So you understand all the stuff up to this point, but there's an issue. You have all kinds of great ideas for a game, and you just know everyone is wrong about how powerful a regular cop is and that vigs are actually cops when you think about it and etc, but you just don't know how to get all of your great ideas into a game. Don't worry! I'm here to help. Hopefully, this primer on my design philosophy will inspire you. If you are running a large theme, you have already designed and run multiple games, so these are LT specific considerations. We are bypassing the part of design philosophy where we learn "well, what *is* mafia" (an informed minority against an uninformed majority, duh) and getting to the good stuff.
Theme vs flavor vs mechanics
Let's get some definitions out of the way first
Theme -
When I say theme, what I am talking about is the setting of the game. Almost without exception, games in this queue are based on source material or an idea that is not just "mafia". Theme can affect the game in many ways that we will cover in this primer.
Flavor -
When I say flavor, I am talking about theme material that is aesthetic only and does not in any way change how mafia usually functions. For example, calling a vigilante a "Teddy Bear" does not change how the role functions mechanically, so it is merely a flavor change. Typically in a theme game, roles will be based on specific characters or ideas from the source material. Flavor can, and often should,
influence
the design of the game. When I show you example role PMs, I can show this more fully.
Mechanics -
This is how the game actually functions. Changes to how Normal Mafia works usually manifest as changes to the mechanics of the game or how typical roles function. If you change how eliminations happen, for example, you are making a mechanical change.
Overall Design Philosophy
The most important consideration as a game designer and moderator is that your players should have fun. It is a game, after all. Players will not have fun if they feel that their actions did not affect the outcome of the game. In your design, ensure that every player can change the game.
In vanilla mafia, the most powerful ability the town has is their vote, and the most powerful ability the mafia has is knowing who the mafia is.
Mafia is a game where an informed minority is playing against an uninformed majority
. The minority uses the means at their disposal to decrease the number of the majority without them realizing who is doing so. The majority uses the means at their disposal to become informed. The information is who is what alignment. The minority does what they can to obscure this information. The whole game votes on who they think is the minority once per day phase and the minority has the ability to remove one player from the majority each night phase. This is vanilla mafia. All deviations from this basic form should bear in mind what the game is at its core.
When designing a mafia game, I think I have succeeded when I put every player in the position to affect the outcome of the game through their actions in the game. Power roles and mechanical changes serve the function of providing new ways for the minority and majority parties to achieve their goals. If power roles warp the way the game functions to the point where it is difference from informedmin/uninformedmaj you should consider either further warping the mechanics to complement these roles or changing the setup.
For example, a vigilante gives one player the ability to unilaterally get rid of someone that they can't otherwise get voted out. Realize that, while it is very satisfying for a town player to successfully kill a mafia member with a vigilante shot, an unlimited vigilante role with no counter allows one person in the town to win the game for their faction despite the mafia hoodwinking literally every other town player and never having a mafia member voted out. If you have a situation where a faction can lose despite outplaying the other faction(s), it is a flaw in the game's design. Try to avoid these types of situations.
Typically, the town should win by voting to remove mafia members and figuring out alignments based on behavior and voting patterns
. Power roles should help the town plan their votes, or help the mafia fool the town. A game of mafia should not be won or lost by power roles.
When considering the size of the game, realize three things. First, larger game sizes do not necessarily require more complex setups. Second, the more players you have in a game, the less agency an individual player will feel that they have. As an addition to the second point, also realize that large numbers of players posting will often cause chaos in the thread while simultaneously causing players to lose interest in the game. This is not to say that there is a magic number after which games are too large, just realize that with each additional slot you add to your game you will exacerbate the issues I listed. Third, larger games will likely last for more nights and therefore will give more opportunities for players to use power roles. When considering the size of the game, realize that power roles affect the game more when they can be used more. Design accordingly.
I like the following strategies for balancing power roles in my games:
Providing direct counters -
When you give information roles to the town, I like to provide ways for the mafia to play around them. For example, when the town is given trackers and watchers, I like to give the Mafia a limited-shot Ninja (untrackable/unwatchable). This serves the dual purpose of giving the Mafia the ability to create plausible deniability around the information provided to town and tells them to be wary of information roles when using their regular night kill. In order for the town to successfully catch the killer, they only have to succeed once. The mafia has to succeed every time in order to not get caught. The direct counter gives them another tool to play around information roles. Additionally, if the town knows that their information roles can sometimes be fooled, it places an onus on members of the town to pay attention to behavioral tells. It also blunts the ability of power roles to shape the game to their will. They are given ammunition to strengthen their arguments rather than a trump card to any opposing view.
Additionally, direct counters can prevent either side from being put in unwinnable situations due to power roles. A Godfather role (always investigates as Town) can prevent an alignment cop from single-handedly wrecking a mafia team.
Limited shots -
Another way of balancing powerful roles, such as alignment cop, is to limit the number of times it can be used in a game. Forcing a player to make choices about who to target while also accounting for the possibility of being night killed or eliminated, benefits the game as a whole. For example, a player who has unlimited shots of an alignment cop might try to lay low and hoard moderator-provided information until the endgame while a player who has only two or three chances will be motivated to take an active role in the game as a regular member of the town. The further into a game you get, the greater a chance for an information role to get helpful information. A player who saves their shots until the end of the game is making a high-risk/high-reward decision.
Indirect information -
The most powerful information you can give the town is direct moderator confirmation of a player's alignment. To balance information roles, you can make the information less direct. For example, Watcher (who visited target player) and Tracker (who target player visited) can give strong indications about who performed a kill and can help catch a player in a claim-related lie without actually confirming alignment. You can weaken Watcher/Trackers by having a multitude of targeted abilities each night, giving the Mafia a Ninja ability, giving the Mafia a roleblocker, etc. When you include information roles in your games, consider how strong the information is by how close it comes to the information being alignment.
Active vs. Passive Powers -
An active ability requires the player to do something to make it happen. For example, a vigilante requires you to chose a player to target. A normal passive ability requires no skill to operate. For example, a bulletproof (cannot be nightkilled) can serve as a direct counter to a killing ability but does not require a player to be good at using it for it to be powerful. Consider this when using them in your game. Having activated abilities (you can become bulletproof for one night) or either/or passives (you can choose to be a PGO or a bulletproof) to create a skill component in passive roles.
