Mafiascum.net

Local suicide bomber is a role though... like if you hit mafia it's technically worth

In post 975, Awoo wrote:Local suicide bomber is a role though... like if you hit mafia it's technically worth

Not really.
If you correctly identify scum and target them with suicide bomber, you get a cop shot at the cost of a life.
If you miss, you lose one full mislynch.
A full mislynch, is like, a guess at the mafia, a la a cop shot. So loyal suicide bomber does not grant a net positive to the town, as far as I can tell.
It does provide a novel gladiator/governor type effect that you are getting that lynch-esque decision without consensus from your town, which can be cool, but those effects aren't generally expected to improve town winrate.

If you miss it's not worth, but if you hit mafia it's worth.

That is to say, overall don't do it. But you know somebody's gonna do it.

In post 969, Karnage wrote:Further adjustments... I'm replacing "the flu" with "sepsis", the infected player inherits the virus role and the original player dies, and patients don't know if they are even/odd commuters. I'm also removing the bulletproof protection from the virus.

Pandemic Mafia

13 Players (1 Cult, 1 SK, 11 Townies)

1

Virus

(Cult Recruiter/Self-Poisoner)
1

Sepsis

(Serial Killer)
1

Town Quack Doctor

(Essentially a Vig)
5

Town Patient

(Odd Night Compulsive Commuter)
5

Town Patient

(Even Night CompulsiveCommuter)

• Each day phase, players must choose a player to receive an experimental treatment. If they are town, they will be killed. If they are anti-town, they are "cured" and removed from the game.
  Lynching is compulsive
  (plurality voting, if there is a tie BOTH players are lynched)
  
• Virus
  - Can infect 1 player each night. The Virus player dies at the end of the next day phase and the infected player inherits the Virus role. If the targeted player is also targeted by sepsis or quack, the infection fails. If the virus is targeted by sepsis or quack, the virus faction survives with an infected player.
  
• Virus wins when the majority of the town is infected or nothing can prevent that from happening.
  
• Sepsis
  -
  Immune
  to being infected. Can target 1 player each night to try and kill them.
  
• Sepsis wins when all other players have died or have been cured.
  
• Quack Doctor
  -
  Immune
  to being infected. Can target 1 player each night to receive the treatment. anti-town targets are cured, patients are killed
  
• Town Patients
  - Are not told if they are even/odd night compulsive commuters.
  
• Town wins when all anti-town players have been "cured"

I think the original "virus" role will receive an "infected" role PM pregame, they don't become the virus until the end of day 1. the "virus" role PM will be part of the virus PT

I was surprised this didn't fill. I expected players to want something simple-ish during these "unprecedented times". I have a couple observations about trying to correct my own predictive thinking:
-Sitewide demand for games in general has been lower, as far as I can tell, during these "unprecedented times". I giggled at the people who thought it would be , but there was an even more dramatic dip than I expected.
-Publicly disclosing a third party seems like it might be a poison pill, with TemporalLich's setup also not doing well (and as an anecdotal thing, being the reason I queued for neither setup based on my personal preferences). It makes me wonder whether closed setup moderators ought to have a come-to-Jesus about whether it's a player-driven choice to use a third party as well. The last time I can remember Open 3p firing was Jingle firing a game with a lyncher nearly a year ago, and I'm pretty sure Jingle has a pretty strong site rep etc.
-Having guaranteed access to a certain kind of game twist seems to do better than having uncertain access to it, at least in the minitheme queue. I think the virus mechanic was interesting, but for those looking at the setup it maybe didn't look interesting enough for players who didn't roll it and instead rolled town to play against it.
I want to make a pointed note that making that speculation is not the same as the game filling, firing, and checking afterwards whether it was actually fun or not. I was part of a closed large with near-unanimous consensus that the SK in the setup made the game unfun. Its a bit disorienting to see an open setup not even fill potentially because a <small 3p faction> was present so soon afterwards, when this one is implemented far better imo.

In post 978, Isis wrote:I was surprised this didn't fill. I expected players to want something simple-ish during these "unprecedented times". I have a couple observations about trying to correct my own predictive thinking:
-Sitewide demand for games in general has been lower, as far as I can tell, during these "unprecedented times". I giggled at the people who thought it would be signficantly higher, but there was an even more dramatic dip than I expected.
-Publicly disclosing a third party seems like it might be a poison pill, with TemporalLich's setup also not doing well (and as an anecdotal thing, being the reason I queued for neither setup based on my personal preferences). It makes me wonder whether closed setup moderators ought to have a come-to-Jesus about whether it's a player-driven choice to use a third party as well. The last time I can remember Open 3p firing was Jingle firing a game with a lyncher nearly a year ago, and I'm pretty sure Jingle has a pretty strong site rep etc.
-Having guaranteed access to a certain kind of game twist seems to do better than having uncertain access to it, at least in the minitheme queue. I think the virus mechanic was interesting, but for those looking at the setup it maybe didn't look interesting enough for players who didn't roll it and instead rolled town to play against it.
I want to make a pointed note that making that speculation is not the same as the game filling, firing, and checking afterwards whether it was actually fun or not. I was part of a closed large with near-unanimous consensus that the SK in the setup made the game unfun. Its a bit disorienting to see an open setup not even fill potentially because a was present so soon afterwards, when this one is implemented far better imo.

I saw the queue needed mods, threw together a game I thought would be fun but oh well ¯\_(ツ)_/¯
I don't think I have the street cred for modding that others do and I'm guessing that played a role in the game not filling.

games that turn the town into an informed majority can be easily broken and tend to degenerate into assassin in the palace-style gameplay where you just don't really say anything and just vote

here's an example of the last game of this type that was run (to my knowledge) where cultofathena had the game solved by page 3. i think that there's basically no way to avoid stuff like that from happening in AitP-style setups

even if you still wanted to run that setup i think that the town bodyguard would be too much, or at the very least it'd be a poor mechanic? there are some weird moving parts when it comes to how the bodyguard probably protects the flagbearer every night and so if the BG dies then scum should know who the flagbearer is anyways and thus if there are 2 scum alive they've already won and if there's only 1 scum alive then all they need to do is survive another day

I think "the community has tried AitP and determined it's not very fun" could maybe be added to BBMolla's common setup traps thing.

okay I'll add it

A really bad setup idea:

Progressive Mafia

2 Mafia Goons

11 Vanilla Townies

Town may lynch up to N times each day, where N is the cycle number.

Mafia may make N factional nightkills each night.

I'm pretty sure anyone actually trying to calculate the EV for this vanilla variant will burst into a phoenix.

Is this fun? I'm working on the assumption 4/8 is balanced for nightless, so I'm adding 2 town to "compensate" for the 2 week deadline + chaotic voting.

Crossfire

4 Mafia

10 Town

Nightless.
Each player may vote for as many players as they like (a single vote on each).
Votes are not reset after a lynch occurs (so combo lynches can occur if someone else was on L-1 with the person who was lynched not voting for that person).
Deadline for the entire game is 14 days. Town lose if the deadline is reached.

Mafia win by achieving parity with the town. Town win when all mafia are dead.

arguably it's more townsided than vanilla 4/8 nightless if you consider shorter deadlines to help town

seems like fun though, yeah. i imagine that it'd be pretty popular.

I'm not sure whether the core idea of giving townies mechanics they can use to game throw, then jacking up the EV is good game design. It probably is I'm just not sure about it.

Well, only having a fortnight to lynch everyone I see as a handicap. Isn't that a little short of a deadline? So multi-voting and not having to replace votes helps gets lynches quicker.
Would it work better as a 5/12 large game?

It's generally presumed that if we tighten deadlines up to a certain point, we reach a different part of the curve where the deadlines are too short and it is favoring scum because there's no time to scumhunt. We don't really have the data to know for sure where that breaking point is, but we know that MU runs setups generally similar to ours with time constraints nearly as drastically different as this and they have better town winrates, so probably this short of a deadline is not truly a handicap. Of course, running the game with a playerbase that might actually post on the 47th hour several times might lead to some differing results, but this setup would be tied with setups that punish activity-based policy lynching the least.

The multivoting and 1-day nature is where the scumsiding is going to come from, and I think it's plenty of scumsiding. This queueable as is, my guess is a little townsided if run on ms.net but not enough for me to advocate removing an entire VT.

While we're on the topic of nightless, how about a pseudo-nightless?
This is in broad terms similar to the Semi-Nightless setup already on the wiki, as the Mafia likely cannot kill early, but can kill late in the game. In addition, it provides a source of information(/paranoia) to substitute for a normal game's Night, partially removing one of the drawbacks of Nightless.

VOTE@NIGHT

13 Vanilla Townie

4 Mafia Goon

(The number of players may vary, but 20% to 25% of players should be scum.)

Days work as in a normal Mafia game.

Every Night, the players vote in secret on one player to lynch. Voting is not compulsory; votes for No Lynch are treated as not voting.
A majority of the votes (not necessarily a majority of the players) is required to lynch; if no player has a majority, there is no lynch.
The number of votes received by each player, but not who voted for whom, is revealed to the players at the beginning of the next Day.

The Mafia do not have a factional kill; however, they have the factional ability to redirect up to two Townies' votes (if two or more Mafia are alive) or one Townie's vote (if one Mafia is alive) onto whomever they wish. The Mafia may choose two unique targets; they do not need to be placing any of their own votes on their target.

Uh. Doesn't the town just agree lynches with accountability are better and vote no lynch each night? Eventually mafia can redirect enough votes to force a lynch at night but they can derail no lynch at the same point they can derail a lynch on a specific player, so this is just vanilla nightless with certain black flag conditions

Even with no-lynching at night effectively banned, I can't see any reason why the town wouldn't just direct the night vote onto a particular player, outnumbering the scum + their redirects in the process. Just like most scumsided voting mechanics, town can get back to regular voting mechanics via pseudovoting and then all massively piling on the same player.

In post 988, Hectic wrote:Well, only having a fortnight to lynch everyone I see as a handicap. Isn't that a little

too

short of a deadline? So multi-voting and not having to replace votes helps gets lynches quicker.
Would it work better as a 5/12 large game?

FWIW votes not resetting at EOD creates some interesting quick hammer potential in the days leading up to XLO.

In post 992, callforjudgement wrote:Even with no-lynching at night effectively banned, I can't see any reason why the town wouldn't just direct the night vote onto a particular player, outnumbering the scum + their redirects in the process. Just like most scumsided voting mechanics, town can get back to regular voting mechanics via pseudovoting and then all massively piling on the same player.

Generally tactics like this being optimal make the game unfun.

I am trying to figure out how to scale Undertale to 3 scum, out of concern 1 scum 9p day2's kind of aren't mafia.
Please halp
https://forum.mafiascum.net/viewtopic.p ... #p11956213
https://forum.mafiascum.net/viewtopic.p ... #p12022732

Narrowed down to this, please can I have some eyes: https://forum.mafiascum.net/viewtopic.p ... #p12022732

random setup made by me for fun, setup name AlternateKP

Assuming Open (given the thread it's in), it isn't obviously broken (and I really like the choice of scum roles). I'm not sure whether it's going to be win-loss balanced (large mostly-vanilla Opens are a genre that we don't have much data on yet, and appear to have substantially different win rates on different sites), but presumably if it isn't, it'd be fairly easy to tweak to balance it (e.g. if it's scumsided, you could add a VT, or make the Vigilante Vengeful, or the like). One issue is that it'll be hard to collect much data on the balance because it'll depend a lot on how skilled a player the Vig is.