I was thinking of saying something to that effect too because of the list of options that you can put your rank in has non European and American associations. But I think the comment might be more because
I think this focusing on AGA or EGF is either because OGS could be seen as a ‘western’ server or because those associations have expressed interest in OGS either to use for ranked games in their associations or some other reason maybe. I’m sure @anoek could clear that up if the details are ready to be shared (it could be that some things aren’t to be shared while there are discussions for instance).
I don’t have an issue with the new ratings decisions, I’m just trying to be helpful.
I’m not that closed mind to not understand that these aga and egf data are the ones disponible and even I guess other data will stay more decorative as anything else.
Not really the fault of OGS if they are missing. But still just a jump over, like we can manage it if we can’t have them.
So I stay with my post idea.
Some ranks like EGF ranks seem to be available. Why still ask the rank for those if the ID is provided. Will the rank be updated once covid allows otb tournaments again?
I’d like to note that an update is pending for EGF ratings that will bump the EGF ratings of players by about half a stone for mid dan to a stone for DDK: https://lifein19x19.com/viewtopic.php?f=47&t=17633
Interesting topic. After a full reading, I think I missed the bump you are talking about, where is it mentioned (and by who?) Thanks for helping
I am already aware of this; indeed I am not the plan maker, but I see that resources and time are invested in something which I feel is bit too much a no way street. Just my feedback as I’m asked to adhere to this.
Since 2 months the daily complaints are about that sometimes you lose rank when winning.
I have an idea that this issue could be solved by blocking this information if it’s just a minor temporary adjustment if it’s so necessary that players can’t stand that. Their feeling will be respected as they can’t know if they miss more wins or if the system is a bit cheating them. If as I understood it’s just a small temporary drawback of a good rating system, it should work.
I didn’t see complaints here in the forum in this last months about the difference between rating on different places, I can do myself being a 2k somewhere and a 1d somewhere else, because what I do care is to play here and there players of my strength.
If instead I could feel less shame when showing OGS to new potential members because of all the weird layout problems, or if tournaments organizers may have better tools or… I’m sure I miss other things on the what-to-do list, in my limited view of a lambda ogs user, I feel there will be much more improvement for our awesome server.
In the commission report it is mentioned that we intend to repair rating deflation.
In posts 16 and 17 of the L19 thread, there is some talk about recalculating the rating histories of all players with the new parameters. Such a recalculation will result in changing the ratings and rating histories of all players in the EGD. For most players this will result in a rating increase. If we don’t do that, it would probably take a decade or more to repair the deflation gradually.
Shortly before the 2020 AGM of the EGF we agreed that we propose recalculation to repair the deflation and the AGM had no objections to this.
Before we do, there will be a communication to all the associations to inform their players of this, as was proposed during the AGM.
You can see what the effect of this recalculation on this website: http://goratings.eu/RatingList, which is a recalculated version of the EGD ratings at 2020-03-15 (though we might still modify it slightly in the coming months).
This suprises me. Is nobody helping him to build and maintain OGS software and fix bugs?
Originally he was a two-person volunteer team with matburt. As our site supporter numbers increased, anoek quit his job and became a full-time developer for OGS (at a huge pay sacrifice). Technically, matburt is still a member of the dev team, but he has been very busy recently with the job that actually pays his living wage, and so essentially anoek is working solo.
The front end (everything you see) is open source on GitHub, and so anyone is free to review the code and submit fixes or improvements for approval, but the back end (everything that makes the site work) is totally on anoek at the moment.
I know this is an older post but is there any update on this?
Update on what exactly? It’s implemented and working as intended as far as I know?
@BHydden probably this one
Oh right, I imagine that would want more than 3 months worth of data, at a guess also anoek has a lot on his plate right now trying to implement the new round of ratings updates and AI improvements (hopefully the final ones of both for a while!) So I imagine until that is finished and implemented and proven in production this project will be on hold? (All just my guesses)
Seems someone else has recently done a similar project, though with less coordination with the other organisations. Still an interesting precursor though.
The relative strength of AGA higher dans is unexpected for me. I asked Stefan about it and he said that the sample size of AGA 4d+ was 22 players. So probably there were even fewer AGA 5d+.
On the other hand, the sample size of EGF 4d+ was also only 39, so those numbers may also be somewhat inaccurate.
I just got to 1 Dan on Fox. Started there at 3 kyu, ranked up three times.
Was 4 kyu KGS but havent played there in 1-2 years. About 4 kyu AGA based on rating comparison with players IRL, though I dont play tournaments.
Struggling at 7-6 kyu on OGS. I used to be 5 kyu here but I’ve lost rank, even though I’ve gotten better and have actually ranked up on other servers like Fox. I don’t know what it is about OGS but it really hates me. 6 kyu opponents here feel as strong as Fox 1 dans.
Cool that there is again research into this subject.
Curious as to how this relates to the research projects of 2015 and 2018, as mentioned in: