Chess rating system

A rating system is fraught with dangers, pecularities, unexpected consequences and simple problems.

First, it doesn't reflect quality of play, only results. Thus, we can't compare players from different eras like Kasparov & Fischer. A 2775 today is probably equivalent to a 2675 from Fischer's era, but who can say for certain.

Second, we don't have a real democratic system where everyone plays in one pool. Thus, you can have kids playing scholastic tournaments and getting a rating of say 1700, but holding their own with 2000s in open tournaments. This has had disastrous consequences in America where thousands of adults have stopped playing in tournaments because their ratings have sunk. Many have also ceased being USCF members and that hurts the organization. Similarly, when was the last time a top player played a 2300 or 2400 or 2500? They play others close to their own rating and thus you have various pools of players who are sorted out well.

Another similar thing to isolated pools of players is the changed K factor which means when you're starting your rating can fluctuate rapidly until you get a statistically significant number of games under your belt. This helps you find your place within the group you've been playing in tournaments. But, if you don't play a widely-rated group your own rating won't mean much. If you played a GM group and lost every game you'd be about 2200, but that would still be meaningless. However, once you get to higher ratings the K factor decreases, so ratings move slowly regardless of your results you find it harder to move up.

Another thing to consider, and this is a big one, is that we humans have an understanding of sport and sorting out winners & losers, so that if you were to draw a series of games against a player with a higher rating you would expect, after a handful of games, you were that player's equal. The rating system doesn't reflect that properly. It might take 50 games for your rating to catch up to the other player's rating. How many games would you have to win in a row before you felt you were clearly better than your opponent (in a match say)? The rating system requires more games...a lot more. But, a mathematical system requires about 25 samples before it can expect to reflect your new rating properly. Beating a GM 4 or 5 games in a row won't change an established rating as much as we would expect.

So, ratings are nice, but hardly meaningful.

every one of us has our own interpretation about the rating system, on what its importance in every individual.ratings is good because it is a measure of your ability and progress, but i believed that in every country though we have a single system in rating players the level of competition is not the same, if a players rating for exemple in the philippines is 2500 but the level of competition here is way below than that in the US or Europe, the much lower rank player in the us can perform better compared to the highest rated player in the Philipines.that is my belief,so never the less rating is good to measure up your ability.

Note: rating here are inflated or deflated- they subtract or add too many points for a win or loss

I think theory of Inflation & Deflation works well with the game ratings!! The controlling force is the expertise of the players engaged on chess board!! Its not like egg first or chick first debate! Here game comes first!!

I hate the rating system. I spend a fortnight's sweat and tears on accumulating a measly 50 points, and lose it all in afternoon!

I think what the rating does though, (and it's probably been said a thousand times on this forum already) is to reward consistently good players 

Which, sadly, I'm definitely not!

Do others find that there are certain times of the day when they win, and certain times of the day when they lose?

I find I play much better in the morning .. then it's all downhill from then on.

Oh!! Yeah its like horse ra ce! I remember at times many people in my place lost a lot and gained a lot while they were at stake!! In all games you shall have to some thing at stake either at evening and morning or at dead of night!! I appriciate your honest endeavour.

the queen is worth 9 points, the king is priceless, and the bishop is worth 3.75 points viswanathan

No kidding.  What is it with all the 1200-raters thinking there's only one scale that's ever been utilized to try to represent piece value?  And of course each one is convinced that the scale he happened to see is the absolutely "correct" one.

Want something to spin your little heads around?  Here are the piece values used by the fellow who designed Rybka's postition evaluation function:

Knight = 3.25
Bishop = 3.25
Rook = 5
Queen = 9.75
Bishop pair = additional 0.5

Values of Knights, Bishops, and Rooks are adjusted slightly up and down depending how many pawns are still on the board; i.e., traffic-jammed closed positions diminish the usefulness of the long-range Bishop and Rook and make the Knight more important, and later open positions decrease the value of the Knight which can only cover one portion of the board without expending turns to re-locate.

EDIT: corrected sp error

Yes, you do go on . . . 

Beats me.  Ask the Rybka guy.

Time management is part of the game.  If you calculate faster than your opponent, that should count in your game, and it does.

Do you think you can defeat Kasparov in a game of chess?  How many times do you think you can defeat Kasparov, enough times as to be statistically relevant?  A rating is a statistic.  

pawn-1

knight-3

bishop-3-4

rook-5

queen-9

king-priceless

I agree. I have to because I know I'm going to lose a lot. :)

Latifa wrote: "Chess is a silent music".

No .. it's a dance. The tango.

NO! Its a meditation!!

no no chess if a fight

Chess is a ravishing dance of the mind mediating against the backdrop of social history. 

...Oh My God..!  :)