Forged Alliance Forever

Re: Rating Distribution

2015-09-19T13:06:23+02:00

One more reason why #of games isnt that importend:
There are some tryhardswho actually watch pro stream/games/casts or analyse own replays and have this way more knowledge at 100 games then a 500+ games player. So i rly wouldnt care much about games played if the nuumber is >80.

Statistics: Posted by Turinturambar — 19 Sep 2015, 13:06

Re: Rating Distribution

2015-09-17T23:53:16+02:00

keyser wrote:
well even if what ionic said is pretty impressive shit.
i don't really understand your reaction... I mean you aren't the only one that is suspected to be a smurf, i've been too. But i didn't experience it at all like you are describing it. I was just able to play higher rated game faster.

I was a bit to rude towards him tbh. But what I wrote is what can happen to you at the moment, and its not nice

Spoiler: show

Statistics: Posted by xvdfhn — 17 Sep 2015, 23:53

Re: Rating Distribution

2015-09-16T21:34:37+02:00

keyser wrote:
well even if what ionic said is pretty impressive shit.
i don't really understand your reaction... I mean you aren't the only one that is suspected to be a smurf, i've been too. But i didn't experience it at all like you are describing it. I was just able to play higher rated game faster.

Community 2 years ago is different than the community now.

Statistics: Posted by KD7BCH — 16 Sep 2015, 21:34

Re: Rating Distribution

2015-09-16T12:15:05+02:00

well even if what ionic said is pretty impressive shit.
i don't really understand your reaction... I mean you aren't the only one that is suspected to be a smurf, i've been too. But i didn't experience it at all like you are describing it. I was just able to play higher rated game faster.

Statistics: Posted by keyser — 16 Sep 2015, 12:15

Re: Rating Distribution

2015-09-16T01:25:48+02:00

Ionic wrote:
Generally we don't have new players, anyone above 500 with a game count less than 100 is a smurf.

Nice self fulfilling prophecy you have there. Its because of people like you that new people stop playing. I nearly did.

You get so much shit when you arrent <1k with less than 30 games its pretty exhausting. People are calling new People names, kick them randomly out of lobby, hell they even do the dot trick to offend you.

And do you know the best part? You can´t do a thing against it. Talk to a mod? pffff. after they talked shit you gonna get kicked out of the lobby and since the game lobby isnt monitored you´re ******. No proofs what so ever. Its a great first look into the community. If i wrote to you what i think right now i probably get banned, so I better stop now.

thank you very much.

Statistics: Posted by xvdfhn — 16 Sep 2015, 00:50

Re: Rating Distribution

2015-09-15T19:43:53+02:00

KD7BCH wrote:
Reaper Zwei wrote:
KD7BCH wrote:However a player with 20 games, 40 games, 60 games, rated 1000 is clearly weaker than a player with 2000 games rated 1000. The system doesn't do well in ranking players with low game values. The game quality metric will treat these players equally when they are not.

This doesn't really make any sense to me unless you mean that the system always overrates players with low amount of games.

The vast majority of players are rated 950 or below so I ask what is the point of distinguishing players who are 1600 vs 1800 when there are so few

Because there's a difference in skill?

It does overrate them because they don't have the experience. Smurfs aside which are not most players with 20 games, a player with 20 games can't possibly have the same ability as a player with 200 games at the same rating. They difference is less pronounced between a player of the same rating who has 200 games versus 2000 games but look at it from the perspective of a player rated 1000 at 20 games or 1000 at 2000 games. The player with 20 games can't possibly know what the player with 2k games does.

In a team game this doesn't affect game quality and it should in order to balance the game more effectively.

Is it possible for a person with little experience to be a much better player than someone with a lot of experience? For example, a person that's rated 500 with 1,000 games vs. a player that's 1,300 with 20 games? If you agree with that, then why can't those numbers meet in the middle? You're perceiving a restriction that doesn't exist, you're incorrectly assuming that all equally rated players are the same, except for experience. One 1,000 rated player could be a naturally better player than another 1,000 rated player, but because he has less experience they are equally matched.

That's why the rating system we have works, only the outcome matters. If it's as you say that one 1,000 is better than another (because of experience or whatever), then naturally they will diverge in rating.

Statistics: Posted by CodingSquirrel — 15 Sep 2015, 19:43

Re: Rating Distribution

2015-09-15T18:59:54+02:00

I agree that inactive players skew the picture, but i cant consider them because i dont have the data (last login eg). I could do that if i could get some data from the server and much more, like map specific stats (2700 is not necessarily 2700 on another map than Setons)

Codepants already started a Survey http://forums.faforever.com/forums/viewtopic.php?f=2&t=10080
much of this data could be answered server-side, and would make things easier and more accurate

Statistics: Posted by ABSTRACT — 15 Sep 2015, 18:59

Re: Rating Distribution

2015-08-01T00:43:07+02:00

Ionic wrote:
All I heard was, I had this rating with this many games. BUT I also played Steam or I also played ladder or I only played GAP.

I am telling you, I have played thousands of team games. If you have played less than 100 actual games, you are terrible at everything except maybe one map and one spot.

My rating was above 1000 before 100 games too, but I played GPG, Thermo (ranked) 4v4 ISIS and Gap.

pfft

EDIT:

Sorry, it's just bollocks.

Statistics: Posted by Gorton — 01 Aug 2015, 00:43

Re: Rating Distribution

2015-08-01T00:18:01+02:00

How has nobody put this in here yet. I'm sure everyone arguing over this knows how the rating system works right? Because it makes sense to argue over things we don't understand.

http://wiki.faforever.com/index.php?tit ... kill_works

Statistics: Posted by Deering — 01 Aug 2015, 00:18

Re: Rating Distribution

2015-07-30T16:07:01+02:00

buildmoretanks wrote:
Here is another factual diagram.

So basically FU to 75% of the playerbase. Sounds about right.

Statistics: Posted by KD7BCH — 30 Jul 2015, 16:07

Re: Rating Distribution

2015-07-30T15:18:17+02:00

All I heard was, I had this rating with this many games. BUT I also played Steam or I also played ladder or I only played GAP.

I am telling you, I have played thousands of team games. If you have played less than 100 actual games, you are terrible at everything except maybe one map and one spot.

My rating was above 1000 before 100 games too, but I played GPG, Thermo (ranked) 4v4 ISIS and Gap.

Statistics: Posted by Ionic — 30 Jul 2015, 15:18

Re: Rating Distribution

2015-07-30T14:48:09+02:00

Here is another factual diagram.

Statistics: Posted by buildmoretanks — 30 Jul 2015, 14:48

Re: Rating Distribution

2015-07-30T11:08:59+02:00

Players coming from GPG era are also better than their game count makes you think. 1000 rating in 20 games is easy.

Statistics: Posted by Col_Walter_Kurtz — 30 Jul 2015, 11:08

Re: Rating Distribution

2015-07-29T22:05:08+02:00

Ionic wrote:
Generally we don't have new players, anyone above 500 with a game count less than 100 is a smurf. These changes would make smurfing even more frustrating.

That, i'm afraid is BS.

bragging

Spoiler: show

Statistics: Posted by Gorton — 29 Jul 2015, 22:05

Re: Rating Distribution

2015-07-29T21:33:50+02:00

keyser wrote:
you completely made those number. I'm pretty sure there is only a few of smurf. (even more with the only 1 account per computer). And there is a lot of people that had way more than 500 in way less than 100 games.

I think I was rated 600 at 100 games, about 700 at 200 games, and then I fell back to 600 for awhile there, then up to 900 where I sit around now, however I play team games and I play almost exclusively GAP, and those team games are not known to be the greatest skill games, they are however the most plentiful to get into and go.

The majority of players are ranked under 950 and the average player is around 750 or so, my point is only that the whole rating system, doesn't really serve to distinguish the lower ranked players from each other all that well.

Statistics: Posted by KD7BCH — 29 Jul 2015, 21:33