FICS have a Surveybot that collects FIDE & ECF ratings of its members, and then produces a formula that converts between them as far as possible. Obviously when two groups of players have overlapping members, some correspondence between ratings is likely - it's not an a priori fact there will be none. Incidentally - my standard rating on FICS predicts my OTB rating very accurately.
Maybe similar statistical work would be interesting here, rather than just pooh-poohing the idea with pure reason?
Originally posted by TommyCYou can surely create a decent approximation, however since they are differnt types of chess!!! it is going to have a much greater rate of failure. For instance I play USCF OTB and USCF CC.
FICS have a Surveybot that collects FIDE & ECF ratings of its members, and then produces a formula that converts between them as far as possible. Obviously when two groups of players have overlapping members, some correspondence between ratings is likely - it's not an a priori fact there will be none. Incidentally - my standard rating on FICS predicts my OTB ...[text shortened]... stical work would be interesting here, rather than just pooh-poohing the idea with pure reason?
Consider player A whom I play a lot because he enters all the same tournies as me.
He is a 1550 OTB player at regular time controls 30 min + by USCF standards.
He is a 1400 quick chess player (at least 10 min game, up to 1 hour, yes there is overlap).
He is an 1700 rated correspondence player.
I also know a kid who is a
1400 regular player
1500 blitz
1100 correspondence.
Finally, I know a third player. He is a
1800 rated OTB regular game player
1770 blitz
1803 corr.
Please tell me how one can possibly make a model that accounts for all three of these with any sort of accuracy??
You make a good point. My case: chesshere 1420, rhp 1560, playchess 1965, ccla 2025, otb 2095 (old…20+ years ago), chessworld 2275. Your call as to which one is most accurate. I don’t play speed chess, but if I did, I’d be rated about 1000…not too quick on my mental feet.🙂
The post that was quoted here has been removedI think the main problem is that this isn't really a normal forum. Most forums are in the format of IPB or Vbulletein. If that was the case then I think this would make the forums alot better. It'll stop the same topics comming up over and over, you'd be clearly able to see staff members, report bad posts, proper search function and loads more which would be beneficial for the site.
Originally posted by zebanoI know it's an approximation, and there will always be some ambiguity and anomalies.
You can surely create a decent approximation, however since they are differnt types of chess!!! it is going to have a much greater rate of failure. For instance I play USCF OTB and USCF CC.
Consider player A whom I play a lot because he enters all the same tournies as me.
He is a 1550 OTB player at regular time controls 30 min + by USCF standards.
He is ...[text shortened]... w one can possibly make a model that accounts for all three of these with any sort of accuracy??
However - my point is, you really don't know what the statistics would say unless you run them to find out.
It's not as if you can't test for levels of correlation.
So why assume you know already the answer before finding it out?
Originally posted by masscatI know what that's like. my stats:
I don’t play speed chess, but if I did, I’d be rated about 1000…not too quick on my mental feet.🙂
1744 rhp
1181 fics blitz
then there's the boy wonder jusuh, who is the polar opposite:
1551 rhp
1990 otb
2050 otb blitz
2300 playchess (if my memory serves me right)
quite many people here 200 pts lower than me are 400 pts higher on fics, so there really is huge variance depending on the type of chess, and the personalities I guess. it's like comparing 100m runners to marathon runners.
RHP around 1600
FICS Blitz around 1500
FICS Standard 2050
UCHESS Blitz around 1600
USCF OTB best approximation 1800 (based only on games against rated players, I haven't been rated)
Yahoo around 1600
pogo 1800-2000 with alot of variance
There's simply no correlation from site to site, nevermind with different forms of chess. Some are rated much higher here than OTB, and some are rated much lower. I agree, this question should somehow get added to a sticky post.