Gauging Interest: Opt In Rankings League

[Jingle]

I've been talking for a few months with individuals (who I won't name unless they want me to, mostly cause I'm not sure I'd be able to remember all of them) and in various threads about a series of games on site where we attempt to objectively rank people's scumplay and townplay.

There are a few barriers to this that would need to be overcome, but I'm willing and able to commit to reading a large number of games I don't play in over a large period of time and keeping all of the data which should solve most of it.

Now clearly this would have to be an opt in service with agreement from whatever game moderators are involved, but before looking at logistics/how to run it, I want to get both a sense of who is interested, to what extent, and open this up to any suggestions as to how it should be run.

Of course, if I do end up going forward with this I'll run it by zor to make sure I'm not stepping on any toes or anything, but I feel like this might be a cool thing for the community as an outlet for people to get their competition/gloating in in an environment specifically designed for that behavior.

The original plan is vaguely to have members of the league submit reads lists for all games they play in the league every night, so that read accuracy as town and ability to be townread (as both) can be tracked long term. Clearly, I would need to not be involved in any games used for this purpose, and I'm fine with that. Also, mods would have to be willing to participate.

Any thoughts? Any interest?

[Mulch]

Bad idea

-Hurts site culturally
- competitive, judgemental
-hurts side gameplay
-something else people are thinking about besides “how can I win” . Discourages changing your mind when presented with new info (huge reason mafiascum towns do badly). Among others problems, some severe
- last thing we need is people getting even more obsessed about their reads in game
Can go into more detail later

[Mathdino]

Weirdly I think mulch hit the nail on the head

This would make things more toxic not less

[Jingle]

I'd argue that competitive doesn't necessarily mean bad, and that this is by necessity an opt in thing to cut down on judgement. I agree with you that making this into a sitewide ranking system would just piss off people who don't want to take part, but offering it for people whose primary enjoyment IS the competitive side of the game might help funnel ego clashes into a forum where they can be more healthily handled. And I think there is a significant portion of the site who would be interested.

Putting aside the cultural impact for the moment, because frankly I'm unsure what the cultural impact of the idea would be and am unsure how anyone could be before it is even tried, I think you're largely misunderstanding the idea.

The only concern in the games WOULD be how can I win. The only thing players would be responsible for in their games is sending a list of reads each night. That's it, no more no less. This, as far as I can tell would actually be a good thing, because reevaluating and changing your mind would presumably make your reads better, rather than worse.

I'm not sure what you mean about it hurting side gameplay, but winning games being your goal should mean that you work with your team to do better. At least, if I'm understanding what you're saying correctly.

As far as people being obsessed with their own reads goes, I actually expect that this might make people more willing to work together when they see how their reads are over time.

Basically, my vision for this would be a service that is as unobtrusive as possible, letting people keep track of how they do overall in games and compare themselves to others, IF AND ONLY IF they actively want to take part. Hell, given a small enough interest pool (or someone interested in coding this into a program with no upkeep), it would probably be feasible to offer without making results public if someone wanted to take part without their results being made public, but that would be considerably more work.

I look forward to more detail though, because I'm sure there ARE issues, and I'm not likely to catch them all myself. Hence, this thread.

[Flubbernugget]

I'm in 1000% support of this over a competitive queue

[Flubbernugget]

The only issue I can think of is "my numbers look nicer so I'm more right than you" type arguments, but I'm not sure if that's a big issue, or something we already deal with in a different form.

[Flubbernugget]

Actually, "your numbers look good but scum didn't kill you" could be a pretty big issue.

[Ranmaru]

Wouldn't it be the same as 'You have paragon but scum didn't kill you'? I wouldn't mind this system. Although I probably wouldn't enter any time soon.

[Flubbernugget]

Paragon is a lot blurrier than a hard number

[Nero Cain]

Haven't read this in detail but my general stance is that "rankings" will only lead to elitism although one good thing is that players might actually try.

[BBmolla]

Sounds needlessly tedious

[Espeonage]

Mafia does not suit a ranking system.

[Gamma Emerald]

I'd rather see a flexible scoring system. Closest I've seen is this system mastina has involving lynches she's been or not been a part of that have been on scum, but that's town only
maybe something like
you start with a score of 1
getting lynched as either alignment loses you points
getting lynched in lylo or lynching wrong in lylo loses you a fair bit of points
failing to kill PRs as scum loses you points
being on the wagon that forces a PR to claim as town loses you points
stuff like that

[Jingle]

It occurred to me that maybe including the impetus behind this might help people to understand the goal.

There's two big pieces. Reading games and threads recently I've seen a lot of arguments along the lines of *player's* reads are good we should follow them. Math's thread about self awareness is a good example that I agree with. I wonder exactly how accurate these arguments of Player has accurate reads really are, whether they have real basis in numerical evidence or are simply perception, etc. This is the numbers bit.

The other half of my motivation is largely based on my personal observation of players in games I've run. I always allow players to PM me reads and talk out their thought processes. I have roughly one or two people who do so per game I run. On the whole, the players who do so tend to be more likely to stick with the game. Very rarely do they replace, need prods, etc. Of course, this is a crazy small sample size, but I think having the ability to send a reads list each phase to someone who won't talk about the game or respond but can be assumed to be reading along and interested might help people stay invested.

Of course, I might be entirely wrong.



As far as mechanically how it would work, my thoughts have kind of led me to this point:

Submit a reads list each phase.
As scum- looking for PRs. I'd track the percentage of accurately determined PRs, possibly weighted by number of PRs in the game, strength of the PR, etc.
As town- I think {list} are town. I think {list} are scum. Tracking accuracy on each numbered day and overall.

Track wins vs. losses.
Pretty self explanatory.

Track scumlynches you're a part of vs. are not a part of as town.
Track townlynches you're a part of vs. are not a part of as town.

Things along those lines. The goal would be to allow the players to compare themselves to others, see where they're improving, where they might need to work on their skills, who they might want to look at as good examples to follow, etc. Not to be able to point to a player and say "That guy is the best at mafia." Because let's be honest here. There's no "that guy" who is the best at mafia. There are strengths and weaknesses. Give me and creature both VT PMs and Creature is far more likely to help town than I am. Give me and Creature Goon PMs and I'm far more likely to help scum (Not to pick on Creature, but I feel that's fair to say given everything I've heard about the guy). That doesn't mean he's better than I am or that I'm better than him, but that we have different areas in which we excel.

I think trying to synthesize it into one rank is a fool's errand, but being able to see where you stand against the community is the first step to strengthening your skills. And I believe it can also be the first step to strengthening the community.

Given the number of times I've heard "MS towns suck" since coming back, I think that's a worthwhile use of my time.

And now, individual responses.

Flubbs: I agree that people basing reads off of the system is a real concern, but policy lynching is already something that people try (and usually fail at, in my experience) so I don't know how much of an issue it would actually be. BoP is real, but I doubt there'd be any more of it than there already is, especially with people like Elli who have been hyped up into gamebreaking boogeymen.*

Nero: I honestly don't think there could be any more elitism than what already exists. In one way or another, there have been gigantic egos clashing on site since before I joined. Will this give those egos a way to continue fighting? Sure, if they really want it to. But I think the possible benefits outweigh that risk. And I don't think you'll ever find a mafia community where there aren't people with gigantic egos who get into pissing matches over who is better. We're a collection of people who like arguing about who to kill for fun.

BB: Yup. I agree that getting this running will be tedious. That's why I'm taking the time to make sure this is going to be something enough people want AND that it will actually benefit the community instead of going out and counting the number of people who say they've caught scum D1. But I'm willing to put in the work and I think this will improve the community. I enjoy reading games, listening to why people think the way they do, and watching things unfold from the background. And if I can help the site while doing so, I want to.

Espy: I am 100% with you on that. Mafia doesn't translate to good/bad or better/worse. It's way more nuanced than that. But there are things (accuracy of reads, etc.) which can be objectively measured.

GE: I think it's a mistake to try to synthesize this using points. Like... at that point you have to make the subjective call of "Is it more important to have accurate reads or get your reads to lynch? Is it more important to not get lynched or to get scum lynched?" I think keeping everything separate and just letting people look at how individual aspects of their games stack up is both more valuable and more healthy in the long run. Oh, I have good reads but trouble communicating them? That's something I should work on. Or Oh, I tend to be lynched early? That's something I should work on. That's more the goal rather than to be able to point at one person and say "Yup, best player on the site there." With that said, I'd definitely love to see more suggestions about what would or wouldn't be worth tracking.


*Citation needed.

[profii]

people will 'reset' when their rank gets bad and this site has far too many alts as it is

[brassherald]

I hate playing Ranked modes in FPSes and MOBAs, why would mafia be any different?

[Kagami]

I had been thinking of creating something similarly named, but with a wildly different implementation.

My plan for MafiaLeague was this, with some details still up in the air:

Players form teams. Teams must have 3 distinct players, but there's no upper limit on size. A player can only be on one team.

Whenever a league player /in's for a game (and prior to receiving a role PM), they can PM a league organizer for the game to count toward their league score. Any non-bastard game on mafiascum is eligible. Moderator permission is not required. If for some reason a moderator explicitly forbids that their game can't be evaluated as a league game, then the request will be respected. Hydras count too, provided at least one head belongs to a team and no two heads are on different teams.

At the end of each league game, a small team will do a post-mortem balance review, and award points to participating league players according to the players' alignments and balance judgement.

At the end of each league season, each team that has met requirements (some minimum number of games, and some minimum number of meaningfully contributing team members) will be ranked by average points per game. Might also be separate town vs scum rankings. Winners get absolutely nothing.

Complete reset each season. Seasons probably last a year or a half-year (summer vs winter).

[Gamma Emerald]

I'd prefer full year to half tear cos activity could be different between summer and winter
honestly Kagami's idea sounds fun

[zoraster]

The fundamental problem with a mafia ranking system is repetition. Even our most active users don't play enough games to give a very good idea of their skill from a purely win/loss standpoint.

[Kagami]

That's why it's a league! With lots of players on a team, playing lots of games.

And it shouldn't really be about a srs biz skill judgment as much as a friendly competition.

[StefanB]

I'd rather see a flexible scoring system. Closest I've seen is this system mastina has involving lynches she's been or not been a part of that have been on scum, but that's town only
maybe something like
you start with a score of 1
getting lynched as either alignment loses you points
getting lynched in lylo or lynching wrong in lylo loses you a fair bit of points
failing to kill PRs as scum loses you points
being on the wagon that forces a PR to claim as town loses you points
stuff like that

I am not that interestead in the ranking system, at the first place. Would perhabs play a game or 2 there if the game seams to be interesting and just ignoring the score.
But I would rather see good stuff rewarded, than bad stuff giving malus.

Getting lynched is somethink that can be your fault or can be the other players fault.
I also think that the lylo-lynch is often the hardest lynch in the game and players feel bad enough for getting it wrong.
I am also disagree that a kill is bad because it lands on a VT.
Getting points for finding PRs as scum would be a different think, because thats a revard for skill.

[camn]

Why would anyone who is actually ~good~ actually opt in for this?

[Espeonage]

Ok my issue with it is how do you score it? Because if you value accuracy of townreads it needs to be percentage based which just makes a bunch of jesters day 2 of scum dies day 1. Otherwise you are valuing living which is inherently anti town. The whole point is that town are expendable and sometime you just gotta die.

[Ranmaru]

Why would anyone who is actually ~good~ actually opt in for this?

Good point, in a tournament for a fighting game, what attracts good players is $. So for good players, the only turn on would be pride and bragging rights. This will give incentive for bad players to become good players. To get good, you need to play good players.

[Psyche]

think should be framed as something less charged than a ranking
it's a way to measure and understand your typical read accuracy within and across games in a way that has more validity than currently available methods
i think that, aside from the rankings thing, could help loads for advancing site meta