[Discussion] DDL's Guide To Hosting Role Madness Mafia Games


Mar 10, 2014
Guide Made by DDL.
Hello, folks.

I have decided to write a guide on hosting role madness mafia games. I don't believe I am the best host in the forum, but I have a lot of accumulated experience in the "field" and I want to pass that knowledge forward, so if I ever quit this forum, it doesn't go away with me.

So, a little about me: I started playing internet mafia in August 2014, and hosted my first game (Order of the Stick) in January of the following year. I'm still hosting games to this day, and so far I have hosted a total of 14 games (4 turbos and 10 full games) and co-hosted another 3, in this site and in Mafia Syndicate. I made most of those setups, copied some from other sites, and almost every single one of them was role madness. So I think I know a couple of things. I will post what I believe are good practices for hosting games, and if some of you want to add your own points of view on it, you are more than welcome to use this thread for it.

There is a belief some people have in internet mafia that role madness is not balanced or competitive, that it is just a casual game to troll and shitpost between generic games and that trying to balance a game with a lot of roles is a futile effort. I strongly disagree with that. It is harder to balance, and it is something you may get wrong even if you have a lot of experience, but there are ways to make a RM game run smoothly, and I'll try to explain them here. Chapters 1 and 2 deal with gameplay theory, while the rest of the chapters are mostly general points on hosting. Some of the points I'll make can be applied to generic/vanilla games too, though I don't specify which ones.

Credits to WolfPrinceKouga for reviewing the guide and adding some extra points.

1- The Paradox of Role Madness (or how to build a list of roles).

(This chapter is the longest one by far, so don't be scared of its size lol)

Role madness games are faced with a paradox, which explains why many of them crash and burn: the game of Mafia doesn’t like roles. The game likes scumhunting, and votes, and lynches and faction kills. Roles attempt to distort this system, by giving town information, or stopping kills, or allowing either faction to bypass the lynch. This explains why generic games are the most popular form of mafia in competitive communities: because they are the most competitive. The roles, in those games, are added in small amounts, surgically, and the game tends to involve players strategizing around them. A single cop in a generic game can potentially dictate the whole discussion of the game.

All of this tells us the following story: Role Madness is a mistake. Yet, somehow, a lot of RM games… don’t crash and burn. They achieve a level of balance just good enough for everyone to have fun, for factions to have an equal chance of winning, and for the game to still feel like mafia. How?

I believe this is due to an effect of “dampening” between roles. Which means in a game with so many roles interacting with each other, a lot of them just cancel each other, and the game ends up having little influence of roles. Furthermore, due to the information-lacking nature of the game, and assuming the host did a good job at limiting the number of sources of investigation, most players (town and mafia included) have no idea who they are supposed to use their abilities on, which causes most actions in each phase to be wasted.

To achieve that state, you must first understand that not all roles are born the same. I’ve identified four main types of roles that tend to appear:

1- Information Gathering (a role that gives a player information about other players or roles)

2- Kill (a role that removes a player from the game)

3- Lynch Manipulator (a role that gives or takes away vote power, or interferes directly with the lynch)

4- Role Manipulator (a role that prevents another role from working, or changes the way they work)

There are a few other types (such as recruiter or reviver) but they are rare enough that I won’t talk about them now.

Now you may ask: where are the doctor, the roleblocker, the rolecrusher, the busdriver? They are all in that #4 – Role Manipulator. All those roles have this in common: they don’t add anything new to the game. They don’t directly make players die, and they don’t make anyone learn anything. Instead, they shut down other roles, or at least change their effect/target. This is very important, because this is where the dampening effect I mentioned before comes from. If most of the roles in a game do nothing but shut down other roles, you may be able to make a game that is not dominated by roles. What you want to achieve is a certain illusion: give the players the feeling they have power, when truth is, they don’t, and what really matters is the voting and hunting part of the game. You cheat the Paradox of Role Madness by making a game that feels like RM more than it actually is.

So, your first job is to make sure none of the other categories above have too many roles. #1 - Information Gathering is the most critical. Mafia is a game of Uninformed Majority vs Informed Minority, and those roles remove the “Uninformed” part from the game. If town knows who they need to lynch, mafia can’t overpower them with their lower numbers. If mafia knows where the town PRs are, the PRs will die before they can do anything. You want those roles in low, controlled numbers. Furthermore, you want them to be easy to get rid of by the mafia, which is why investigative roles in my games almost never have bulletproof shots. Now, for the number, I usually have 1 cop per 20 players, and 2-3 of the following: tracker, watcher, (non-public) lie detector, mason duo. This is not a rule, and can be tweaked. A cop can be made into a faction-specific one that only finds one faction. Roles can be weakened by X-Shot limits, and/or broken into multiple players. A mason group is way stronger than a mason duo, and a public lie detector is way stronger than a non-public one (the latter is a weaker cop, but the former is considerably stronger), so those can replace the cop. Tracker is a fairly weak role that can replace cops in a pinch and reduce the amount of information available (the same doesn’t apply to watchers, as they are much stronger). Finally, give mafia ways to directly counter them: godfathers, millers, but also role cops, roleblockers and super-kills (aka strongman kills).

#2 - Kill is also tricky. Vigs can be deceptively strong, or deceptively weak. They give town the power to bypass the lynch and decide who should die by themselves. They may break through the thread presence mafia took so long to build and expose a player who was well hidden. And they may also make town kill each other and hand the game on a plate to the mafia. But kills are also very fun, so you may want a few of them. I tend to be more liberal with vigs than with cops, but generally, no more than 1 vig per 15 players. A few extra one-shot or every-other-cycle vigs can be added because of flavor, but don’t go wild on them. Consider having only one or even zero dedicated vigs if you are going to have a lot of part-time ones, even in a big game.

Mafia, on the other hand, NEEDS kills. A mafia with kills that don’t work or are hard to access is a mafia that can’t deal with townies who are doing PoE (process of elimination), and a good sign that you screwed up as a host is seeing mafia losing to PoE despite having most of their members still alive. If there is only one mafia team, and the game has more than 25 players, consider giving this mafia access to extra kills. Not necessarily kill twice per night, because this may be OP, but conditional. This can be done in a number of ways: X-shots, game mechanics that let you buy abilities, reflexive kills, etc. Make sure your kill condition is not too hard to achieve, too. Now, if there are multiple scum factions, you may get away with having each faction with only one kill. On the other hand, you now need the mafia/town ratio to be higher, because the mafias will kill, and vote, for each other. If a game with one scum faction enjoys a mafia with 1/4 to 1/5 of the player total, one with two requires something closer to 1/3. Favorites V, which had 3 mafias and 10 indies, only had 50% of the players as townies.

Finally, kills dictate how long your game will last. A long game where nobody dies is not only boring, but may help the town too much. The more lynches there are to analyze, the harder it is for mafia to hide. On the other hand, a super short game probably means your vigs and indies are monopolizing the gameplay. To balance this, what I usually do is calculate the number of deaths I expect to happen per cycle (including the lynch), and divide the player number by that number. The result is then multiplied by a factor (I usually use 1.5, but it’s not very precise), related to failed kills, and vigs/indies who die too early, which nets me the predicted length of the game. If I’m okay with the number (between 7 and 12 days is what I call “okay”), I go on.

Next, #3 - Lynch Manipulator. Those are surprisingly forgiving, depending on their type and alignment. In town’s hands, vote power is very weak, so you can be liberal with it and give it to multiple roles. In mafia’s hands, it’s stronger, but you can limit it by not letting mafia boost their own vote power during the day (they have to activate it in the previous night), or give them other drawbacks. Actual lynch stoppers and lynch redirects are extremely strong in the hands of mafia, and still somewhat weak for town, but even then they can make the game frustrating, so you should stay away from those in most games.

#4 – Role Manipulator, like I explained above, are your dampeners. Be liberal with them. Of course, don’t give town 6 doctors in a 25p game, because they may actually start stopping the faction kill consistently, but you can usually get away with 2-3 and watch as town protects the wrong person (just remember to give mafia a super kill/roleblock to make sure the docs can’t make a confirmed player immortal). Likewise, roleblocks are super easy to relate to most characters, and don’t hurt the game too much (again, don’t let town control the game by having a million of the same role, see doctors above). Ability redirectors require some caution: while they don’t add anything new to the game, they tend to make players die in unlucky ways that they didn’t really deserve, and make the game feel like a troll game. Furthermore, they are a pain to process in high numbers. Don’t be too liberal with those.

Finally, a note about passive abilities: be aware they are intrinsically better than their active counterparts, because the player doesn't have to think when to use it, they act whenever they are needed. Bulletproof can be problematic, especially in the hands of town, since it messes with mafia's ability to kill consistently, as explained above. My townies almost never have more than a one shot of bullletproof, and if they do, it's either conditional, or the role has nothing else, and they always have some clause that makes them lose vote power after they stop a kill, so it doesn't break the endgame. Mafia and indies are allowed to have more bulletproof, since town's tool is the lynch. Other passives may be used depending on the need and creativity, as long as you add counters to all of them. Abilities that remove passives on the whole, such as rolecrushes (you can make a rolecrush that only lasts one night too, if you want) can be useful in that regard.

So you add 2-3 doctors here, then 2-3 blockers there, some vote manipulators, 1 redirector, some roles that remove passives, some roles with cool passives, and when you see it, you have filled your role list, without giving town a bunch of cops and vigs. Mission accomplished.

2- How to make Cults, Serials and Indies that don’t suck.

So you come to a mafia forum for the first time and sign up for a big RM game with 40+ players. Then you watch as an indie player raises hell and fire upon the puny townies and scum. Then you read the role after the game ends, and it’s like an angel wrote it. You make a decision: “that’s the kind of role I wanna make!”. So you write your 25 player setup, add 5 indies, put it on sign ups, get players and start the game.

Two weeks later:

1- Your serial killer was taken down by a vig on Night 1.

2- The non-hostile indie sided with town.

3- The cultist failed to recruit all but one person, then was lynched, and left that person alone with no chance of victory.

4- Your flavor-specific role that had to hunt for items never interacted with anyone and stopped posting by Day 4.

5- That role the entire setup depended on was modkilled after two failed replacements.

All the above have happened to me, or way or another.

First, let me teach you the Paradox of Role Madness : the game hates these roles except the Serial. This is because the Serial, at the end of the day, is just another mafia faction, the difference is that they are alone. So they integrate naturally into the game’s logic, by threatening to destroy all other factions, and needing to do that to win. When you add any level of ambiguity to it, you give the players a choice to do something other than play mafia, and it’s where things get tricky. They start siding with other factions, or doing things that are against the game’s spirit. Now, that doesn’t mean you should never make other types of indies, but you need to understand the problems of each role and design ways to overcome them.

The most important thing about these roles, in my opinion, is that they have no obvious incentive to side with other factions. That’s why they are seoerate from the town and mafia. A good role like this threatens and is threatened by other players, even if they don’t need to eliminate them all. A survivor or another non-hostile indie is a problem role because they are can win just by claiming and being left alone. That doesn’t mean the survivor is impossible to balance, but you may want to add a twist to it, like giving players a prize for taking it down, or punishing other players if they are left alive too long.

Next, be aware that most of these roles are weaker than they look. A player who is alone not only needs to survive (except the jester), but also accomplish something else. They are extremely vulnerable to stray kills. Consider giving these roles generous amounts of bulletproof, even lynchproof depending on the game (except the survivor and the jester, they are already very strong without that). Try to run endgame situations in your head, so that you can see if your indie won’t be stopped by a single bulletproof opponent, for example. Make sure they have the tools to get over those, even if conditional or limited. It may also be wise to give these roles defences against some permanent effects, like being made vanilla, since it can quickly prevent them from having a chance of winning. And remember that these roles strength is very dependent on the number of players. A cult in a 15p game is a juggernaut, a cult in a 50p game is not. Balance accordingly.

Finally, make sure these roles are fun to play with. Maybe hunting the seven chaos emeralds sounds like a great time to you, Sonic fanboy, but not to that person who came to play mafia instead. Their wincon, no matter how original or flavored, should be related to mafia in some way. It should involve hunting specific roles, or taking down specific factions, for example.

3- You are a computer (avoiding host interference during the game).

Mafia is like a videogame. You input actions and votes into it, and it gives you deaths and lynches. In a videogame, you have a computer program that processes all of that automatically. Mafia, being a board game of sorts, has no program. The one who does the processing is the host. So the host must act like a program.

This means you don’t get to interfere in the game after it starts. You don’t change roles, you don’t help the faction that is losing, you don’t give hints to people. You don’t make events during the game that help certain players, and you certainly don’t have a role in the setup. You are a computer. You receive input, and give output, based on a logical system you created.

Sometimes this will not work in practice. Sometimes two roles will interact in a way you didn’t plan ahead for, so you have to decide how it will work. Sometimes a player will spot a loophole that makes their role more powerful. In that case, you must choose how to interpret the mechanics. You can pick the solution that keeps the game as balanced as possible, or one that abides by the original philosophy of the role. And there are rare cases where you must edit a role after the game starts, usually to add a clause to prevent a specific situation from happening, before the players spot the loophole. In all cases, use your common sense, and try to find the solution that you think players will feel is the fairest.

Either way, do everything you can to keep the game as close as possible to a computer program. When writing roles, make it clear what is a day role, what is a night role, how often they can be used. Add clauses to predict how abilities will work in certain tricky situations. Simulate role interactions and decide in advance how they will work. Sometimes you get to find the loophole while you are still designing the game, so you can fix it before it’s too late.

Some hosts (*cough*CR*cough) intentionally avoid this rule and make a career out of interfering in their games. That’s fine if you want to adopt this style, and those games can also be really fun, but let players know that you are this type of host. Otherwise, it you want to make balanced and competitive games, you need to make sure they will be able to run without host decisions as much as you can.

4- How to deal with role claiming.

Role claiming is one of the most polarizing topics of internet mafia. Some people believe they should always be allowed, and that if you ban them, it means your game wasn’t well designed to begin with. Some believe they are a shortcut for people who can’t scumhunt or defend themselves properly, and prefer it to be banned every time. Personally, I think both sides are right to some point.

Role claiming adds an extra dimension to the gameplay. Now you are not only figuring out if the player is scumhunting or not, but you can also figure out whether their claim fits into the setup or not, and whether their posts fit with the actions they are claiming. And mafia must learn how to make claims that people will believe in.

Role claiming adds a huge problem, too. If the roles are all named, and the players know which names are good or evil, then townies can become confirmed just by claiming those names. Mafia has no way of fighting against that. This is even worse if the game is open setup, where players can just claim every role in the list and solve the game like a puzzle.

There are, however, a number of ways to get around that. The most obvious one is to have generic roles. But you are hosting a RM, so you don’t want those. Maybe you can, however, have generic roles that get to use abilities through some common game mechanic (i. e. the generic roles from WPK’s Zootopia game).

You can also add game mechanics that punish people for claiming. For example, a Death Note mechanic, which is basically giving mafia extra kills that only work if they know the player’s role name. Or janitors which let them fake claim a town’s role. But beware that janitors and other specific roles can depend on luck to work (i. e. the janitor not getting lynched on day 1 and hitting their kills), so if they are your main plan to stop claims in a game that is very vulnerable to it, you need to make sure they will be powerful and abundant (but not too abundant so that town has no info at all, you need to balance that out).

Some setups are less vulnerable to claiming. Original flavors that you made yourself. Mash-up setups where good characters can get evil roles (i. e. Favorites). Setups with loads of characters where players can’t predict who got in (i. e. Game of Thrones), though in this case it’s advisable to provide the mafia with a list of claimable roles. On the other hand, open setups are the most vulnerable to claiming, and should ban claiming in most cases. Another factor that you must take into account is how much information you are showing in write-ups: only the deaths, or other types of abilities? And do they show the alignment of the role using them? Depending on those, a game may become more or less vulnerable to claiming.

At the end of the day, use your common sense, and don’t be afraid to press the “ban role claim” button if you are afraid your game will not work otherwise. But make sure you make this decision while you are still writing the setup, not when a player tags you during the game to ask if it’s allowed. Because your entire game design will be affected by what decision you make.

5- How to deal with inactivity.

Inactivity is one of the worst things that can happen to mafia, and yet it happens nearly every damn game. People have lives, and those get in the way. Sometimes we sign up for more games than we should, or one game delays to start and collides with another so now we don’t have time to play both. Some people are just irresponsible and don’t care. You can and should try to improve your community by educating people on the importance of not flaking, and maybe blacklisting those who do it a lot. But when you are a host, and you have an inactive in your game, it’s your job to deal with them yourself.

There are two ways to deal with an inactive: modkill and replacement. There is also the option of not doing anything, and leaving them in the game. All those options have problems. Generally, I follow this priority system:

1- Early game replacement (using a new player)

2- Early game replacement (using a dead player)

3- Modkill

4- Late game replacement (using a new player)

5- Late game replacement (using a dead player)

6- Leaving the inactive alone

This is very subjective and most hosts don’t follow the exact same system. I have also broken it myself before. Again, use your common sense, and pick the option that harms the game the least.

I will now explain the pros and cons of each option. Replacements are usually harmless, but the later they happen, the more they affect the game, because other players now have less time to read the new guy’s alignment. Most of the time you should modkill instead of replacing at the end of the game. Using dead players to replace is viable, but it can distort the game a little because mafia now has an incentive not to kill the best scumhunters in town early, since they will just come back as replacements. Try to avoid it if possible. Modkills are clean and nobody will hate you for doing them, but they change the setup’s numbers randomly and remove roles that you worked hard to make. I like to take my time before doing them, and go around hunting for replacements for some time. Leaving the inactive alone is my least favorite option, because it leaves the burden of inactivity to other players. Now they have to figure out how to read someone who’s never posted. A lot of the time mafia will leave the inactives alone and only kill the actives, which means your game will become really imbalanced near the end. I try to avoid letting that happen.

One thing that helps you a lot is being pro-active. You don’t want to learn about inactivity when a player tags you in Day 4 and asks you what are you going to do about it. You want to be aware of it, and be hunting for replacements before the modkill point comes. Know who the inactives are, check when they were last online, see if you can contact them, know how many replacements you have and what are your options. Sometimes this means sending PMs to potential replacements at 6 AM before you go to work. It takes work to do, but it’s worth it.

Finally, there is the option of modblocks. Those are not likely to bring the inactive player back, but they deal with a different problem: mafia being able to use abilities for inactive teammates. It’s common to let mafiosos send abilities for each other, because it facilitates their teamwork, but if the owner of the ability is not posting, they are likely to coast the entire game while the townies avoid them, and now you have given someone an award for not posting. So, I modblock players who don’t post, and forbid them from using abilities while I am looking for a replacement. This isn’t meant to be a punishment, but a way to keep the game fair for everyone.

6- Take it easy when you are starting.

A common mistake hosts make, and that I’ve made myself, is to try to create an amazing, complex game in their first try. This ties together with the topic about Cults, Serials and Indies, as they tend to be the crazier game mechanics that new hosts get excited about.

Go slow.

If this is your first game, your lack of experience will come to bite you at the times you will least expect. You probably lack the intuition necessary to know if a role is too powerful or not, or if a mechanic will be fun or not. You are also more prone to making in-game hosting mistakes, because you are dealing with a bunch of new information at once that you are not used to. In my first game I didn’t know mafia cops were supposed to reveal roles so I had to edit it in during the game, after the mafia team asked. Yes, it’s stupid, but I had started playing online mafia 6 months before. My second role madness game (Favorites V) had a huge number of in-game mistakes.

These mistakes (both related to hosting and game design) will happen. They are part of the experience. But you can reduce their effects by not trying to chew something too big in your first game. So, keep the player number reasonable: anything above 30 is probably too much, maybe even above 20. Don’t add too many indies. Make sure most roles have a low number of abilities, and try to focus the role power on a few strong roles, or a single game mechanic for example. And if you want to create new game mechanics, don’t add too many of them. It’s always tempting to add every single fantastical aspect of your favorite manga into the game, but each new game mechanic is a thing whose effects you have predict before the game starts, and trust me, experience helps a lot in that. Maybe you should save your favorite theme for your 2nd or 3rd game instead, if you really want to make a big game from it.

Heck, if you really want to be safe and learn the basics of hosting, start with a generic game. It’s hard to get them wrong. If you want to start with role madness it’s fine though, just be careful not to make something too big for you to deal with.

Finally, consider asking a veteran host to review your setup. Or even someone who’s hosted only one or two games before. They may help notice things you didn’t, point out if one of the factions is too powerful, and maybe you can even invite them to be your co-host! Just be aware that if your game is not fully open setup, you won’t be able to allow them to sign up for it.

7- Getting through the sign-up phase.

This is a subject that has probably caused nightmares to many hosts before. How to make your game fill up. Well, there isn’t an easy answer for this, if there was, we wouldn’t have so many games dying before they even start.

There is one thing you can do that helps a lot, though: predict how many sign ups you are likely to get, and set the required player number according to that.

There are a few factors that play into that. The ones I think are most important are theme, season, and host reputation. The theme of the game may be more or less attractive to players. For example, this is a forum about Naruto. So Naruto is the winning theme that will get you any sign ups you want. Feel free to make a 40 or 50 player game with it, people will come out in droves, provided you advertise it in other sections. Next, other anime/manga related themes, especially popular shounen from the same generation of Naruto, like Bleach and One Piece. After that, come other well-known popculture themes, like Marvel, Game of Thrones, Zelda, etc. At the dead end are themes from types of media most people don’t even consume, like webcomics. If you want to host those, be aware that you are probably not getting any people to sign up for the theme alone. The ones who do sign-up, will do it because they like mafia anyway.

Finally, host reputation. This is probably a little unfair, but people will be more willing to sign up for your game if they already played (and liked) your games before. This ties up together with the previous point nicely though: don’t make a game too big in your first try, not only because it will help not to screw up with making it, but also help you get all the required sign ups. And if the game is good, help you get even more players in the following game.

Now that you have predicted the number of players, some tips on actually getting there. First, advertising is the soul of business. Tag people. If it’s a big event game, maybe even ask the staff to put out a forum-wide announcement. Also, a tip: people are more likely to sign up if you invite them personally, rather than add their name in a giant invite list. Go to their profile page, or address them in the convo thread and ask them if they want to play. You will be surprised how well this works, but it does. I guess people feel flattered when they are individually told they are wanted in a game. The downside of this is that you have to do it slowly, so it doesn’t look like the aforementioned giant invite list.

Another point: the presentation helps. Make sure your sign-up thread doesn’t look like shit. Don’t use words like “I guess” and “whatever”, and don’t give people the impression you are half-assing it. If you are good at art, make a cool banner. If there are any cool game mechanics you can show, show them. This is tricky if it’s a fully closed setup game, but sometimes there are things you can show, so consider doing so in the sign-up thread instead of just when the game starts.

8- Design your game to match your free time.

Free time is one of the most critical assets of modern civilization. That’s why people retire from mafia: they get jobs, and families and social lives, and realize all those hours they are spending in an online game every day are not worth it. This is fine. But every adult needs a hobby. And hosting big and complicated mafia games can be that hobby if you want, even if the number of hours you have for doing it is limited.

The trick is to know how much time it will consume while you are still making the game. Do not try to repeat that monster of a game you made during university that took 4 hours a day to host. You have no time for that now. Instead, plan on the game’s size and complexity ahead. If you can’t go online multiple times a day, don’t have day actions. Don’t make night actions with obnoxious mechanics like counting words or things that force you to do a lot of work. Don’t make a game too big and crazy that you know it will take two hours a day just to process actions.

Have systems in place to do things faster, too. Spreadsheets where you keep track of players status effects and X-shot abilities, and whatever other mechanics your game uses. A list of all actions in the game, saying whether each action worked or not, and why. Threadmarks. Role PMs with the name of the role on them (I don’t do that usually, but Tiger did in CotH1 and it’s awesome). A list of rules you can copy from previous games, a write-up format you can copy from previous write-ups, and so on. The more shortcuts you can take, the better.

Finally, take your time to design the setup. Opening the sign-up thread for a game that starts in a month whose roles you haven’t written yet, and then rushing it out of the door may work for some people, but not for me. I don’t open the sign-up thread until the game is like 90% done. Even if it takes me months to get there. I don’t care. You don’t flake on hosting if you only open sign-ups when you are ready to host it.

9- Balance is your king, but fun is your god.

The game must be fun. Otherwise, there is no point to it. This may seem obvious, but many hosts get this wrong. In an attempt to make legendary roles, they create something that is not entertaining for anyone except the guy hosting it.

This ties up well with the indies chapter. You want roles that are fun to play with. Roles that give people incentive to scumhunt, or pretend they are scumhunting. In the night, their abilities should have some practical, tangible effect, like giving a chance of learning something, or stopping a kill, even if the player can’t see it easily without some puzzle solving. I know some of you will read this and scream “but the best role is vanilla!” Yeah maybe but people sign-up for role madness to get role action. Don’t betray their expectations.

Even more important is that the abilities don’t ruin the fun of their target either. Abilities that make people unable to post or vote for long periods of time ruin the game for them, and should be avoided. Abilities that make people lose roles permanently should be rare. Abilities that force townies to work for the mafia can seem fun at first, but make sure they don’t last too long. And please, be careful of post restrictions. Unless the words “troll” or “bastard” can be seen in your sign-up thread or you are famous for making those, do not add them in your game. They piss players off more often than they entertain anyone. Next, don’t go crazy on redirects. While a carefully aimed redirect can be hilarious, a game where half of all abilities keep getting redirected is just frustrating for the players.

Finally, lynches. They should work. The game is about lynches, if all the death is only happening in the night kills, this means most players are not participating in the gameplay. Think twice about making an ability that stops or redirects lynches, and if you insist, make it a one-shot. Maybe even a conditional one. Likewise, mafia will get bored if they can’t kill anyone. As a rule of thumb, if town overwhelmingly thinks a player is scum, they should be able to get rid of them with little trouble or delay. And if mafia wants to get rid of a player, they should be able to kill them with little trouble or delay. Follow this rule, and your game will flow much better. Remember, you want to make a game that feels like role madness more than it actually is.

10 – Short topics that don’t deserve their own chapter.

- Co-hosting: This sounds like a good idea, but make sure you have the necessary organization to make it work. One host alone should be responsible for processing the actions, otherwise you two will just be confused at what is happening with them. The other host should help designing the game, and then help with manual labor like counting votes or making write-ups (provided the actions have already been decided).

- Make rules clear: make sure the game’s rules are known before it starts. This includes things that vary between games, like whether role revealing is allowed, posting requirements, tiebreaking policy, etc. If anyone asks you about something that is not included in rules, add it, and use it as a template in your next game so you don’t miss it again. I also like to make a glossary of terms that may not be obvious to site newcomers, like rolecrushes and super-kills.

- Show cults and jesters: this ties up with the above. Unless it’s a setup where those are expected (i. e. Favorites), those roles should be advertised in the sign-up thread. If not, at the earliest, in the game thread for the jester, and the first time someone is recruited, for the cult. Jesters are ridiculously easy to win with if people do not expect them, and a surprise jester flip is annoying at best, and infuriating at worst (when it causes the game to end). Cults can be very hard to fight against if players are not expecting someone to be culted. This point may also apply to any other indies that get too powerful if their actions are not known by other players, so think about that when you are desiging the game.

- Pseudo-investigations: a pseudo-investigation is an ability that was not designed to be an investigation, but ends up working as such due to a design oversight. This usually includes ability copycats and mind control powers. If you will implement those, either be aware they are investigations and account for that in the balance, or find a way to nerf them, such as limiting the amount of information the mind controller learns, or forcing the copycat to pick multiple players to copy and giving them a random one, or making them passive abilities.

- Don’t depend on players being gullible for your game to work: while mafia is all about players fooling one another, be careful about expecting players to be fooled at very specific times for certain key roles to work. This includes designing roles that make offers to other players and expecting them to fall for them, or expecting players to relinquish their vote power or abilities. There is a famous expression at NF that is “town gonna town”. I have my own version: “town gonna town, unless the host needs them to town”. Townies are notoriously bad at noticing when other players are fooling them, but they seem to be great at noticing when the host is trying to fool them. It’s like they have a sixth sense for it. So don’t taunt it. If you are really making a role that depends on players being gullible, at least make sure that role isn’t the one thing the game’s balance and fun factor rely on.

- About day abilities: most people like day abilities in RM games. Provided you are available to process them (refer to the chapter about free time), they can be a good idea, but if you are not making a huge game, be careful with having too many of them. Day abilities are automatically stronger than night ones since they are harder to block, and they distort normal mafia system that is lynches at day, abilities at night. Also, if you are going to do day kills, it may be tasteful to not allow them on Day 1, so players can have at least one day phase to post and defend themselves.

- Don’t be an asshole: the game may get rough sometimes, but make sure you are the mature side of the conflict. Don’t argue game balance with players during the game, apologize for host mistakes, don’t criticize players’ performances before the game ends. You have an advantage since you are the only person who has access to all the information, so don’t use that to bully the players. Be civil and fair as much as you can, and players will be more likely to ignore the mistakes you’ve made and remember the fun things about the game.
Last edited: