KGS vs. EGF rank

[Ares89]

Hello,

I’m having this debate with my friend about Go ranks. To be specific, EGF ranks contra Internet ranks.
In my country we have limited with players, and we have this shift in ranks around the dan mark.
What I mean by this is; people above shodan(KGS) down rank themselves 2 or 3 ranks for tournaments, to match EGF ranks.
However kyu players do not, so this creates a “wall” between kyu and dan ranks.
Of course this have made me think about the difference in KGS/EGF ranks.
(Example of a problem that might occure: One 2kyu KGS player register as 2kyu, 2dan KGS player register as 1kyu, these players meet in an even game.)

Most of my fellow Norwegian players do seem to think it is a 2 stone difference.
However some makes the case that it is only an 1 stone difference.

There are some reasons for why I’m interested in this topic.
1. It is simply an interesting topic.
2. I find it important that when people which attend European tournaments they are not under or over ranked.
3. In tournaments people who rank less than 1dan have no way of winning the tournament by default.
4. If everyone in Europe could have similar rankings everything would be easier, now that we are so few it is easier to fix the system if it is wrong.

Some random stats: (wish to keep them anonymous)
# 3dan EGF, 6dan Tygem.
# 2dan EGF, 5dan Tygem.
# 1dan EGF, 4dan KGS.
# 1dan EGF, 3dan KGS.
# 1dan EGF, 3dan KGS.
# 1kyu EGF, 1dan KGS.

Personally I have seen more players with a 1 stone difference (KGS vs. EGF) in Europe, however not sure what is more correct.
If I remember correctly most 7dan EGF = 8dan KGS as well.
Also when checking up on these Youtube players most have either same KGS and EGF rank or an one stone difference.

TLDR: what is the difference between in rank between EGF and KGS?

- Ares89

[paK0]

Depends on the rank it seems, but for the higher levels something about:
EGF = KGS + 3

Not sure if this is 100% accurate, though.
Source: http://senseis.xmp.net/?RankWorldwideComparison

[tapir]

Self reported = biased toward higher KGS ranks.

1 dan players with a 1 dan or even 1 kyu account on KGS won't harp about it, but of course the local 2k will tell everyone he reached 1 dan for the first time even if only on KGS.

I personally would subtract more from KGS kyu ranks than from KGS dan ranks to arrive at an initial EGF rank, which fits with the observation that KGS ranks are closer to each other than one stone. I completely fail to understand how the previous post manages to claim EGF = KGS + 3 (or do you mean DDK with higher levels?) and then link to a page (incl. even the mainly self reported rank comparisons on a subpage) where none of this is supported.

On a related note: Please do not force players from other servers (notably OGS) to downrank when starting in an EGF tournament.

[paK0]

I completely fail to understand how the previous post manages to claim EGF = KGS + 3 (or do you mean DDK with higher levels?) and then link to a page (incl. even the mainly self reported rank comparisons on a subpage) where none of this is supported.

Example: Europe = 7K, Kgs = 4K
7 = 4 + 3

For dan ranks you would have to subtract the number. But in retrospect writing higher ranks was rather unprecise.

[RobertJasiek]

EGF rank = KGS rank plus EGF-rank-dependent shift plus country-dependent shift plus-minus 5 ranks uncertainty. IOW, from either rank, you cannot convert to the other rank well for a particular player.

[Gerlige]

Yeah. The level of EGF rank is quite country dependent. For example in Czech Republic is known that our kyu players are level or two stronger than German players with same rank.

Btw I'm EGF 12k and KGS 6k, deal with it.

[RBerenguel]

Yeah. The level of EGF rank is quite country dependent. For example in Czech Republic is known that our kyu players are level or two stronger than German players with same rank.

Btw I'm EGF 12k and KGS 6k, deal with it.

EGF 8-9k, KGS 3-4k.

[alejo]

I'm Egf 10k, kgs 5 kyu... and it gets worse with the lowest ranks. EGF 20 k may be a 13k at KGS... which did happen at the Barcelona tournament a few months ago.

[RBerenguel]

I'm Egf 10k, kgs 5 kyu... and it gets worse with the lowest ranks. EGF 20 k may be a 13k at KGS... which did happen at the Barcelona tournament a few months ago.

Bullsh*t 10k EGF, next tournament you have to enter as 8k Sandbagger!

[Ares89]

So simply it varies from country to country.
So our tournament leaders, which also is the leaders in GoNorway,
may change the ranks to whatever they wish. Within reasonable levels.

I guess it will be hard to figure out the correct rankings.
However won't this affect tournaments like European Go Congress?
Two players playing each other in an even game might in fact be 2-3 stones apart.

- Ares89

[tapir]

Yeah. The level of EGF rank is quite country dependent. For example in Czech Republic is known that our kyu players are level or two stronger than German players with same rank.

Btw I'm EGF 12k and KGS 6k, deal with it.

Taking pride in a broken rating system, oh the beauty of sandbagging. (Ever tried taking 9 stones from Czech 3k?)

There is this wonderful study by Andre Engels about kyu ranks. Basically showing that self-ranking has higher predictive value for the outcome of a game than EGF rating. Sample were kyu games in Netherlands as far as I remember.

[shapenaji]

http://www.europeangodatabase.eu/EGD/Pl ... y=10213203

http://www.gokgs.com/graphPage.jsp?user=Sum

I fail to see significant differences in the ratings.

EGF kids are kinda snarky about how weak all of our dan ranks are...

[Mef]

Trying to find a direct comparison between KGS and EGF will be challenging, if not impossible. EGF is for most players a tournament rank. That means the games are played in a controlled environment, with long time controls. Players have been preparing specifically for those games, more likely to be well rested, focused entirely on the game etc. In short it's a measurement of the upper tail of a player's abilities. Some players may choose to replicate this experience on kgs, but many don't. Its sample includes games where your cat jumps on the keyboard, the neighbor is playing loud music or maybe when you came home late from the pub and decided you had a hankering for some blitz games. KGS is more likely to be measuring the average of your games rather than just the upper tail. Seeing two players with similar EGF ratings and very different KGS ratings likely tells you something about how they (or perhaps their opponents) approach online games differently.

[hyperpape]

As you will soon learn, Robert's games are played under circumstances that allow him to claim they underestimate his strength.

[hyperpape]

Amusingly, the European Go Database includes records and ranks for some major international tournaments that feature European players. So you can filter for US players, or learn that Cho Cheolhan is the best in the world, ahead of Gu Li.

For many reasons, not just the lack of data, you should use this solely for amusement. But I must note that Shapenaji is a respectable 2594. So much for weak US dans.