Quote:
Originally Posted by ScottTK
Interesting stuff.
Just out of curiosity, for those of you guys rated UFCF/FIDE, how does it compare? (vs your online rating, chess.com or otherwise)
For me personally it lined up very well. My chess.com ratings of 1419 blitz, 1518 standard, 1333 bullet, were all within a couple lines of each other, and my 1806 online rating was just a little out of sync with them. Overall those four ratings, on his table, suggest a USCF rating of ~1550. I'm actually only 1447, but my recent sample size is small, and over my last few tournaments the rating has been trending upwards, so I'm pretty sure I'm still rated a bit below my "true strength" right now.
So it's certainly not a perfect measure (because everything Sholar said is dead on). But it's not a terrible place to start for some rough guesstimation.
Quote:
Originally Posted by Sholar
One important source of variance here is how seriously *you* play online. The more you do, the more these calculators will overestimate your strength, since pretty much everyone plays FIDE games seriously, but many people just coffeehouse online.
(Also, linear regression is absolutely not the right way to model this...)
In addition to all of the above, which is 100% right, it also matters how MUCH you play online. I suspect *most* players have a much larger sample of recent online ratings than they do of recent OTB games. So if someone is in the midst of a phase where their "true strength" is rapidly changing (usually this will be an upward trend, particularly for younger players) there's likely to be a lot more lag in their OTB rating (where you can easily stay underrated for a long time, if you don't play too many tournaments) than their online ratings, which should normalize pretty rapidly if, like many people, you play a lot of online chess.