[MARCH CHALLENGE]

[callforjudgement]

Microtown

• Microtown is a 9-player game using the Smalltown mechanic: there are 36 possible setups, each of which has the same roles, but a different assignment of alignments to roles (each assignment of 7 town and 2 scum is possible).
  
• Each player's role (but not alignment, obviously) is confirmed by the moderator at the start of the game. (This means that the roles can be picked via WIFOMdraft, if desired, rather than randomised.)
  
• Mafia members automatically get the Multitasking modifier, allowing them to perform a kill and a night action at the same time.
  
• Normal phase sequence and kills (i.e. Day Start, scum have a factional kill).
  
• Scum win by killing the entire town, and vice versa; play beyond 50% is possible if town have a power role combination that could potentially allow them to eliminate all the scum first, despite not outnumbering them.
  
• The roles are as follows:
  • Night 3 Vigilante
    
  • 1-shot Roleblocker
    
  • Personal Redirector (investigative roles get "your target"-style results, so won't know they've been redirected; targets both the player being redirected and the player being redirected onto, thus can't redirect onto the Ascetic)
    
  • Rolestopper
    
  • Ascetic Alien Babysitter (by analogy with Alien Jailkeeper): all actions aimed directly at the Babysitter's target fail (not just kills), but if the Babysitter dies overnight, so does their target
    
  • Night 2 Informed (informed of the alignment of a randomly selected living townie, not counting themself, at the end of night 2)
    
  • Traffic Analyst (always gets "can communicate" on the Messenger, thus will likely investigate someone else)
    
  • Unlimited Messenger; can send an unlimited number of private messages to other players (via the moderator), day or night, but the recipients cannot respond privately (they'll have to respond in-thread or not at all)
    
  • Doublevoter (the variant which casts one vote and it's given double weight in the voting)
• Action resolution is Reasonable Action Resolution. That gives the following resolutions for loops of manipulatives:
  • Rolestopper, Babysitter target the same player: both actions fail, other actions targeting that player succeed
    
  • Redirector attempts to redirects actions onto a target who's rolestopped or babysat (but not both): redirection fails, actions target their original target
    
  • Redirector attempts to redirect actions of a target who's rolestopped or babysat (but not both): redirection fails, actions target their original target
    
  • Roleblocker attempts to block a player who's rolestopped or babysat (but not both): roleblock fails, that player's actions go through
    
  • Redirector attempts to redirect the Rolestopper or Babysitter onto a new target: redirection succeeds, unless the other rolestopping role targeted one of the Redirector's targets
    
  • Redirector attempts to redirect the Rolestopper or Babysitter onto their current target: the redirection has no effect
    
  • Redirector attempts to redirect the Roleblocker, who is roleblocking someone else: redirection succeeds
    
  • Redirector attempts to redirect the Roleblocker, who is blocking the Redirector: redirection fails, the Roleblocker blocks the Redirector and not anyone else

I've been wanting to make a good open Smalltown for a while, and this contest has finally spurred me into action (as the mechanic is conceptually one in which the game is fundamentally semi-open and everyone is Named). Besides, I make setups with nonstandard win conditions all the time, and aren't the monthly challenges a good time for a change?

To balance the various subsetups of a Smalltown, the sum of how useful a role is for town and for scum should come to a constant; thus, roles should be good for town and useless for scum (e.g. N2 Informed), or good for scum and useless for town (e.g. Doublevoter), or mildly useful for both (e.g. Rolestopper in a setup where it's known to exist). I haven't been able to do that exactly here, but it's close.

Balancing a 9p with 9 power roles is also difficult; you have to err considerably on the side of "useful for scum" (thus the large number of manipulatives, and the use of rolestop effects for protectives). It wouldn't surprise me if this were unbalanced, or even if some subsetups were broken, although I've tried to keep things under control. (If it's unbalanced, it's probably townsided.) It might have to be run to work out where the balance lies. (Monthly challenge setups tend not to be all that balanced anyway, probably because it's hard to balance a game that uses unusual mechanics without having some experience with them.)

[callforjudgement]

Are duplicate toppings allowed? (Note: the setup is broken whether you say yes or no, but in a different way in each case. Probably yes is more interesting.)

[callforjudgement]

Oh right, I missed that everyone claiming exposes the cop as well as the scum.

In that case, yes, it basically just turns into a 2:9 if everyone claims. Town can do better by delaying the massclaim to D2, though, in which case there's a small chance we end up in a vanilla 3:10 and a much higher chance that two scum get caught D2.

[callforjudgement]

Setup B is Matrix6 subsetup C (44.7% town win rate within Matrix6). However, the town win rate would be much higher in this setup because scum have no viable fakeclaims regardless of the subsetup, giving town a substantial advantage compared to Matrix6.

[callforjudgement]

The 9V version of the setup is pointless: with 9V, there won't be any way for townies to survive kills, thus the Loud modifier on the Goons does nothing. There are probably other issues like that in the setup; I haven't looked at it in detail.

I like the idea, though; it's more that the details need working out.

[callforjudgement]

Oh, I missed that 0 of a power role letter didn't create a vanilla setup. In that case it probably works.

[callforjudgement]

Scum seem to have no reason not to camp the bomb, here. Town doesn't know where the bomb is, so they're as likely to pick the bomb's location as any other.

Also, I expect this to be very townsided. 7:2 nightless is townsided as it is, and town are hardly hurt by any outcome of the defusal mechanic (it leads to an occasional town death, but the information gained from it is probably more valuable to town than that, and the deaths, despite being deaths of townies, are town-directed).

[callforjudgement]

Ah right, I hadn't thought about the situation after scum deaths. There is some WIFOM involved when there's only one scum left.

I'd be inclined to choose 8:3 as the correct numbers rather than 9:3; I think you have the right idea by changing the numbers round though.

[callforjudgement]

I'd be inclined to choose 8:3 as the correct numbers rather than 9:3; I think you have the right idea by changing the numbers round though.

Is that with the 1-shot NK or nah?

Without.

With the kill, 9:3 is probably correct. (I missed that you'd changed more than the numbers and location count.)

[callforjudgement]

If you make it non-nightless, you can't give away the weapon with which the scum kill is made, or else the game will end up with really large swing (a lone scum will get caught by PoE very quickly). The nightless version is, therefore, probably more interesting.

[callforjudgement]

From best to worst:

Microtown: I still like this. I'm not sure it's balanced, but it'll be hard to balance without playtesting.
CYOA: I think the basic idea works, and there's lots of space for tweaking to avoid unbalanced combinations; additionally, there seems to have been quite a bit of effort put into avoiding unbalanced combinations in the first place.
Lost Heirs v3: There are lots of other setups playing round in this sort of space, so I know the basic idea works, and it's likely to be close to balanced. It hasn't really caught my imagination, though.
Pick 5: Some of the categories are unbalanced against others (e.g. the scum roles in other categories aren't strong enough to tempt scum away from disabling the first category for town). Scum will just go for the most scumsided setup, which I suspect is (Encryptor, JOAT). That said, given the additional information gained from the semi-open nature of the setup (in particular, that there are exactly three town power roles), this is probably close to balanced if the scum pick well, and scumsided if the scum pick badly.
I'd Take a Bullet: This is probably close to balanced, but I think it could suck to play as scum: if there are one or two very strong town players, town can arrange protection for them very easily.
Leveler Mafia: High is townsided, Special is probably townsided, so townies should pick H (perhaps with a small chance of L, as scum will probably also pick H). There are also swing issues in setups with a high number of town-controlled roleblocks, such as the High setup here, as they make it almost impossible for the last remaining member of the scumteam to win.
Martyrs Updated: This is surely townsided, isn't it? You have a 7:2 with two roles usable as named townies, plus a high degree of control over the scum nightkill, plus the potential to use the town power roles for good.
Jingle's: Setup B is probably scumsided (you can treat the two town power roles like named townies, making it closer to balanced, but that seems like a weird use of them), and I don't see any particular advantage from combining these two subsetups into one (there's no interaction that influences claim outcomes, for example).
PYAE: Setups like this have been run before, and they never turned out remotely balanced. This one has a nonzero chance of producing a balanced setup, but it's pretty close to zero.
Inventions and Stability: I don't really like the idea behind this, and also suspect it's scumsided if it's scum with the power roles (probably also if it's town with the power roles).
Pizza Mafia: I don't like the idea behind this one either.
Counter-Strike: I don't think this is really semi-open, given that the subsetups are identical up to permutation of location names.