Aight. I have some downtime, so here goes.In post 5139, lilith2013 wrote:Thank you for all the feedback and appreciation <3
@Bingle (and, well, everyone) the original plan was to run a series of these with similar mechanics (ie the room exploration, the night actions, etc) but with an entirely different Haunt and set of win conditions each time. In that sense, I think the mechanics themselves can be adjusted for balance, and separately the Haunt storyline can be more balanced as well. I’m looking for more specific feedback on the former and more general feedback on the latter. Here are the types of things I’d put in each category.
mechanics
- should stats be private?
- should the sender of items be private?
- should you not receive the identity of your attacker?
- should you not receive the results of your own actions?
- which dice roll(s) should you see the results of, if any?
- should rooms be harder to find or have a different distribution?
- should actions be unlimited or set to a different threshold?
- should stats have a different scale/limit?
- do stats need to be easier or harder to change?
- were items too overpowered?
- should the number of players be adjusted (eg 3v8)?
Haunt balance
- were labs too easy to destroy?
- should scum have been able to destroy them in order to prevent confirmability?
- should the distribution of labs have been higher?
- should the labs have been less difficult to activate?
- should doctors have had additional factional night actions? what type (ie extra attacks, extra lab activation, extra open actions for whatever they want, or some combination of those)?
Loose modding philosophy (This, of course, is just how I see things and as such is entirely subjective. Feel free to disregard anything you disagree with.):
Unless there is a reason to provide information, don't. Generally speaking, if you leave things vague and only give hints as to what happened, people have more to figure out, which is a major portion of the draw of hidden information games imo. Don't skimp on knowledge like "Voting is normal." or "There are no nightkills." because those are central to understanding if you're going to enjoy the game.
Example: Scum could not destroy labs. There was absolutely no reason, however, for town to KNOW scum could not destroy labs. Scum had to roll an X to activate a lab. There was absolutely no reason for town to know scum had to roll an X to activate a lab. Maybe scum had to pick up a corpse and then they could activate a lab. Maybe scum had to kill someone and that activated a lab. Room for speculation adds fun both to the people who can speculate and gives scum room to argue their way out of corners.
Approach to design:
When designing a game, decide what your core setup is. Do you want to build a game based around the dice mechanic being 0-2? Do you want to build a game around "stats matter"? Do you want to build a game around the exploration function? What is most important and what can be sacrificed in the interest of making the game more fun. It really helps if you only have one or two things that HAVE to be present so that you can tweak everything else to improve the quality of the game. I'm not saying you can't have a game that has 15 different mechanics and they all mesh well, but in my experience, the games that do that well start with a core concept and then branch into those mechanics as polish.
Approach to balance:
There are two separate things to consider when balancing a mafia game. Balance and swing. Neither is necessarily required to have a fun mafia game (see: literally any fun bastard game ever, my recent fearvig games), but generally equally balanced and low swing games tend to be more enjoyable for more people. From a balance perspective, I have a checklist:
How many days are required for scum to win? (For a mini, you want this to be roughly 3-4 days)
How many days are required for town to win? (For a mini, you want this to be roughly 3-4 days)
How many days is it possible to take for the game to end? (If this is ever more than half the player list size, that's a major red flag. Games should have a feeling of an inevitable scum win.)
How many miseliminations are required for scum to win?
How many miseliminations are required for town to lose? (If these are ever less than two, that's a major red flag. Note: there are setups where this answer is 0 that are both playable and fun (EHOBANHAR, for example))
How many confirmable town can be generated? (This should at most be 1/3 of the town, and even then that's a fuckton of conftown. Generally speaking, you want this to be 2-3 for a mini. This does not just refer to cop clears, btw, but also successful doctor protections without ambiguity as the the reasoning, vigs, claims that are obviously town, etc.)
How many players can be rendered unkillable? (Roles like BP or commuter and protectives)
What is the overlap in the previous two numbers? (Any situation in which a player is both confirmed to be town and unkillable is demoralizing to scum. Any situation in which there is more than one player who is both conftown and unkillable is just mean.)
Are there any roles in your setup that cannot be answered? (BP IC, investigating hider, Elite PGO, cop without a way to cast doubt on the results/find the cop/stop actions, etc. Basically, anything that can't be answered is bad, because if someone who is particularly on point that game rolls it scum just can't answer that role and the game is over at role distribution.)
Are there any broken interactions in your setup? (Follow the cop, etc.)
You might notice that most of that boils down to: look at the best possible case for town and the best possible case for scum. You want those cases to feel very similar, and you want the spread of cases in between to be roughly symmetrical. It's also a very good idea to make that range as narrow as possible, because the closer a game is, the more satisfying it feels to win and the more fun it is to go down swinging.
Swing is a little harder to address. You have some central mechanics that are going to naturally make games like this very swingy. The exploration will naturally favor one alignment or another regardless of what you do. The dice are going to give wonky results sometimes. The career mechanic is going to create fluctuations if you keep it as a smalltown distribution. It's just the nature of the beast. There are mitigating factors you can add (the career of pick a type of room to explore, the suggestion of a flat benefit to scum while attacking, having a pool of careers tagged for scum only or town only) but games like this one are naturally going to be swingy.
Game Size:
The larger a game, the easier it is to balance. I know, it sounds backwards. Doesn't make it any less true. More slots means more room to work to iron out interactions. The difference between a 19 player and a 21 player game is much smaller than an 11 player and a 13 player. If you put in the effort, it's actually easier to give yourself the wiggle room to account for interesting mechanics with a complex setup. The larger the game, the less forgivable the swing factor is. There's just a huge difference between the commitment to playing a micro and a mini and a large. If you lose to swing in a micro that took 3 weeks, you're going to feel a lot less salty about it than a large that took two and a half months.
As far as specific applications to this game: Everything town can do, scum should do better. Basically, what I mean is that while town can have unique individual careers that give them a power, any action that town can take should also be able to be taken by scum. The scum toolbox should always be the VT toolbox plus additional tools. From a game theory side of things, speed is disproportionately powerful. There is a reason everyone chose to go to the gym N1. More actions = more power. I would suggest instead of having speed give more actions, you might want to make it a function of might. Strength could be your offensive stat, speed your defensive, and the combination your health. That leaves knowledge as the defacto interaction with the house stat (the stat primarily responsible for events/haunt actions/misc), although obviously strength and speed could be used some amount of the time.
Exploration, also, doesn't necessarily have to happen at night. Just like with items, you could leave it as an option to explore during the dawn phase and just have it be a the first x players who explore each dawn find a new room. This gives you control of how many rooms can pop up in the game and makes night actions less leashable. You still have the option of permanent rooms which can be used at night, fighting, looting corpses, and any special abilities from items/careers, but there doesn't really need to be a sacrifice of those abilities to progress the game. Also, anonymous scum activities like scum can anonymously explore one room per day and scum can anonymously attack one player per day could be attributed to the Monster of a haunt, and as such wouldn't need to be public at all.
tl;dr: if you run similar setups in the future, feel free to reach out for review/balance help or a player.
