Comprehensive Modding Guide for 2010
Introduction and Disclaimer
This is a guide intended to help first-time moderators make their modding experience as successful as possible.
Yes, it's long. You don't have to read it all at once. The guide has been split into a number of small sections for your convenience.
This guide is primarily intended for use on mafiascum.net. However, the contents should apply to other sites except where flagged.
This guide was initially written up by Vi and revised by Xylthixlm, two mods who have a tendency to be dogmatic. You may not agree with some of the advice in this guide, and that's fine - as long as you understand the concepts behind it and you can defend your alternative opinion.
There is a small subset of mods who are adept at selectively breaking the rules laid out in this guide in some of their games. We affectionately call them "bastard mods." While some of them are very successful and create enjoyable and innovative games, just as many of them fail spectacularly - often to the extent that they don't consider trying to mod another game for their own sakes.
Mods that think that the rules in this guide do not apply to them at all, on the other hand, are simply referred to as "bad." If you are one of these people, this IS a patronizing judgment against you.
Are you ready? First-time mods should definitely read Chapters I and IV; experienced mods are encouraged to read the rest of it.
I. Setup Design and Preparation
- mafiascum.note: This entire chapter does not apply to Open games.
Setup creation is one of the most gratifying parts of moderation - you can design a setup exactly how you want it, with all the individuality you want to add!
Because of this, this section will focus mostly on common pitfalls that will cause people to hate your pet game design.
A. Getting an Idea
An interesting setup starts with a core concept. This can be a theme, a new set of mechanics, the interplay between some roles, an entirely new role, exploiting the power of a familiar role, or even just wanting an "ordinary" game with none of the above trappings.
If you can't think of a concept you like, don't force yourself to choose one - a good idea for a setup will come eventually. What's important is that you like the setup you're designing and have thought it through carefully. If you reach the top of the mod queue and don't have any ideas, do everyone a favor and /out to mod. Playing and modding Mafia games are long-term commitments that shouldn't be taken lightly.
- If you are opting for an "ordinary" game, you can move on to the next section now.
If you are trying to be creative with your setup idea - especially in the realm of game mechanics - make sure you know how it can be applied to Mafia. Not all of your cool mechanics-based ideas can translate smoothly into Mafia games, or even at all. Think of how your mechanic would affect gameplay - can it be abused? to the point where it can break the game? Has this idea been done before, and if so, how did it go? Would it be a good idea to make roles that could take advantage of the new mechanics? Explore how much the mechanic will affect gameplay, and how heavily it will factor into the game. And most importantly of all, do it in the Theme queue (see I.B).
From here, you can build your setup around your concept. You can do that on your own, although if you would like a place to start there are tips provided in later sections. If you're playing on mafiascum.net, you'll need to make it fit into the site's constraints, which you can read about in the next section. When you're done designing the setup and you've looked it over for balance a few times, move to I.F.
B. Normal vs. Themed, Mini vs. Large
- mafiascum.note: This section only applies to games on mafiascum.net.
At the time of this writing, discussing "Normalcy" is asking for a firestorm of controversy. Understand that the loose definition of "Normal" involves a setup that is not deviant from a basic game of Mafia. As for what a basic game of Mafia looks like and how far one can deviate from it... everyone has their own ideas, and while there are vague guidelines with plastic enforcement, any setup that isn't terribly basic is virtually guaranteed to make someone upset.
Therefore, if there is any doubt that your game is not Normal, ask the appropriate list mod. It's that simple. Be prepared to be told "no". Be prepared to change your setup if that happens.
Note that "Normal" does not connote "boring". Your setup, especially your first setup, does not have to be revolutionary; but there are still ways to make the game interesting. For example, inviting a good player list will accomplish the same effect (see Chapter III). After all, a more complex game means a higher chance of messing something up - and a higher chance that the screwup will be as legendary as the setup you're trying to run.
If you still wish to show some unique flair in a Normal game, there are still several of good options available within the limits. You are allowed to have one unique role in your setup as long as it's not too far removed from the other Normal roles, and altering an existing role somewhat is a good way to put one together. In this game, the unique role was a role that Trackers would always read as having targeted someone who died that Night. Alternatively, you can try breaking a relatively benign meta. The players in this game got a surprise when someone claimed Jailer and was then counterclaimed... and both of them were in fact Town Jailers. So it's entirely possible to run Normal games that are interesting as long as you have a sense for where the limits are and don't get too close to them.
The other constraint on your game is size. Minis are for 13 players or fewer; Large games include everything else. Consider that it's more difficult to fill Large games. If you really want to run a 30-player game, you're going to need to do a tremendous job of advertising to fill it. There are a limited number of talented and available people onsite at any given time; the larger the game, the smaller the proportion they can possibly make up in your game given that you recruit them. The rest of the game will likely be filled with untested unknowns who have a nasty habit of flaking or lowering the level of play. In addition, it is notoriously difficult for majority votes to materialize in the early Days in Large games, making them very painful at the start. A decent ceiling for the size of a Large Theme would be about 19 if you only plan on having a lynch and a kill or two to whittle the population down each Day/Night cycle. Games that are larger than that will probably be better served with four kills per cycle. Note that since kills can fail, this will not necessarily make the game run more quickly; your very Large games are going to be longer-term commitments than Minis regardless.
C. Xyl's Setup Design Crash Course - Flavorless
- Decide how many players you want in the game, and what the scum faction breakdown will be. The standard is for scum to make up anywhere from 25% to 33% of the number of players based on how powerful the Town is.
- Figure out what power roles you want in the game.
- If it's a smaller setup, include a maximum of two information roles, including no more than one sane cop
- A maximum of one each for vigilante, protective role, town roleblocker.
- Count a pair of masons as an information role.
- Don't use both a sane (or insane) cop and a regular doctor (an established broken pair leading to "Follow the Cop") without including a Mafia Roleblocker.
- If you have a larger setup, you can add a second role in one or two of those categories. You don't need a role in every category; it's often good to have one or two empty. Backup roles like nurse and deputy should count as power roles here.
- How many power roles are there? If at least 1/2 the town is power roles, you put in too many; take some out. If at least 1/4 the town are power roles, the mafia need power roles. Put in a mafia roleblocker or godfather. If you violated any of the rules above (like putting in an extra doc, say), give them both.
- Now you can add any minor roles that don't fall into the above categories. You can split a major role between two players (such as even and odd night vigs), or replace a role with multiple one-shot versions. If you gave many townies minor powers, you should give the scum some too. Try to give the scum powers that would make just as much sense on townies.
- mafiascum.note: Games with few or no vanilla townies should be run as Theme games even if they don't have a real theme.
- Consider what would happen if the town mass-claimed day 1. Assume one scum claims doctor or tracker, and the others claim townie. Does the town have a really good strategy to help them win? If they do, put in some way for the scum to stop it.
- Look at how fast the town can clear people. Masons are automatically cleared. Cops and most other info roles clear one person per day. Assume the scum never kill an info role or cleared player, and no player is cleared twice; how long will it take for more than half the living players to be cleared? If it can happen before day 4, you've got a problem; go back and take out an info role.
- Compare the game to other setups you've seen. Does the town seem much more powerful? Not powerful enough? You can either go back and change the number of scum and/or power roles, or you can add or remove vanilla townies until it looks right. If you change anything make sure you go back through the list.
D. Vi's Quick Tips
Regardless of whether setup is Normal or Themed, it should not be more complex than it has to be. Decreased complexity lightens the workload for everyone, and more importantly does not necessarily take anything away from the setup. If you have lots of ideas, spread them among multiple setups. Putting them all in one setup decreases the value of each of them while increasing the complexity of ensuring that they all work as you want them to, and makes you work harder to gather new ideas for future setups.
Each of your Role PMs should come with a Win Condition and precise descriptions of what each ability does. That sounds wordier than it is in practice. If you just say
- "You're a Town Doctor"
there are a number of questions you'll have to answer - what do you have to do to win? Can you protect anyone, even yourself? Can you protect against daykills? Are you guaranteed that protecting is ALL you do? Do they stop multiple kills on the same Night? These are all questions that you'll have to answer eventually; you may as well spell them out in the Role PM--
- "You're a Town Doctor; at Night you can protect a living player to prevent one kill that targets them that Night; you win when all the non-Town are dead; confirm in-thread".
Do not run a time-sensitive game (i.e. one with short deadlines or precise action resolutions) unless you are absolutely certain that you are going to be present and on-call throughout the game.
Check your Win Conditions so that they are precisely what you want them to be. Generally speaking, the Mafia wins when they comprise half of the living players because they can force No Lynch each Day and NK their way to victory. However, in some more complex games, the Mafia is not guaranteed a victory even if they comprise half of the living players. For another example, if the Town is told they win "when the Mafia is gone", and you include an SK in the game, technically they do not have to kill the SK to win.
E. Night Start vs. Day Start
Most commonly (on mafiascum.net, at least), games start during the Day. An alternative is Night Start, where the game starts at Night (obviously). The advantage to Night Start is that power roles have an opportunity to influence the game before it begins, possibly giving the Town insight as they enter Day 1. The disadvantage is that a player who signed up to play almost always dies before they get the chance to do anything material, and the previously-advertised insight does not necessarily materialize in practice. It's generally best to start during the Day unless you seek a lot of emphasis on power roles or are concerned that the game will run too quickly for your information roles to be of too much use (due to many people dying at a time, etc.).
F. Writing a Ruleset
You're going to need a ruleset to run the game with. Most rulesets tend to be very similar, but each one has a few peculiarities. Your best bet is to take the ruleset from a game you enjoyed. You can either copy it exactly or make minor modifications to suit your preferred style. Either way, make sure you know exactly what the rules are and why they're there.
As an example, many games reserve a color that only the mod can use when editing a player's post. This can be abused by a player faking a mod edit to their posts. That is blatantly fraudulent, against the rules, and warrants a modkill. However, if you choose a "common" color like red or green, occasionally someone will slip and use it without intending to look like they were impersonating the moderator. Modkilling for that is silly and extreme; only a warning is merited (the first time). It is up to you to judge rule infractions.
- mafiascum.note: This section only applies to games on mafiascum.net.
If you are running your game in a "normal" queue, consult the list mod before including any unique or original rules. Most (but not quite all) of the time, games with unusual rules or mechanics must be run as theme games, even if they don't actually have a theme.
If you DO use unique or original rules, make sure you know all the potential implications that can follow from them. Players that hunt for breakable setups LOVE poorly thought-out rules. In addition, make sure that you can live with these rules throughout the game. You DON'T want to have to retcon one of your initial rules because you realize midgame that it was a terrible idea.
G. Getting a Reviewer or Back-Up Mod
Getting your setup reviewed should not be considered optional, regardless of your setup, your experience level, or the minimum requirements set by the administration. Your idea of a balanced setup does not necessarily match someone else's, and you can get some good ideas by pitching the idea to people with different perspectives.
If a similar game has been done before, find the moderator of that game and ask them to review yours. If you're not so lucky, find your favorite experienced moderator and ask them to look your game over. Posting in a review request thread does not guarantee your request will be picked up; you should be proactive in seeking a reviewer to ensure that you actually get one. When asking someone to review your setup, it's considered common courtesy to simply ask if they want to review in your first PM, rather than just sending them the setup and saying "please review this". It's also considered common courtesy to tell your reviewer in that first PM if your setup is unusually complex.
During the peer review process, make sure you ask if the central idea behind the setup will work and if any similar games have been run to their knowledge. Having a sample game to work from takes a lot of the guesswork out of balance tweaks.
Your reviewer may ask or tell you to make some changes, and why. While you may see it as a boorish infringement on your glorious idea (to which you have probably become attached by this point) you should generally either accept their criticism and make the changes or discuss why you would like to keep what you have as is. If your reviewer tells you to start over, consider a second opinion... but again, you wouldn't be told this without good reason. If your reviewer doesn't have any comments at all, it doesn't mean that your setup is perfect, it means that you need to find another reviewer.
It would be ideal to get more than one reviewer for your setup if it's anything more than basic. Different people bring different levels of expertise along with their different opinions. You don't want to start the game, have an issue come up, and facepalm in unison with your single reviewer.
Getting a back-up moderator is not as easy because it requires a lengthy time commitment where the backup is anchored to the site and may have to pick up your game at any time. Generally speaking, your backup should be one of your reviewers, and it will likely be one of your reviewers that actually likes the setup. Make sure it is someone you can trust to take over the setup responsibly in the event that you have to abandon it.
Note that a back-up moderator is insurance against you leaving the site - not an excuse for you to leave the site. It is your responsibility to send your back-up mod all of the information that you have to process during the game (night actions, for instance).
A good standard volley of review questions is below.
H. Vi's Review Checklist!
- Is there any way, no matter how unlikely, that this game can be "broken" (made so that one faction is highly unlikely to win even with optimal play)?
- Is there any way, no matter how unlikely, that this will go to a Kingmaker scenario (made so that one player is guaranteed to lose but chooses the winner)?
- Is it possible for a "tie" to occur (i.e. two players from two different Mafia groups alive at endgame)? If you do not want a joint win, then who has priority and wins the game?
- Is there an alignment Cop? with a Doctor? without a Mafia Roleblocker?
- Is there a Cult? with normal recruitment mechanics? (This is a good way to get people to hate you and your entire family)
- Is there a Jester or Survivor? (Again, this may cause a legacy of hatred)
- If the game starts with at least 9 players, can the Town lose without mislynching three times?
- Are there a high number of "confirmable" roles? (i.e. a one-shot Vig can be confirmed by shooting; Masons can confirm each other; etc.)
- Are there "mod notes" or other facets of the setup that are completely unknown to every player?
- Are there any unclarified role interactions where the results could become fuzzy? (Bus Drivers, mutual Roleblockers, and multiple kill flavors are the usual suspects)
- Are there any abilities that can be abused? (i.e. Inventors that can ask for anything they like; Town falseclaim abilities)
- Are more than half of the players power roles?
- Are any of your players given debilitating post restrictions?
- If you were playing in a similar game and received any of the roles from this setup, would you be upset with that game's moderator?
If the answer to any of these questions is "yes", your setup needs a second look - at least.
So once you've gotten your game sufficiently reviewed, it's time to move on to the next chapter.
II. Setup Design for Theme Games
- This section is intended only for games with flavor themes, obviously. If you aren't using one, you get to skip this whole chapter!
Everyone has a favorite non-Mafia passion, be it a video game or a TV show or a book or a time period. So it stands to reason that you'll want to merge that into a Mafia game, right?
It's not quite as easy as that, unfortunately. There are a few hurdles you'll have to clear.
First and foremost, you will be faced with the prospect of a massclaim. Themes where the goodies and the baddies are obvious and unambiguous are extremely dangerous to work with. If you try to include all of the main characters in the setup, everyone can simply claim their role and lynch whoever claims an antagonist or an extremely minor (safe) character. Death Note and Chrono Trigger are examples of popular themes that are easily susceptible to massclaims simply because there aren't many good-side characters to go around. Similarly, games where there are an abundance of power roles are well-suited for massclaim, as the Town can reduce the game to a logic puzzle if it succeeds.
One of the two most popular precautions taken against massclaims is to arrange a "parallel universe" game where alignments are not guaranteed to be as they are in canon. There are two ways to arrange this kind of setup. The first is to cleverly select who you want to be differently-aligned first and then design the setup around that. The other way is to decide on a ratio of alignments (8 Town, 3 Scum, 1 Neutral, for instance), assign each role a number, and let random.org do the sorting for you. (It's cheating if you don't like what random.org returns the first time and you roll again!)
The upside to this is that you can include any characters you like; the downside is that players that have knowledge of the theme but do not know that alignments are no longer the same as usual are actually punished and deceived for knowing the flavor in a game that they likely joined BECAUSE of the flavor.
The other preferred precaution is to deliberately leave out main characters and use them as safeclaims for the scum groups. Here you can preserve the canon alignments at the expense of not having the whole cast in the game. If you're having trouble staying objective about which roles you should include, you can list all of the flavor names you can think of including in the game into a list, use random.org's sequence generator to make a list of numbers that you can pair with the name list, and then only use the names that drew numbers 1 through 12 (or however many players you expect in the game). Every other name on the list is a safeclaim you can dole out at your discretion.
Here is a game that demonstrates both of these. Eleven of the twelve roles were arbitrary characters from the source flavor, and two of them were randomly chosen to be Mafia; this is Precaution 1. In this post, Cobalt claims the main character and expects to become automatically confirmed Town for it; in a game that utilizes Precaution 2 well, this is an appeal to mod-side WIFOM and should have the potential to be a dangerous scum tactic.
Note that post restrictions, preventing people from talking about their Role PM, or otherwise making it against the rules to attempt a massclaim are terrible ideas that should not be attempted. This is essentially pushing the mod-side problem of making a setup that can be broken onto the players. Townies will get very frustrated when they learn that the tactic they can use to solve the game through flavor analysis or Night actions (which ordinarily they can do) is specifically unavailable to them, and clever Townies will find ways around whatever arbitrary post restrictions you set. An interesting subversion in making massclaiming a player-side problem is including a player-side solution. In two of the extremely role-heavy Mind Screw games by Tarhalindur, a non-Town player was given an ability that enabled them to kill players that had revealed any of their ability titles inthread. In the case of the linked game, the "Punisher" role went on to win the game.
This is a tangential subject to massclaims. If a role's alignment is as expected in canon - regardless of whether you are applying the precautions from the previous minisection - your scum players will need safeclaims.
The first rule of a safeclaim is that it has to be believable. The previous discussion has already touched on scum getting easily caught after claiming characters so minor that they are unlikely to be in the game when all the main characters are present. Precaution 2 from the massclaim section is the conscious and deliberately planned way to deal with this; you can do the same thing to a lesser extent in any game provided you ensure that the safeclaims you give the scum are not as obscure as a significant number of the actual Town roles.
The second rule of a safeclaim is that it has to fit the abilities the original role has. Safeclaims that don't fit the roles the scum are likely to claim are useless. Ideally, a scum player should be able to fullclaim each of its nonfactional abilities with different flavor names and be able to pass a spot check for flavor. This requires you to think about which Town roles would have equivalent abilities to your scum roles, and match them together in safeclaims.
Scum roles that are protagonists in-flavor do not necessarily need safeclaims. Town roles generally do not need safeclaims regardless of their canon alignment; however, this could lead to a meta where deliberately claiming "scum" roles would be a Town-tell. Feel free to mix this up.
C. Flavor Knowledge
Believe it or not, there are people who haven't heard of your pet flavor before. How accessible do you want this game to be to people who have no idea what the background for your flavor is?
Obviously, you can make a more immersive and interesting game by heavily applying flavor. Many of the issues that come with massclaiming haunt this approach. First, the character-specific abilities would be confirmable - and therefore cannot be faked or falseclaimed out of. In addition, an amateur scum player would not know this and would be very likely to commit a gaffe when trying to falseclaim. And of course the flavor discussion would heat up to the point where people who try to play your game who DON'T know the flavor will become frustrated. Thus, the game becomes exclusive to people who have experienced your flavor - and depending on what that flavor is, that will cause the pool of people who are able to functionally play the game to dwindle.
The other direction is to make it so that players don't need to know the flavor at all to fully enjoy the game. The upside to this is that it serves as a more casual introduction to the source flavor for people who have not encountered it before, and it also keeps the pool of potential players open. The downside is the same as it was in Precaution 1, two minisections ago - flavor knowledge actually becomes a liability. Consider a scenario if someone claimed a flavor name they didn't know but you recognize as a flavor antagonist. What would you do? What would you do if you had previously been told that you didn't need to know the flavor to get full enjoyment out of the game?
It should be evident by now why Themes are more difficult to run well than standard games, and why mafiascum.net doesn't allow first-time mods to run them. With that said, the freedom the Theme banner gives you is enormous and it's highly suggested you learn the ins and outs of what to do with the potential.
D. Xyl's Setup Design Crash Course - Themed
- Go ahead and do the first few steps of the flavorless game checklist, up to adding the minor roles. You might want to keep in mind what sort of characters are in your game; for example, if there's a town character who logically should be a cop, you probably want to include a cop.
- Now, assign the big power roles to characters. Try to make them fit thematically; if you really have to stretch, that's probably a sign you should go back and pick a different set of power roles.
- Look at each character. Is there any iconic ability they should have? If it's a major ability you should have assigned it earlier. If it's a minor ability, go ahead and give it to them. If the character does have a major ability, consider modifying it to fit the character's flavor. Make sure you don't nerf it into uselessness or make it gamebreakingly powerful, though. Most characters should have at most one ability, although major characters from the source material might have two. Simple abilities with small effects on the game, such as weak one-shot powers or minor immunities, work best here. It's perfectly okay to leave some characters as vanilla. The scum will appreciate it.
- If you have an awesome idea for a really unique ability, it should replace the ability of a power role. Pay very careful attention here; you're going to need to understand how the role will likely affect the game. If you don't understand, find a simpler version that you do understand. You can always use the full version in a latter game once you've seen the simple one in action. Don't put more than one of these in a game. (Xyl: Do as I say, not as I do.)
- Don't forget to give the mafia minor abilities too, as necessary. Unlike town roles, mafia minor abilities should be in addition to their major abilities. Try to make the minor ability something that would make sense for the character as town. If you went with the safe claim method, it should also fit with the safe claim character as town. You can sometimes do an ability that fits both the villain and the safe claim with different flavor, but if you do, you should make sure you provide the player with the right flavor for their safe claim; don't expect them to know the source material.
- Take a look at how many active abilities there are. If more than 2/3 the players have active abilities, or there are many abilities that modify other abilities such as roleblockers, you're going to have a problem trying to resolve the role interactions; cut back some.
- When unique roles come into play, you need to double-check all of your role interactions. If any two roles can even possibly be used in tandem to create a gamebreaking effect or fuzzy results, you need to fix those issues before they inevitably become a problem once the game has started.
- Go through the rest of the flavorless game checklist. Be especially careful when considering the day 1 mass claim scenario.
III. The Queue
- Once again, this chapter is most important to mods on mafiascum.net. On smaller sites, there is only one queue, only one game running at a time, and a very limited number of players to choose from. In that case, while it may be helpful to read through this chapter a lot of the material won't really apply.
As mentioned earlier, Mafia games are long-term commitments and experienced players generally do not haphazardly join them. So it's YOUR job to get those players to come to you! And to do that, we have to turn to one of the more interesting (yet traditionally more sleazy) parts of any good production - marketing.
A. Advertising the Game
So you're midway through the long line to mod a game, and you're really excited about what's coming up, right? ...Right? No? Then get out of the mod queue. Aside from raising the question of why you would willingly put yourself through a long-term commitment you don't really want to do, your players will likely be as excited to play the game as you are to mod it - if not immediately, then during the game when your lack of enthusiasm creeps in.
Still reading? Great! It's your job to get other people excited about your game - excited enough that they pre-emptively join it (pre-in). So, get ready to hype your game as much as you can without being obnoxious. If your setup has a draw to it - a unique theme or mechanic, for instance - definitely get ready to mention it. Make sure your setup has a cool or clever name - people may join based on that alone if you're witty enough. Put a line in your signature telling people that your game is coming up. After you've done these, it's time to really work your charm.
As a general rule, people who would otherwise be ambivalent about joining your game will usually not say no to you if you specifically ask them to join your game via PM. Find some people that you like - or who like you - and send them a simple message detailing that you're going to run a game, what kind of theme or mechanic is in the game (if any), that you would like to see them in your game for [personal reason here], please and thank you. It's important that you don't come on strong or heavy-handed here - you should gear your message toward how your game suits your invitee's interests, not your need to fill a game. Remember: The best advertising doesn't feel like advertising.
Dropping obvious hints that you are running a game that you want to drum up interest in around discussion forums is a somewhat more indirect approach to recruitment. The effectiveness of this tactic depends on how well you play it. If a hot subject comes up on its own in some topic and your game just happens to be geared around that subject - or you decide based on that topic to build your next game around that subject - it's okay to say as much in a single post and answer any subsequent questions about it. If you do this well, you can get people to pre-in that you may not have thought to recruit on your own, as well as score some notoriety for your game that will hopefully make it the kind of game that people will refer to in future discussions.~ To contrast, if you spam propaganda about your game to the point of trying to hijack a thread (or creating your own thread about the game), you will probably just annoy a lot of people.
The best marketing you can get is your prior modding experience. If you have already modded a game on this site and people liked it, by all means call your former player list back and ask them to play again. If the game attracted outside attention, people will start coming to you asking to join your next game. Games have been known to be filled from pre-ins solely based on the moderator's reputation - without the mod having to do anything except say they're going to run another game.
B. How to Refuse Players
You have the right to refuse to let people play your game for any reason. It should go without saying that while you're trying to recruit people to play your games, the flip side is that you don't want some players to join. Rejection is not easy to do - these are people who want to play your game, after all - so here are some common-sense tips for how to do so effectively.
When some mods have a setup they really want to run at full potential, they set experience requirements for joining, like needing to have completed at least two games. This is the easiest and safest way to safeguard your setup, as there's little negotiation to be done with that kind of restriction and it pretty clearly hints that your setup is not for beginners. However, it is highly suggested that you not try this with your first game - the number of games you require your players to have played before joining your game should not be more than the number of games you have modded to completion.
There are going to be some specific players you don't want to join your game, though. If you're going to refuse them, you had better have a good reason for why and be prepared to tell a very confused and disconsolate potential player why you don't want them in your game. Unless the reason for denying them is very obvious and non-negotiable, you should be willing and able to enter a conversation via PM in regards to arguing the merits of whether they should be able to join your game in spite of your initial judgment. After all, people do change... sometimes. Many times the player will consider the re-evaluation conversation too much of a hassle, though, and will simply not bother to pursue the issue - and you get your wish.
Incidentally, do not write up a formal blacklist unless you plan on filling it with people you are unequivocally NOT allowing into your games. You are not in elementary school; you do not need to carry a Ledger of Hate around with you to point to and say "oh sorry, you're on The List". If you don't remember why you don't want someone to play your games, their "ban period" has expired.
Some players will have their own lists of players who they refuse to play with. Generally this isn't a problem, because they will simply not join your game if a player they don't like is already in it. However, if the player with the list joins first and then a player they don't like joins later, you'll have to decide whether you will allow in the new player and have the old one leave, or not.
Barring special circumstances, any player who has flaked out of your games before should not be given a chance to flake out of your games again. Some mods extend this to any player who has flaked out of a game they have played in.
Letting newbies - players who do not have a reputation to research - into your games is a mixed bag. As a general rule, you should ask a few intro questions ("You know how to play and you know games are going to last ____ long, right?") and make a decision based on that. Newbies who really want to play will likely answer the questions at all; you don't have to worry about the ones who don't. Newbies that use proper English get preference.
Are you being asked you to judge people? Absolutely!
While it's unfortunate that you may have to turn away people who would otherwise want to play your game, it's more important that your player list will not cause you trouble later (by flaking, etc.). On mafiascum.net, they can join another game as easily. On smaller sites, that's actually not the case, so the player group must decide on its own whether to allow some players in.
IV. Facilitating the Game!
Okay! The game's full and you've been given the green light to start the game! Now to finally realize what you've been wanting to do for quite the while now...
...carefully. It's a lot easier to screw a game up than you would think. That's why this section is very important - now that you have players relying on your moderation, you need to make sure this part of your job goes well.
First, you need to determine who gets what role. Regardless of any promises or cool hypothetical situations you may have made up before now, you need to make the role assignments random. Random.org can do this for you if you like; or you may use this tool, which is specially designed for this purpose. Not doing so will result in mod-side WIFOM - people outguessing your personality to attempt to figure out who you would give your power roles and Mafia cards to, which most of the time is NOT what you want involved in your games. (And if you don't randomize your role assignments, don't tell people as much.)
Next, you need to send your Role PMs out. It's easiest to create one tab for each PM you send, paste one Role PM per tab, fill in the recipients, check them, double-check them, and send them out. You DON'T want someone to receive the wrong role, especially in a game with unique flavor. It's easiest to place the recipient's name somewhere inside the PM for extra security (or if nothing else errors will be reported to you sooner). Make sure you send your Role PMs in the order of your player list - games have been broken because the scum Role PMs were sent at a different time as the Town roles.
If you have a back-up mod, send them the complete setup (if you haven't already) and role assignments now. Don't leave them to guess later if they have to take over.
When players confirm, they are acknowledging that they have received their Role PM, they understand it, and are ready to play. The on-paper best and most secure method of role confirmation is to have the players PM you back their role while the thread is locked. Do not accept merely picking up the PM as confirmation, as that just shows that they have read the PM but have not made a final, active decision on whether they want to play. Many mods allow players to confirm in-thread so that they can engage in pregame silliness or openly ask the mod to clarify rules, etc.; this occasionally leads to problems but is acceptable, popular, and fun at the same time.
As there's a nasty tendency for one or two players to need to be replaced pregame for not confirming, it's best not to wait for every player to confirm before starting the game. Good numbers of confirmed players to wait on before starting the game are 8 out of 9 players, 10 out of 12 players, 12 out of 15 players, etc. These numbers are chosen to make it very likely that at least one scum player is actually playing at the beginning of the game.
B. During the Game
The mod's job once the game has started is fairly hands-off. You provide vote counts, prods, replacements, flavor, deadlines, kill scenes, action resolution, and answer questions via PM. I'll go into each of these.
i. Vote Counts
Contrary to popular belief, vote counts are not difficult to do and should come at regular intervals. There are lots of approaches to doing them, as well. This guide recommends you make a separate post for each vote count you do, as they are easy to spot in isolated searches and generally stick out well. Editing vote counts into posts, even at the tops of pages, makes them difficult to find and generally ends up being more redundant than the effort put into writing them is worth. Keeping a single vote count in the topic post is strongly recommended against, as they make for an absurdly long topic post (or the continuous deletion of vote counts) and force players to keep the first post displayed on every page for every topic they read.
As long as you post vote counts on a regular schedule, you should only have a few posts to skim through for bold text. Find a time each day when you know you can be on the computer, navigate to your previous vote count, and quote your old vote count in a new tab. Now scroll through the posts looking for bold text (it should stick out to you), and everywhere you see a vote change, change your vote count post accordingly. When you reach the end of the posts, just swap over to the vote count tab and submit the post.
If at all possible the votes in your vote count should be listed in the order that they were cast. Exactly when someone joined a wagon is a fairly common scumhunting tool; listing the people on the wagon in any other order hinders your players while providing no noticeable benefit to you. This also doubles as a check - if you make a mistake in your vote count ("When did ____ vote for ___?"), listing votes in order will give you a window of time to search.
Each time you post a vote count that does not include the deadline or how many votes it takes to lynch someone, a puppy comes down with terminal cancer. This is not an option or a nicety; if you don't do it you WILL be peppered with questions "When's the deadline?" "How many votes do we need?" until you do.
When a player has not posted for a significant length of time, it is the moderator's responsibility to "prod" them - send them a PM saying that they haven't posted recently; please come back. If the player does not post after a certain length of time after you send out the prod, you need to replace them.
Your method of prodding is, well, yours. It's easiest to set a specific length of time (on mafiascum.net, three days works well) and prod anyone who has not posted inthread during that length of time, even if they're on V/LA (they'll need the prod when they come back to the thread~). Some moderators prod only upon request; this is also acceptable. You do not need to announce unrequested prods, although it is generally helpful so you don't get any questions from the players asking what you're doing about someone's inactivity.
Replacements are for people who request it or people who are so inactive that you have to kick them out. The recommended standard for that is five days without posting on mafiascum.net - almost a week. The rate of play onsite is slow, but not that slow. Ideally, you should have someone in mind that could potentially be ready to replace into your game - an e-friend, someone who couldn't sign up for your game quickly enough, etc. If you don't have anyone immediately available to replace in, you're going to need to make a public notice that you need a replacement. All of the advertising from Chapter III comes back here - you need to make your game look more interesting than the ones around it so people will want to join in. You can also let your players try to recruit replacements on their own, although there's a certain factor of trust that goes into that - you don't want your players to recruit people by telling potential replacements their role, for instance.
After making your public notice that you need a replacement, you may find that someone you don't particularly like wants to replace in. Unless someone else also requests to replace in really soon after the fact, take them and deal with it. It's more important that every player slot is filled once the game is underway. This is especially true at Night; if at all possible every player slot should be full when you start a new Day, and you especially cannot have more than one empty slot when Day breaks.
Sometimes a replacement will not come immediately. You need to keep advertising. At the very worst, you may need to make an executive decision and modkill that player slot if it does not look like a replacement is going to come at all.
A good rule to have is that if a player wants to voluntarily replace out of a game, they need to find their own replacement. Player-side problem, player-side solution.
iv. Flavor and Mod Commentary
Flavor is just a snappy script that goes along with your vote counts and lynch scenes. It should certainly be fun to read, but as a general rule you should not tell players more than you have specifically allowed in the setup. While some sites and mods drop clues to help the Town along (i.e. someone saw the killer go by and caught a flash of blue from its avatar), again, this turns the game into a contest of outguessing the mod. However, it is certainly possible (especially in Theme games) to throw in flavor clues that may help the Town without derailing the game. Kill flavors - where each dead player is listed as killed by a certain method according to who killed them - are an easy way to do this. For instance, Mafia and Vigilantes typically kill with a gun, SKs tend to stab their victims, vampires drain their prey, and zombies/werewolves shred and maul their targets. While you can use this to distinguish which faction killed whom, it doesn't directly implicate a player, so it's okay... although you may not want to make kills distinguishable for balance reasons. It's up to what's best for the game, and when in doubt you should simply not bother with it.
Similarly, official mod commentary through the game should not be revealing. One easy mistake to make deals with time meta: If a player was not online at all during the Night phase, they obviously could not have submitted any actions. Whether it's right or wrong, you should try to avoid this if at all possible. For that reason, it is highly recommended that you not cut Nights short because all actions have been submitted.
v. Deadlines and Lynches
Deadlines vary from moderator to moderator and site to site. Three days per Day seems to work outside mafiascum.net, and two weeks to work ON mafiascum.net. I have seen mods put in fixed three-week Days on mafiascum.net and week-long LyLo Days off the site, and in both cases the Day has needlessly stagnated every time. Don't try that sort of thing; it helps neither Town nor scum and makes for momentum-stopping boredom where there was once an active game.
It is acceptable to add an extra half week on mafiascum.net for the first Day or two of Large games. If you are not going to use a fixed deadline policy - and it is recommended you don't when possible - your deadlines should scale directly with the number of living players. Setting Day length to something like (3 + # of living players) RL days is a good example of a dynamic deadline function.
When setting and announcing a deadline, it's very helpful to specify which time zone you are in so there is no ambiguity. You should also stick with your deadline - if someone tries to vote after the deadline passes, even if you're not able to access the site at the time of the actual deadline, you should not count the vote. Of course, this is an easy problem to fix if you set deadlines for when you know you'll be online.
Unless there is a special mechanic that specifies otherwise, if there is no majority at deadline, there should be No Lynch. Deadline lynches that do not require a majority of the Town to be voting for someone facilitate Town laziness. While Towns tend to be lazy anyway, this just lowers the quality of play and consequences of absence.
You should generally not extend deadlines unless you are personally not going to be around for the deadline or you are missing at least two player slots as the deadline gets to be very close. The players will inevitably ask for an extension if THEY cannot make up their minds. The deadline is therefore THEIR problem, and not YOURS, therefore YOU do not need to move it.
vi. Deaths and Modkills
Deaths are pretty easy to do. Check and doublecheck the votes and actions, post that so-and-so died, their role, how they died (Lynched, Killed, or whatever kill flavor you have), and then un/lock the topic. Unless there is a very specific mechanic that says otherwise, you should avoid including extraneous game-related flavor in your kill scenes to prevent mod-meta.
Modkills, on the other hand, are not so easy to do. Each player counts for game balance, and by modkilling someone, you're heavily influencing the outcome of the game - so don't do it except as a last resort or if rules have clearly been broken. What this means is that you should avoid modkilling for things like inactivity or general misanthropy unless you are absolutely sick of the offending player and cannot find a replacement. On the other hand, modkilling for the integrity of the game - like for deliberately quoting a Role PM or illegally communicating with other players - should be reflexive. In the case of an accidental rule infraction - see I.F for an example - a modkill is not necessary so much as a warning. Use your best judgment with determining the intent of the infraction as a guide for whether you should modkill or warn.
Modkills on Town generally end the Day. Modkills on scum generally do not. The underlying principle is that modkills are meant to disadvantage the modkilled player's faction.
Whenever you are making an official mod ruling, you should lock the topic and put up a post briefly explaining why the topic is locked. (This can be as simple as "Mod scene forthcoming".) This will save you from getting barraged by PMs "The topic's locked for some reason, in case you have no idea why".
vii. Action Resolution
By this time you should already have established an order of resolving Night actions. If you have not, Natural Action Resolution is available and is generally what is used, albeit not always consciously. It is important that you have some order of resolving actions so that you don't leave anything out. Using a spreadsheet or master list of actions is a good way to keep actions organized.
There are two schools of thought as to what to do with power roles that do not submit any actions. The most common response is to simply treat it as "No Action". Some moderators will randomize the target of the action if they do not get a response; this is most common with compulsive roles (roles that are required to submit an action).
If you have a back-up mod, always send them the Night Actions and their results. Without knowledge of previous events, their hands may be tied when it comes to taking over your setup.
viii. Giving Away Information
As a moderator you will likely be asked questions about the game. If the questions are about clarifications to the rules, those are obviously safe (although if the players are trying to break the game you may not like the answers you have to give). When questions start getting into subjects like confirming roles, you need to err on the side of whatever caution you can take without contradicting any of your earlier policies. Remember, some players will try to bring the mod in to help in breaking the game toward their Win Condition. It is imperative that you not allow yourself to be drawn into a question-and-answer session about the minutae of the rules or game mechanics.
When in doubt, say as little as you can get away with.
Along these lines, remember that unless you gave yourself a Role PM you are not a player. You do not need to post like one with extraneous information, hype, or invitations to modWIFOM. It's most professional to include all of your own comments as an aside with one of your regular mod posts. You're welcome to answer direct questions, but do not involve yourself in the game.
ix. Dealing with Mod Error
Okay, so you screwed up. Or you inherited someone else's setup, and they screwed up. Obviously, you want to limit the overall upward screwing so the game mechanics don't cascade into catastrophe.
As a general rule, you do not want to make ANY setup alterations after the game starts. So it turns out the setup was unbalanced toward Town, or you sent someone the wrong Role PM. Resist the inevitable urge to try to retcon what you sent out. Altering the setup is at worst a form of favoritism toward one side and at best flying by the seat of your pants in hopes of producing a modified setup that *hopefully* isn't messed up. You would be better served keeping to the original plan as much as possible most of the time.
Let's start from the beginning - sending someone the wrong Role PM. In closed setups, you may not be able to simply reroll and start over. The easiest way to continue is to forcibly replace the player who got the wrong Role PM (sending their replacement the correct Role PM) and continue.
Suppose you found a broken role interaction after the game begins. There's nothing you can do about it.
Suppose you resolved Night Actions incorrectly. Your course of action depends on what happened, precisely... If you sent incorrect results (i.e. a Track result on the wrong target) you may be able to get away with informing the player that you screwed up and who they actually targeted ("When I sent you the "not going anywhere" results I thought you were targeting Player A"), although if you foresee the extra result making a difference at any point in the future you'll have to replace that player. If you send the wrong person results (i.e. a Track result to someone other than the Tracker) your best bet is to replace that player outright without public explanation. If your mod error relates to role flips (i.e. someone died when they should not have, or someone did not die when they should have) you may be able to get away with telling the party that did the saving/killing that there was a mod error and that the errored party should have lived/died. If you reveal the wrong player's role, that's it - you need to abandon the game.
Suppose you put a Jester or Cult Leader into the game. It's not too late to start writing apologies for postgame.
Another general rule is that your problems are necessarily worse after someone talks about them inthread. In the case of someone dying when they should not have, if the Doctor claims "hi I protected this person and they died anyway so I think the mod screwed up", the ruse is over.
If your error relates to rule enforcement, there are two cases to be considered. One is if the rule has already taken effect, and you realized after the fact that it was a bad idea. Here is a game that demonstrates this with the two modkills D1 for discussing ongoing games. Both of the modkills were objectively terrible, no question - but consistency is the goal in these cases. The other case is where the rule was only intended to be temporary. This game provides a great example of a temporary rule - the moderator, noticing nonparticipants during Day 1, made a temporary rule that each player should send in a PM or be prodded at the beginning of the next Day. But in announcing who was prodded at the beginning of the Day, by the same logic as time meta those players could not have sent in the Mafia kill or any other actions. This small disclosure nearly broke the endgame on its own. In cases like these, consistency is decidedly not the goal - whatever it was you did, don't do it again.
Similarly, when inheriting a game and discovering a previous moderator's error, your goal should be to stay consistent with the previous moderator's rulings. However, any secret mechanics that have not yet come into play can be modified if necessary.
Finally, remember: It's much, much more important to be correct than to be on time. Never "rush" mod scenes, and always double- and triple-check what you're about to post. Your players can wait.
So you've finally made it to the end of the game! There are only a few steps left to make your game complete!
It is expected that you will post the complete setup (each Role PM and who got which role) and all night actions at the end of the game. You won't get any suggestions or much praise about the setup until people know all the details of it. This sounds like an obvious check on your to-do list, but it's surprising how many times it just doesn't happen.
It's okay to let discussion of the game go on after the game is over, and even for you to participate in it as well. Do not be condescending to the players even if they were particularly boneheaded; you're speaking from the viewpoint of someone who has had perfect knowledge of the setup throughout while the entire point of the game was for people to try not to be wrong. You should not lock the topic for the last time until discussion has obviously ceased for a while - cutting off discussion hinders your players from learning from the experience or even establishing friendly relationships with each other.
If you are ever in doubt, don't be afraid to ask for help. The inconvenience of asking for help is vastly preferable to messing up and/or running an unsatisfying game.
With that in mind, this guide was typed up to help you. Please read it!