Mafiascum.net

Add two VTs to switch too, make it 16p.

Thanks for the input BBMolla.

POLYGAMIST
Yeah this is basically a hydra version of Lover's Mafia. Adding another townie in a nightless would break it in favour of town. 71% winrate on random lynch.

SWITCH
Yeah I think basically every nightstart game is horribly outdated. It's kinda hard to boycott nightstart as a whole when they hold up the open queue for a week. I'm gonna recommend overhauling to daystart and seeing how it plays.

DUCK DUCK GOOSE
I assume by 2 VTs you mean 2 1-shot PGOs? In that case yeah sounds good.

HARD BOILED
I'm not convinced this is 100% broken yet... but this is a discussion that needs to be had for sure IMO.

PICK YOUR POISON
My understanding is that Stack The Deck was specifically designed to counter Pick Your Poison's design flaws? Why is Pick Your Poison still running?

In post 50, BBmolla wrote:Add two VTs to switch too, make it 16p.

I think it's gonna have to be either/or. Either add VTs or daystart. Honestly I think daystart will fix most of this setup's issues. I might run that to see how it goes.

I mean the inherent problem here is that if mafia and SK repeatedly fuck each other's switches over, town has a hilarious inordinate amount of power. It can't be saved from that inherent swinginess in the concept, but I think it can be saved by the swinginess of N0 kills.

In post 49, BBmolla wrote:Play Dota 2 it's better

You're not wrong.

In post 52, Mathdino wrote:

In post 50, BBmolla wrote:Add two VTs to switch too, make it 16p.

I think it's gonna have to be either/or. Either add VTs or daystart. Honestly I think daystart will fix most of this setup's issues. I might run that to see how it goes.

I mean the inherent problem here is that if mafia and SK repeatedly fuck each other's switches over, town has a hilarious inordinate amount of power. It can't be saved from that inherent swinginess in the concept, but I think it can be saved by the swinginess of N0 kills.

A typical mini has 3 mafia, 10 town

so 3 mafia, 1 sk, 10 town is silly

it should be 16.

In post 51, Mathdino wrote:POLYGAMIST
Yeah this is basically a hydra version of Lover's Mafia. Adding another townie in a nightless would break it in favour of town. 71% winrate on random lynch.

Would it help if a townie pair was added and the mafia quad was split into two mafia pairs? That might defeat the whole point of the setup but it looks like the quad being together is what makes random lynching so prevalent.

In post 51, Mathdino wrote:DUCK DUCK GOOSE
I assume by 2 VTs you mean 2 1-shot PGOs? In that case yeah sounds good.

I do.

In post 51, Mathdino wrote:HARD BOILED
I'm not convinced this is 100% broken yet... but this is a discussion that needs to be had for sure IMO.

It's just a bit aimless at the moment. Too late to look at now, will look in depth later.

In post 51, Mathdino wrote:PICK YOUR POISON
My understanding is that Stack The Deck was specifically designed to counter Pick Your Poison's design flaws? Why is Pick Your Poison still running?

no idea

POLYGAMIST
Then it would be 2:4 nightless or 2:5 nightless but with lover pairs. Basically an involuntary hydra game.
Not sure this is salvageable on a conceptual level.

HARD BOILED
I'll add that to the setup list. I remember it was broken years ago, and ABR added a townie to make it 13p and less broken.

What's the random lynching winrate for Polygamy if town is just made into 8 VTs?

That increases town winrate for random lynching doesn't it.

mm.

In post 57, BBmolla wrote:What's the random lynching winrate for Polygamy if town is just made into 8 VTs?

probably very townsided

Lovers and Losers

9 Players:

3 Mafia Lovers

4 Town Lovers (2 pairs)
2 VTs

• Daystart
  
• Nightless
  
• Only one VT may get lynched and only one Town Lover may get lynched.
  
• If someone attempts to lynch a role that has already been lynched, the game ends in a mafia victory.

Lovers and Loners

9 Players:

3 Mafia Lovers

4 Town Lovers (2 pairs)
2 VTs

• Daystart
  
• Nightless
  
• There are two public groups, "Lovers" and "Loners." The two VTs are in "Loners" and the four Town Lovers are in "Lovers." Pregame, Mafia choose one player to be in the "Loner" group and two to be in the "Lover" group.
  
• Only two members from each group may be lynched
  
• If at any time mafia outnumber or equal the number of town, they win.

Eh?

I'm bad with EV stuff but I think random lynching is unideal in both these setups. Might be the right direction?

4 Mafia Lovers in unworkable due to random lynching I think, but 3 is more fun anyway.

I think the top EV is something like 17% town winrate so probably the bottom is better.

Actually if I did the math right, I think town has a 77% chance of victory on the second setup soo

Someone confirm that, I'll think on it to see if there's anything I can do to find something in between with a much more reasonable winrate.

Edit: Ignore me I'm doing some seriously wrong math. Someone calculate the winrates on those for me.

Is it top 55% and bottom 77% for town win?

(assuming optimal strategy for the bottom one which is lynching VTs first.)

55% is at least a little better than 60% eh?

I'll think on it more.

Polygamy For All

14 Players:

4 Mafia Lovers

10 Town Lovers (5 pairs)

• Daystart
  
• After every lynch, mafia choose two of their players to make publicly lynch immune. (So town gets two lynches total)

42.9% town winrate, there, done.

Polygamy For You and Me

16 Players:

4 Mafia Lovers
1 Mafia Double Lover

11 Town Lovers (5 pairs)

• Daystart
  
• Mafia Double Lover is in Love with a Townie as well as all his Mafia buddies.
  
• After the first mislynch, the 4 Mafia Lovers become publicly lynch immune.
  
• After the second mislynch, if the Mafia Double Lover lives, mafia win.

48% town winrate boom, happy?

4 Mafia Lovers itself is so insanely townsided it's hard to even balance around it lmfao

Btw I saw the namecall and wanted to respond to it--

In post 19, Flubbernugget wrote:Anecdotal evidence for what it's worth, but when I first came to this site I remember prs having a bit of an aura of being a necessary evil. Hearing some of mastina's experiences with the nrg sound like they're now being treated more like a commodity. I wonder if that's contributed to setup swing at all.

The two are not mutually exclusive, and in fact they both apply.

Power Roles are still every bit a necessary evil as they were, if not moreso. Towns inherently require some level of power to counter the innate natural advantage scum players hold. This level of power is such that the town must be able to hold a playing field with the scum: they must be able to overcome not only the inherent scum advantages, but ALSO the inherent disadvantages town players possess even as PRs. (That being, as the uninformed majority, town power roles are uninformed and thus usually perform sub-optimally compared to theoretical statistics for what random targeting would give.)

In a Closed Mini Normal, the magical number is 3-4 power roles of moderate strength (and when on the heavier side, usually with a scum role also of moderate strength). However, therein lies moderators' usage of power roles as a commodity: they are in fact a material which, in a manner of speaking, are "bought" or "sold" depending on the moderator's intentions and design for the setup.

It is possible to run a Mini Normal with less than 3-4 power roles. (Selling them.) For instance, a 10:3 Open with a JK and a Cop is reasonably close to balanced in spite of it being less than the typical, thanks both to the strength of the roles and to the nature of the setup being disclosed from the onset.

It is possible (and increasingly common) to run a Mini Normal with more than 3-4 power roles, and/or for those power roles to instead of being moderate strength to be the five big PRs of incredible strength (Vig, Cop, Watcher, Jailkeeper, Doctor, in approximate order of how strong I'd rate them from a design perspective most to least). Buying them. To accomplish this, a moderator needs to provide compensation to the scumteam in order to keep the playing field level.

The magical formula has been shown to work countless times to provide respectively balanced games that are reasonably low on swing--diverging from that mold in EITHER direction will thus introduce a certain degree of uncertainty, but moderators have the right to do so if that is what they choose to make. This degree of uncertainty is where swing is introduced. You need to take your best guess for what is appropriate compensation.

We mostly get it right, but we can't guarantee success, especially when things are more swingy, because the more swing inherent in the design of the setup, the more room to error in judgement there is as far as townsided/scumsided goes--it's harder to keep the level playing field when you are dealing with additional variables.

Personal opinion, I think that overall, moderators are for the MOST part doing well. They are providing an incredible level of diversity in setups, and yet most of them are reasonably close to the balance mark even if the swing is higher than I personally would prefer it to be. For the most part, I feel that moderators are doing their jobs in providing an experience where players are placed in situations where at the beginning of the game, both sides had loosely equal chances of winning.

When you play a Normal, you aren't able to put the pieces of the puzzle together until late in the game if you can put them together at all. (In other words, you can't break the game open with a massclaim D1; you won't have the answers until much later for what the mods did.) I think this is a good thing, and should in fact be encouraged.

I would, however, prefer to see more 'conventional' setups, i.e., less-crazy ones with a moderate level of strength run. Not to the point where they'd dominate altogether (because too many of these being run can lead to the aforementioned "can break the game on D1" problems not to mention it becoming boring and monotonous), but it'd still be nice to have a fair amount of RECENT setups I could point to and say, "These are setups which were low on swing and high on balance", which is something a bit difficult to do because while I can point to plenty of setups I thought were balanced or close-to-balanced nearly every single one of them comes with the caveat of 'albeit swingy'.

I think an important factor is making games fun to play and to mod. If it's too boring then people will flake or replace out more. I'm not as concerned about observed win rates -- IMO any game can be won or lost and the percentages are nothing more than a guide. Challenges are good...

In modding opens, I like to have something to do other than just vote counts.
The more potential interesting night action the better.
So nightless and lover type games are totally uninteresting to me. Straight mountainous and white flag are barely tolerable to mod.

In playing opens, I like to have a puzzle to solve. My ability to read people on posting nuances is limited at best. To that end the games with claims to evaluate, like x9++ and pick your power x/y are my favorites, with anything MB next.

Personally i think the best open setups are the ones where, not only is the setup mostly different from game to game (ie c9++ or stack the deck), but there's also some player control in how things play out – i think pick your power x/y is the kind of setup that new open setups should be modeled after. Especially the whole drafting mechanic adds a lot to the game in terms of player influence and strategy that other variable setups don't really seem to have.

PYP X/Y needs to be hosted more often IMO

Polygamist should be reduced to 3 sets of town lovers.

Alright let's revive this a bit.

I'd like to run an experimental open setup soon, whether it be one of the setups under review or a modified version of a problematic existing setup.

BBMolla's Polygamy For All seems actually really cool. Lot of players but not with the game length that coincides with larger games. My hesitation is in whether people would actually be interested in playing it given how Polygamist went.

Are there any less mechanical setups you guys think we can play around with?

MASONS & MONKS

2 Mafia Goons
2 Werewolf Goons
2 Masons (not Mafia)
2 Monks (not Werewolf)
5-7 Vanilla Townies

- Masons/monks are confirmed to each other as not mafia and werewolves, respectively.
- Masons cannot both be werewolves, and monks cannot both be mafia.

Got a request for this setup, which has a 100% town lossrate on the wiki.

The problem with this setup (within the constrains of multiball) for town is that it can't actually confirm anyone. (Please no snarky comments about how the problem is multiball, I get it)

I wonder if this would be balanced just by converting a townie to an activated innocent child?

Maybe add one doctor?

If we take Fire and Ice as roughly balanced, a doctor would turn it into Fire & Ice with the additional ability to confirm townies when one scumteam gets killed.

Thinking along the lines of bulletproof.

In post 72, Mathdino wrote:If we take Fire and Ice as roughly balanced, a doctor would turn it into Fire & Ice with the additional ability to confirm townies when one scumteam gets killed.

Thinking along the lines of bulletproof.

You should note that Fire & Ice can turn two kills in the same nightkill equivalent to doctor.

In post 49, BBmolla wrote:

In post 2, Alisae wrote:Hard Boiled - atleast give scum a roleblocker, daytalk, and maybe a Rolecop? Otherwise its really townsided.

Force town to choose Tracker (it's gamethrowing to do otherwise), give Mafia a Rolecop. Or, more interestingly, force town to choose Vig and add a VT and keep Mafia as is.

A lot seems to depend on D1. The tracker/vig dichotomy allows adaptability for town. Plus, there's still an optimising strategy for if T/V goes vig. Hell, the 12p version was broken in half by making T/V go vig.

You know, in keeping with the theme, I actually think giving scum a Jailkeeper would be the most thematic and balanced way to fix this. Can be used defencively (maybe scum fakeclaims vig) or offencively.

I'd be very willing to run this with a scum JK I think.