Live Action Mafia

Feel free to add your own!

The main concepts that we started with for a game (well, okay, they were Sammy's ideas ) were:
(1) inception theming (the game is just a dream within a larger game)
(2) lots of trap-based mechanics
(3) some werewolf abilities that caused role-swapping
After a while we realized it was better to have two roles so that you weren't swapping away your mystery role that allowed you to do swapping in the first place, and so that people weren't trying to trap mystery roles that they'd never heard of. And I pointed out to Sammy that if we wanted people not to immediately realize that the dream world wasn't the whole of game, we'd have to make it complicated enough to plausibly be a mystery game on its own. And things sort of got crazy from there...


Things we tried to implement that people mostly liked:
- Dead players got to keep playing.
- There were enough public rules that people had some ideas of how to get things done.
- People had clearly defined teams so they knew who to work with.
- People had quite a bit of power (two roles plus items) to work with.
- We policed people for intros. (We didn't do a perfect job of this, but we built game for 26 players with 27 signed up and kept one FLEX spot that we could delete, so that we could send out roles to all but the last few people who hadn't posted, and bet on at least one of them posting an intro so we'd have 26 or 27. The FLEX spot had a valid role/goals, it just wasn't a game-critical one.)


Things people didn't like so much:
- Rules for dead people communicating with the real world were vague and mostly came down to what Mods were feeling at the time. It's not clear how we could've made a consistent and useful but not broken policy.
- No-kill day rules incentivize mafia to kill only every other day and let doppels/dead people/lynches do their job for them.
- Many dead players stopped playing once they died, which unbalanced teams.
- As GMs, we hated how complicated this game was to implement; having a web app or keyword system like previous GMs probably would've been smart. Day end alone took a couple hours each day. This led to sad delays for players.
- Werewolf theming (though cool) made it somewhat obvious when players who didn't realize the werewolf pattern tried to make up special role claims.
- Investigations were underpowered (we think this is because there were too few of them / too much manipulation, and not just because of monty hall investigations)
- Mafia kills came almost entirely from traps, which meant they had to plan ahead a lot more than they should've needed to and decide who they wanted to kill in advance.
- Some players goals' were in conflict (warlock ksedlar needing to share role info with mage lzahray, police antelope francesca wanting elephant mafia ahkim dead)


Things that spontaneously broke:
- We realized early on that if doppelgängers and mafia teamed up they'd be SUPER OP (Mafia could get plenty of role information and doppels had lots of kills given role info) so we took away mafia's disarm power to avoid this.
- After that, two mafia immediately died to doppelgängers for no good reason except randomness (and maybe somewhat careless role sharing? but not more than most people). RIP. In an attempt to somewhat rebalance, we added a Conscription Potato (the flavor text was excellent btw). This was unfortunate because dead ksedlar and jakob believed they provably knew the set of mafia and didn't realize there could be a conscript. However, we believe it was the right decision for balance; even though mafia are now in a position where they'll very likely win, they've played extremely well to get here, and townies have been dropping the ball recently (investigators and others not using their powers, priest and prophet who lost their powers not attempting to trade roles with someone to get another priest/prophet, basically no townies using items even if obviously extremely beneficial to them, no vig kills). That said, town has also made some excellent plays, so we'll see how game finishes.
- Various rules were unclear or just incorrectly written (e.g. internet stalker didn't say that you couldn't repeat targets); we tried to implement the thing that was fair but also not game-breaking, but this was sometimes frustrating for players. (We're sorry!)
- Mafia ended up getting way too many kill items at some point. It turns out having a warlock on their side, plus the fact that they got items from people they killed, was strong. We toned down the number of items that we were inserting into game / POIs a bit. I think the Live Action Potato mechanism of looting 1 item per body would've been better.
- Spy team 355 achieved their goal on day 2, because ksedlar knew too many roles. We gave them a bonus goal, but this was a bit sad for both them and the other spy teams. In general, players sometimes had achieved all their goals before game end; it's better to have goals that need continued attention.
- People got many copies of one item (since items were role-correlated) which would have been unbalancing, so we added a rule that you can only use one copy of an item per day.

This is aok posting here. I'm too lazy to give myself access to the right forums.


The ideas for game seemed pretty cool.

I like this, although it's hard to come up with ways to do this in most games.

Agreed. Although it's really fun for players to figure out game structure, it's sad when they have no idea about game structure.

I've realized after my mystery game that people at least need to be given contacts. Even if it's just "You have tea every thursday with these players", it helps people stay involved in game.

I like this. It's hard to balance power for 3rd party (proper balance means more power, especially info) with balance for TvM (proper balance means less power, especially defense)

Not a huge fan of this, but it does cut down on inactives.

GMs coming up with stuff on the fly is not recommended for mystery games.

Given the number of kills that were happening, I don't think this is that much of a problem actually.

Eh. It happens in mystery games.

That's a lot of work, unfortunately. There's no real solution, except for pravinas' idea of a mega-webapp flexible enough to run any mafia-like game (FWIW, I don't believe in that)

Figuring out stuff like this can be part of game. It's an especially fun part of a mystery game, because people will make correct and incorrect guesses.

Kind of agreed.



It happens. As a GM you can avoid cases where players are very likely to cripple themselves on accident, but in many cases it's up to the player to recognize this (like ksedlar did). That's why I prefer giving players clearer preferences, such that they know what to do if they feel like their goals are in conflict (prioritize one, try to force through both, ignore one completely, etc).

It's hard to see this coming (See: Crazy alliance from potato mafia)

Randomness inevitably ruins mystery games. Having more killing factions exacerbates this, especially if you have multiple killing factions killing indiscriminately. Many mechanics were pretty self-balancing in this respect, though (if there are fewer mafia, they are less likely to get caught), so it's not that bad.

This always happens in mystery games. Don't worry about it.

Mathematically balancing stuff is hard.

Agreed. I have a strong preference for living players never having 100% or 0% to succeed goals. It's ok to have some goals have corner cases, which you can have secret rules to get players out of if they get into them by accident (see: cultist in my first MM) or not (see: potato mafia where you could make a kill as a townie who needed to survive game). But goals that care about end-game state, or have alternatives ("Complete 2 of 3: Beat the opposing team in a race, having someone on your team survive, kill everyone on the opposing team") are my preference for this reason.

Seems reasonable.

_________________
DO AS I COMMAND!

Thanks aok! I agree, all this stuff is hard, and it's nice to hear that from you, the awesome mystery game designer

Hopefully I'll learn stuff from this and maybe write another game someday.

Yay, thanks for the comments

This is also my first time running a mystery game; some of the ideas (like most of the non-werewolf special roles) I'd had since summer 2015, so I'm glad they seemed to actually fit in decently with the werewolf and real world parts of the design.

Jackie covered all the main issues to think about that I can think of, but I'll add more things if I think of them.