[OLD] Open Setup Discussion and Nominations

[mith]

We will soon have a new administrative game category, Open Games.

Rather than being Moderator designed, Open Games will be chosen by the players. The signup thread will have a small number of games forming at any given time, and when a game is full the first Mod on the waiting list will take that game, and another setup will take its place.

This thread is for discussing potential setups and nominating them for the Queue thread.

Setup Guidelines

In order to be eligible for the Open Game list, a setup must fit the Normal Game definition.

In addition, Open Games will have standardized rules and role PMs (more on this later).

When a setup is designed, it can be posted here, and a list of setups will be maintained.

Nominations

Users may nominate posted setups for inclusion in the signup list. Whenever a new setup is needed, the setup with the most nominations will be taken, and nominations for it will be dropped.

If you are posting a new setup, please nominate in a separate post.

Initially, the Vanilla setup will have a "permanent" place in the signups; when a Vanilla game is ready to run, a new Vanilla signup will take its place on the forming games list. As such, the Vanilla setup does not need to be nominated. It is possible that other setups will be given permanent places as well, in addition to or instead of the Vanilla setup.

Setup List

To be revamped - 11/03/07

[mith]

Example Setup:

Vanilla Mafia, 13 Players

2 Mafia
11 Citizens
Day Start

[Thok]

C9, 7 players

2 Mafia
0-1 Doctor
0-1 Cop
3-5 Vanilla Townie
Day Start

[Trojan Horse]

Basic Twelve Player,

12 Players

3 Mafia
1 Cop
1 Doctor
7 Citizens
Night Start

[Kelly Chen]

I would like to see

Nightless Vanilla, 11 players

4 Mafia
7 Citizens
Day phases only

I could see 12p with an 8th townie also.


As far as the standard vanilla setup: I think 2:7 day start or 2:8 night start will be more viable. Yes, they're skewed against the town, but I think they'll hold more interest. We already play some 2:5 vanilla newbies... And my criticism of these isn't so much that the odds favor scum, as that the scum can win before there's even been much of a game.

I also think that the lower the scum:town ratio, the greater the feeling that there isn't much genuine information to work with. I think this hurts the town's motivation to do a good, thorough job. I also think it makes the game take longer, which seems like a concern if these games are intended to be modded by newbies.

Consider also what becomes of a game where one of the scum is new and inept and gets lynched D1. From there it's essentially an 11p game with 10 townies and one SK. To me that seems like a scenario to be avoided at all costs!

My two cents.

[Trojan Horse]

Assuming all lynches are random, a 12 player nightless game with 9 townies and 3 mafia would be fair (50-50 shot). Taking into account the town's ability to find mafia... maybe 7-3 or 8-3 would be best.

7-4? My money's definitely on the mafia with that one.

[Kelly Chen]

With 3:8 nightless, you're asking mafia to get the town to lynch

at least

five townies. To my ear that's like saying the town lynches the five scummiest players (with

five days

to talk it over), and none of those five are the scum. I don't think it's fair.

The vanilla/mountainous games that have been played here, with the 2:10 night start setup, require scum to get the town to lynch at least 4 townies. And with 1/6 scum, the town has less information.

Even with 4:7 nightless, it isn't going to be a cakewalk for scum. They need at least three bad lynches. This is comparable to a standard mini. Minis usually start out with 3:8 or 3:9, meaning a random lynch is less likely to hit scum than it would here.

[mith]

The 13 Player was given as an example. While the Vanilla setup will be "permanent", we will probably run it with different numbers of players and scum.

[mith]

But for the sake of argument:

I also think that the lower the scum:town ratio, the greater the feeling that there isn't much genuine information to work with. I think this hurts the town's motivation to do a good, thorough job.

People saying there isn't "genuine" information to work with in a Vanilla game = pet peeve.

Anyway, maybe, but isn't this a problem the townies need to deal with, rather than the mods or the setups? Players not playing their best because they aren't motivated is not just a problem in this type of game, and part of my hope with these basic games is that players will discover/rediscover the "pure" form of the game, rather than relying on the usual things they take for granted.

I also think it makes the game take longer, which seems like a concern if these games are intended to be modded by newbies.

I disagree. For one thing, part of the benefit of Open Games is that if a newbie mod flakes, it's easy for someone to step in. For another, I think if anything it's important that new mods learn right off that games on this site can take a while.

(There should probably be a standard deadline system in place, though, to keep these moving. Don't want someone stuck as a "new mod" for a year because of an indecisive town.)

Consider also what becomes of a game where one of the scum is new and inept and gets lynched D1. From there it's essentially an 11p game with 10 townies and one SK. To me that seems like a scenario to be avoided at all costs!

We've had a 1 scum game before, at least once, as a closed setup. It ran just fine.

Now, I'm not saying that our goal should be to take away as much information as possible and make the town rely solely on reading people individually. But sometimes the town does well to start the game, and sometimes there's not a lot to go on other than reading people, and that's not a bad thing. Having to deal with situations like this once every six games... that's not a big deal to me.

Even with 4:7 nightless, it isn't going to be a cakewalk for scum. They need at least three bad lynches. This is comparable to a standard mini. Minis usually start out with 3:8 or 3:9, meaning a random lynch is less likely to hit scum than it would here.

Three bad lynches is not that many in a nightless game. The town needs four good lynches, after all. And with over a third of the players scum, lynches certainly won't be random. There's a lot of room for the scum to steer the first lynch or two away from each other without being obvious about it.

[Kelly Chen]

But for the sake of argument:

I also think that the lower the scum:town ratio, the greater the feeling that there isn't much genuine information to work with. I think this hurts the town's motivation to do a good, thorough job.

People saying there isn't "genuine" information to work with in a Vanilla game = pet peeve.

Anyway, maybe, but isn't this a problem the townies need to deal with, rather than the mods or the setups? Players not playing their best because they aren't motivated is not just a problem in this type of game, and part of my hope with these basic games is that players will discover/rediscover the "pure" form of the game, rather than relying on the usual things they take for granted.

By "genuine" information I'm talking about scum interacting with each other. If 1/3 of the players are in the same scum group, a townie has a much better chance of picking up on interactions between scum, than if say just 1/6 of the players are scum. If there are few scum, more interactions are townie-townie and not actually meaningful.

Even with 4:7 nightless, it isn't going to be a cakewalk for scum. They need at least three bad lynches. This is comparable to a standard mini. Minis usually start out with 3:8 or 3:9, meaning a random lynch is less likely to hit scum than it would here.

Three bad lynches is not that many in a nightless game. The town needs four good lynches, after all. And with over a third of the players scum, lynches certainly won't be random. There's a lot of room for the scum to steer the first lynch or two away from each other without being obvious about it.

I agree, but I'm not sure this is damning. If it turns out that scum don't have to be particularly good to win in this way, I'd go with 4:8.

[HezLucky]

#1 -
Mafia x3
SK
Townie x4
Pick 3: vig/cop/doc/roleblocker/tracker (guaranteed at least 1 of cop/doc)
either one of the above/another townie with some sort of confirmable power ie. double vote

#2 -
Mafia x2
Mafia x2
Doc
Cop
Townie x4
Pick 2: Roleblocker/Tracker/Vigilante/Double Voter

[Max]

I've got 1

3 mafia
3 werewolves
2 docs
2 sorcerors
2 bulletproof townies
2 wolves bane
2 cops
2 seer
1 cultists
2 lovers (1 chosen randomly from above)(1 just lover)

making 20players and 6 scum

[Fiasco]

I'd put Nightless at 3:9 or 3:8. Lynches could be worse than random as well as better than random.

I think the definition of C9 should include the probabilities of cop and doc (50% each). Mods shouldn't choose the setup themselves.

Some nominations:

"Strawberry III" (13 players)

3 mafia
1 cop
9 citizens
night start

Pie C9 (7 players)

1 mafia roleblocker
1 mafia goon
1 cop
1 doctor
3 citizens
day start

Cop C9 (7 players)

2 mafia
1 cop
4 citizens
day start

dethy (5 players)

1 mafia
4 cops (sane, insane, naive, paranoid)
cop head start

dethy^2 (12 players)

3 mafia
9 cops (2 sane, 2 insane, 2 naive, 2 paranoid, 1 retired)
night start

Vengeful Mafia (5 players)

(not sure this would be considered simple/normal enough)

see http://www.mafiascum.net/cgi-bin/wiki.p ... ul%20Mafia

C9++ (12 players)

(even less sure this would be considered simple/normal enough)

see http://mafiascum.net/forum/viewtopic.php?p=389384

[Max]

dethy^2 (12 players)

3 mafia
9 cops (2 sane, 2 insane, 2 naive, 2 paranoid, 1 retired)
night start

Maybe make it a bit stranger like


3 mafia (1 GF, 1 goon, 1 RB)
1 miller
8 cops (2 sane, 2 insane, 2 naive, 2 paranoid)

[Kelly Chen]

@Hez - I'm a little concerned if there are only and exactly 4 townies.

@Max - You didn't give the mafia and wolves anything to claim in that first setup. As far as dethy... I think the point is to come up with setups we might run multiple times, not strange setups.

@Fiasco -

I'd put Nightless at 3:9 or 3:8. Lynches could be worse than random as well as better than random.

That is totally wrong. Do you think that if there's no cop, the town may as well flip a coin?? Scum need to get 6-8 townies lynched under 3:9. Doesn't that seem ridiculous to you?

Here's some scum win odds, with average length of game:

79.8%, 8.9 days - 4:17 day start (i.e. Norinel's Himalayan Mafia)
77%, 2.5 days - 2:5 day start (i.e. a newbie with no power roles)
72.7%, 6.3 days - 4:7 nightless

<---- I propose

66.6%, 7.7 days - 4:8 nightless
64.7%, 4.5 days - 2:10 night start (i.e. current mountainous minis)
60.5%, 5.4 days - the 2:11 day start mith gave as an example
50%, 8.4 days - 3:9 nightless (town can be guaranteed half the wins no matter who's playing)

The longer the game, the more town should be expected to gain an information advantage.

[Twomz]

Wait, is this for mini normals or reg normals, because if it's reg normals, Texas Justice, Hospital, NYPD, ect need to be included (all vig, all doc, all cop). Also, won't they need to be 13+ player games?

[Kelly Chen]

Twomz, what? It's for open (or semi-open) Normals of apparently any size. The "Normal Game" definition excludes the setups you listed.


Another setup I'm interested in:

3:3:6 nightless

: 18.9% town, 37.9% each scum group, 8.5 days avg.

It's about 5% to end with an equal number of both scum and no town.

I think this compares well to Himalayan, considering that there are three teams here.

3:3:5 drops to 14.35% town and 7.3 days, and I'm a bit uneasy about the scum majority there.

[Trojan Horse]

Okay, I've given it some further thought. I agree, we need to skew things percentagewise towards the mafia in a nightless game, since the mafia can't get rid of not-so-scummy people (a big disadvantage). I'm now leaning towards either 3-6 or 4-8 as the best nightless setup.

As for the two scum group nightless game... there's a weird twist that arises here. Let's say we are down to a single townie and two equal sized mafia families. For simplicity, let's say it's 1-1-1 (2-2-1 would be similar). In a regular mafia game, we'd have a no-lynch, followed by a prisoner's dilemma situation. But here, there aren't any no-lynches or night killings. So what happens?

If I was the townie, I'd say this: "I'm the townie. Don't you DARE vote for me. If you do, I'll vote you back. I'll throw the game to the other scum." Whether they believe I'm the townie or not, it seems to me that we might end up with a "happily-ever-after" scenario, with no one daring to vote for anyone else. What do you guys think about that?

[Kelly Chen]

I would think the townie has to play kingmaker. If it's a problem, and the townie won't do it, I would call this n:n:1 scenario game over.

If that rule change is made, then under 3:3:6 nightless, the odds of nobody winning go up to 18.3%... A little high, isn't it? Town's odds aren't affected of course, and the day is only shortened by .18 days...

Perhaps you could break this sort of tie by total post count

[Fiasco]

That is totally wrong. Do you think that if there's no cop, the town may as well flip a coin??

If the scum are reasonably competent, lynching isn't much better than coin-flipping. It's true that in a 50% game, the town has a strategy that wins 50% of the time. But the mafia also has a strategy that wins around 50% of the time; there's no reason they need to actually defend each other rather than behave like a typical townie, though of course they can sometimes do so if they think it makes their odds even better.

Admittedly the strategy that gives the town a 50% chance is slightly easier to play than the strategy that gives the mafia a 50% chance, so games should slightly favor the town (edit: that should say "mafia") given random lynching. 3:8 would probably be more balanced than 3:9, but 4:7 would be very skewed in favor of the mafia.

Himalayan was also very skewed in favor of the mafia. The first two lynches were mafia, yet the mafia still won.

[Yaw]

Question we need to answer: Should setups without fixed numbers for each role be considered open setups? For example, newbie games used to be a clearly open setup, with exactly 2 scum, 1 cop, 1 doc, and 3 townies. Because a breaking strategy was found, it was changed to the current C9 setup, where there's probability involved in the doc and cop roles existing. Since two C9 games could have different exact setups, is this open?

[Kelly Chen]

That is totally wrong. Do you think that if there's no cop, the town may as well flip a coin??

If the scum are reasonably competent, lynching isn't much better than coin-flipping. It's true that in a 50% game, the town has a strategy that wins 50% of the time. But the mafia also has a strategy that wins around 50% of the time; there's no reason they need to actually defend each other rather than behave like a typical townie, though of course they can sometimes do so if they think it makes their odds even better.

Admittedly the strategy that gives the town a 50% chance is slightly easier to play than the strategy that gives the mafia a 50% chance, so games should slightly favor the town (edit: that should say "mafia") given random lynching. 3:8 would probably be more balanced than 3:9, but 4:7 would be very skewed in favor of the mafia.

It is not just "slightly easier" for town to agree to lynch randomly using the Dice function, than for ALL the scum to be so "reasonably competent" that they're no more likely to be lynched than a townie. Not that many players are that competent. If the game has a scum team like that, the scum should win a hell of a lot more than half the time, and not worry about being thwarted by use of the Dice function.

4:7 gives town a 27.3% chance if they play using Dice. Put differently, town has 27.3%, at best, if the scum are all no more suspicious than anyone else. These are both situations where the town should LOSE. The 50/50 mark doesn't go where one side is sleepwalking or the other side plays a flawless game.

The stats for 3:8 nightless are 54.6% and 7.4 game days. This is not a good use of the scum players' time.

[Fiasco]

It is not just "slightly easier" for town to agree to lynch randomly using the Dice function, than for ALL the scum to be so "reasonably competent" that they're no more likely to be lynched than a townie. Not that many players are that competent.

Are you sure? It's much harder to avoid lynching if you're actively trying to protect your scum-mates than if you don't care whether or not your scum-mates get lynched. We're talking about the latter situation; or if not, we need to take into account that scum are harder to lynch because part of the town is scum.

I'd guess that in a 75%-chance-of-scum-win game, a reasonably competent scum team can win 70%-75% of the time just by throwing away their role PMs and not talking at night, against

any

town. That shouldn't happen; ideally, given equal skill, each side should win half the time.

I guess scum play is more skill-dependent than town, so games with skilled players should favor the town more than games with unskilled players.

[mith]

Yaw, I'm fine with "Semi-Open" games, as long as the randomness in the setup is simple and easy to understand. (And in the case of C9, I agree that we need to make it clear what the percentages are, whether 25-25-25-25 or something else. And to ensure the choice is random, I will likely just do all semi-open randomization myself and send the setup to the Mod, unless I happen to be playing.)

Fiasco, I don't think Vengeful is too complicated, and it fits all the current "Normal" definitions. C9++ on the other hand, I'm not so sure about. It's only Semi-Open in the sense that the setup generation is public knowledge, but there's so many possibilities that the play would likely be more similar to a Closed game; I'd prefer to see it run as a Mini Normal.

(Though the setup generation or something like it could also be used to generate purely Open setups.)

[Kelly Chen]

It is not just "slightly easier" for town to agree to lynch randomly using the Dice function, than for ALL the scum to be so "reasonably competent" that they're no more likely to be lynched than a townie. Not that many players are that competent.

Are you sure? It's much harder to avoid lynching if you're actively trying to protect your scum-mates than if you don't care whether or not your scum-mates get lynched. We're talking about the latter situation; or if not, we need to take into account that scum are harder to lynch because part of the town is scum.

If you actively defend your scum partners and get suspected for doing so, you're not actually doing that good of a job. If you can defend your scum partners and not get suspected for it, you're even more competent than what we're talking about here.

I'd guess that in a 75%-chance-of-scum-win game, a reasonably competent scum team can win 70%-75% of the time just by throwing away their role PMs and not talking at night, against

any

town. That shouldn't happen; ideally, given equal skill, each side should win half the time.

Replace the 75%'s with 50% or 1% or 99% and the first sentence is still true. If the mafia play blind, town skill is hardly being measured.

In my opinion, playing blind is basically cheating, and a waste of the other players' time.

I guess scum play is more skill-dependent than town, so games with skilled players should favor the town more than games with unskilled players.

But the odds statistic doesn't reflect a situation of equal skill. It reflects either a town that is horrible, or scum that is either skilled or cheating to the point that town skill is irrelevant.

So you can't use the odds statistic to ensure that skilled towns get a fair shake, without royally screwing over skilled scum in every other game.