The Art of Game Creation

[Ookla de los Cuervos he/him]

2 minutes ago, Araris Valerian said:

I think era 4 Scadrial would be an appropriate setting for this, if you wanted to Sanderfy these rules. We've had a few cyberpunk themed games run already, and that's the route they've taken mostly.

oh, thank you!

[DrakeMarshall he/him]

hehe wouldn't it be funny if there was a game where you didn't know your own role

Spoiler

Core Rules

Spoiler

Standard V/E setup.

Two vote minimum. Ties decided randomly.

48 hour days, 24 hour nights.

PMs are closed by default.

Death Rattles

Spoiler

You may submit to the GM a 100 character maximum message to be revealed on your death. As long as you are alive, you can update what you want your death rattle to be whenever you want. If you fail to submit any last words, I reserve the right to invent something colorful.

Surges

 

Spoiler

Players of either alignment may have access to a single Surge.

The total numbers of each Surge in circulation will be made public at the beginning of the game.

However, you are not informed what Surge you possess, if any.

You can only find that out by trial and error and process of elimination.

Regardless of whether you have any ability or not, you may submit an order to visit another player each night. (Self-targeting not permitted.) If you have access to a Surge, you will use it on that player. If you do not have a Surge, nothing will happen. Unless otherwise specified, you will not receive any feedback for actions.

ADHESION: The target is added to a group PM with you and those you’ve targeted.

GRAVITATION: If your target is attacked, you kill the attacker.

DIVISION: If you die in the next cycle, you take your target down with you.

ABRASION: You and the target gain an individual PM.

PROGRESSION: The target is protected from an attack.

ILLUMINATION: You are notified if the target is Unmade and only one Illumination action targeted them.

TRANSFORMATION: The target’s vote is changed to match your own vote the following day.

TRANSPORTATION: The target is swapped with you, actions that would hit one hit the other.

COHESION: The target’s action is blocked.

TENSION: The target is made immune to all actions except attacks for the night.

Unmade

Spoiler

The Unmade team has access to a google doc in which to collaborate, and a factional kill they may use every night. In order to use it, one member of the Unmade chooses to carry out the kill in lieu of performing other actions.

The Unmade team also has access to some number of Unmade powers. Unmade powers are factional abilities that are not carried out by any one team member, don’t require an action, and are all single-use. Unlike Surges, the distribution of which Unmade powers are available will not be publicized, although of course the Unmade will be told.

ASHERTMARN: Redirect a player to a new target.

BA-ADO-MISHRAM: Learn the exact role of a targeted player.

CHEMOARISH: Make the nightly kill unstoppable.

DAI-GONARTHIS: Gain a PM with a dead player for a cycle, and compel truthful answers to yes-or-no questions that have a clear right and wrong answer.

MOELACH: Rewrite the role flip and/or death rattle of the nightly kill target. This ability is not used up if the nightly kill target doesn’t die.

NERGAOUL: Designate a player during the night. If they are executed the next day, the earliest voter falls victim to the Thrill, permanently converting to the cause of the Unmade.

RE-SHEPHIR: Send midnight essence to carry out the nightly kill. This means it doesn’t require an action from a player, and that there can be no reprisal from Gravitation.

SJA-ANAT: Specify a member of the Unmade team. All villagers who visit them with Surges that night are corrupted, and will unavoidably die after one cycle.

YELIG-NAR: Allows one member of the Unmade team to use any Surge of their choosing with their action.

the end

 

[Kasimir he/him]

On 3/28/2024 at 11:27 AM, DrakeMarshall said:

hehe wouldn't it be funny if there was a game where you didn't know your own role

 

Says the guy who collaborated on a game where you don't know your own alignment

[DrakeMarshall he/him]

On 3/29/2024 at 10:16 PM, Kasimir said:

Says the guy who collaborated on a game where you don't know your own alignment

surely not knowing your own role is sufficiently cruel and unusual

even I'm not so deranged as to make a setup like the one you're describing smh

this is libel and calumny and all manner of unproven and baseless accusation

[RoyalBeeMage he/him]

On 4/1/2024 at 11:11 AM, DrakeMarshall said:

surely not knowing your own role is sufficiently cruel and unusual

even I'm not so deranged as to make a setup like the one you're describing smh

this is libel and calumny and all manner of unproven and baseless accusation

What about a game where you don’t know alinement or role?

[Experience he/him]

3 hours ago, RoyalBeeMage said:

What about a game where you don’t know alinement or role?

Idk, because generally you want informed minority vs uninformed majority, and in that case you just have uninformed minority vs uninformed majority. 

I mean, maybe it could work if there's pretty simple ways of figuring out for yourself what your own alignment is.

[Araris Valerian he/him]

I think a game where you don't know who has died yet is the pinnacle of missing information.

[Ookla de los Cuervos he/him]

Just now, Araris Valerian said:

I think a game where you don't know who has died yet is the pinnacle of missing information.

especially with inactives. albeit you could probably tell whos dead by their activity.

[Experience he/him]

And you'd want to make it so players can only post once per cycle in order to make finding out who died simply by who posts much harder

[Araris Valerian he/him]

13 minutes ago, TheRavenHasLanded said:

especially with inactives. albeit you could probably tell whos dead by their activity.

Here are the rules: https://docs.google.com/document/d/1-6F8d9LW9q8lSaZQawfGEwB-Skp5TPA5USZmbudNqI0/edit?usp=sharing

I have already posted these rules a ways back in the thread, but the idea is that "dead" players aren't removed until their entire faction has been killed (or notified that they've been killed). There's plenty of vote manip and no official vote tally.

[Ookla de los Cuervos he/him]

hmmmmm interesting.

[DrakeMarshall he/him]

2 hours ago, Araris Valerian said:

I think a game where you don't know who has died yet is the pinnacle of missing information.

everyone is schrodinger's cat

they are in a state of uncertainty most of the time, even if they get attacked or the like, except when you specifically check whether or not a player is alive, then you either get a flip and they are dead for reals or you don't and they're still alive

[Araris Valerian he/him]

Given the low player counts we’ve had recently (TKN’s game is at 4 right now, though I’m guess that’s partly end of semester syndrome), I was thinking about what kind of mechanics make for a better game with 8-10 players.

Some general question that may be food for thought:

• Are there things we can pull off with smaller games the would be relatively infeasible with more players?
• A game with 8-10 players would typically last 3-4 cycles. What mechanics should we avoid to keep from pushing this lower, and what ways could we boost game length without frustrating players or otherwise compromising play?
• Are there any previous rulesets that would be suitable (possibly with minor modification) to our current player counts?
• Since we are a Sanderson forum, what settings might be particularly suitable to the sorts of mechanics/game lengths/game sizes that come from the above?

I have my own thoughts on some of these but am curious to hear from y’all.

[DrakeMarshall he/him]

15 hours ago, Araris Valerian said:

Are there things we can pull off with smaller games the would be relatively infeasible with more players?

I’m not sure. There are definitely things that you can’t do if you keep adding players. I definitely noticed it was getting hard to meaningfully read and engage with every other player in the most recent AG due to how many players there were (…at the beginning at least, coincidentally the only part I was around for, but regardless). And something like a 100-player game just sounds unplayable.

Going down to the other extreme of really small games, it seems like you get more opportunities to focus on and engage with the fewer players you have.

Which is a pretty good segue into the next question, because if the game ends too soon, then you don’t really have the full opportunity to take advantage of that.

15 hours ago, Araris Valerian said:

A game with 8-10 players would typically last 3-4 cycles. What mechanics should we avoid to keep from pushing this lower, and what ways could we boost game length without frustrating players or otherwise compromising play?

If the paradigm for progress & winning is based on player deaths, and the rate of deaths is fixed, then small games will be shorter. I see no way around this.

Therefore, if you want longer games, there are basically two categories of ways to do that. You can either adjust the rate of deaths or change up the paradigm so that player deaths aren't the de-facto measure of progress & victory.

Adjust the rate of deaths.

The low-hanging fruit here is that we can and sometimes should probably avoid having there be a bunch of extra kills flying around in small games. Including tertiary kills (ie coinshots, serial killers, etc) above and beyond the standard execution+NK duo found in most games is a Choice, and one you can reasonably just say “no” to.

Beyond that, we could take a more proactive approach to fundamentally lower the rate of deaths below the traditional level of 2/cycle.

A naive solution might be to give everyone an extra life. I think this is obviously quite a bad solution, and I will duel pistols at dawn with anyone who seriously thinks it’s a good one. I would find it very frustrating. Functionally speaking, it would just make everything in the game happen twice as slowly.

• However, on further examination and soul-searching, I don’t think the problem with “give everyone an extra life” is actually that it makes the game slower. I don’t think I actually fundamentally have any beef with it taking longer for players to die. The problem with the idea is really that it’s entirely fricking boring and unimaginative, and utterly devoid of interesting tactical or strategical implications. Other than slowing things down, it changes nothing about how you play the game. The optimal approach is literally just, grit your teeth and kill each person twice, rarely-if-ever deviating or turning aside once you’ve committed to killing someone.
• To put it another way, my issue with “give everyone an extra life” isn’t that it slows down the game, it’s that it JUST slows down the game. That’s it. It doesn’t do anything else, it doesn’t add anything, it doesn’t increase the number of meaningful decisions you get to make over the course of the game. It’s incredibly lazy. If you’re going to stretch a game out to twice as long as it was, you’d better double the amount of content and strategy and meat in your game as well, or else the pacing is going to be shot to hell.

<shameless advertisement>As an example of one approach to lowering the rate of deaths in a way that’s actually interesting, I’d like to talk about my Reckoners game. One of the fundamental premises of the game is that many of the players are high epics that possess prime invincibilities, making them very hard to kill. There’s quite a bit of diversity in how different invincibility powers work, ranging from “you always die without any flip and then reincarnate a cycle later“ to “you control two accounts, and cannot be harmed unless both accounts are tapped at the same time” to “everyone around you is subject to a compulsion preventing them from deliberately taking hostile action against you.” The rate at which players die in my Reckoners game is likely going to be noticeably less than average, due to all these crazy powers. But of course, it’s not impossible to kill high epics. Each of these powers has logical drawbacks and potential blind spots. On top of that, each epic has a weakness that makes them completely vulnerable. What this means in practice is that if you try to kill someone and it doesn’t work, well then you did it wrong. It’s a puzzle you need to solve, and one that likely requires you to make social deductions and educated guesses about other players. If you’re clever enough, you can take down a high epic on your first try with no backup, and on the other hand if you turn your brain off and just try to brute force everything, you are going to be less effective. Compare this to the “everyone’s a thug” game where if you fail to kill someone, the only conclusion is “oh well, guess I need to do that exact same thing again.” I would argue that the difference is night and day.</shameless advertisement>

Here’s another random thought. I have no earthly idea how this would work, but I bet there’s some kind of interesting SE game premise where each player controls multiple pieces, and a player isn't necessarily knocked out from just losing one of their pieces. Coup is a social deduction game that more or less achieves exactly that, if any of you folks have played it. Of course, Coup is also a free-for-all game, whereas I’d want to try something that preserves the V/E uninformed majority vs. informed minority dynamic common in SE.

Another option here, which is pure sleight-of-hand but arguably still pretty effective, is to have mechanics that allow players that have been killed to keep participating somehow. Kas seems to be a fan of this angle, given LG92 or the last BT. Letting dead players keep participating probably won’t make a game actually any longer, but it might still feel longer, because your experience of the game won’t be cut short by dying. The thing is, I would contend that making the game feel longer and more substantial is honestly mostly just as good as making it actually longer.

Change up the paradigm so that player deaths aren't the de-facto measure of progress & victory.

Imo, this very loosely means imitating games like Avalon or SH. These types of games are still about figuring out who is trustworthy and who is a traitor, but progress is measured by something other than “killing people on the enemy team.” Because of this, you have more flexibility with how long you want the game to last, and smaller games don’t necessarily need to end quickly. I honestly think there’s a lot of depth to these kinds of setups and that they probably haven’t been explored very fully. For that matter, I really don’t believe Avalon/SH is the pinnacle of what games of this type could do.

I would be remiss if I didn’t reference Fifth’s Alice in Wonderland game. The game was actually three short-form games in a trench coat, and… It worked. It was pretty great! Imo, that was because each of the three minigames was fresh and creative and really fun and not particularly because the overarching point system was intrinsically very engaging… but it didn’t need to be. This is a practical example of a game where death wasn’t the final measure of progress, and this allowed it to subvert the usual game length constraints.

Here's another angle of approach. While the time loop in the BT Kas and I ran was still basically a game where winning was contingent on player deaths (for the village, there was a specific villain that needed to die -- all other deaths got reset, but the loop was still structured around killing and flipping players, with the goal of killing the correct one), one could conceive of other time loop games that don’t work exactly like that. The only hard requirement in a time loop game is that there has to be some way to break the loop and end the game, but nobody says that it has to involve killing someone. If you had a time loop game with a different objective, then there’s no reason it couldn’t last as long as you wanted. The trick of course is coming up with an objective that’s interesting and fun in practice.

Conclusion

At the end of the day, the real challenge isn’t contriving a way to make the game last longer. The real challenge is making sure players actually have something to do with the extra time. To these ends, I think it’s worth asking, how much hidden information is there for players to find in your game?

Information asymmetry is what makes SE tick. It’s what makes any social deduction game tick. I think a valid way of defining the “small player count” problem is actually just that small population games have fewer role/alignment slots, and thus there’s fundamentally less information, less stuff for players to figure out and solve. This is why “give everyone an extra life” simply falls flat in addressing the problem, because it doesn’t change how much information there is to figure out. If you’re going to write a game that runs for a long time with a smaller number of players, you need to address the fact that small games tend to suffer a deficit of hidden role/alignment information. You need to make up for the lack by adding back hidden information in some way, shape, or form. Maybe that’s adding crunchy epic power interactions and secret weaknesses. Maybe that’s breaking the game down into minigames and giving each minigame a separate and fresh distro. Certainly there are other ways to do it that I haven’t covered or even thought of! But I’m quite certain there always has to be something.

The kick in the teeth, of course, is that all of this is actual work. The standard formula that works for bigger SE games is well-explored and well-liked, it can be used as a starting point with relatively little effort and yields good results. It’s a genuinely good thing and it’s easy to take for granted how convenient it is, until you try to do something new.

P.S.

To clarify, I am not saying that I hate Thugs. I think games with 1-2 Thugs can create interesting tradeoffs and tactics, where the Thug is incentivized to be baity and draw more attention from the opposing team. But if everyone’s a Thug then no one is, and you don’t even get any of those juicy tactical considerations.

 

...Anyways, all 4 of your questions were quite good, but I'm tired and that's enough talk from me someone else can share their thoughts on the last 2 questions ig.

[Kasimir he/him]

1 hour ago, DrakeMarshall said:

I’m not sure. There are definitely things that you can’t do if you keep adding players. I definitely noticed it was getting hard to meaningfully read and engage with every other player in the most recent AG due to how many players there were (…at the beginning at least, coincidentally the only part I was around for, but regardless). And something like a 100-player game just sounds unplayable.

I think there's the flip side which I've commented on on a few occasions: a lot of the volume can be illusory because you get pity sign-ups or overenthusiastic players who don't think through if they can commit and then that plays merry hell with balance and the game dynamic.

1 hour ago, DrakeMarshall said:

Here’s another random thought. I have no earthly idea how this would work, but I bet there’s some kind of interesting SE game premise where each player controls multiple pieces, and a player isn't necessarily knocked out from just losing one of their pieces. Coup is a social deduction game that more or less achieves exactly that, if any of you folks have played it. Of course, Coup is also a free-for-all game, whereas I’d want to try something that preserves the V/E uninformed majority vs. informed minority dynamic common in SE.

Everyone is a lich with a phylactery...

Or HP system I guess. Ah soddit.

1 hour ago, DrakeMarshall said:

Another option here, which is pure sleight-of-hand but arguably still pretty effective, is to have mechanics that allow players that have been killed to keep participating somehow.

I think there's a game we discussed that might work, the one that reminded you of Gunnerkrigg Court, where the goal was to reach a specific informed equilibrium point or something, so dying doesn't end your work in the game. Dunno.

1 hour ago, DrakeMarshall said:

Kas seems to be a fan of this angle, given LG92 or the last BT.

They were both BotC expys, what do you expect  

1 hour ago, DrakeMarshall said:

A naive solution might be to give everyone an extra life. I think this is obviously quite a bad solution, and I will duel pistols at dawn with anyone who seriously thinks it’s a good one.

IT WORKS THANKS

Your place or mine?

j/k but okay I am referencing a very specific Meta game, which, as all Meta games/distros, was designed to break certain assumptions SE had about how the game was played and that was why it was so effective/fun. It's just not really appealing for a normal game.

Sorry, life's been tiring lately and I don't have much significant to add. I think something that can also be considered is replayability - if a game is short but fun, it should be ok for it to be designed to be replayed twice. I'm sort of thinking of something like Araris's Aracle QF. 

1 hour ago, DrakeMarshall said:

<shameless advertisement>

Joining in on this - I think (I don't know if successful) - my Star Wars game takes a slightly different approach, though I'm still trying to adjust a number of interactions. I designed it for 8-12 players but the idea was to take the same walking wincon model from MR2 and MR61 and to use that along with the potential for players to go SK to realistically keep things circulating a bit. The initial idea was to try to create unique and distinct roles that had character, flavour, and interacted and role madness so that even if it ended fast, players would hopefully have a fun experience with a role they enjoyed. I feel like embracing the shortness but packing extra oomph is a valid tack too - at smaller player counts, it's possible to put a bit more care into the roles I think.

[Araris Valerian he/him]

I like the idea of having extra information to solve for. My next LG is a semi rerun of the PM Unsafety game, and I think I’m going to retool the locations mechanic to be like a mini One-Night game every cycle. Then take away most of the other roles and require knowing location for votes to go through.

One thing that might be worth considering adding to games is some kind of anti-hammer mechanic. In large games, hammering is fine (though it can problematically compound with vote manip) because 7 cycles versus 8 isn’t a huge relative difference. But dropping from 4 cycles to an effective 3 is a lot more impactful, and in my opinion cuts out a bunch of otherwise fun solving/gameplay.

[DrakeMarshall he/him]

6 hours ago, Kasimir said:

I think there's the flip side which I've commented on on a few occasions: a lot of the volume can be illusory because you get pity sign-ups or overenthusiastic players who don't think through if they can commit and then that plays merry hell with balance and the game dynamic.

Yeah, agreed. I mean, we've both said it before

I think designing a game with some assumed number of players in mind is something to be avoided where possible. Although uh. I definitely do it sometimes.

It creates more issues than just the illusory volume thing.

But enough about that.

On 4/11/2024 at 5:06 AM, Kasimir said:

j/k but okay I am referencing a very specific Meta game, which, as all Meta games/distros, was designed to break certain assumptions SE had about how the game was played and that was why it was so effective/fun. It's just not really appealing for a normal game.

I-- I don't want to duel Meta

On 4/11/2024 at 5:06 AM, Kasimir said:

Everyone is a lich with a phylactery...

Nalthis

3 hours ago, Araris Valerian said:

One thing that might be worth considering adding to games is some kind of anti-hammer mechanic. In large games, hammering is fine (though it can problematically compound with vote manip) because 7 cycles versus 8 isn’t a huge relative difference. But dropping from 4 cycles to an effective 3 is a lot more impactful, and in my opinion cuts out a bunch of otherwise fun solving/gameplay.

Huh. That's an interesting thought. I hadn't really considered it.

Anti-hammering seems pretty doable if you're okay with having variable-length cycles... Just say that the day can't end if the majority train shifted in the last X hours.

Doing it with fixed-length cycles (which are preferable in a number of ways) seems... Slightly harder.

Oh, here's an idea maybe. What if you ruled that votes have to be spaced out? All voting has a cooldown. You cannot cast a vote if anyone else has voted in the last 10 minutes.

To prevent it from getting tedious, let's just rule that it's actually perfectly fine to post a vote less than 10 minutes after somebody else voted. Your vote will still be counted normally, unless the day ends before the 10 minute window expires. If multiple people are "in line" to have their vote counted, the order their votes is processed is decided randomly, and each one incurs a new 10 minute cooldown.

To prevent silly exploits, let's also rule that if one player has multiple votes, you just throw out all the ones but their most recent. Because that's the only one that matters anyway, and people shouldn't be able to DoS other people's votes.

With the pacing of most cycles, this is not even a noticeable restriction. You won't ever have to think about it if there are multiple hours left in the day, no matter how much voting activity is happening. It only really affects things if you're trying to do a lot of coordinated votes and you're very close to EoD. Of course, this applies equally to last minute village flash trains as it does to last minute elim coordinated hammers, so there's that to keep in mind.

I wish it were simpler and more intuitive, but I think it maybe does what you want? Can you see any problems with this

[The Unknown Medallion he/him]

So, I don't really have a good answer to the idea of low player counts, though I do have one main thought that hasn't come up yet. 

The biggest problem with sub-9 player counts, is it's hard to rationalize 1 elim. It's too fragile and luck dependant, but multiple elims end the game too quickly. It's really hard to find a way to balance it. 

I happened to stumble across this episode of No Rolls Barred today, where they house rule Clue(do) into a mafia-esque game. It solves this problem fairly well, and worked just fine with only 5 players, which SE seems to always be able to find so far. (My game has 4, 5 if Araris didn't need to IM). 

That's a pretty good example, though other ideas would also be good. I'm tinkering with a Fused game where they can body hop unless certain items are used.

[Aeoryi she/her]

On 4/10/2024 at 11:37 AM, Araris Valerian said:

Given the low player counts we’ve had recently (TKN’s game is at 4 right now, though I’m guess that’s partly end of semester syndrome), I was thinking about what kind of mechanics make for a better game with 8-10 players.

Some general question that may be food for thought:

• Are there things we can pull off with smaller games the would be relatively infeasible with more players?
• A game with 8-10 players would typically last 3-4 cycles. What mechanics should we avoid to keep from pushing this lower, and what ways could we boost game length without frustrating players or otherwise compromising play?
• Are there any previous rulesets that would be suitable (possibly with minor modification) to our current player counts?
• Since we are a Sanderson forum, what settings might be particularly suitable to the sorts of mechanics/game lengths/game sizes that come from the above?

I have my own thoughts on some of these but am curious to hear from y’all.

There are easily more ways to adapt to the smaller player counts we see more often. I think the most important thing to keep in mind is to have a plan in place for lower player counts. Just like how a 34 player game would not be acceptable, a game that requires 14+ players is also not acceptable. The easiest solution would be to have a role setup pre-planned for your game. For instance, if I was GMing a Tyrian falls game, I might have a setup planned like:

Quote

 

7 players: 2x Village smoker, village soother, 2x villager, elim smoker, elim soother

8 players: Village coinshot, Village thug, Village lurcher, 3x Vanilla Villager Elim Lurcher, Elim Seeker

 

Having a plan will make your game able to run, even with low player counts, generally no matter the ruleset. (Although the 7 player setup looks sketchy). You don't have to share the distro with the players, since that's not really a thing here, but it will be beneficial to have something planned for most occasions.

Here's another thing: Don't expect games to last long. Nerfing the amount of kill potential or making kills less likely to hit is not the best thing. I think the better course of action is to simply embrace it, or build a ruleset around it. If Exlo/LYLO is at cycle 3 on average, so be it. Obviously, don't make the game be over by cycle 2, since that's lame and unsatisfying, but don't try and make your game last longer by tampering with the eliminator kill or execution. Here's why tampering with the two kills that should always happen is bad:

Say I have a 7 player setup where players have a 50% chance of dodging a kill. What does this accomplish? Well, it makes my game longer, yes, but it also means that people will not be enjoying it when the execution fails for the third time in a row. It means the eliminators will be cussing at me from their doc because they can't kill Kas by cycle 3. It means that people will blame the outcome of the game on RNG, and that's not how the game should be decided. 

Now, you might say, “Well, I would never design a ruleset like that, that sounds terrible!” Well, let me modernize it a bit for you, again using tyrain falls:

Quote

8 player setup: 2x village thug, 1x village lurcher, 2x village vanilla, 1x village smoker 1x elim thug, 1x elim tineye

What's the problem here? Well, if you look closely, the eliminator kill has a default of a 50% chance to fail, and the execution doesn't even kill the eliminators half of the time. In other words, this setup is fundamentally the same as the 50% survival chance setup. So there will be the exact same problems with both. I know the lurcher doesn't really count as protection, but you get the idea. Don't try and make a smaller game larger by extending the duration through protection. Most successful small games work because they aren't a large game optimized for a low player count. They succeed because they use the player count to their advantage.

Now, let's move on to araris’ questions.

1. There are some setups that work better with less players. I can't name any right now, but I so think that smaller games can have setups that a larger game couldn't have. 

2. As I have said previously, I don't think prolonging the game is a great strategy to combat low player counts.

3. I think some good things to convert old rulesetups are: 

Limiting the role usage (I.e charges, both non-replenishable (like lg98b) and replenishable (like lg99)) for chronically powerful roles. This can help it be less RNG. For instance, if you want to add a coinshot, but don't want the game to be decided by c2, maybe give the coinshot one shot (flavored as a vial or something) and you'll still be able to have 3 kills on one cycle and 2 after that. But let's say you want something more. Don't feel afraid to add other limitations. Let's return to the coinshot in your setup. So maybe you don't want the coinshot to become a vanilla after cycle 1. Maybe you want to make them only able to shoot on odd cycles, which gives them some relevance to them having that role. Maybe you want the elims to have some resistance against your coinshot, so you give them a thug ability that only works on even cycles, solving the 50-50 issue somewhat. Or maybe you want the coinshot to choose when they shoot, but not be able to shoot every cycle. Give them the limitations of having to shoot every other cycle there after. Or maybe you have a target scanner but you don't want them to become a seeker after the only other village role dies. Give them a modifier that makes it so they can't scan after the other role dies, or when there's only one elim remaining. You can use these limitations just as much as you can use roles, so feel free to use them to balance old setups. Just give people a heads up that certain modifiers might be applied to roles. 

Lowering role density. Having a few key roles instead of many general ones will make it so that you don't need more players. Take a smoker, for instance. Let's say you want to design a ruleset around village smokers. Instead of making the elim team have vote manipulation, and the town have a seeker, and the elims have a mistborn, maybe just make it so it's just an elim soother versus 2 town smokers. That'll make the game more controlled, and ultimately come out the way you want it. Now while not having cool roles as often as before may seem wrong, just remember that in the books, magic tends to be scarce (mistborn are very rare in the final empire, why are there 3 mistborn in some random town in tyrian ). Keeping a game simple will make it more viable for lower player counts.

4. I don't really understand what this question asks.

Anyways, for those who may be struggling to understand how I am implementing this, I made a ruleset designed for low player counts (7-10) as an example:

Spoiler

Game: Out of the Cognitive realm

Rules:

48hr cycles. All actions happen simultaneously, and then the execution and subsequent kills. Inactivity filter of 2 cycles.

Alignments:

Odium aligned: standard elim faction. Wins when they reach parity with the non-odium aligned

Non-odium aligned: standard village faction. Wins when all odium aligned are dead.

Roles:

Splinter: Once per game, target a player to make them become a cognitive shadow. You may not target yourself.

• Cognitive shadow: target a player. That player's alignment will be revealed when you die.

Shade: target a player to kill upon your death

Moash: Target a player to make them lose their role. You may only use this 2 times per game, regardless of if the target has an effective role or not.

Anyways, while Drake and I probably have overlapping points, this is what my suggestions are.

[Araris Valerian he/him]

13 minutes ago, Aeoryi said:

Here's another thing: Don't expect games to last long. Nerfing the amount of kill potential or making kills less likely to hit is not the best thing. I think the better course of action is to simply embrace it, or build a ruleset around it. If Exlo/LYLO is at cycle 3 on average, so be it. Obviously, don't make the game be over by cycle 2, since that's lame and unsatisfying, but don't try and make your game last longer by tampering with the eliminator kill or execution. Here's why tampering with the two kills that should always happen is bad:

Say I have a 7 player setup where players have a 50% chance of dodging a kill. What does this accomplish? Well, it makes my game longer, yes, but it also means that people will not be enjoying it when the execution fails for the third time in a row. It means the eliminators will be cussing at me from their doc because they can't kill Kas by cycle 3. It means that people will blame the outcome of the game on RNG, and that's not how the game should be decided. 

Now, you might say, “Well, I would never design a ruleset like that, that sounds terrible!” Well, let me modernize it a bit for you, again using tyrain falls:

Quote

8 player setup: 2x village thug, 1x village lurcher, 2x village vanilla, 1x village smoker 1x elim thug, 1x elim tineye

What's the problem here? Well, if you look closely, the eliminator kill has a default of a 50% chance to fail, and the execution doesn't even kill the eliminators half of the time. In other words, this setup is fundamentally the same as the 50% survival chance setup. So there will be the exact same problems with both. I know the lurcher doesn't really count as protection, but you get the idea. Don't try and make a smaller game larger by extending the duration through protection. Most successful small games work because they aren't a large game optimized for a low player count. They succeed because they use the player count to their advantage.

Drake basically brought up the same point about artificially inflating game length, and I agree with both of you that it's bad. But I like his suggestion about adding extra substance to the game in a way that makes it longer (his example was finding Epic weaknesses before being able to kill them).

1 minute ago, Aeoryi said:

I don't really understand what this question asks.

This question was getting at something you partly addressed in one of your earlier points about using "vials" as flavor for limited charges for a role. Hence Mistborn makes a good setting if you were going to use this mechanic.

I hadn't aimed that question just at mechanics, however. If we have a, say 9-player game, one story to tell that fits would be a skaa thieving crew trying to root out ministry spies while pulling off a heist. But that same story doesn't work as well with 20 or really even 15 people.

Basically, folks can go anywhere on the internet to play mafia. But our mafia is Sanderson flavored, and with some variation we tend to like to lean into the flavor when designing mechanics (in fact, I think plenty of folks go the reverse direction, starting with a cool Sanderson story and getting mechanics to match it). So I think any discussion of how we design our games should keep that in mind.

Thanks for leaving your thoughts!

[Araris Valerian he/him]

I'm going to put this in a separate post since it's a different thought, but I was brainstorming and came up with this outline for a game:

Spoiler

The goal of these rules is to accommodate 8-10 players in ~3 different scenarios. The village needs to discover which scenario they are in to pursue their win condition. Players can choose to explore the 3 different floors of the house to impact the direction the game takes. Day 1 is RP only, with a vote to elect a leader of the group. The effect of the leader changes with each scenario. During Night 1, players explore the house, gaining roles and setting which scenario happens. During Day 2 forward, there is a standard execution, and from Night 2 forward, there is a kill and players can continue to explore the house to find clues about the scenario

 

Scenario weights are Player Count + 2*Explore Count

In a 3 player game, where 2 players explore the ground floor and 1 player explores the basement, the scenarios would be weighted at Standard - (3+4=7), Hopping - (3+2=5), FFA - (3+0=3).

 

Scenario 1: Standard Eliminator game

• Weighted more heavily by players exploring the ground floor during N1
• Elims are more likely to be chosen out of players that explored the basement during N1
• Village roles are more likely to be given to players that explored the upstairs during N1

 

Scenario 2: Free for All with Smoking Gun

• Weighted more heavily by players exploring the upstairs during N1
• The kill is more likely to be given to players that explore the ground floor during N1
• Other abilities are more likely to be given to players that explore the basement during N1

 

Scenario 3: Body-hopping Elim(s)

• Weighed more heavily by players exploring the basement during N1
• Elims are more likely to be chosen from players that explored the upstairs N1
• Village roles are more likely to be given to players that explored the ground floor N1

The idea is to have a semi-blackout game by not telling the village which of the 3 scenarios they are in. And each of the scenarios is supposed to handle information differently, so not knowing the scenario is a significant handicap. It's just a skeleton without concrete rules, but I'm curious if anyone has feedback on the concept. Also I was thinking of this as a LG by default because this is loosely based on the original Awakening game by Wyrmhero, but another format might work better.

Edited April 14, 2024 by Araris Valerian

[Aeoryi she/her]

5 hours ago, Araris Valerian said:

Drake basically brought up the same point about artificially inflating game length, and I agree with both of you that it's bad. But I like his suggestion about adding extra substance to the game in a way that makes it longer (his example was finding Epic weaknesses before being able to kill them).

This question was getting at something you partly addressed in one of your earlier points about using "vials" as flavor for limited charges for a role. Hence Mistborn makes a good setting if you were going to use this mechanic.

I hadn't aimed that question just at mechanics, however. If we have a, say 9-player game, one story to tell that fits would be a skaa thieving crew trying to root out ministry spies while pulling off a heist. But that same story doesn't work as well with 20 or really even 15 people.

Basically, folks can go anywhere on the internet to play mafia. But our mafia is Sanderson flavored, and with some variation we tend to like to lean into the flavor when designing mechanics (in fact, I think plenty of folks go the reverse direction, starting with a cool Sanderson story and getting mechanics to match it). So I think any discussion of how we design our games should keep that in mind.

Thanks for leaving your thoughts!

Ah, okay I thought you meant mechanical settings lol

I think good settings can be anywhere, really. Things like a Reckoners' cell tearing themselves apart trying to find a spy, or as mentioned, a scandrial setting works well. I do think that settings that don't Invision like, the village as an entire village work better. For instance, instead of having tyrian falls village faction be thematically an actual place, maybe just making it a group of explorers or settlers would make it seem appropriate to the setting. And as a huge Nalithis fan, I think places like the court of Gods won't work well as the village faction when compared to something like a group of priests dedicated to a specific returned (the GM! :P). There are definitely some ways to make the game not feel devoid of players by simply adjusting the setting.

 I think the path to take depends on the GM, and ultimately while I do have opinions on which method is better, I think both work just as well for games.

[DrakeMarshall he/him]

16 hours ago, Aeoryi said:

Just like how a 34 player game would not be acceptable

[Araris Valerian he/him]

On 4/14/2024 at 12:31 AM, Aeoryi said:

Just like how a 34 player game would not be acceptable, a game that requires 14+ players is also not acceptable.

I'd forgotten that I've played several games with this many folks. LG12, MR10, LG26, LG28 (this was the House on a Hill game), AG3, LG37, and AG4. Of course, there were plenty of inactive players in these, but aside from the KKC games they were decently well-designed.

[Mat he/him]

MR42 was close to that as well! It didn’t have much inactivity either, if I recall correctly.