rankings are up

[dakar]

and I'm at rank 20 (though why am I behind #19?). *feeds a witch to his dragon for a job well done*

In any case, if you're up on the list, you might as well boast and insult everyone else who didn't make it here. I'm sure Dragonfire and OutKast would like to make use of the publicity.

[K-84]

I am #53.

And can somebody tell me how this ranking thing works?

I am 121 6 17 and the guy in front of me is 63 0 0.
How?

[Snork]

Originally posted by K-84
I am #53.

And can somebody tell me how this ranking thing works?

I am 121 6 17 and the guy in front of me is 63 0 0.
How?

Because you have a lot of losses, it doesn't have any yet.
It's not just about total wins, ratio figures into it somehow as well.

[Sir Vilhelm]

#12 woohoo!


Long live Something Wicked"

[A.F.I.]

#4

man dakar u should to join my clan

[kyrios24]

#5, stupid disconnects...

K-84, thats the stupidest thing I've ever heard. Never mind. But still...ratio is key. The system is very win-friendly as it is. By your standards, you would get rewarded for playing the most, not being the best. If I were in charge, that 63-0-0 would be a whole lot higher.

By the way, has anybody figured out what formula is being used for the rankings?

[Siquo]

#31, straight from out of nothing. Not bad for a week of playing

[Shaman]

It would be fun playing my friends, since we could chat and BS and have a lot of fun jabbing each other. But I know I would beat most of them 100% of the time, just because they are not strategist. Would that be a viable rank?


The problem I see with rankings here is that I dont see a way to filter out the person who just drops with his friend and beats him up 100 times in a row.

How to fix it? I dont know.

Maybe servers could track opponents and not count the same opponent more than 5 times in a 24 hour period or something.
It wont stop the person who creates dummy accounts for wins, but it woudl cut down on rank jumping (accidently or not) for people who only play the same person all the time for wins.

[canopus]

#23

I have the impression that the more you play overall the better your rank will be. Dont be afraid of playing folks better than yourself because that is the only way to learn. Losses are calculated at a much lower value than wins.

In other words it could be something like (wins*5)/losses to determine your percentage.

I hope this interesting form of ranking encourages the good players to play each other and not to avoid possible losses, as would a straight win/loss ratio system.

Ranking is of course still potentially broken since people can fake games to add wins, and it doesn't rate wins by skill of opponent, like people have suggested (from chess ranks).

I can't complain tho.

-canopus

[xyxaxyz2]

My formula is located in this post:

http://www.tacticsarena.com/forum/sh...&threadid=1361

Not completly accurate, but extremely close. However, this was tested with the GL rankings weeks ago, so I don't know if it still works.

[yasli]

im number 11 with 79 wins and one loss--and the one lost was against a gold before i was gold--i didnt think i would do well enough for it to matter in the long run so i didnt change names--heh but oh well--its fun

[QuicksandSlowly]

I wonder how this will all break down once the ratings system is installed eh? It's already up and working on Legends Server. I'm #12 out of 50 there currently.