IS 3000 RATING POSSIBLE IN CHESS.COM?

If you have been following the list of top players in chess.com, you may have noticed a continuous climb in their ratings. I was told that just a year ago, a rating of 2400 would be enough to land on page 1 of the players list (good for a top 50 ranking). Now, that will only be enough to land on page 4 (for 151st-175thth place).

Two of our players have surpassed 2800 so far (Pelger and at one brief time, Gonnosuke). Do you believe we are going to see people breaking 2900 and maybe 3000 on this site? If so, what do you think are the factors that will make this possible?

Although chess.com’s rating system perhaps is somewhat different from that of FIDE‘s, it will be interesting  to note that so far, only 4 players in the history of chess have exceeded 2800 namely Kasparov, Kramnik, Topalov and Anand with Kasparov achieving the highest rating of 2851.

Chessnetwork has a rating of 4000 on live quick chess.

The reason why we have people achieving over 2700 so easily is that they're outliers in the distribution of chess.com members. We don't have enough people around their playing strength so the rating is proportionally higher than if they were in a population with the strength that FIDE has.

 Could it be that the top players do not play against each other often enough which if they did would result in more draws and therefore bog down their ratings?

I think all the rating are inflated here due to lots of new players who start 10 games then leave the site forever.

their rating will go down and they gift all the other player rating points.

Fide doesn't suffer this problem because when a rated player stop play it doesn't gift points to the other player.

Here's the thing. Comparing ratings in one closed system (Like Fide) to another (like Chess.com) is hardly ever a good idea. A more important factor would be the difference between the ratings of the two players. Don't know if it's still the case, but used to be that a 400 difference in Fide should virtually give a hundred percent success rate for the stronger player. In another closed system that may or may not be the same, depending on the formula they use.

Is 3000 rating possible in chess.com?

Yes, I think that's attainable given that the rating system in this site are much less stringent (where friendly, fun and sometimes crazy games are included in the ratings). 

In my approximation, turn-based chess ratings should be AT LEAST 500 higher than OTB because of the luxury of time. This means that a player with USCF or FIDE rating of 2000 should have at least 2500 rating in turn-based chess. If we factor in the less stringent rating system here, the same player could even reach 2800-3000 in chess.com! (Nice numbers but utterly meaningless, at least for me).

On-line Chess ratings aren't as reliable as FIDE ones. Obviously a FIDE rated 2000 player will be much stronger and play much better OTB than a 2000 rated on-line player. Also, OTB games introduce things that on-line chess doesn't have; such as psychology.

I think that the true measure of a player's strength will come from his OTB, FIDE rated games, not from chess.com.

Yeah exactly, a rating of 3000 on chess.com means nothing in terms of fide ratings, and you cannot compare across them. The small distribution of players, as pointed out by someone, is a prime reasons for the large ratings you see today. And btw, chessnetwork is a computer programme. I totally agree with Daniel3.

I agree 100% with your post, Daniel3!

Yes of course chess.com ratings are inflated, and not nearly as representative as OTB ratings for many reasons. For one, I fairly often play a move after a couple of seconds of thought on here, whereas if I was playing a 1 1/2 hr each OTB game, I'd obviously play within those time limits.

Put another way, as good as the top players on this site are, even if one or more players got to a 4000 rating, put any of them up against Kasparov or Topalov and I know who I'd be putting my money on.

 I don't understand why  turn-based chess ratings should be at least 500 points higher than OTB because of the luxury of time, both players have the same luxury.

One thing you have to keep in mind is that in OTB games you usually play opponents who are roughly the same rating.  In online chess, that's usually not the case so in many cases a rating is the result of playing lots of weaker opponents.

Also, in online games there are all sorts of tricks that players do to ensure that they maintain a high rating.  For example, Pelger plays all games at 14 day intervals and postpones his negative results as long as possible.

In most cases a stronger player makes better use of time to create an idea, devise a plan, or formulate a strategy (Ex. 2500 -vs- 1800). A weaker player can have all the time in the world but because of his relatively limited understanding of the game he's not as capable of making efficient use of it.

I recall Kasparov wrote in his book 'Life Imitates Chess' that if a grandmaster is given more time his rating should increase. (I don't recall exactly how much extra time and how much increase in rating Kasparov mentioned in that book). I think his reasoning is the same as what I wrote above. By the way, the 500 points increase over OTB I mentioned in my previous post is just my rough estimate.

Thanks for your post, Ed.

Saying Chess.com ratings are inflated compared to FIDE ratings is like saying football scores are inflated compared to soccer scores.

 This isn't true. Soccer and football are different games. Here we are talking about the same game being played differently.

Hi guys,

I think Chess.com should rework their rating system to bring these numbers down.  Every other day someone starts a new topic about how Chess.com's ratings do not match up with other ratings systems.  They (Chess.com) have mathmaticians available that could do this easily (and fairly).  Of course, everyone will belly-ache, but I think it would be best in the long run.

 Soccer and football are the same game: get the ball to the other end of the field. In football you can use your hands (opening books/databases), and there are more breaks in the game (longer time controls). Maybe it isn't a perfect analogy, but the point remains: OTB and correspondence are different enough that trying to compare the ratings between the two (especially with different populations) is meaningless.

 Your Avg. Opponent Rating is what I think is your real rating, at least after a 100 games or more, and for us having a rating between 1200 to 2000, which is what most of us on chess.com has.

Who cares about the ratings they all were created to give comparision among players and thus creating many a negative things like jealousy, egocenteredness, false convictions and diseases of addictions to net and chess  and others which results in false personality and which has bad impacts on our relationships.