Balancing a Game Philosophy

[Jake The Wolfie]

The purpose of this thread is to talk about the philosophy of decisions made before and when balancing a game.

For example: How often should town win against 1 scum faction?
The answer might seem like 50%, however some people (from other sites) have argued that the town should win more that 50% of the time, as if they lose, then more players lose overall.

Some questions of my own:

How often should town win against X scum factions?

How often should a particular neutral (such as a Survivor or Jester) win?

What would technically be considered normal, but is debatable not normal?

[Something_Smart]

For example: How often should town win against 1 scum faction?

50%.

The answer might seem like 50%, however some people (from other sites) have argued that the town should win more that 50% of the time, as if they lose, then more players lose overall.

Sure if the other sites have different metas that's fine. Here people generally want the games to be fair.

How often should town win against X scum factions?

1/X imo, because if you say 50% then the scum players are going to be facing a real uphill battle. But I don't think there's a consensus on this and the type of people who like multiball games generally don't care much about balance, so it's not a huge deal.

ow often should a particular neutral (such as a Survivor or Jester) win?

40%-70%, imo. Has to be hard enough that they felt like they deserved it, but not so hard that the odds were heavily stacked against them. Also, I'm not sure this is a productive question because when designing a third party there are several things that are WAY more important than its chance to win.

What would technically be considered normal, but is debatable not normal?

- Unexpected numbers of scum
- Scum/scum neighborhoods
- Curveballs like surprise ninja or surprise multitasking
- Degenerate shit like bulletproof IC's
- Gotcha roles like disloyal fruit vendor (this one's just my opinion, but come on that's so mean)

But really, "technically normal" is a dumb phrase, because part of being normal is falling within reasonable expectations. That's why we have normal reviewers!

[Isis]

The purpose of this thread is to talk about the philosophy of decisions made before and when balancing a game.

For example: How often should town win against 1 scum faction?
The answer might seem like 50%, however some people (from other sites) have argued that the town should win more that 50% of the time, as if they lose, then more players lose overall.

~42%. I think it's silly to say the answer is 50% for an asymmetric game. Baseball is designed for one side to lose 73% of the time (about 73% of innings are scoreless for the batters and runners, which is the absolute worst they can do). And it'd be a less exciting game if they changed the rules until that came up to 50%.
Scum get lots of interesting lines of play when they can afford throw away game equity, townies typically get less interesting lines of play when they can afford to throw away game equity (like, consensus vigging, and more abstractly regarding the social deduction, play patterns based on constructing a townblock rather than a focus on posing interactions that can lead to incrimination and solves).

How often should town win against X scum factions?

Probably still 42%, but I might not be able to follow the thought exercise as well because of not liking multiball, I could see the answer being stuff like 70% because pitting the 2 scum factions against eachother and deliberately trying to harm the larger faction is fundamentally interesting without the help of difficulty, or something like 20% because a low observed winrate still doesn't create many substates that are hopeless in and of themselves due to the constant possibility of cross shots.

How often should a particular neutral (such as a Survivor or Jester) win?

I think it should have a floor of like a quarter or a third, and winrates of up to even 80 or 85% for special roles like "guess each night who the scum is" is acceptable. It depends on what the 3rd party is doing and how it works and fits into the setup though. A 1-shot BP sane cop survivor has a more fair winrate than a vestless cop survivor but the vestless cop survivor is probably (?? many questionmarks for third party ??) a better designed role.

What would technically be considered normal, but is debatable not normal?

Traitors are part of the Normal queue right now and are generally thought of as a pretty normal thing to include in theme games, but they seem to always upset player expectations somehow even when moderators go to great lengths to make them not upset player expectations, so I've heard people discuss that. As of today they're still normal, and the sample size on traitor setup designs is small enough it seems pretty inconclusive to me, but it sounded like you wanted to hear about something people ponder for normalcy. People ponder that.

[Albert B. Rampage]

what's game equity?

[Isis]

I maybe should have said EV. What I meant was, the scum taking actions that decrease their probability of winning if you assume that from the point after they take the action, everyone starts playing randomly after that point. So, bussing would be something that always tanks your EV, but if you have a good assessment of the game, it can benefit you in the ways the game deviates from EV enough to be worth it (and the less EV the setup makes you sacrifice when making this judgment call, the more realistic, defensible, and interesting it is to consider that judgment call).

An example of a setup that doesn't give scum choices for how they want to choose to discard EV for other gains would be White Flag, where bussing the second-to-last scum simply isn't allowed as an option. (white flag is a fine setup, but it's very illustrative). Setups with enough town power to reliably put town at 50% winrate tend to have narrowing effects similar to this white flag extent, it can feel unsafe to bus at all in normal modern setups because 2 mechguilties in normals is totally possible, and to a lesser extent it can be more difficult to do other things, like commit one of your goons to a deathtunnel on a townie that makes it clear either the goon or the townie is scum, but suffer from a high probability the recipient gets mechcleared town.

[Isis]

I feel more strongly that it should have nothing to do with 50% than I do with my opinion that observed town winrate should be <50% . There's lots of arguments for >50% that make sense. I have trouble buying into "it should be 50%" though.

[Something_Smart]

~42%. I think it's silly to say the answer is 50% for an asymmetric game. Baseball is designed for one side to lose 73% of the time (about 73% of innings are scoreless for the batters and runners, which is the absolute worst they can do). And it'd be a less exciting game if they changed the rules until that came up to 50%.

Huh? Baseball is symmetric by virtue of both sides having the same number of chances up-to-bat. A single half-inning is asymmetric, yes, but that's why both sides get the same number of them.

A better example of a game that isn't fair is chess. But chess has (frequent) draws, so there's something for the disadvantaged side to play for besides just hoping their opponent screws up, and a balanced version of chess will never catch on because the game has such a long history.

But all else being equal, I would think that people want the game to be fair? Maybe that isn't the case, but, like, fairness (or getting as close to fair as you can) is a standard game design principle and the burden of proof is on the person saying it shouldn't apply.

If the setup is designed to constrain scum's options significantly (for instance penalizing bussing by adding a lot of investigative power), then that's a different issue entirely.

[PookyTheMagicalBear]

Every Mafia game is like a story - the actual percentage does not matter in so much as every player should feel that they can win, and possess the agency to influence the outcome towards their victory. The most exciting games are ones where there is a real chance for either side to win and the stress and tension mounts towards an exciting conclusion. Not every game will reach lylo or even a game state where both sides feel they have a chance to win, but the most memorable ones tend to be a real nail biter to the end.


I dislike multi-ball games in general because it feels like doing the right thing could end up setting you back rather than moving you forward - they are notoriously difficult to balance from a fun perspective and engage players in terms of agency. By increasing the number of sides you reduce each player's agency in the sense that their decisions matter less overall in terms of setting the game state. I think 3P and individual kill roles such as SK are more fun than having multiple teams of killers as it is more of an individual challenge.


I think the win percentage of a 3P should be balanced so it does not materially affect the win percentage of the main parties in the game - aka mafia/town. I think of 3P wins as something special and incredibly difficult - a memorable experience because they are not

expected

to win so when it does happen it should be through tremendous adversity.

[Something_Smart]

Every Mafia game is like a story - the actual percentage does not matter in so much as every player should feel that they can win, and possess the agency to influence the outcome towards their victory. The most exciting games are ones where there is a real chance for either side to win and the stress and tension mounts towards an exciting conclusion. Not every game will reach lylo or even a game state where both sides feel they have a chance to win, but the most memorable ones tend to be a real nail biter to the end.

This is very true. Actual chance to win is important-- but perceived chance to win is even more important.

(Of course, they're related, so you should still aim to balance actual chance to win because that will maximize the likelihood of getting a good perceived chance to win.)

[Isis]

~42%. I think it's silly to say the answer is 50% for an asymmetric game. Baseball is designed for one side to lose 73% of the time (about 73% of innings are scoreless for the batters and runners, which is the absolute worst they can do). And it'd be a less exciting game if they changed the rules until that came up to 50%.

Huh? Baseball is symmetric by virtue of both sides having the same number of chances up-to-bat. A single half-inning is asymmetric, yes, but that's why both sides get the same number of them.

A better example of a game that isn't fair is chess. But chess has (frequent) draws, so there's something for the disadvantaged side to play for besides just hoping their opponent screws up, and a balanced version of chess will never catch on because the game has such a long history.

But all else being equal, I would think that people want the game to be fair? Maybe that isn't the case, but, like, fairness (or getting as close to fair as you can) is a standard game design principle and the burden of proof is on the person saying it shouldn't apply.

If the setup is designed to constrain scum's options significantly (for instance penalizing bussing by adding a lot of investigative power), then that's a different issue entirely.

I'm fine being told I have a burden of proof for a deviation from 50% and I haven't satisfied someone that I've shouldered it. I'm only bothered if someone said a proof couldn't even be considered and they're basing that on a gazillion symmetric games.

[Menalque]

How often should town win against X scum factions?

As often as they’re able to play better than scum.

[TheGoldenParadox]

How often should town win against X scum factions?

As often as they’re able to play better than scum.

i mean, just intuitively, it seems really hard for town to "play better" than scum, even as much as that is possible to quantify in an asymmetric game. scum have the advantage of information, buddies that they can consult, knowing exactly who is on their side. town don't have that, and so saying town has to "play better" than scum to win is a fundamentally flawed requirement that seems like it would make scum wins far easier than town wins.

[Jake The Wolfie]

I think that, between two opposing forces, whether they be town v scum, or two neutrals in a 1 on 1, each should have an equally probable chance of winning.

[Infinity 324]

Scum get lots of interesting lines of play when they can afford throw away game equity, townies typically get less interesting lines of play when they can afford to throw away game equity (like, consensus vigging, and more abstractly regarding the social deduction, play patterns based on constructing a townblock rather than a focus on posing interactions that can lead to incrimination and solves)

Huh.

How much do you think this applies for mountainous setups? If they’re more townsided (close to 50/50), town does get to just PoE the hell out of the game sometimes, but you avoid the “scum got guiltied twice lol” games which are a lot worse. The other thing is that in more scumsided setups you have games where town is totally lost and never finds scum, which isn’t a much better play pattern imo.

[Menalque]

and so saying town has to "play better" than scum to win is a fundamentally flawed requirement that seems like it would make scum wins far easier than town wins.

Yep. Towns need to suck it up and get better

[Isis]

A 4:3 mountainous eLo has 6 viable options for every player in the game, which is lots of agency. A 6:1 day phase with 3 clears has just one thing for the remaining scum to be aiming for and roughly half as many choices for the town as well. So it seems obvious to me that that first play pattern is much better.

I guess it's possible all the best games are Nightless Vanillas with the playercounts adjusted to reach 50% observed winrate, though. Because the sample "town is winning" "scum is winning" endgames for that universe results in a lot more agency for both sides.

[Alisae]

and so saying town has to "play better" than scum to win is a fundamentally flawed requirement that seems like it would make scum wins far easier than town wins.

Yep. Towns need to suck it up and get better

This is a really unfair take lmao.
Not every game that is designed is played out the way it was intended to be. In a lot of the games I’ve ran, a really important role has always died first.

Also if towns were able to just get better, mountainous wouldn’t be scumsided.
A core part of being town is that you just don’t have information. When you don’t have information, its easy to be wrong. Just saying get better invalidates the fact that this game is really hard.
Towns will suck. Sometimes they’ll be good but the cosmic dice may decide that a town will suck. It is what it is.

[Alisae]

It’s also really weird that you think towns suck and I think mafia teams suck because most players want to play town and not mafia.

Its really weird to say that most people do not want to kill people in the game. Hell, this is actually the appeal to PLAYING mafia in more casual games like ToS. A newbie will naturally gravitate to wanting to roll mafia because you get to kill people.

Like people get excited about being able to kill people in a game. But in this game, or maybe rather on this website, people do not get excited about killing other players. I mean some do but most would rather just be town.

I’m now writing this and realizing we’re talking about something completely different holy moly

[Alisae]

and so saying town has to "play better" than scum to win is a fundamentally flawed requirement that seems like it would make scum wins far easier than town wins.

Yep. Towns need to suck it up and get better

tbh most scum players who win do get a good roll from the cosmic dice.

And to an extent, town should win if they played better then scum and punished their mistakes. They can also get extremely lucky with actions which can happen but generally most games town wins they played better then scum. And the games that scum should be winning are games where they played just a little less badly then town. Sometimes scum can outplay town significantly, but also sometimes they could just get extremely lucky with actions. It happens.

[Alisae]

I’ll add something to contribute to the thread as well ig as the head of the church of lolbalance. I always design close except when I decide to design an open setup that I never finished.
I never design multiball games because I hate multiball and no one deserves to play multiball. The only reason why I would run a multiball game is to sadistically torture my players and I know exactly just how to do that.
In 2017/2018, as the head of the church of lolbalance, I would say that I have an extreme hate for third parties. Truth is, I actually have a guilty pleasure for serial killers. It’s mostly just a fantasy but generally speaking, I won’t really put any 3rd party in game except I am memeing or if the game was designed around the third party. You can’t splash in a third party into your game just because you feel like it. Actually Third Parties are a whole different topic.
I will always hate third parties that are done poorly.

My mindset has always just been to come up with a few roles that are like “WoW tHiS sOuNdS lIkE a CoOl RoLe” and put those in first, since generally I always have some idea of what I want to actually put in my game.

Now generally for me this is step 1 and I make sure that the shit thats in it isn’t too broken/unfair to play against/unfun to play as. I don’t really proceed unless I’m trying to make a game that takes itself seriously. For games that are meant to be taken a bit more seriously, step 1 would be to design the core, step 2 would be to design things that interact with the core, and step 3 would be to trim out the bullshit and make it fun.

[Something_Smart]

Yep. Towns need to suck it up and get better

This is gatekeeping. We're balancing for enjoyment. You're saying that people only deserve to enjoy the game if they reach a certain level of skill.

[Jake The Wolfie]

How often should town win against X scum factions?

As often as they’re able to play better than scum.

That's answering "how often do town win against scum?"

[Menalque]

I guess it's possible all the best games are Nightless Vanillas

[Menalque]

and so saying town has to "play better" than scum to win is a fundamentally flawed requirement that seems like it would make scum wins far easier than town wins.

Yep. Towns need to suck it up and get better

This is a really unfair take lmao.
Not every game that is designed is played out the way it was intended to be. In a lot of the games I’ve ran, a really important role has always died first.

Also if towns were able to just get better, mountainous wouldn’t be scumsided.
A core part of being town is that you just don’t have information. When you don’t have information, its easy to be wrong. Just saying get better invalidates the fact that this game is really hard.
Towns will suck. Sometimes they’ll be good but the cosmic dice may decide that a town will suck. It is what it is.

and so saying town has to "play better" than scum to win is a fundamentally flawed requirement that seems like it would make scum wins far easier than town wins.

Yep. Towns need to suck it up and get better

tbh most scum players who win do get a good roll from the cosmic dice.

And to an extent, town should win if they played better then scum and punished their mistakes. They can also get extremely lucky with actions which can happen but generally most games town wins they played better then scum. And the games that scum should be winning are games where they played just a little less badly then town. Sometimes scum can outplay town significantly, but also sometimes they could just get extremely lucky with actions. It happens.

In my experience at least, scum wins are often in spite of the cosmic dice rather than because of them. I've seen/had games that felt like everything was an incredibly uphill struggle against setups that were just deeply setup to favour town. Correspondingly, I've had town games where I've had to do basically nothing to win because we had one or two PRs who got lucky and cleared out half the team.

Don't disagree about the game being really hard, but I also don't think many people make much of an effort to actually improve at the game, partly because they don't have to because the glaring weaknesses in their play are buttressed by town power.

If I'm town, I'd rather lose to scum if I couldn't catch any of them on dayplay than win because of TPRs.

Yep. Towns need to suck it up and get better

This is gatekeeping. We're balancing for enjoyment. You're saying that people only deserve to enjoy the game if they reach a certain level of skill.

The question was about how often town

should

win, that's a moral question. I personally don't find games balanced for enjoyment because I don't enjoy empty wins and I don't enjoy spending almost every scumgame fighting uphill.

Just achieving so that people win 50% of the time isn't necessarily balancing for enjoyment either.

How often should town win against X scum factions?

As often as they’re able to play better than scum.

That's answering "how often do town win against scum?"

No, it's not, although they're similar. Your question here doesn't account for times when town win despite playing worse than scum.

[Ircher]

But the question is, in those games, did town /really/ play worse **/GIVEN/** their information disadvantage? I'm sure it's true in some games, but what about for the majority of games town wins? I think that's a really subjective call, and the better you are at the game, the more likely it is that you will feel town did a poor job. That's simply due to a mismatch in skill level.