistics for five days in a row to prove why it's best to not lynch me today, day one. If that points to day two, yay, I was right. If not, meh. It still won't point to day one.
So which is it? Are we having an argument about whether to lynch you day 1 vs. day 2 or are we having an argument about whether to lynch you at all? Because this is important.
It's both.
Again,
When to lynch me, if at all.
Don't you see how this is a negative result for town?
Not really. I, personally, consider that the chance of my target being night-killed removes that much percent from the chance of me getting a lynch on my target. (Which would leave less than 3% chance of me getting my target lynched total, and over -1% chance per day of me getting my target lynched.) I failed to mention it due to the fact that others might obviously disagree with me on that matter.
Of course, there are more scum around, so the likelihood of lynching scum is higher, but a 19% chance of losing the game outright over five days?
And in the given scenario with one scum, over 20% chance of winning the game in five days. Increase the number of scum (I was about to do that math, too!), and you increase the percentage drastically, yet keep the lyncher percentage the same.
do we really want to try and foil Mastin every day we want to lynch?
You already know my target. I've made it quite clear that Red is my target, and given evidence to support it.
Last, consider his charge that my logic was bad because of my math.
THAT is a twisting of my words.
Your math is good.

The SPECIFIC SCENARIO is NOT, though.
Then he uses, to prove that the differences are insignificant, an unrealistic setup -- which, by the way, reaffirms that my math was right and holds true whether you're talking 1 scum or 4+sk+vigil.
1: I said I'd do all the math--no matter how unrealistic the setup would be. I intend to follow through.
2: It didn't, if you were paying attention, reaffirm your math. It casted severe doubt on your math being accurate, for the reasons I listed throughout that post.
Rather than show all the math, Mastin, why don't you show us the math in the scenario that you think most helps your case (so long as the setup is within the realm of possibility)?
Because the only way to know what scenario best supports my case is to do the math for said scenario. And if I were to do that, then I'd be hypocritical, anyway, for I'd be using math that ONLY supported my argument, and none that did NOT. And I made it rather clear that to do so would be scummy.



How many more times must I defend myself from this?
I'm back! Well, kind-of.
No Access on Weekends
. :/
Advid reader/contributor to MD, as I'm far better in theory than I am in reality. :P

True to my word, I'm retiring. Totally not me. :P
User avatar
Mastin
Mastin
She/Her
Unabridged
User avatar
User avatar
Mastin
She/Her
Unabridged
Unabridged
Posts: 1622
Joined: October 7, 2008
Pronoun: She/Her
Location: Scumread Inc.

Post Post #492 (ISO) » Fri May 22, 2009 9:35 am

Post by Mastin »

Zor wrote:The discussion about whether to lynch you at all is a very different discussion and involves the question of "is it worth having a higher chance at getting scum if there's also an infinitely higher chance of losing by lynching lyncher target?"
You've got it backwards.

Every day, the chance of hitting scum DRASTICALLY increases. Every night, the chance of my target being night-killed EVEN MORE DRASTICALLY increases. Look at the math again--

The Math, 23.13% chance of lynching scum.
19.28 % chance of lynching my target,
17.72 % chance of my target being night-killed.

23.13 / 5-->4.63% chance of hitting scum per day.
19.28 / 5-->3.86% chance of hitting my target per day,
17.72 / 4-->4.43% chance per night of my target being shot.

The chance of my target being shot, per night, is just below the chance of us hitting scum every day, and WELL above the chance of me getting a successful lyncher target lynch every day.

4.63,
3.86 (LOW!),
4.43 (.2 lower than hitting scum).

And that's only with one scum.

Again, with more scum, the chances of hitting scum DRASTICALLY increase, while the chances of hitting my lyncher target DECREASE.
Getting a 5% advantage to getting a scum in exchange for a 19% chance of lynching your target is less than optimal.
No. It's OVER 20%, not five.
And I'll argue the existence of a lyncher alive only makes that chance of finding scum smaller.
How so?

I'm confirmed not scum, which is AT LEAST one less person.
My claimed target, if believed, is semi-confirmed (my statistics did not take this into account), making Red drop out,
And the person who outed is not counted as confirmed, but only semi-confirmed, as well.

These were for later on.
Later-on math in the possible scenarios.

But added together, even just one (yet alone, two probable and one confirmed!) person not scum is a SIGNIFICANT advantage.
Anyway, I didn't cherry pick my scenario.
Nor am I. I'm going to show ALL of them, if it takes me a week to do. We're far from a deadline; people could wait if they were truly pro-town and wanted to see the math.
To my mind, much more likely than the 1 or 2 scum scenario you've painted.
Let's bold, italicize, and underline it with caps, one more time, to make it clear:

The math I've posted so far is a
FRACTION OF ALL THE POSSIBLE SCENARIOS
, AND I MADE IT RATHER CLEAR THAT I INTEND TO DO
ALL
OF THEM, AND THAT WHAT I POSTED WAS JUST A MERE
SAMPLE
OF THE SCENARIOS
.

My time could be spent elsewhere.
Instead, I'm debating this with you, and readying myself for perhaps the greatest math project in MS.net's history.
Simply coming up with the most helpful numbers for your case in at least a plausible setup can do wonders for the town's understanding of the probabilities. That includes my understand as well, by the way.
Yet it also shows a bias on the matter. It means I'm only giving the numbers that support my cause, SOMETHING I ACCUSED
YOU
of doing.
I'm not going to be a hypocrite, even if it decreases the chance of me being lynched.
I'm back! Well, kind-of.
No Access on Weekends
. :/
Advid reader/contributor to MD, as I'm far better in theory than I am in reality. :P

True to my word, I'm retiring. Totally not me. :P
User avatar
zoraster
zoraster
He/Him
Disorganized Crime
User avatar
User avatar
zoraster
He/Him
Disorganized Crime
Disorganized Crime
Posts: 21680
Joined: June 10, 2008
Pronoun: He/Him
Location: Belmont, CA

Post Post #493 (ISO) » Fri May 22, 2009 9:40 am

Post by zoraster »

Because the only way to know what scenario best supports my case is to do the math for said scenario. And if I were to do that, then I'd be hypocritical, anyway, for I'd be using math that ONLY supported my argument, and none that did NOT. And I made it rather clear that to do so would be scummy.
Well, there are trends. For example, I know that if you add another mafia to your calculations, the numbers will shift, but it won't make it a statistically better call to lynch you d1 versus d2. If you're going for your best case, I doubt the right way to look is simply to add more mafia to the situation.

But if you find that by calculating the chances of cross shooting between, say, 2 scum teams of 3 each and 1 sk, makes it come out so that it's actually advantageous to lynch you d2 versus d1, then you might start to look at whether adding a second SK will make that difference even bigger.

Then, if when you post it people say something like "well, that's just an isolated example" you can say, "no. it's also true to a lesser extent in Setup X, Y, and Z."

Frankly, it's not hypocritical. If the charge against me is that I handpicked my example, and then you turn around an show even one plausible example that shows the opposite result (rather than just a smaller similar result as you've done so far), then my statistics are weakened considerably. This is because my argument is that my statistics are representative of a universal result.

---
And in the given scenario with one scum, over 20% chance of winning the game in five days. Increase the number of scum (I was about to do that math, too!), and you increase the percentage drastically, yet keep the lyncher percentage the same.
Yes, I agree with this. But that's a higher percent chance that we catch ONE scum. Not that we catch ALL scum. The issue is that catching one scum does not win us the game. The 20% chance that you get your target lynch loses us the game.
.
User avatar
Mastin
Mastin
She/Her
Unabridged
User avatar
User avatar
Mastin
She/Her
Unabridged
Unabridged
Posts: 1622
Joined: October 7, 2008
Pronoun: She/Her
Location: Scumread Inc.

Post Post #494 (ISO) » Fri May 22, 2009 9:47 am

Post by Mastin »

Red wrote:Yeah, the argument keeps being brought up. And?

This fear of lynching a townie is such a scummy sounding argument to me. The town cannot be afraid to lynch, period. We'll lose if we let the fear of lynching power roles keep us from taking risks.
Quoted for truth. ;)
it's preservation based on the sake that me and Dev are the closest things the town has right now to being clear.
Actually, Red, that's me.
Dev could be a scum rolecop,
And you could be scum who I happened to have the luck of targeting. (Extremely doubtful, though)

I, however, am 100% guaranteed not scum, making me 100% clear. ;)
I think few people on the Mastin wagon disagree with the fact he's a Lyncher, because it's likely that he is. Obviously we don't want to lynch a town member, but lynching someone else, even if they are town, will cause them to flip. If we know whether or not said person was scum, we can base further opinions we have tomorrow and the day after that on what that player said and did today.

If we lynch Mastin, and he comes up Lyncher, all we'll get is that Dev is basically a confirmed JOAT, and the same back and forth about who Mastin's target was.

If we lynch someone else, we not only get the information from that lynch, but we create a chaotic WIFOM situation for all anti-town roles at night by leaving Mastin alive.
This, I am also quoting for truth. I've made this opinion clear several times, and I am frustrated at how it has been ignored.
I don't know if that's the reason, but I agree with this logic 100%.

The sheer number of positive possibilities of leaving you alive outweigh the negatives to such a degree that I'm confident the scum/third-parties are rooting for a Mastin lynch today. zora can flash all the statistics he wants, there's no way to account for the sensibilities of players with killing powers that have to make a serious decision during the night. You can't simply say mafia has X percent chance of shooting someone, we don't know who the mafia is, we don't know if they're risky or conservative, we don't know whether or not there is a Mason, we don't know.

How do we aquire the knowledge to understand these things more clearly? By lynching people who aren't investigated by the JOAT.
And this as well...

Yea, are you seeing the picture?

It isn't just my opinion, anymore. Others have weighed in, and agree with me on the matter.
No one is saying that Mastin will be for sure town tomorrow, and I'm certainly not saying Mastin should be kept alive indefinitely.

The idea that we won't be able to lynch Mastin tomorrow, or any other day, because of potential reports, is ludicrous. Assuming we have a Cop/Tracker/Watcher, assuming they get a guilty report, assuming the supposed guilty doesn't counter them... so many assumptions.

In any case, not getting any new information from the D1 lynch could seriously hinder this town, especially after we may have just lost our JOAT's usefulness.
And this.

Like I said, roleblockers.
Mafia love 'em.

They override any other actions, based off of the order Jebus posted in the first post. (Unless they're unblockable, of course.)

And if not, redirectors, godfathers, possibly no doctors (though the last one is extremely doubtful) might cause this to happen.
Empking wrote:Mastin: We have better odds on hitting mafia day 2 than day 1.
Only by the math thusfar posted by Zoraster. Given time to type it up, I can prove how Zor is wrong in this matter.
Knight wrote:Anyway, I'll do a big ol' summary post tonight
You better be DARN sure not to put me any closer to a lynch. I'm already at L-2, due to people ignoring half of what I say to support their own arguments.
King wrote:But we DON'T have to take that risk today because we have someone who definitely is NOT A TOWNIE. How was that not obvious?
Thing is, we can take the risk today, and lynch me tomorrow, if I survive the night.



Done with page 19, and I'm sure by now, there's a response on page 20 waiting for me.
I'm back! Well, kind-of.
No Access on Weekends
. :/
Advid reader/contributor to MD, as I'm far better in theory than I am in reality. :P

True to my word, I'm retiring. Totally not me. :P
User avatar
zoraster
zoraster
He/Him
Disorganized Crime
User avatar
User avatar
zoraster
He/Him
Disorganized Crime
Disorganized Crime
Posts: 21680
Joined: June 10, 2008
Pronoun: He/Him
Location: Belmont, CA

Post Post #495 (ISO) » Fri May 22, 2009 9:48 am

Post by zoraster »

Quote:
Getting a 5% advantage to getting a scum in exchange for a 19% chance of lynching your target is less than optimal.
No. It's OVER 20%, not five.
I said 5% advantage. That's the difference between 23% and 18%.
I'm confirmed not scum, which is AT LEAST one less person.
My claimed target, if believed, is semi-confirmed (my statistics did not take this into account), making Red drop out,
And the person who outed is not counted as confirmed, but only semi-confirmed, as well.
We're simply not going to resolve this at this point. I don't believe you've stated your real target. I don't think any amount of "proving" it at this point will make me believe it. Therefore, RC is (sadly) not close to clear.
.
User avatar
Mastin
Mastin
She/Her
Unabridged
User avatar
User avatar
Mastin
She/Her
Unabridged
Unabridged
Posts: 1622
Joined: October 7, 2008
Pronoun: She/Her
Location: Scumread Inc.

Post Post #496 (ISO) » Fri May 22, 2009 9:59 am

Post by Mastin »

Zor wrote:Well, there are trends. For example, I know that if you add another mafia to your calculations, the numbers will shift, but it won't make it a statistically better call to lynch you d1 versus d2.
Disagreed. VERY strongly disagreed. The changes in the statistics can be HUGE with a simple adjustment to the setup, like adding a roleblocker, subtracting a mafia, adding a serial killer, etc.

Even simply adding one mafia will change the stats.
If you're going for your best case, I doubt the right way to look is simply to add more mafia to the situation.
Of course not. But it does give a fair comparison.
But if you find that by calculating the chances of cross shooting between, say, 2 scum teams of 3 each and 1 sk, makes it come out so that it's actually advantageous to lynch you d2 versus d1, then you might start to look at whether adding a second SK will make that difference even bigger.
Or a vig, or two vigs, or two scum factions, or maybe even three...
You get the idea.

ANYthing can change the statistics.

And I intend to show them all.
Then, if when you post it people say something like "well, that's just an isolated example" you can say, "no. it's also true to a lesser extent in Setup X, Y, and Z."
Well, why wouldn't I just give the setup, and then Setups X, Y, and Z in the same post, then?
Frankly, it's not hypocritical.
It is to me. And I'm the one typing up the math, so...
If the charge against me is that I handpicked my example, and then you turn around an show even one plausible example that shows the opposite result (rather than just a smaller similar result as you've done so far), then my statistics are weakened considerably.
A-hem.
I DID post a scenario with the opposite result (although it's not a lynch me day two one). I posted some similar results to yours, sure, but also gave reasoning as to why they're not an accurate measurement.
This is because my argument is that my statistics are representative of a universal result.
If that were the case, then I'd already have proven you wrong.
The 20% chance that you get your target lynch loses us the game.
1: It's 19,
2: You yourself said that the town doesn't lose--you say that you'd play for second place, an opinion I disagree with. (I say that a lyncher win on MS.net is just one extra person who wins, who wouldn't win, otherwise. One extra player winning doesn't really mean anything. On Epicmafia, I can understand wanting to get rid of any anti-town role who could steal points and cause you to lose, but MS.net isn't point-based.)



This will get old, fast. But it's not like I have a choice, or anything. You'll NEVER see me EVER just curl up into a ball and die. I'll prove that the statistics are wrong, even if it's the last thing I do. (...Considering I'm at L-2, this might hold true. :P)
I'm back! Well, kind-of.
No Access on Weekends
. :/
Advid reader/contributor to MD, as I'm far better in theory than I am in reality. :P

True to my word, I'm retiring. Totally not me. :P
User avatar
Mastin
Mastin
She/Her
Unabridged
User avatar
User avatar
Mastin
She/Her
Unabridged
Unabridged
Posts: 1622
Joined: October 7, 2008
Pronoun: She/Her
Location: Scumread Inc.

Post Post #497 (ISO) » Fri May 22, 2009 10:05 am

Post by Mastin »

Zor wrote:I said 5% advantage. That's the difference between 23% and 18%.
Ah. I thought you were thinking of the 4%/day compared to the 19% total.

Still, a 5% increase in the chance to win, when the increase of a chance to lose is less than 2% (calculating in the chance of Red being night-killed being subtracted), is WELL worth it.
We're simply not going to resolve this at this point.
If we do not right now, we never will, due to, well, you know, ME BEING DEAD. <_<
I don't believe you've stated your real target.
And I've given all the proof I can as to why he is.
I don't think any amount of "proving" it at this point will make me believe it.
If Red isn't my target, then WHO IS?

Give someone who you think is my lyncher target, or I'll call it BS.
Give someone who you think is my lyncher target, so I can prove (again) how they are NOT my target.
Don't give someone who you think is my lyncher target, and--as I said--I'll call BS and basically, strawmanning me. For I can't defend against an imaginary attack, for it wouldn't exist.
Therefore, RC is (sadly) not close to clear.
Hence, "semi-clear", like Dev is.
I'm back! Well, kind-of.
No Access on Weekends
. :/
Advid reader/contributor to MD, as I'm far better in theory than I am in reality. :P

True to my word, I'm retiring. Totally not me. :P
User avatar
zoraster
zoraster
He/Him
Disorganized Crime
User avatar
User avatar
zoraster
He/Him
Disorganized Crime
Disorganized Crime
Posts: 21680
Joined: June 10, 2008
Pronoun: He/Him
Location: Belmont, CA

Post Post #498 (ISO) » Fri May 22, 2009 10:22 am

Post by zoraster »

Disagreed. VERY strongly disagreed. The changes in the statistics can be HUGE with a simple adjustment to the setup, like adding a roleblocker, subtracting a mafia, adding a serial killer, etc.

Even simply adding one mafia will change the stats.
Change the stats, yes. Change the result, no.
A-hem.
I DID post a scenario with the opposite result (although it's not a lynch me day two one). I posted some similar results to yours, sure, but also gave reasoning as to why they're not an accurate measurement.
Of course the chance of us hitting scum, at least in a vacuum, is higher if we don't lynch you. But the chance of hitting your lyncher target is much, much higher as well. This is why I have the duality listed. D1 vs. D2. It's why the discussion whether to lynch you, period, is a far different question.
Or a vig, or two vigs, or two scum factions, or maybe even three...
You get the idea.

ANYthing can change the statistics.

And I intend to show them all.
That's fine, but given the time it takes, I'll continue to push your lynch. You've shown me no data that has flatly contradicted my point: lynching you D1 is preferable to lynching you D2. THIS is why I'm trying to get you not to waste your time and ours by giving us your best possible case.

Additionally, don't expect everyone to suddenly fall behind you because you've put a lot of work into this. If the statistics show something startling, yes. If it's just a bunch of statistics that you spent days on, I don't think you're going to get much.
Quote:
The 20% chance that you get your target lynch loses us the game.
1: It's 19,
Sorry.
2: You yourself said that the town doesn't lose--you say that you'd play for second place, an opinion I disagree with. (I say that a lyncher win on MS.net is just one extra person who wins, who wouldn't win, otherwise. One extra player winning doesn't really mean anything. On Epicmafia, I can understand wanting to get rid of any anti-town role who could steal points and cause you to lose, but MS.net isn't point-based.)
The town does lose. Playing for second place is consolation, nothing else. Perhaps there's some fun left to be had playing for second as we try and track down scum. But while I still have a shot for first, I'm going to play it that way. Put in a hypothetical situation where I have two choices: one that gives me automatic 100% chance at second and one that gives me a 5% chance at first and a 95% chance at third, I'll take the latter every time.

This theory that a lyncher win doesn't make town take second place at best is self-serving hogwash. You win, we do not. It's as simple as that.

My goal in this game is to win. Not to defeat scum, not to lynch you. not to do anything else. It's to win within the parameters of the game itself. If there's a lyncher, then I do not win if I hit the lyncher's target. The only question I can ever have at the core is "is this the best option to get me to win?"

Finally, I'm not interested in EpicMafia.
.
User avatar
zoraster
zoraster
He/Him
Disorganized Crime
User avatar
User avatar
zoraster
He/Him
Disorganized Crime
Disorganized Crime
Posts: 21680
Joined: June 10, 2008
Pronoun: He/Him
Location: Belmont, CA

Post Post #499 (ISO) » Fri May 22, 2009 10:27 am

Post by zoraster »

If Red isn't my target, then WHO IS?

Give someone who you think is my lyncher target, or I'll call it BS.
Give someone who you think is my lyncher target, so I can prove (again) how they are NOT my target.
Don't give someone who you think is my lyncher target, and--as I said--I'll call BS and basically, strawmanning me. For I can't defend against an imaginary attack, for it wouldn't exist.
In the immortal words of Goldfinger: I don't expect you defend, Mr. Mastin. I expect you to die.

More seriously, I don't understand why feeling fairly certain that RC is not your target and not knowing who your target is otherwise are inconsistent. Yes, it's hard to impossible to defend this charge, Mastin. I don't believe there's any way for you to do so or I probably would have pushed you to prove it in that way.

But it's precisely because you can't prove it that we need to lynch you today.

Oh, and I've mentioned twice now that while I'm really not sure who your target is, my best guess is Dust. But again, it could be any number of people... perhaps people you have yet to mention.
.

Return to “Completed Large Normal Games”