FIDE rating changes: Are they working so far?

The Lowest injection point is a bit hand-wavy because ideally, you'd also have to account that recent new players face players who are relatively closer to the rating floor after the compression technique.

An iterative method is indeed the best method to assess relative strength. On top of having a lag as small as theoretically possible, it can also rate accurately semi-isolated pools against each others.

The issue with URS however, is psychological. Your URS changes spontaneously all the time, even when you don't play, because of factors seemingly independent from your action. The best solution might be to use the insight of URS to introduce inflationary or deflationary measure, different in each federation,reevaluated on a yearly basis. Such measure could be rules surrounding initial rating, rules for rating won/lost against unrated players (you could be awarded points after they get one).

Maybe one simple elegant such measure would be to compute Elo variation of any person based on their Elo relative to the URS of their opponent; or some weighted average of of their Elo and URS, even 2% URS should have Elo converge close to URS in no time while barely being noticeable by any individual player on a singular game.

@Cheshire_the_Maomao said in #12:
> The skills and attention a player is paying to a game is different. Gukesh will not play the championship games with Ding as attentive as he play at some domestic casual show-offs. For the championship, someone may prepare half a year for some decisive tactics that he will absolutely not use on casual games. FIDE should not consider all games equally - and how to rate the "importance" of a regional tournament will be even more difficult too.
>
> A better approach will be using the Glicko or other more advanced system, as they can somehow converge much faster.

The importance of game is equally not accounted for in any current calculation system.

About Glicko, it converges much faster than Elo, but not than URS. URS does not converges, it straight up sits at the best possible strength approximation given informations at hand.

@Cheshire_the_Maomao said in #10:
> The problems is that the rated games distribution is by nature NOT normal distribution. People who is likely to win a match will be more likely to play it, and vice versa. Especially when the ticket fee and rewards are both high these days....
>
> Elo will not work for such systems I fear.
>
> Better game arrange system or continuing to Glicko probably will work...

This study is about players, not games. It's irrelevant that good players play more. Also, the rating budget is on a per player basis (at least in conservative system like elo), not on a game basis.

A better argument might be that good players play longer (started earlier, retire later), but URS shows that it doesn't matter that much. As far as I understand, URS does not try to fit people onto a bell curve, it just so happen to do it while trying to give a rating with the best predictive powers on recent games.

"without a birth year"

I wrote beginning of this year an article about identityfraude see schaken-brabo.blogspot.com/2024/03/identiteitsfraude.html

Fide is not serious about cheating when something basic like a birth year is not mandatory on somebody's profile.

I will try to summarily answer questions posed in the thread, without delving too deeply:

1) Is URS still maintained?
Yes, the lists are still updated, but not in a very timely fashion. Sometimes they will batch update 2-3 months instead of releasing monthly lists.

2) What is happening to adult players rated 1400-1450?
Most of them are dropping out of the list. In fact between June and July we had a net decrease in the number of FIDE rated players (classical only)

3) What is happening to adult players 2000-2050, are they getting closer to a title?
This is a neat question that will probably be part of my next investigation, thanks for providing it!

If I missed anything, feel free to tag me in the thread. Alternatively, for following all of my updates, (short form and long form included), the best place to interact is probably on Twitter: x.com/vlad_chess. Feel free to follow, but only if the content is interesting to you.

@mvhk said in #15:
> How does the URS handle titles? It sounds strange to wait for a computer to do some complicated calculations to know if you obtained a title or not. Players like to know in advance what still is needed or how do you see that? I couldn't find anything in the faq about that.

URS is not a chess federation so it does not handle titles ==> not needed in FAQ. If similar system would be used I doubt anyone would care whether it takes a month or not. There are lots of delays anyway. One has to apply via national organisation and the either org or player needs to pay fee and then at some point you get the title. For IM and GM more often than not player achieves top Elo required way earlier than the norms so there would not be any delay.

@f4xel said in #21:
> The issue with URS however, is psychological. Your URS changes spontaneously all the time, even when you don't play, because of factors seemingly independent from your action.

I think the problem is not the changing, it is the delusion that our individual rating represents something about our ID that is some kind of chess skill linear ladder of "strength" (which is also something kind of social but not human social, more like wolf social, or territory constrained chimpanzee social, scarcity of resources (or percieved as such).

if it were some kind of learning measure or quality of play measure, it would fluctuate and change anyway even dally, or upon alea of the human individual physiology and variable immune system energy consumption (and all the ecosystems within dynamics, all possibly affecting mind resources).

That illusion of social ladder competion might be the problem to root out. ("the enemy within..." lol). yep I might be crazy.

> It’s so beautiful, that sometimes I wonder why it hasn’t been met with widespread adoption among chess players. Old habits die hard, I suppose...

I for one simply didn't know about it! The charts are gorgeous, I'm switching to URS completely for my OTB preparation.

@Vlad_G92 question about URS: how are rating differences in this system interpreted?
With Elo, 200 point advantage translates into EV of 0.75 in a single game. Is it the same/similar with URS?

Dear @Vlad_G92 ,

This is a very nice and instructive article, that is easy to follow and confirms a lot of things everyone believed, but few had actual proof for. Good work!!

I have just one issue with it
At the very end, you advertise your coaching, and you do it via "(geared for adults rated under 1800 FIDE or equivalent)"
Surely that goes against everything you taught us in the article prior!!
It should be something like "(geared for adults from Germany rated under 1800 FIDE, or for adults from India rated under 1500 FIDE)"!
Please fix

Thank you