Maybe this is why I don’t progress. I thought the point of playing was the joy of the complexity of the game, the always learning thing and the analogy to so many aspects of life. Rank is just an indicator of playing strength within a given pool of players enabling matching. I don’t see it so much as the point of playing as a tool to enable good games.
14 Likes
Higher ranks certainly don’t mean the game is more enjoyable.
I had the most fun when I was around 15k-10k.
Playing “for rank” is why people bot and cheat and shit.
13 Likes
Not if the source is quoted. A Fox 6d is not a 6d, it’s a Fox 6d. An EGF 6d is not a 6d, it’s an EGF 6d. Ranks only ever have true meaning inside their own rating system.
A Fox 6d is, by definition, a very realistic Fox 6d. It is no more, no less. There is no such thing as a “true” 6 dan strength.
A little piece of me just died. I may as well stop playing, as I have long since ceased to care about my rank.
Oh, maybe I should keep playing, this sounds much more like fun!
18 Likes
I still don’t understand why I demoted from 1k to 2k with 75% win rate in 12 automatic handicap games (I won 9 straight games as 2k before promoting so the 15 games window shouldn’t have that much of an impact) T.T
@topazg Hey, good to see you again. Hope everything’s going well!
4 Likes
You too, was delighted to see that name pop up in the thread 
Getting ranks is the whole point of playing? I feel saddened by this kind of thinking. Personally I would like to advocate the idea that the point is to challenge your mind and have great fun! Also, for each his own, but for me getting to understand the game better and getting stronger is great fun. Worrying about ranks, one is just worshipping some numbers spat out by some random computer algorithm.
7 Likes
There’s been like 5 posts piling on S_Alexander. I think we get the point now 
I don’t think it’s that bad to use ranks as an indication of improvement (or just enjoying having a higher rank). No one’s being hurt by this.
7 Likes
I understand people wanting a “badge” for something they have worked on (and I guess all the people worrying about ranks aren’t useless 30kuys, they devote effort and time to their rank).
Also, work reasons (very important, no matter how we each perceive it) and ability to participate in tournaments and other official functions are valid as well.
The point being missed, I think, is that the numbers are arbitrary. The fact that they have been historically this or that way doesn’t mean much here and now, if the here and now dictates adjustment (true in things more vital than Go). And it’s only logical for this to be amplified the more Go reaches other countries (with different cultures about achievement, from school even to any other public display of accomplishment, with the number of players growing exponentially, with new heights reached in the game for so many reasons.)
The more people get used to this, the more it will be acceptable that the system (kyu/dan/pro) is one, but open to interpretation, because all achievement systems are open to interpretation, according to current conditions, context and participants.
I’m pretty sure other competitive sports work like that, we’re just used to it by now.
I believe that, subconsciously, a seasoned player can place themselves and their opponent on any ladder (if tomorrow KGS said ranks are now 100-1, people would pretty soon adjust to that frame), regarding comparative strength.
What is important for me, is the data to be able to show that normally a 1D beats a 2k. Since I believe there is absolutely zero chance for one single system to prevail on everything, correct placement within each system is what matters.
7 Likes
@anoek This could be a helpful one to inspect, because it would shine light on how better to answer the question “why did my rank go down after a win”. In this case, there is no drop at the -15 game point in the person’s history, so it must be something else…
4 Likes
Interesting, I’d expect it to trend towards the 13k starting point, but the fact that it trends towards the boundary between 9k&10k is interesting
3 Likes
I noticed that a few players don’t have recent live games in rating history.
E.g. my last game is 25043950 and not in the table https://online-go.com/termination-api/player/109488/glicko2-history
I think Great Glicko is upset with us.
2 Likes
I understand what you mean, but when we say that ranks have no meaning outside each separate system, then why does OGS even try to maintain some alignment between their ranks and EGF/KGS/AGA ranks? OGS could just ignore other rating systems, like Foxy and Tygem seem to do.
I was pretty sure that we had shifted away from that around the time Glicko-2 was implemented and we stopped allowing players to choose rank…
In fact, a lot of the motivation for our rank metric has been to align ranks properly to the amount of handicap between ranks first and foremost, and keep certain min/max constraints for familiarity.
4 Likes
I am confused with the issue of the ranking, I see that after playing it remains the same even if I won many games, from now on we have to wait for the ranking to be modified, do you have any idea, or did the ranking freeze for any new change?
I agree, this is a critical question. It’s clear OGS is aiming to be internally consistent. The site has demonstrated having a non-linear relationship between rank and rating with an attempt to maintain an accurate handicap per rank, and base “appropriate handicaps” managed by internal win/loss statistics to keep it as accurate as possible across its own playerbase. The “first handicap stone is worth half a rank of difference” is a well made point, but it will be covered by the collection of statistics. It looks like their efforts are being successful if nearly all of their “expected/actual” results are as close as the stats we’ve seen on this thread.
Some discussions that I have not yet seen concluded are “do we change the cap at the bottom and top end of rank display?”, and “does OGS choose to align the ranks to somewhere? If so who and why?” Even aligning is not straightforward, as you will plenty of situations where player A and B are the same in one and two or three ranks out on the other. There are a lot of balances to be made between what different people are looking for (why is this the one server I can’t get dan on? what about my teaching prestige? having 12d looks stupid! etc etc).
There is also a question of how best to address people’s confusion over the rolling ratings issue and people with unexpected rank adjustments after a single game. The accuracy implied by the current system appears fairly implicit to me, but how much importance do we place on clarity for the player = comfort for the player? If it’s the kind of thing that’s big enough to decide between “playing here or not”, I’d say it’s pretty important, and clearly there has been unrest in here and chat about things not working correctly, mostly it seems because their rating went down on a win (or the increase from a number of wins was cancelled by a single loss that wasn’t to someone low enough rated that such a big swing was justified). How to address it is not straightforward.
7 Likes
Well, the EGF doesn’t align to any other system. It’s basic premises are that 7d EGF should align with 1p (or at least the lower end of pro level players) and down from there, handicaps should align with ranks (when corrected for a half stone handicap deficit).
These premises aren’t routinely being monitored, but some way or another, those premises seem to hold pretty wel over time when analysing the data.
There exist some important contributions to keep the system consistent, though people may not realise this:
1: Newcomers to the system get to declare their own rank. They don’t start at some fixed rank that is the same for all newcomers. So the system is constantly being fed by declared ranks that are roughly correct in the context of European ranks, assuming that those newcomers derive their declared rank from club handicaps against stronger players in their European club, which used to be how club ranks were determined in Europe.
2: Also, players can reset their rating by making a double promotion. For example, when a 15k hasn’t played a tournament in 6 months and he knows from handicap games in his club that he has improved 3 stones, he can declare 12k in his next tournament and his rating will be reset automatically to 12k. These rating resets are another external data source to keep the system aligned with handicaps, even though only 10% of tournament games themselves are handicap games.
But in recent years, the EGF system is not only being fed by European club ranks anymore. It is becoming more exposed to online rating systems, because more and more people join real life tournaments, who only played online, without ever having played in a European go club.
So when the organiser asks about their rank, these newcomers will say they are 2018 OGS 12k, 2012 IGS 8k, 2020 Tygem 2d or whatever. And then it’s up to the organiser to try and convert that to a reasonable EGF rank. This is becoming a real challenge.
6 Likes
When 1p isn’t an actual strength, isn’t this somewhat hard? How many freshly qualified top of their class Korean ex-Yeongusaeng-now-1p do we have to set it against (and bearing in mind players like Oh Chimin and Cho Seokbin arrived, playing at 1p+ level in Europe despite not being professional, implies that the top crop of new qualifying 1p players in Korea are stronger IMO than EGF 7d - or at least they were back then), and how confident are we about comparing their ability to other pros who have been around a lot longer and have a higher rank but are probably less strong (for example teaching pros such as Guo Juan but have not been playing active top level tournaments for some time)? Even if we drop the rating per rank difference to 30 (which I believe is what the EGF do?), it’s still not really justifiable as professional ranks are honorifics, not strength determiners.
Coming from the UK, and I don’t know how much this is still true, but when I was an active player a Dutch or German 1 dan was not the equivalent of a UK 1 dan. Of course you can mostly be confident that the ranks are within a couple of stones. The problem is that almost all of the EGD submissions from UK tournaments were for players who basically only other played UK players (and in quite a lot of cases only played in 2 or 3 tournaments, and the same ones, each year). You can never have a cross-country accurate system without a lot more matches between player groups, and travelling internationally for tournaments was always very much a minority of players. Internally consistent data, but not well matched across country borders.
I genuinely sympathise here, it is a difficult position. On the other hand, without a clear tendency or bias towards under-ranking or over-ranking individuals, it’s probably easiest to just take grades at face value and let the system itself sort it out over time. The only thing that seems to be strongly worth avoiding is a large net inflation or deflation of the system, and that would have to require a relatively systematic bias when entering new people.
3 Likes
Yeah, I got points for losing just now, never happened under the old system.
When you train a dog it’s important to give it a treat when it does what you want. Rating points is a treat. You play good, you get the treat, you see the number go up.
8 Likes
I know how to fix it.
update rank in profile only if rank increases after win and if rank decreases after lose
ranks still will be real, its just users will (rarely / sometimes) don’t get update of information, so no more complaints
8 Likes
Yes, there are differences between different regions and different time periods. Becoming a pro in Seoul as a male in the 90s was much tougher than becoming a pro in Nagoya as a female in that same period.
Still the handicap between them won’t be that large (on average maybe 1 stone or perhaps 2 stones in more extreme cases). Also, the 7d EGF range still matches quite well with Europeans who became pro in Asia or Europe. Again, the error seems not much more than a stone or so.
Also between European regions there are differences. Yes, a 2005 UK 1d was probably weaker than a 2005 FR 1d. But again, I don’t think the average French 1d could give the average British 1d a 3 stone handicap back then.
Also there are differences of perhaps 1 or 2 stones between individuals of the same rank from one country or even one club.
But overall, the margin of error is seems limited. It works as well as can be reasonably expected with all of this variation.
3 Likes