Idea about the difference of Fide rating and chess.com rating for a player

i just want to know,do you have any idea about,how much a Fide rating differ from chess.com online rating for Avarage. . .e.g . Any1 has the online rating 1800 here ,what should be his probable rating range?

And if any1 has got the FIDE rating 1600,then what should be the scale of his Online and Live standard rating here. . ?

Due to the mathematics involved, they are not comparable. There is not even certainity that given two players, if one's rating is higher than the other on chess.com the same thing is true for FIDE ratings.

Also, remember that only the ELO difference between two players is well-defined, while the ELO itself is not, therefore an arbitrary reference level is chosen (pretty much like potential energy in physics). In other words, increasing all ratings by, say, 10,000 points doesn't change how ELO works at all.

though i have enough idea about potential enery in physcs,but still i am unable to understand your 2nd para pretty clearly

my estimation is that FIDE and live standard might be somewhere in the same ball park but I dont think online correspondance can be judged mainly because people spend different amounts of time analyzing and some dont analyze or use databases at all in online correspondance so peoples ratings vary massivly based on these variables, (based on my limited knowledge)

yes! I also think so

I guess Fide rating would be 200 points lower then chess.com rating. Eg. If Fide 1500 points would have 1700 to 1750 Chess.com live rating.

Since player @ chess.com which we encounter daily basis are hobby player and say around 40 to 45% players are actually chess professionals so according 2 me probability of winning more games increases here as compared to OTB.

+1.  I too feel that correspondence rating is very inflated.  I've only been playing chess for 6 months maybe and my correspondence rating is in the 1400's I think, but my live standard is around 800, which while I still think is too generous, is much closer to reality.

Just given the fact that we know, unfortunately, that there are cheaters online, you can't draw any kind of parallel.

Because among the population of those who seek engine assistance, their online score is going to be light years better, while among others it will be considerably less so.

Unless we were operating in an environment of perfect information where we knew who was cheating and could sift them out of the system, there would be no way to correlate scores meaningfully.

(ETA: this assumes that the prevalance of online cheating via super-strong computer engine is going to be far higher than its OTB equivalent.)

You don't define potential energy, you define the difference in potential energies between status A and B as the (negative) work it takes to get from A to B. Likewise, you don't define ELO, but the difference in ELO score between players A and B as the amount of points required to get from the "statistical skill" of A to the one of B.

This means that ratings can be crazy absurd, for example increased by 10,000 points, as long as the point difference between any pair of players remains constant.

The problem with comparing different rating pools relies in the fact that the ELO formula assumes that players' skill is normally distributed; the peak of that curve, the "average" skill is then defined to be some arbitrary value (chess.com uses 1200, but I think FIDE uses 1600?): from that baseline differences are calculated and ratings are inferred as games are played out.

The problem is that the thing you call 1200 (i.e. the chess.com average) might not be the same the FIDE rating calls 1200. For this reason, ratings are self-consistent but not inter-consistent.

There are other issues regarding finesses such as rating inflation, deflation, handling of inactive players, volatility of ratings, K values, scale factors and whatnot, but that is, in a nutshell, the main problem from a mathematical point of view.

yes! Online correspondence rating is one kinda valueless. . But live standard may be a compared. . Though differences can't be measured i think. . And now got an idea about why.

Baloney. My standard rating on chess.com is between 1300 to 1400. If what you say is true, then I would have a fide rating around 1100 which is unlikey. Chess.com is for enjoyment, nothing to be taken seriously. Cheaters can affect your rating, and this site seems to have a cheating problem. They have a growing list of cheaters being banned.

I think that live standard games here on chess.com are taken seriously by certain people and teams (myself being one of them) the DHLC (dan heisman learning center) for one arranges long live games 45/45 or 90/30 and I think these in-team tournaments are very similiar to real life otb tournaments they recreate them really well. we play long games 2-over 4 hours trying to finish the game with little time left then we analyze with our opponents ect and this is I think very similiar to otb play and can either substitute otb play or prepare you for it, so that is why i think that live standard rating(for those serously taking there time on long games ect) can be somewhere in the same ball park as FIDE or equivilant country's ratings (here in ireland we have an ICU rating(Irish chess union)),

Do people think live standard is a good estimation of FIDE because they also have FIDE ratings?  Or are they just guessing because they think long time controls makes them similar?

Just guessing?  I thought so.

Attow points out why this is silly.  Basically that ratings don't measure skill, they measure your performance relative to others.  i.e. not the individual person, but the gap between that person and others.

Also, just from checking people's profiles now and then, it seems no one really has over an 1800 live online rating.  I see titled players who have 1500 live standard ratings for example.  I see people with 1800-1900 ratings with no losses and they play like houdini...

There are too many variables to make a comparison possible.

That's not how ratings work at all.  Ratings don't measure a person's skill.  It's always results relative to others.  I mean it's not even knowledge or good vs bad moves.  It's just results relative to others in that pool.  As attow points out everyone could be rated in the 10,000s and it doesn't change anything.

Online ratings are comparatively higher for at least two reasons.  The pool is weaker and new players make an account, lose some games, and quit, which injects rating points into the pool and artificially inflates it.  I believe this happens because comparing chess.com's reported users (7 million or something silly like this) vs actual active players.

I have looked at at least 5 acf cases, including my own, and this estimation is almost exactly spot on

The ratings are not comparable at all. Not only it's a different pool of players, but the rating are also calcolated by using a slightly different matematical system (chess.com uses glicko rather than elo). Even more important, the seriousness of the player when playing online if very uneven: whet playing in a real (and often expensive) FIDE tournament you always do your best, but online some people just plays for fun, others take the game seriously. As a consequence the relation between chess.com ratings and OTB ratings is very inconsistent (the same goes for every other website of course).

In my case, i'm a 1870 fide (possibly a bit underrated, since i played just a few tournaments and i'm still discounting an horrendous 1500 performance in my third tournament); my standard live rating is 1700 and my blitz rating howers around 1600, while my online rating is slighly over 2000. 

Junging from my case you would say that live chess has deflated ratings and online chess inflated ones; but some people have totally different stats. I met a few 2100 fide players with a 1900 online rating, and a few 1500 fide with a 2200 rating. Live chess is a bit more consistent but still i know of a 2000+ fide who has a mere 1500 live chess rating and a 1600 fide who is well over 1700 live. It really depends on how much effeort you put in your games, and this makes the comparison between fide and chess.com games rather irrelevant.

@chessph

Above opinion was expessed by DanHeisman in one of Q & A session @ chess.com. ...and I believe it should b correct. ...

And cheating is yes one of d issue but I don't think below 2000 anybody concerned about their rating.

According 2 me its just a number n even if someone cheat and get rating of 2700+ his not gonna get something from chess.çom and at d end of d day ultimate talent of that guy would remain same.....2 what he deserves.

hahaha

FWIW, my USCF rating after my last active tournament was 1618.  I would be rusty in timed tournament play at this point, but my understanding of chess is stronger now.  My ceiling here so far seems to be ~1800.