Hello, enjoying learning a bit of Go and thank you for this site. When I click on my profile, I can see (25k) next to my name. Scrolling further down to the Game history table of wins/losses, under User there is always a number like (69k, 61k), and this changes per game with an opponent.
Why are there 2 numbers, and what do they indicate? Thank you!
Welcome to OGS!
Hope you enjoy your stay here.
It would help if you posted a screenshot, so that we know what your are talking about.
Attorante is right, a screen shot is always good, but in this case I do see what you’re asking about.
It’s a little bit untidy of us, but this is the situation:
Your rating can be lower - 25k is about 500 glicko points (rough!). Your rating is presently 141 glicko points.
The maths says that 141 glicko points would be 61k, but this makes no practical sense.
Below 25k, the skill level of all players is about the same: beginner.
So we show 25k everywhere, and pair you with other beginners.
Except in that history table, where you can see wins and losses changing the number.
If it were otherwise, you would go for ages without seeing any change: that also creates questions 
Hello and welcome at OGS!
Thank you, this helps!
Also if you’re not familiar with how go ranks work.
I have never seen those ranks before. Like @GreenAsJade said, even though kyu ranks do not go below 30 (25 in case of OGS) it helps to notice any progress while getting acquainted with the game. I started playing on Dragon Go Server (DGS) a few months before I played my first games on OGS so by that time my rank was already somewhat stable at around 20k. Just remember: every skill is the result of many hours of practise, so play, play play. And enjoy!
@anoek is it ui glitch?
Welcome to GO and OGS.
I hope this bad experience doesn’t put you off playing go.
I just want to share some handy resources that helped me at the beginning. May be it will help you too.
And if you want to read theory (it didn’t exist when I was learning)
This has just been posted on the OGS forum, highly recommendable:
It’s not a glitch.
As I mentioned, and Issa_T follow up also underlining, it was put there to answer the question “why does my rank never change, I keep beating the other person”.
It’s “between the devil and the deep blue sea” as they say.
Without that, a person has to look at and understand the graph, instead of the game history, to see any progress, and to know how far they are from “moving up past 25k”.
From your screen shot, we can see that the OP needs something like 10 wins vs a 23k to reach 24k.
That said, I wonder if we might help understand it with a shaded area/line on the graph at 25k 
I’d look into that, only … the graphing code is some serious whoa.
That’s more confusing for me. History shows real number only for user profile, not the opponent. If opponent rank is below 25k it is shown as 25k except in their profile history.
When your rank is below 25k (for example 60k) and opponent is below your rank you point gain is negligible. It doesn’t affect rank. The only hint for this in history table is shading result background.
It is working perfectly. Every tiny changes are noticeable.
I think it is easy to track progress on the graph.
For people who know to look there and know how to read a graph, sure.
Classic case of “it’s fine for me so it should be fine for everyone”. 
I’m not saying “it is not confusing”.
I’m saying “it’s fundamentally confusing that we have a floor rank, and it’s tricky to implement a clean way of showing this, and showing progress, that works for everyone”.
On this point: this has nothing to do with the rank floor.
It is always the case that when you defeat someone well below you in rating the point gain is negligible.
Since 25k is a floor, the only reliable way to increase rank is to play someone better than you: 24k or better.
In games between people of 25k in our system, rank is not a successful predictor of outcome.
Analysis has led to the conclusion that games at this level are effectively random.
That’s why we have the floor, and why it makes sense that 25k defeating 25k makes little difference to rank: it doesn’t tell us anything about skill.
On the previous point, the graph is not as helpful as the table in deciding how many more games you need to get past 25k.
I was able to see at a glance in the OP’s screenshot that they are at 60k equivalent rating, and that a game vs 23k changed this by 4 ranks. Since they need ~40 ranks, they need ~10 games like this.
All approximate, because the exact numbers change as the player’s rating changes, but much easier to see that than look at the graph and find a slope and extrapolate… in fact I had a try, and can’t even see how to achieve it in practice on the graph.
I wonder though was that analysis done assuming that that all of the players involved were 25kyu, when they clearly have a much broader set of ratings which correspond to 30kyu, 40kyu etc. Like you said if that’s a rank floor, it’s a big bucket. If you pooled all of the 10-20k players into a bucket called DDK the results might also look random, if you don’t take into account rating since you’ve now lost rank information.
I thought the main reason was simply that the correlation between rank and handicap stones stopped working well past 25kyu. That is a 25kyu giving 5 stones to a 30kyu (or the appropriate adjusted handicap for 9x9) wouldn’t be an even game.
I’ve heard both arguments - I think they are both valid, but it wasn’t me that did the analysis (which is why I used the passive voice there).
In the thread where that topic came up I recall contributors saying “I can tell the difference between a 30k and a 35k”, so not everyone agrees.
Well I think the rating should still be a okay predictor.
It’s adjusting after each game so that it should become a good predictor.
I think there are a lot of differences between new players and how quickly they grasp some basics, and prioritise moves. So there’s definitely scope to have person A beats person B almost all of the time while still being below 25kyu.
But I think the rank floor being 25kyu is more so because when you go into an automatch game with handicap defaulting to on, and you’re told you get one stone per rank difference, that will be false beyond 25kyu.
I think that’s a road block there to just displaying the ranks below 25kyu, it moves the confusion to another aspect of the code, which probably won’t change without further analysis.
Some relevant quotes, threads
~2017 Glicko2 system
~2020 Glicko2 analysis
Mostly focusing on win-rates with handicap stones.
~2021 updates
You can see what I mean that a lot of the focus is on aligning rank and handicap.
I think this was an update that gave lots of players a bump in rank, so it might’ve redistributed ranks a bit more.
Since then we’ve also had the pick your starting rank update. I’m not sure what the impact of that has been on any of the analysis.
I can see this is a complex topic! Thank you all for the replies, insight and links given, am making my way through the beginner links in this thread.
