Posts: 1,297
Threads: 83
Joined: Mar 2007
Reputation:
16
Your ranking score (points) is the combination of your ELO-rating, penalty and your activity.
points = ELO - penalty - ( floor{ days_of_inactivity } / 7} * days_of_inactivity )
That means as long as your are not activly playing your points will be reduced. The reduction will increase per week - first week will reduce your points by zero per day of inactivity, second week by one per day of inactivity, third week by two per day of inactivity and so on.
Playing again will nullify this inactivity reduction
Marvin Wrote:The first ten million years were the worst and the second ten million years, they were the worst too. The third ten million years I didn't enjoy at all. After that I went into a bit of a decline
Posts: 25
Threads: 5
Joined: Nov 2010
Reputation:
0
Hey guys,
hope this thread is ok for my request. I observed that my elo differs around like 100 points once in a while. it is similar for other players. That means that the ELo system isn´t very good yet. We need to fix it. I know there are a lot of influences u can´t avoid, like leavers or corruption or luck and lets say kinda daily form, but 100 Elo points difference is way too much.
First of all. To compensate a high Elo there allways bad players in a team who probably feed. That is a nice idea. But for being good u currently get punished twice by what u call "chance of winning" penalty. The chance to win shall be 50 % for both team. It has to be equal for all members of the same team for obvious reasons.
It would be the first step in the rigth directions, but it actually won´t solve the problems.
Posts: 332
Threads: 56
Joined: May 2011
Reputation:
3
you will never ever be able to create a ranking system in a PUBLIC TEAM BASED game.
too many capricious factors. thats a fact... and as long there are leavers, kicked people, manipulators out there, this will not change.
Posts: 1,494
Threads: 32
Joined: Sep 2007
Reputation:
5
And still we created one with the currently best player being top ranked (even with 3 accounts). A miracle!
Posts: 332
Threads: 44
Joined: Dec 2008
Reputation:
4
(2011-07-11, 16:42:12)Prog Wrote: And still we created one with the currently best player being top ranked (even with 3 accounts). A miracle!
I still remember the BTNK vs BUG match we did not too long ago - and though bug definitely had the players which where regarded as being "better", the nearly lost. Only reason they didn't was because our BTNK team played together for the first time in their life, and as we had 4 infernals vs 1 infernal in the other team, and were standing at their base' door, no one dared to drop, and thus we lost (I'd have if i wasn't the one who did the ups and thus didn't have the money for an infernal).
Either way, there is no 'best player' in the first place in btanks, because there are just way too many factors to weigh in when thinking about how "good" a player is.
Teamplay, timing, speed & reflexes, tactics & strategy, communication skills, farsight, looking-at-the-map, .... and dozens of other things are important.
And there's also the hardware aspect like ping, monitor resolution, fps, ... which may also account for winning a game ( probably everyone has had the moment where they died or just didnt get a kill because of a ping spike, and sometimes those are even server-related and you can't do anything about it ).
Also it's possible to just join a game with 2 accounts to increase the chance to win. Or have friends help with winning.
The ranking doesn't say much about a players skill in fact, and as such, you need not care about it.
Just look at the top 100 playernames:
How for example could RaptorXI ever be behind stibi if the ranking really told you about skill? Can't imagine him ever loosing out in a 1on1 fight - and still, at the time of writing this, RaptorXI is ranked 78 while stibi is ranked 72.
Or High_Bob on 81 - that one should probably somewhere around 200 or so regarding player skill.
I don't care about how people are ranked, and I don't care much about my own ranking, but I care about totally imbalanced games, as those aren't much fun - and as we all know the ranking is based around points which directly come from ELO - penalty.
And gozo is totally right about the rest - it's just not possible with 3 leavers on average each game, dozens of people who try to give up before it's over, afk'ers and stuff like that.
Getting used to the Sand everywhere. At least it brings us map updates.
Posts: 542
Threads: 52
Joined: Jun 2008
Reputation:
8
Elo is a measure for skill - not perfect, but still good in my opinion. Players who reach 1700+ are definetly top. Problem is more that most league games features are massive feeders, noobs, leavers... simply said public players meet clan players, who have a totally different look onto this game (which you mentioned already "Teamplay, timing, speed & reflexes, tactics & strategy, communication skills, farsight, looking-at-the-map").
Me i stopped playing league games so seriously (was top 10 too for quite some timewith almost elo 1700 before the elo distribution was changed and some of the best players left league), now i just flame hardcore if i'm in the shit team (i use waiting time in lobby for figuring new speeches). 
Posts: 25
Threads: 5
Joined: Nov 2010
Reputation:
0
(2011-07-11, 20:09:27)RaptorXI Wrote: now i just flame hardcore if i'm in the shit team (i use waiting time in lobby for figuring new speeches).
Wow - ur so cool . And u even look too old to think it´s cool to flame.
Yes the Elo system works. And even with all the problems - the miracle is done. Best players are top ranked. Good work! Really.
But pls telll me why u think this factor called "winning probability" is for single players in the same team different. That doesn´t make any sense. Orange has the same winning chance like grey. Cause grey can only win if orange does. That seems to be weird in this formular. If i find some time i´ll think about it and suggest another formular.
regards
Posts: 29
Threads: 1
Joined: Nov 2010
Reputation:
0
(2011-07-18, 18:49:09)El_Polacco Wrote: guys what's wrong with league ranking when someone leaves/drops is kicked are only decreased penalty points to ranking and not for won/lost game.
I just got disconnected for unknown reason i never had such problems and it added besides penalty points got also lost game points.
On my first account when i left game league ranking added me win points . Is it all right with this system league ranking ?
I think it works like if you leave game, but your team goes on to win the game, then your ranking points are unchanged; however if after you dropped your team loses, then you lose ranking points too.
Posts: 27
Threads: 3
Joined: Feb 2011
Reputation:
0
In general, I think ranked games are somewhat balanced. I love that. However, when you look at my ELO compared to that of other players and compare our performances, it's obvious that the system needs some fine tuning. I admit that I suck. I'm usually in the top 4 worst players in each game. Often I'm in the top 2. Players who are clearly much better than me often have a lower ELO. Perhaps it's because they leave games and I almost never do (I've been kicked sometimes, though). I don't have any suggestions yet, but knowing there is a problem is the first step needed to fix it.
Posts: 1,297
Threads: 83
Joined: Mar 2007
Reputation:
16
I've tried a couple of tweak and system the last week and recalculated all games (which requires about 15minutes) but i haven't got an satisfactory result.
i would like to have - a better distribution, currently most players just moving around 1500
- zero as start value
- a somewhat stable value independent of winning/losing streaks
Marvin Wrote:The first ten million years were the worst and the second ten million years, they were the worst too. The third ten million years I didn't enjoy at all. After that I went into a bit of a decline
Posts: 542
Threads: 52
Joined: Jun 2008
Reputation:
8
Maybe bigger influence by personal stats?
And what if you removed balance and distribute points only due to team elo?
Posts: 164
Threads: 22
Joined: Oct 2010
Reputation:
4
(2012-08-14, 00:50:24)RaptorXI Wrote: Maybe bigger influence by personal stats?
And what if you removed balance and distribute points only due to team elo?
I think that idea about infuence of personal stats is rly good. How about to not decrease "good" players ammount of points so hard after loosing a game where he was lets say the only one with positive stats or something like that?
Posts: 1,494
Threads: 32
Joined: Sep 2007
Reputation:
5
That is bad because it encourages selfish play of all kinds. We had this discussion already way too often.
Posts: 542
Threads: 52
Joined: Jun 2008
Reputation:
8
I was rather speaking about teamwise-related stats like support points, assists, upgrades made (not even shown at the game info at the moment), control points captured than just kill/death ratio. I have no idea how that would encourage selfish playing.
Posts: 127
Threads: 13
Joined: Mar 2007
Reputation:
2
2012-11-12, 17:35:30
(This post was last modified: 2012-11-12, 17:39:49 by ICE2607.)
also something like damage dealed/taken ..
LR-gays do not take alot dmg because they are still heli on min 40 with ironhull 
they do not defend CPs if capture is in progress because they die instantly
they do not share any dmg with the team
lets say, they are useless if ur opponent is superior
~=| LuVs |=~
Posts: 164
Threads: 22
Joined: Oct 2010
Reputation:
4
(2012-11-12, 12:54:10)Prog Wrote: That is bad because it encourages selfish play of all kinds. We had this discussion already way too often.
Ok but i dont consider my self as selfish-ego player. And i never was called that way. But i loose many points caus of bad luck for the teams....
Posts: 1,494
Threads: 32
Joined: Sep 2007
Reputation:
5
2012-11-12, 19:35:01
(This post was last modified: 2012-11-12, 19:35:29 by Prog.)
I did not claim that any specific person falls under this description, but I certainly know that I would killsteal as much as possible with more individual stat influence of, say, k/d.
I have no problem with increasing the influence of team-specific stats, e.g. support/assists.
Posts: 16
Threads: 1
Joined: Nov 2011
Reputation:
0
Hi all,
BTanks League RANKED #38436
map version: 878b
gameID: 38293
winner: Dark Force
started: 2013-07-04 01:58:01
ended: 2013-07-04 02:54:51
length: 56m50s
Admins, please look at stats of this game and tell me if it's normal to get the lowest delta in my team and why???
Thanks
|
