More on A Curious Case Study in KGS Ranks

[RBerenguel]

IIRC, the win rate was ca. 70%, anyway enough to improve a rank, because I WAS promoted a rank.

And so, what is the problem?

[Mef]

IIRC, the win rate was ca. 70%, anyway enough to improve a rank, because I WAS promoted a rank.

Then what you are recalling is a 100 game period, not a 300 game period. Your 100 game moving average win rate across your entire kgs career was posted...why on earth do you think you can keep making up stories that never happened?

[RobertJasiek]

And so, what is the problem?

Explanation see before.

Then what you are recalling is a 100 game period, not a 300 game period. Your 100 game moving average win rate across your entire kgs career was posted...why on earth do you think you can keep making up stories that never happened?

Do a proper web search, and you will find what has happened NOT for just 100 games.

[Mef]

Do a proper web search, and you will find what has happened NOT for just 100 games.

Why would I do a web search when I (and others) have your entire win/loss history on KGS pulled into a spreadsheet?

[RBerenguel]

Do a proper web search, and you will find what has happened NOT for just 100 games.

Why would I do a web search when I (and others) have your entire win/loss history on KGS pulled into a spreadsheet?

CSV file actually. This morning I wanted to find out how to do moving averages with R (then I realised I had before but I tend to forget stuff) and re-did ez4u's moving averages plot. The win rate never raises that much, and when it does, coincides with rank jumps. There are also apparent a lot of streaks of sub-optimal play.

[Bantari]

I want that the rating system acknowledges my change from a non-serious period to a serious period within ca. 100 successive games (69 wins : 31 losses) instead of ca. 400 successive games (276 wins : 124 losses).

Ok, again - fair enough. This is what you want.

I still don't see why 100 is a better number than 400, and why wouldn't you then complain that you want the change to take place within 50 or 20 games, but this is beside the point. You are profoundly unhappy with KGS because it sees you as 4d (based on your play), while you want it to see it as 5d (based on your "peaks") - or maybe you just want your "peaks" to count for more.. Ok, got it.

The one point I can make here is that what you are talking about is not a "failure" but a "design" I think. I am sure parameters can be set to make the system behave in a more sensitive ways, but decision was made to make it behave as it does. I don't know for what reasons, but I am sure there were some. And you should investigate the reason, because this is really what you have a beef with, not some imaginary "failure of the system". And it is this reason you have to argue agains and hopefully change, not complain about needing "superhuman effort" - which is such load of gooey.

In the meantime, since you cannot get what you want on a silver platter, there are several *easy* options open to you, as already pointed out:

• create multiple accounts for various "modes"
• play unrated games when not in best form
• switch to a server that offers what you need

There are probably more such easy options if you think hard. I see all of the above options as no-brainers, each one solves your problem easily, and that's what any reasonable person would do in your shoes. But apparently you find each of these options "inconvenient", and you rather suffer the terrible fate of being a 4d than do any of that. Well, fine, not really my problem, just trying to help here.

All I can say - keep knocking on doors, maybe somebody will open. But if you wish to argue, argue smartly, not make a nuisance of yourself. Find a reasons for why things are the way they are, and then find arguments why these reasons are no (longer) good or valid.

PS>
Oh, and one more thing, because I really need to drive home this point: If you play strongly enough, you get to 5d and beyond. Trust me, it is a proven fact, I have seen it happen. Many times. Goes for any rank, actually, except maybe 9d. I am saying that in anticipation of having the same discussion with you about 6d when you finally reach 5d.

Other than that, I pretty much said what I wanted to say, so I will probably bow out of this discussion or all that is left for me is to keep reiterating what I have already said.

[RBerenguel]

Let's do it another way. Robert, I downloaded the same file ez4u used, which is a log of all your game results in the account. I computed the rolling average of your win-lose:

Code: Select all

library(zoo)
a<-read.csv("kgs-sum.csv")
r<-rollmean(a$Result2, 50)

I picked 50 items in the window because you like fast changes, or so it seems. I can use 10 or 100 if you prefer, but whatever.

Code: Select all

> sum(r<0.55)
[1] 8400
> sum(r>=0.55)
[1] 5974

I wanted to see if you had good averages (~50% WR) No. Most often your WR is below 0.55

Code: Select all

> sum(r<0.45)
[1] 3032

So, in 3000 of these data points you were "losing" rating (ballpark figure: that's 1/4th of it).

Code: Select all

> sum(r>0.65)
[1] 1646

In 1646 you were "wining" rating (somewhat more than 10%).

So, in summary, in a high volatility rating it's quite likely you'll be 4d far, far more often than 5d. And the average of your rating would just put you as a strong 4d:

Code: Select all

> sum(r>0.75)
[1] 92
> sum(r>0.75)/sum(r>0)*100
[1] 0.6400445

That's 0.6% of all data points.

[RobertJasiek]

I still don't see why 100 is a better number than 400, and why wouldn't you then complain that you want the change to take place within 50 or 20 games,

Because 100 does not require frequently playing players to devote ALL their effort and time on KGS rank improvement. 100 is a number of games that can be played seriously in succession day after day. 400 is not.

Ok, got it.

No. Read my earlier explanations again.

decision was made to make it behave as it does.

Other failures of this system:
- occasional manual global rating adjustments
- chaotical ratings or incidents of chaotical jumps of ratings of part of the players
- unclarity for the players what the likely rating effect will be of handicap games or games with 2+ KGS-ranks difference

you should investigate the reason

Eh? You mean the accounting of the history of a player's games?

it is this reason you have to argue agains

I did. Several times. Several years ago.

and hopefully change,

It was in vain.

each one solves your problem easily,

No. It does not SOLVE the problem. It CIRCUMVENTS it.

[RobertJasiek]

in a high volatility rating it's quite likely you'll be 4d far, far more often than 5d.

(Supposing the same rank scale.)

No. In a high volatility rating system, I see the immediate success of a win and there is no frustration, so I am immediately motivated in every game. I would thus frequently play seriously and produce very other results.

[Mef]

It sounds like the ideal rating system for Robert is L19's: Type whatever number in the box makes you happy

[RBerenguel]

in a high volatility rating it's quite likely you'll be 4d far, far more often than 5d.

(Supposing the same rank scale.)

No. In a high volatility rating system, I see the immediate success of a win and there is no frustration, so I am immediately motivated in every game. I would thus frequently play seriously and produce very other results.

Wishful thinking.

[Bantari]

in a high volatility rating it's quite likely you'll be 4d far, far more often than 5d.

(Supposing the same rank scale.)

No. In a high volatility rating system, I see the immediate success of a win and there is no frustration, so I am immediately motivated in every game. I would thus frequently play seriously and produce very other results.

Maybe you should find a better motivator than rank chasing then.
Personally - I see an immediate success of a win. Its a win. Not sure why this is not enough.

Just enjoy good games and good moves, try to make more of them, and when you play like a KGS 5d, you will be a KGS 5d, I promise. And because it is hard, it will be an accomplishment to be proud of, rather than some fleeting bleep on a volatile rating radar.

[RobertJasiek]

I see an immediate success of a win. Its a win. Not sure why this is not enough.

Because... KGS does not reward a win properly as a win for frequently playing players. KGS only rewards a win properly for frequently playing players if they invest so much successive effort and time in creating enough wins and a sufficiently high winning percentage that they have to have that great amount of effort and time available, because they need not invest any effort and time in a job or other essential activities of life. A frequently playing player can make himself a slave to the KGS rating system and devote all his life to fitting its requirements, or he has no good chances of reaching the rank, at which he meets a distribution of opponents against whom he would win ca. 50% even real world games.

"A win is a win" is enough only for infrequently playing players, because they need not become the slaves of the KGS rating system in order to reach the rank, at which they meet a distribution of opponents against whom they would win ca. 50% even real world games.

[daal]

Robert - I can understand your sense of frustration at the perceived sluggishness of the KGS ranking system. What I don't understand is why you tolerate this frustration instead of applying the obvious fix: reserve the "Sum" account for unhampered games (where you are not tired etc.) and make another account for when you want to play under less than optimal circumstances.

Despite all of your criticism of the KGS system, you obviously find enough things to like about it - otherwise you wouldn't play there so much. Since your criticisms of the ranking system are regularly rejected by those in charge, why continue to waste energy repeating them? I can't see why you wouldn't want to remove the frustration and have more time to write, play and study.

[RobertJasiek]

The reason is: I prefer one person - one account. Therefore, I have tried to convince that KGS should be like that. Since obviously KGS wants the opposite, I guess some time I need to do what I dislike - create two or several accounts and possibly discard and restart accounts.