Chess.com Online rating system

Hi everyone,

I'm going to enter a FIDE rated tournament next month, and I am just curious to know if chess.com follows any OTB rating system like FIDE, USCFor BCF?   I have noticed that different sites have different rating systems all though I play the same way in all the sites.

Thanks!

i think this thread is what you are looking for:

http://www.chess.com/forum/view/general/chess-rating-system

specifically the 3rd post.

 Thanks!

online and otb ratings are as you appreciate different animals.

if you are ungraded fide, uscf, bcf, when entering otb tournaments its usual to provide information on any game history with your application. I'm not sure what weighting the tourney controllers would apply to online chess but I think its best to mention so they are aware.

:)

 I agree! It seems that chess.com is following the Glicko rating system!

http://math.bu.edu/people/mg/glicko/glicko.doc/glicko.html

yes thanks, I wonder what Glickman thinks about the application of rd at chess.com. One player reached the top of the online standard tree with 2800 rating after only twenty games. 

 Yes! And in the normal FIDE or ELO rating, even a great GM is only 2700!

Chess.com grades are not close to OTB grades - keep in mind that chess.com is a different thing.  On here you get to study a position for days, can refer to books or databases etc... in OTB you can't.  Higher grades in online chess are often those who are the most patient, not necesarily the strongest OTB players.  For that reason the online grades should never influence how you are positioned in an OTB event imo.

I wonder if Chess.com would ever revise it's rating system, which seems to be inflating fairly rapidly, if I am interpreting this observation correctly:  My rating has increased from 1800 or so to 2014 in the last six months but my percentile ranking has remained in the 95% neighborhood....seems to indicate that EVERYBODY is winning.....?

One possible reason for a raise in everyone's ratings is the timeout floor. On their last day on the site many leave unfinished games.

After five consecutive timeouts a timeout floor is established. This means more points are put into the system than if all remaining games were resigned.

It matters not though, these extra points are redistributed slowly to everyone in the pool. In the future you still can estimate the strength of your opponents whether you are rated 5026 and they 6076 at chess.com. If you want to try to compare your chess.com rating to FIDE or USCF or ..., just convert with whatever formula applies at that time.

As for your ranking consider at the higher levels there are fewer players to overtake so it is reasonable your ranking does not change quickly even if you did increase your chess strength.

http://www.chess.com/echess/players.html

Do you know of a system that doesn't have inflation? Because I've never seen one. Even ELO inflates. The fact of the matter is that one person's rating, all by itself is meaningless, in any system. It's not like a chess IQ score. The only meaning a rating has is in relation to the other players being measured by that same system. So, as long as everyone rating is inflating at roughly the same speed, the inflation doesn't mean much.

 This is true, but as TadDude must have presumed, I was hoping to correlate the chess.com rating, roughly, with a USCF rating. If one inflates more rapidly than the other, than this becomes more difficult to do. But it does not matter in any essential way...if one enjoys the site, and I do, and enjoys the game, and I do, then it's all good. Peace, thanks for the replies.

There are lots of issues beyond variable inflation rates that will interfere with any attempt to correlate USCF and chess.com ratings: differing rating pools, differing skill sets, differing play dynamics. I would actually expect inflation rate to be the least problematic, but your post makes me wonder. I will put redoing my chess.com rating quantiles on my to do list, which might give some idea of what's going on with the inflation. Of course, it could also be affected by new players entering the system. Anyway, I'm heading out of town this weekend, so it may be a while before I get to it.

Its back again to the mutual relations argument. Apart from the fact that chess.com updates ratings after each game whilst official governing bodies publish less frequently, there's absolutely no reason why the grades from the various dot.com sites or ogb should not be comparable. But regrettably as we all know in some cases they aren't.

>:)

 There are plenty of reasons whey they should not be comparable. I stated several of them in the post you quoted.

The points made are not reasons, simply excuses. Wake up and smell the coffee.

 The points made are reasons, based on the underlying mathematics of the rating systems and the reality of the situation. If you don't like it, take some math classes and come up with a better system that can, among other things, make apples and oranges comparable. Because that's what you're insisting should be possible.

The excuses you are highlighting as justification for saying there's no correlation are entirely anecdotal and lack any science.

 They're based on better than science. They're based on mathematics.

Would you care to reference this science or should I say mathematical research which has given conclusion.