What do you consider 'Normal'?

[Nexus]

If you sign up for a Normal game - what are you expecting? Are you expecting a 50/50 chance of winning? Are you expecting to see the usual roles (Cop/Doc/Vig/JK/Tracker etc.)? Day talk?

Help us to shape the Normal Queue and make it better.

Make the Normal Queue Great Again

MtNQGA

[RadiantCowbells]

To me the point of playing a 'normal game' is to be playing a game that:

1) Is balanced or close to it. I've played some really dumb, really imbalanced theme games and even the more balanced theme games are often swingy and have very questionable edge cases. Normals should contain none of the above.
2) Focused on day game play. Don't want to see role madness or games principally decided based on night actions. Decent number of vanilla townies.
3) Night actions and PRs are mafia classic and predictable. Cop innos are innos, trackers/watchers know their actual results, relatively simple set of roles that someone can read through, no roles that massively tamper with the game like bus drivers and redirectors, etc. For the record I am fine on principle with stuff like Novice and Loyal being added (and loyal ought to have been added a very long time ago.)
4) Reviewed in a comprehensive manner. The standard that the NRG holds itself to while passing a game is far cleaner than most theme games are held.

So it's a mix of quality control, wanting to play a game without a ton of frills, wanting a specifically day-game focused setup, and wanting it cleanly balanced.

As a counterpoint to Ether I think the tendency to just straight up NK the skilled/experienced players is very unhealthy for the site meta and I think that it should be an expectation that you be punished for making an easily predictable nightkill. I'm sick of seeing every LyLo contain the least invested players who rarely have much inclination to analyze the rest of the game. I've seen so many mechanically confirmed scum win LyLo situations off of the fact that everyone who would reread the game to figure it out being dead. As long as it's easy for scum to get rid of the people actually playing the game town winrates will never be good unless setups are just broken. Many of the 'experienced' players read pre-2011ish I've talked to say that they don't try early on in games because they don't want to be nightkilled; that being a site meta is extraordinarily unhealthy for the game.

[mastina]

The ideal Normal game:
-Has balance where each side at the beginning of the game had an equal chance of winning
-Has a setup where REALISTICALLY, rather than just technically, the above is true
-Has a low swing factor such that the above two can continue to be true even if a key town/scum role is eliminated early
-Has about half of the town (at minimum) as VTs (is not super-role-heavy) and is light on scum power
-Still produces something which is interesting, unique, and maybe even quirky, while following all our established rules.

While not every Normal game will have all of the above factors, the more the better.

[Ether]

It isn't the NRG's job to moralize about how well towns should be doing. Striving for 50/50 balance is really the only objective thing we can do. That doesn't mean we'll get there, but at least we'll get closer than if we throw up our hands and say, "Fuck it, a crappy town doesn't

deserve

to have cops."

Normal games should basically give you the experience you're expecting to have. Major curveballs and gimmicky setups should really just go to the theme queues. (I really wish we wouldn't lionize normal setups just for being unique--the normal queue isn't really the place for unique,

and that's okay

.) Normal games shouldn't require people to be good at setup speculation to do well, or let roles--either people's actions, or just speculation over who has what--be more important than how people are playing. Mafia should be able to safely claim vanilla town without being swarmed.

An individual normal game shouldn't be a meta-shifting statement unless it can be done in a way that doesn't screw over either side. (I'm looking at BBmolla's game with a mafia vig here.) People don't necessarily have to be rewarded for what is normally seen as optimal play, but they shouldn't be overtly punished. (This is also the main reason I dislike watchers so much--I don't think they're overpowered, but punishing scum for making obvious kills bothers me in a way that just making the kill fail wouldn't. If mafia know that there's a watcher in the game, then I wouldn't really care, but unfortunately the "Informed" modifier isn't currently whitelisted.)

Swing shouldn't be too major. Most players should be vanilla, but they should still get to feel like they had influence over the game.

Normal games evolved a certain way on Mafiascum, but some unrelated site could have developed with totally different meta expectations. Our current conventions aren't gospel; it's just important to

have

conventions, and these are what we've got. Having a "redlist" with roles that can only be used if they're announced in signups to be in the setup would be cool--my problem with roles like godfathers and tailors isn't that they give cops bad results, but that town has no way of preparing for them. Things like the redlist and more leeway for semiopens* would give normal games more design space while still letting players know what they're getting into, which is ideal.

*The kind with a narrow publicly-given list of roles, not the crazy Xylbot kind.

[Firebringer]

I agree with pretty much everything Ether just said.

[FakeGod]

If you sign up for a Normal game - what are you expecting?

A cheesy "unique" setup run by a new mod who rather be modding something else new and exciting, but alas, this queue is all he has, so he toed the line as best he can while trying to sneak as much "fun" across under the noses of draconian

TSA

NRG workers.

[Creature]

I expect games can go smoothly without players saying like "oh these clears aren't really clear, there could have a godfather".

[Kmd4390]

I don't even know anymore. It's always changing.

[mastina]

(I really wish we wouldn't lionize normal setups just for being unique--the normal queue isn't really the place for unique,

and that's okay

.)

Counterpoint: the normal queue is not the open queue or even the semi-open queue, and having unique setups is one of the main ways we tend to keep it from being that way.
It is possible to run interesting games that are Normal.
It is possible to run games that have unusual roles in a Normal.
It is possible to run games with a role platter that's not often seen in a Normal.
And all of these things give Normal Games versatility and variety--things which should be encouraged.

Normal doesn't mean monotonous to me--at least it shouldn't.
While there are standards I feel Normals should be held to, and there are RULES that Normals must follow (the two having huge overlap incidentally), two normal games can and SHOULD feel different from one another, and that is something which we should not only allow but also encourage, within reason.

I mean. If you're running a mini normal where 8/10 of the town are power roles I'm gonna tell you as a reviewer "fuck no, that's not a normal game", and if you have a single role which is a "daycop nightvig", I'm gonna shoot it down as being beyond what is reasonable to consider being a normal role. But that's why I say "within reason". The worst which can happen with an unusual idea is your review team shooting your idea down and explaining to you why they are shooting your idea down. (...Okay, so technically the worst which could happen is an absolutely terrible idea being passed through the reviewers, but that's why we have three of them, AND the listmod, and between all four, GENERALLY SPEAKING, we're pretty good at stopping that from happening.)

Normal games shouldn't require people to be good at setup speculation to do well

Speaking from experience from actually reading the normal games we have had recently: 90% of Normal Game players are absolutely shitty at doing setup spec. They either think the standard/reasonable level of town power is "OMG town so OP!!!", or they think the standard/reasonable level of town power is "this way too weak!".

But I don't think this is something reviewers can really do much about, aside from educating the players postgame on what actual standard/reasonable level of town power is and why it is standard/reasonable; players are going to do setup speculation no matter what, and that's not something going away any time soon. They do it, often wrong, sometimes backfiring, other times winning in spite of being wrong. Setup speculation is part of the game of mafia--on any site which doesn't use an open, it's going to happen. It happens in semi-opens. It happens in normals. It happens in themes. It's a fundamental part of the game.

Mafia should be able to safely claim vanilla town without being swarmed.

Thus my stance that at minimum half the town should be VTs and even then that's uncomfortably high in town power levels. (Five PRs is a little bit high in a mini. Ideal level is 3-4.) The actually larger problem is the inverse: mafia typically have trouble CLAIMING a PR and having it be believable, specifically thanks to town power already being too high. (That being said, see above--about half of towns are shitty enough at setup spec that the scum get away with it anyway, so. Not really a huge problem, though it is *a* problem. Whether it's one we want solved is another matter since there's the question of whether mafia SHOULD be given that ease of claim and that's a hornet's nest I'd rather not disturb.)

[callforjudgement]

I agree that towns are generally terrible at setup speculation. This can make games rather hard to balance, as a setup that's balanced if the game has nobody who can setup-speculate can become very townsided if the game does have someone who's good at speculation (and they're believed by the rest of town). The only way in which I disagree with mastina is that she's assuming that setup speculation doesn't happen in Opens; it does, and it's often worse than in Normals, because players have the setup in front of them and thus feel that they have more information with which to draw conclusions. Some of my very simplest Opens have inspired pages of discussion about what optimal play in the setup is.

I would expect the number of town PRs in a Normal to be in the 3-5 range. (The number of scum PRs is less important; in most cases, scum PRs affect the power of town power roles, rather than helping out scum directly.)

Redlisting is something that doesn't currently exist in normals (except for daytalk), but which I feel would not be out of the spirit of what I'd expect from a Normal. The games don't all

have

to be interchangeable (although I wouldn't mind if they were!); different mods will have different rules for things like deadlines, and a different general style in setup creation.

And to answer the original question: I expect a Normal to be a game that isn't horribly broken or unfair in any direction, which has a setup that can be immediately understood from the signup post + a list of roles for each player, and that won't try to do anything that fundamentally changes the way the game plays. It's a place of safety from redirectors, fake dayvigs, and having to deal with potentially stupid theories like "maybe all the kills are coming from an SK and all the groupscum look innocent on investigatives". Or to summarize even further: it's a type of game in which you can reasonably expect that player skill will have a large effect on the outcome, as opposed to things like correctly guessing what the mod's idea was, or luck in who happens to draw each role.

[mastina]

The only way in which I disagree with mastina is that she's assuming that setup speculation doesn't happen in Opens; it does, and it's often worse than in Normals, because players have the setup in front of them and thus feel that they have more information with which to draw conclusions. Some of my very simplest Opens have inspired pages of discussion about what optimal play in the setup is.

Setup speculation is part of the game of mafia--on any site which doesn't use an open, it's going to happen.

It happens in semi-opens

. It happens in normals. It happens in themes. It's a fundamental part of the game.

I mean I guess it could happen even in a game where you literally know every role is 100% in the game as a full-open if it's a game which involves scum lying/counterclaiming?
I wouldn't really consider "who is the godfather, and who is the goon" in a vengeful game to really count as setup spec though.

[callforjudgement]

I've seen pages worth of setup spec in setups as simple as Vote for Town Mafia or Desperation Day. People arguing about whether it's better to lynch very scummy or mildly scummy players first, for example (that's in Desperation Day; in Vote for Town, the same issue happens with townies).

[mastina]

I don't consider talk of optimal usage of mechanics to be setup spec, so much as I do mechanics utilization. (Well I guess we don't really have an established term for mechanics exploitation, but it's not what I think of when I think of the term setup spec; I think of setup spec as being speculating about the exact composition of roles in the game--

maybe

potentially extending that to who holds said roles, but not necessarily.)

[Nero Cain]

town being bad is not a reason to stack town power.

[mhsmith0]

If you sign up for a Normal game - what are you expecting?

A cheesy "unique" setup run by a new mod who rather be modding something else new and exciting, but alas, this queue is all he has, so he toed the line as best he can while trying to sneak as much "fun" across under the noses of draconian

TSA

NRG workers.

As someone in NRG who's actually been a reviewer in a bunch of games already, I agree with this statement

I would expect the number of town PRs in a Normal to be in the 3-5 range. (The number of scum PRs is less important; in most cases, scum PRs affect the power of town power roles, rather than helping out scum directly.)

The bulk majority of mini normal games that are 10v3 have 3-5 town PRs. 2 is too few, 6 would require something very odd in order to work

The number of scum PRs is utterly unimportant on its own; what matters is balancing town power against scum power. You can pretty easily balance 3 goon, 2 goon 1 encryptor, or 2 goon one sort of OK scum PR setup, and you can also balance scum PR-heavy setups (though it's harder!).

[mastina]

town being bad is not a reason to stack town power.

And we don't, which is why the optimum town power is ~25% of the town playerlist altogether. (Does not scale perfectly, but it's about right. Think, number of scum in the game, then maybe slightly more.)

That's the ideal level of power in a game (3-5 roles in a mini), and it's the fairly typical standard.

If you think that's too much, play a few mountainous games where the town has literally no PRs at all and tell me what your opinion is after having played a few. (Needless to say there's a damn good reason we have power roles in the first place.) Oh and don't forget the obligatory CLOSED mountainous game where the game has no power roles but people don't KNOW the game has none. Have fun with that.

[callforjudgement]

There is actually something of a problem with balancing, though, which is about whether you try to balance for the players being competent. The best example is probably Vigilante; it's comparable to (slightly weaker than) a Cop in theory, but it tends to be misplayed so badly in practice that it ends up being borderline negative utility. So how do you balance a setup with a Vigilante in, given that it's likely impossible to account for both the case where it's played correctly and the case where it's played terribly?

[mhsmith0]

wrt vig I'd say

1) You don't have any other super-powerful town roles (so vig/cop is probably bad), so that if the vig is REALLY bad there's limit to the damage that can be done
2) You presume it's played with a baseline of at least semi-competence; it's a powerful role that is GENERALLY town-positive barring substantial evidence to the contrary (I'd need data that actually suggests vig shot rates are utterly atrocious to believe otherwise), and balance with that in mind.

FWIW, in general I try to balance with the idea that players are not completely terrible, and/or just use a data-based approach of what success rates reasonably similar setups have had in the past.

[Creature]

Vigilante's only problem are town PRs.

Otherwise, they work as an extra lynch.

[mastina]

There is actually something of a problem with balancing, though, which is about whether you try to balance for the players being competent.

I've brought this up in the past, as the two separate definitions of balance:
Do you balance off of what the THEORETICAL NUMBERS say a setup is...
...Or do you balance off of what pragmatic play actually suggests?

Ideally a setup has both (and it's not impossible to be both), but most setups are going to strongly have one be dominant: "This was a balanced game in theory, but realistically, this game was never being won by *alignment* short of a fluke." (Usually said alignment being town.)
"This game might be balanced given the results, but theoretically, it really favored *alignment* and it really took the game's results for it to be anything other than that". (Usually said alignment being town.)

A game with neither of the two problems is the optimal, perfect ideal of balance, but 95% of setups have a bias towards one of the two, and a problem we have is that neither one of these is explicitly "right" or explicitly "wrong", and reviewers tend to not strongly favor one over the other overall. Both appear about equally, with a near-even 50/50 split.

[RadiantCowbells]

There is actually something of a problem with balancing, though, which is about whether you try to balance for the players being competent. The best example is probably Vigilante; it's comparable to (slightly weaker than) a Cop in theory, but it tends to be misplayed so badly in practice that it ends up being borderline negative utility. So how do you balance a setup with a Vigilante in, given that it's likely impossible to account for both the case where it's played correctly and the case where it's played terribly?

I /did the math/ and vigilantes are actually far more likely to hit scum than lynches.

[chamber]

I think its more important to say what isn't normal than what is. A focus on what is is very self reinforcing.

Alignment changing isn't normal.
Non-sane cops aren't normal.
any explicit lying isn't normal.
too high of a power role density isn't normal.
and probably (though this one may just be me) night action redirection isn't normal.
Non mafia themes aren't normal.
edit: Hydras too, though I personally don't mind them I know many do


I admittedly haven't played the game in a while and am unlikely to, so in many respect my opinion should be weighed less, But I feel like you could black list those things and remove the white/grey list weirdness entirely and end up with a functional queue that produced normal games.

[kuribo]

mafia typically have trouble CLAIMING a PR and having it be believable,

Those players are weak and will be the first against the wall when the bus comes

[Randomnamechange]

I think its more important to say what isn't normal than what is. A focus on what is is very self reinforcing.

Alignment changing isn't normal.
Non-sane cops aren't normal.
any explicit lying isn't normal.
too high of a power role density isn't normal.
and probably (though this one may just be me) night action redirection isn't normal.
Non mafia themes aren't normal.
edit: Hydras too, though I personally don't mind them I know many do


I admittedly haven't played the game in a while and am unlikely to, so in many respect my opinion should be weighed less, But I feel like you could black list those things and remove the white/grey list weirdness entirely and end up with a functional queue that produced normal games.

important to remember the distinction between flavour and theme (just a general comment). flavour is fine as long as it doesn't affect mechanics

[ChrisOrmie]

Normal should be a fairly standard setup (like Newbie games) imo. We need balanced roles, and a setup (or set of setups) that everyone knows going into to it. This then pits "player skill vs player skill" instead of a role you didn't expect giving one side or the other an advantage. We have other game types for that kind of thing, so let's get normal back to a "default" or "standard" game that provides little in the way of surprises.

Let these games simply be people trying to read each other and using strategy, not crazy abilities and setups which can break the game.