Tracking Karma (manually as of now)

[Cabd]

So I'm trying to use my user title to do something kind of positive for the site, and here's where I'm at: Tracking Replacement Karma.

Karma here is defined as a number generated by the number of replace ins and replace outs a player has done. I plug in all the numbers and it spits out your score. Replace outs scale in difficulty to make up for as you replace out more and more. Replace ins do not scale. Players replacing out of newbie games are given an extra penalty, and players replacing into newbie games are given a bonus, in order to make this project have the side effect of encouraging public service. This system is entirely optional, of course, but I plan to make the data available for moderators to use. (As in, players with less than X karma score are forced to sign up as a replacement and not a player slot)


Here's some sample data from some users self-reporting in sitechat:
The format is: Playername | Ins | Ins (newbie) | Outs | Outs (Newbie) | Karma Score


Cabd 7 5 0 0 4.12
Kazekirimaru 0 3 3 1 -2.55
Bert 17 10 9 0 -2.92
FakeGod 3 1 0 0 2.24
Fferyllt 6 5 2 0 2.00
Aegor 4 0 0 0 2.00
Pitoli 0 2 4 1 -4.00
Ms Marangal 10 4 2 0 2.24
Natirasha 3 1 1 0 1.24
Cephrir 4 0 0 0 2.00
MattP 2 2 4 0 -1.55
Brian Skies 0 2 2 0 0.00
Wisdom 3 15 1 0 4.74


Your karma score is calculated by a simple formula:

SQRT{(replace ins)+(2*Replace ins newbie games)} - {Replace outs + (2* (Replace outs newbie games))}

The goal will be to provide and host this data on the Wiki, updating it once a week or so.

[Aegor]

Thank you! This is a great idea.

Are you starting this system by taking requests from mods with games in signups?

[Cabd]

I'm manually going to be feeding data from completed games starting with January 2013 completion dates, but I'd love it for users to self-report their own data. I know i'll be asking for data when I take signups.

[Psyche]

<3

[talah]

1 regular replace-in
1 newb replace-in
1 regular replace-out
2 newb replace-outs

The above does not take into consideration in any form the games I joined as a non-replacement, so in addition..
2 regular joins
1 newb join

^ignoring this,

1+2x1=3
1+2x2=5
3-5=-2

Which makes me -1.414, well, at least in the irrational sense

However, if I do the sums with the games which I've joined in the first place included, rather than just treating them as null...

3+2x2=7
1+2x2=5
7-5=2

Which is the same number but with a + in front of it, funnily.

The calcs feel slightly clinical (no reason is good enough to replace) and as if they're missing something - which I guess, is that they don't appear to take into any account whether you've played a full game from scratch and not replaced out.

Not sure if I'm missing the intent here either but replacing into a game, and replacing out of a game, seem like two completely different vectors. Like as if the formula has a predisposition to encourage replace-ins for those who've previously replaced out.

[Cabd]

Like as if the formula has a predisposition to

encourage replace-ins

for those who've previously replaced out.

This is entirely the intention.

is that they don't appear to take into any account whether you've played a full game from scratch and not replaced out.

This is also intentional. The system is designed to track and reward players with pro-community replacement patterns and identify those causing harm.

[Natirasha]

I worry that putting a number next to everyone will lead to some sort of egotism or whatever.

Isn't there a reason we don't have everyone's post count show under their avatar?

[Cabd]

to some sort of egotism

It's the same as tracking win rate. Public data. Although in this case, what if somebody decides to be the very best at this system, so they..... replace into lots of games and don't replace out of games! Oh wait, that's awesome.

Edit: my response in XKCD strip form

[Natirasha]

Well, I more mean, we hide cumulative post counts and pretty much every 'ego' feature seem on most BBCode forums on purpose to prevent their use in-game. I'm not opposed to a karma score, but having it displayed publicly will taint games I worry("Don't listen to him! His karma score is -5!).

[talah]

I think you need an inverse joined&!replaced calc on the end there then to be entirely balanced.

But m'okay. I kind of get it. Except for the fact I don't generally associate naming and shaming with encouragement ^^

[Cabd]

It's the same as a win ratio, so the argument of "Don't listen to natirasha his win ratio is like 40%" is equivalent and I'd laugh that one off as an idiot trying to make it, soooo.... (Well actually win ratio is more skill equivalent than the helpfulness of replacements)

[talah]

Although in this case, what if somebody decides to be the very best at this system, so they..... replace into lots of games and don't replace out of games! Oh wait, that's awesome.

Well actually something like this I can see, as far as a game playing contribution index goes.
You'd have to formulate it though so that it came up from an asymptote (no number below zero).
Really, I think joined games contribute to the experience as well. And games you leave are kindof subject to interpretation, that's the other big thing.

ed: as in not all replacements are equal, but how is a legitimate reason assessed?

[Natirasha]

It's the same as a win ratio, so the argument of "Don't listen to natirasha his win ratio is like 40%" is equivalent and I'd laugh that one off as an idiot trying to make it, soooo.... (Well actually win ratio is more skill equivalent than the helpfulness of replacements)

I think you're missing what I'm saying.

[Katsuki]

Time to try and replace out of as many games as possible

[Vi]

It seems like you're going to have your list clogged with newbies who replace out of their first and only game. Those who want to come back may be discouraged because they're on a Wall of Shame.

If you wanted to make something community-service oriented, you may want to consider whether the game is 50 or 100 pages long at the time of replacement.

The biggest problem you're going to have is widespread acceptance of this. This is not at all the first time karma has been proposed, and I'm having difficulty thinking of a time when it was actually used.

[Espeonage]

@Above. I've replaced in three or four(?) times since I came back which was a month ago. And my aim is to never replace out of a game again pending a personal disaster precisely because I know a lot of players and moderaters hated the fact that I used to be a chronic flaker. I was a liability to mods to have in their game. And it is something I am not proud of. I think that given my history, it was imperative for me, on coming back to mafia here, that I put that right and it has driven me to replace in to games I never would of. 100+ page replace in? Yep. Before, I'd be unlikely to post even if I'd been in the game from the start.

Obviously most people wouldn't want to go through the effort to get into the 'good books' but I feel those players are going to be a part of the problem for a long time, so in a way it doesn't matter. If they aren't going to join games, they aren't going to flake from those games.
Whereas those that want to be liked will want to turn that around. They'll want to be in the green with their karma.

I know that being put on a blacklist for my actions was a wake up call for me in terms of activity. And now, given my recent history, that Mod is giving me another chance. All of which is specifically because I got called out for having a horrendous karma.

[Espeonage]

It's the same as a win ratio, so the argument of "Don't listen to natirasha his win ratio is like 40%" is equivalent and I'd laugh that one off as an idiot trying to make it, soooo.... (Well actually win ratio is more skill equivalent than the helpfulness of replacements)

Win ratio is hilariously skewed by the fact that people often join games because players they like are in and if one of the group has an impact on the game, it will look like they each independently have that same impact.

[ObsessedWithCats]

Given that there will only ever be as many possible replace-ins as there have been replace-outs, shouldn't they either both be linear or neither be linear? There's more incentive to replace in if it's going to have a more appreciable effect on your karma score and you might find more people think it's worthwhile to try and get rid of a very bad score if it's actually possible.

Also like Vi said, replacing into a 100 page game takes considerably more immediate commitment than replacing into a 3 page one - maybe add some page-number categories for the effort involved in the replace? It'd probably make the formula a bit complicated but something like...

Score = (

ins<25pages

+ 1.5*

ins25-100pages

+ 2*

ins100+pages

+ 2*

insnewbie

) - (

outs

+ 2*

outsnewbie

)

...with SQRT for replaceins if you want to keep those might account for effort without overdoing it. Maybe bump up the multiplier on newbies to keep it shiny relative to the others, though as a serial newbie-game-replace-iner I might be a bit biased there :3

I think increasing the loss from replacing out at higher page numbers might encourage people to replace out earlier than they need to to minimise the effect on their karma, when in fact they might have worked through it without much effect on the game relative to a replacement, but aside from complicating the calculation I don't see much problem with rewarding replacing into a game that's got loads of pages. I can't find any newbies with much more than 50 pages so scaling those to size seems a little unnecessary.

Maybe very specifically not count a new player's very first newbie game, to give them a second chance if they did the sensible thing and started in the newbie queue?

[Cabd]

It seems like you're going to have your list clogged with newbies who replace out of their first and only game. Those who want to come back may be discouraged because they're on a Wall of Shame.

If you wanted to make something community-service oriented, you may want to consider whether the game is 50 or 100 pages long at the time of replacement.

The biggest problem you're going to have is widespread acceptance of this. This is not at all the first time karma has been proposed, and I'm having difficulty thinking of a time when it was actually used.

As far as I'm concerned, this system is more opt-in; with the note that if you sign up for my game or the game of a mod using this optional system, you agree to have your data ran and published. I certainly don't plan on using it for newbie players. I was planning on using it for players that would play my potential games in Coney Island, Theme Park, New York, and Little Italy.

Weighting the size of the game might be something viable down the line, but at the same time makes things needlessly complex, KISS is kinda the goal here.

[Cabd]

The above does not take into consideration in any form the games I joined as a non-replacement

As far as I'm concerned, signing up to play a game is a promise from you to the mod and the other players that you'll see the game through. Why should I, or anyone else, reward you just for doing what you promised you'd do?

[4nxi3ty]

What about players who replace out because of irl emergencies/issues? Or when a player gets really into it with another and it is healthier for everyone involved to replace out? Are those going to be punishable?

never liked the idea of a karma system. Just don't like an objective, automated criteria enforcing how I should behave. (I think we already handle excessive flakers pretty well on a personal level.)

If you are interested in helping a replacement drought, replacing in yourself has an immediate effect on the problem.

[Psyche]

shaming people in the way you advocate seems like it'll make site flakes more permanent

[Zachrulez]

I'm sure there's a handful of times I replaced out of a game vs the many more times I've replaced into one.

Does your system count this game against me?

I had no choice because I would have played the game knowing for fact who one of the scum was.

[Cabd]

I'm not counting replacements due to moderator error such as spoilers given on accident, similar to my leaving New York 165



Remember that the end goal is to have players self report the numbers and spot check them, untilods start working with me.

[Aegor]

What about players who replace out because of irl emergencies/issues?

That can easily be addressed on a case-by-case basis. Do you know any mods who would not be willing to consider such extenuating circumstances?

Or when a player gets really into it with another and it is healthier for everyone involved to replace out? Are those going to be punishable?

That scenario should not arise in the first place.

never liked the idea of a karma system. Just don't like an objective, automated criteria enforcing how I should behave. (I think we already handle excessive flakers pretty well on a personal level.)

I am not so sure -- the majority of my games have featured more than one replacement. There should be no replacements in most games.