[OLD] Open Setup Ideas and Discussion

[shaft.ed]

in most cases end game will be day 2, and it is all about solving the logic puzzle resulting from the inevitable massclaim. No lynch might be optimal for day 1.

Well let's see how this would play out:

I'd say the Miller's have incentive to claim D1, one goon has to counter or else they are cleared so you have:
2 Millers + 1 fake Miller, town has 33% chance of hitting scum. This is slightly worse than a random non-self lynch so the town may not pursue it, however the benefits outweight the very slightly lowered odds. so the town either lynches a Miller (A: 2/3rds) or scum (B: 1/3rd).

A Miller lynch: Scum can't NK the other miller. Town is left now with a 50% odds lynch from the remaining two millers. If the cops survive the night (50% chance), either cop may catch the remaining scum (1/4) outright with a night investigation, or clear a townie/cop (2/4) or accidentally clear a GF (1/4). Town's best play is most likely to decide between the two millers unless a cop hits a guilty, which is a decent probability. If that doesn't pan out there's about a 50% chance they lose here.

B Scum lynch: Town's in much better situation here. The scum is likely forced to kill one of the Millers since they are now confirmed innocents. Cops both survive with investigations because of this each having a 25% chance of picking up the lynched players scum buddy. Town is not likely to lynch an innocent investigation, which may help the GF.

tl;dr: 2/3rds chance scum have a very likely chance of winning
1/3rd chance town are in a very strong position.

Anyway I kinda like the set up, sorry for being an ass. Please add a townie; 6 v 3 is a bit crazy don't you think?

[Adel]

Nope, 6 vs. 3 isn't crazy in this case. It is a logic puzzle. A correct massclaim at the correct time swings the game from being very scum balanced to very town balanced.

Day 1 miller claim + day 1 lynch of claimed miller probably isn't the dominant pro-town tactic. Millers claiming day 1 greatly increases the scum's chances of hitting one of the cops.

[shaft.ed]

Nope, 6 vs. 3 isn't crazy in this case. It is a logic puzzle. A correct massclaim at the correct time swings the game from being very scum balanced to very town balanced.

Day 1 miller claim + day 1 lynch of claimed miller probably isn't the dominant pro-town tactic. Millers claiming day 1 greatly increases the scum's chances of hitting one of the cops.

But forcing the scum into a D1 miller claim makes the logic puzzle start much earlier (this could be arbitrary). It also gives a narrower pool from which the cops need to investigate and has a huge payoff should the town lynch correctly.

[Xylthixlm]

Mutually Assured Destruction

4 scum, 8 town.
Day 1, the players to decide who to lynch normally. Then the lynched player gets a kill before they die, then the killed player gets a kill, etc, until the remaining players are all town or all scum. The player can choose not to kill, in which case it goes back to lynching. Players (including the player about to make the next kill) can talk at any time.

[Adel]

Nope, 6 vs. 3 isn't crazy in this case. It is a logic puzzle. A correct massclaim at the correct time swings the game from being very scum balanced to very town balanced.

Day 1 miller claim + day 1 lynch of claimed miller probably isn't the dominant pro-town tactic. Millers claiming day 1 greatly increases the scum's chances of hitting one of the cops.

But forcing the scum into a D1 miller claim makes the logic puzzle start much earlier (this could be arbitrary). It also gives a narrower pool from which the cops need to investigate and has a huge payoff should the town lynch correctly.

how about we make it a cop head start?

[Mr. Flay]

Mutually Assured Destruction

4 scum, 8 town.
Day 1, the players to decide who to lynch normally. Then the lynched player gets a kill before they die, then the killed player gets a kill, etc, until the remaining players are all town or all scum. The player can choose not to kill, in which case it goes back to lynching. Players (including the player about to make the next kill) can talk at any time.

I'd put a limit on how long they can wait before killing; maybe 12-24 hours, to keep things snappy.

Also, you probably want a different name, as Thok(?) is running a series of nuclear-themed games with that title.

All Fall Down

?

[Xylthixlm]

Chain Reaction

The no-time-limit was because, honestly, I don't expect people to ever decide to go back to lynching. But adding a time limit might add some incentive to do so.

[ac1983fan]

I thought up this set-up, and it seemed pretty neat in my mind:

Skum & scum

1 Goon
1 Serial Killer
4 Townies
1 Cop
1 SK Cop

Each cop has one-shot NK immunity.
Serial Killer is NK immune
Both cops are sane
Cop finds out if the investigated player is anti-town or pro-town (i.e. gets guilty on both Goon and SK)
SK Cop finds out if the investigated player is the SK or not.
If the SK Cop finds the SK, he gets a 100% success rate kamikaze daykill for the SK only, simply by posting

daykill: (sk's username goes here)

in any post. Both the SK and the SK Cop will die. If the SK is lynched, the SK cop becomes a standard sane cop.

Thoughts? It's probably far from perfect...

[Max]

Chain Reaction

I've run a game called that what about:

Everybody's Vengeful

[Xylthixlm]

Chain Reaction

I've run a game called that what about:

Everybody's Vengeful

How about

Popcorn

?

Also, rethinking the math I think 9-3 is better than 8-4.

[shaft.ed]

Nope, 6 vs. 3 isn't crazy in this case. It is a logic puzzle. A correct massclaim at the correct time swings the game from being very scum balanced to very town balanced.

Day 1 miller claim + day 1 lynch of claimed miller probably isn't the dominant pro-town tactic. Millers claiming day 1 greatly increases the scum's chances of hitting one of the cops.

But forcing the scum into a D1 miller claim makes the logic puzzle start much earlier (this could be arbitrary). It also gives a narrower pool from which the cops need to investigate and has a huge payoff should the town lynch correctly.

how about we make it a cop head start?

Yes good.

Nominate

really like to see a cop fake a miller claim

[Adel]

Don't Cut the Red Wire!

1 Serial Killer
1 Mafioso
1 Bomb (when targeted for kill, takes killer with him)
1 (ONE-SHOT) Bulletproof Townie
1 Townie

Day Start
Lynches Compulsory
Nightkills Compulsory

It creates an interesting dilemma if the townie is lynched.

was nominated by two people:

I'm noyt sure if its balanced but:

Nominate: Don't Cut the Red Wire!

seems logical. I think I would like

don't cut the red wire.

caboose was on the right track by raising the question:

So what happens here:
Townie Lynched D1
Bulletproof Killed N1
Mafioso Lynched D2

Who wins?

Also, is the SK nightkill immune?

It looks like it would be better for the town NOT to massclaim.

by pointing out some possible conflicts caused by the must-lynch and must-nk mechanics.

These issues, unfortunitly, were not resolved before the game was run:
http://www.mafiascum.net/forum/viewtopic.php?t=10486
it was over in 3 pages. result: 1 scum player and 1 town player left alive in D2 with no way for a lynch to occur. The mod ruled it a tie.

from the postgame:

Sorry for wasting your time folks, but this entire game was to prove that it is possible to create an intriguing setup with a fatal flaw that can be easily exploited by anyone paying attention, and then push it through the open game queue without anyone noticing as proof that the open game system is flawed in that it does not check for setup brokenness. There is no way to actually salvage this setup without bastardizing the concept behind it.

<_<

I only chose it because people nominated several times in the Open Game Nomination thread. Next time, mod the game yourself to prove a point.

Note that I did have queasy feelings about how the game would end, and especially why the author of a setup join his game. I guess I should have known better, but me getting hoodwinked by this setup really just points to my noobness as a mod as much as it is the open game system (in which most mods are first-time mods).

Which is why I did this, in fact; to ensure that the first-time mods that populate the open queue get the modding experience they deserve.

Well many wanted to play the game. I think I know a way to fix it but JD in the future how about a little warning to me hmm?

Can't. That would have ruined the point, which was to show that there needs to be more oversight for obviously broken setups in the open game department.

my thoughts:
1. the only real problem with the setup (other than JDodge's lack of sincerity) was in the application of the must-lynch and must-nk special rules. Many of the probable outcomes of the game would result in a draw or otherwise lack a clear winner. Kudos to Caboose for actually catching on, and shame on the rest of us for not following up on his insight.
2. No rigorous analysis of the setup and its probable resulting game states occurred. JDodge (probably for the last time in this thread) was granted the benefit of the doubt for being one of the most experienced mafia players on this site. I figured that the point behind the setup was some sort of prisoner's dilemma for the scum during N1, and that was his "interesting dilemma if the townie is lynched".
3. I don't think that farside22 (the listmod) or SilverPhoenix (the game's moderator) did anything wrong.
4. While there isn't anything we can do about people insincerely suggesting setups, I would like to offer the following change in process:

current process:
a. setup presented
b. setup commented upon
c. setup nominated
d. list mod adds setup to queue if she approves it

proposed alternative:
a. setup presented
b. setup commented upon
c. setup nominated
d. list mod lists the nominated setups she is considering adding to the queue, and asks for players to "

certify

" the setup
e. as a quality assurance step, players who take to time to do a careful review of the setup can "

certify

" the the setup is free of significant errors and oversights
f. listmod adds certified setups to the queue if she is confident of the competence of the certifying scummers.

my list of lessons learned:
1. the informational shortcut of "if the setup is suggested by a really experienced scummer than it must be sincere" can be false
2. the informational shortcut of "if the setup is suggested by a really experienced scummer than it probably isn't broken" can be false
3. so many setups are suggested that it is impossible for one person to deeply analyze all of them. If we had a little warning before a setup is added to the queue then it would be easier for interested scummers to take a deep look and identify problems before they crop up in-game.


~~~

@JDodge, now that you've proven that you can push a broken setup of your design through the review process:

1. What lessons do you think we should learn from this?

2. Which broken Open Game setups motivated you to try your experiment?

3. What other ways did you try to address the problem before you resorted to your experiment?

4. Do you recommend this kind of "destructive testing" to other scummers who think they have identified a problem with some aspect of the site?

[Adel]

Nope, 6 vs. 3 isn't crazy in this case. It is a logic puzzle. A correct massclaim at the correct time swings the game from being very scum balanced to very town balanced.

Day 1 miller claim + day 1 lynch of claimed miller probably isn't the dominant pro-town tactic. Millers claiming day 1 greatly increases the scum's chances of hitting one of the cops.

But forcing the scum into a D1 miller claim makes the logic puzzle start much earlier (this could be arbitrary). It also gives a narrower pool from which the cops need to investigate and has a huge payoff should the town lynch correctly.

how about we make it a cop head start?

Yes good.

Nominate

really like to see a cop fake a miller claim

massclaim letter nine

1 godfather
2 mafia goons
2 cops
2 millers
2 townies

cop headstart (scum can talk during N0 but can not kill)

this is a sincere setup, and I fully expect that it will yield an interesting game for players who enjoy logic puzzles and fakeclaims and don't mind playing "follow the cop".
I suggest that the mod uses short deadlines, like

# of living players * 1.5

(rounded up) with a minimum deadline length of 10 days.

[shaft.ed]

Another current problem with the Open Queue is high demand while no one is nominating "tried and true" set ups. So basically less rigorously discussed experimental set ups are getting pushed through to meet demand. For example, Red wire appears to have only had two noms where three was the previous protocol.

I suggest more running of established games and more discussion of new setups.

[JDodge]

@JDodge, now that you've proven that you can push a broken setup of your design through the review process:

1. What lessons do you think we should learn from this?

2. Which broken Open Game setups motivated you to try your experiment?

3. What other ways did you try to address the problem before you resorted to your experiment?

4. Do you recommend this kind of "destructive testing" to other scummers who think they have identified a problem with some aspect of the site?

1. We need more oversight of the open game queue.

2. None, I just saw the flaw and wanted it fixed.

3. I did not, as I felt other methods of addressing the problem would just be met with a resounding "THIS IS A PROBLEM THAT DOESN'T EXIST".

4. Only in extreme cases, and when done with extreme caution.

Oh, and thanks for ruining my chances of blitzing MD with three thought-provoking threads sure to earn me the Professor Mafia scummy.

That's

all shot to hell now, isn't it?

[Darox]

Oh, and thanks for ruining my chances of blitzing MD with three thought-provoking threads sure to earn me the Professor Mafia scummy.

That's

all shot to hell now, isn't it?

ITT We learn what JDodge's real motivations were.


Something that would help a lot with this is establishing a list of successful Open setups that are easily accessible (Say, on the front page with links to wiki articles explaining each setup) to prevent them from fading into obscurity and grab more nominations, and a screening process for any suggested games not part of the list. At the very least this would involve running hypothetical scenarios of all the possible outcomes to ensure there aren't any obvious flaws.

[/obvious solution]

[Oman]

3. I did not, as I felt other methods of addressing the problem would just be met with a resounding "THIS IS A PROBLEM THAT DOESN'T EXIST".

Assumption is the mother of all fuck-ups. This experiment loses a lot of credibility because of this point.

Oh, and thanks for ruining my chances of blitzing MD with three thought-provoking threads sure to earn me the Professor Mafia scummy.

That's

all shot to hell now, isn't it?

Well done.

[JDodge]

3. I did not, as I felt other methods of addressing the problem would just be met with a resounding "THIS IS A PROBLEM THAT DOESN'T EXIST".

Assumption is the mother of all fuck-ups. This experiment loses a lot of credibility because of this point.

Explain how this experiment loses credibility when:

1. I came up with my hypothesis (the open queue does not have enough oversight)

2. Ran an experiment to test my claims

3. Empirically proved my hypothesis true

Learn scientific process, bitch.

[Oman]

Yeah, I understand the scientific method. I really do. Yes I agree that your experiment supports your hypothesis (supports, not proves, i.e. "Learn scientific process especially in regards to repetition of experiment"), but I'm saying that you lose credibility in the application of the experiment. Really, if you had complained about this, I'd be completely on your side, and actually congratulating you for the well performed act. Instead I say, yeah you proved your point, but it shouldn't have been done like that.

[JDodge]

Yeah, I understand the scientific method. I really do. Yes I agree that your experiment supports your hypothesis (supports, not proves, i.e. "Learn scientific process especially in regards to repetition of experiment"), but I'm saying that you lose credibility in the application of the experiment. Really, if you had complained about this, I'd be completely on your side, and actually congratulating you for the well performed act. Instead I say, yeah you proved your point, but it shouldn't have been done like that.

How does that affect my credibility at all?

You agree that my hypothesis is credible, then argue that because of the experiment proving it credible, I have lost credibility. That makes no sense at all.

[somestrangeflea]

Oh, and thanks for ruining my chances of blitzing MD with three thought-provoking threads sure to earn me the Professor Mafia scummy.

That's

all shot to hell now, isn't it?

I don't think highlighting a flaw in the Open Queue system is a particularly good contribution to

Mafia

. Behind the Scenes award, maybe, but no Prof Mafia award for you.

[Darox]

ITT people manage to completely miss the point.

It doesn't matter how JDodge showed the problem or any other non-sequitur phrases.
It's about the problem and how to fix it.

[crywolf20084]

Yeah, I understand the scientific method. I really do. Yes I agree that your experiment supports your hypothesis (supports, not proves, i.e. "Learn scientific process especially in regards to repetition of experiment"), but I'm saying that you lose credibility in the application of the experiment. Really, if you had complained about this, I'd be completely on your side, and actually congratulating you for the well performed act. Instead I say, yeah you proved your point, but it shouldn't have been done like that.

How does that affect my credibility at all?

You agree that my hypothesis is credible, then argue that because of the experiment proving it credible, I have lost credibility. That makes no sense at all.

Not to mention with approving of the hypothesis, then disagreeing with the way he went around to prove it, is hypocritical.

[Korts]

I don't see what the big fuss is about. JDodge didn't really ruin anything, he just pointed the flaw out in practice instead of raising his concerns. I'm fairly sure that if he had done the latter he would have been ignored. And anyway, the setup he proposed was clearly designed to end quickly while still proving the point.

BTW I agree with shaft.ed. More tried and tested games, less experimental unless thoroughly discussed and approved.

[Shanba]

I don't see what the big fuss is about. JDodge didn't really ruin anything, he just pointed the flaw out in practice instead of raising his concerns. I'm fairly sure that if he had done the latter he would have been ignored. And anyway, the setup he proposed was clearly designed to end quickly while still proving the point.

BTW I agree with shaft.ed. More tried and tested games, less experimental unless thoroughly discussed and approved.

I agree with this. Actually, I thought open games were supposed to be relatively simple, not all this jazz with enablers, death millers, cops of weird sanities, etc. etc.

I accept that things like Dethy and Vengeful defy that in practice, but those sorts of setups are impressive mostly in their simplicity, their elegance of design and their clarity of goal. Vengeful, for example, is an excellent way of giving some depth and interest to a five player setup, by giving the town an excess of kills, whilst keeping a proper scumgroup and therefore allowing for useful scum hunting. I dunno. I just think this current stuff is all a bit weird.

I've seen xyl's setup in scmchat before, run by Skruffs. The name "STD mafia" always seemed interesting to me :P