There is a forum that answers Illyra's question and I couldn't find it, but I have read it. I want someone else to find it for me so I don't have to search through 150 forums with the word "Rating" in them...
The best place to send him is probably the FAQ section on how ratings work, that can be found here: http://support.chess.com/index.php?_m=knowledgebase&_a=viewarticle&kbarticleid=21
(in this topic http://www.chess.com/forum/view/help-support/ratings4)
If you really want an interesting article on the rating system read the article about Mark Glickman in the 2006 October Issue of Chess Life. It does an excellent job in explaining the system (Glicko and Glicko 2). It gives you the essential formula used for calculating ratings. The equation looks very simple. However, when you start to get into refinements, etc, you see how difficult creating a rating system really is. The article has some good information on the development, from a not to difficult mathematical standpoint.
It's not how well you play the game, it's if you win or lose!
someone can explain about blitz games? and long games?
Yeap, it's the case for me. I'm good at short games, but I always lose in long games.
i mean the diference between them
sorry viswanathan, but a queen is actually worth 9, (not ten) pawns.
turtle, the general points system followed is as follows:
pawn - 1pt.
knight/bishop - 3pts.
rook - 5pts.
queen - 10pts.
of course points are not everything... the position of your piece also matters.. for example you might not mind losing a bishop or rook to save a pawn on the 7th row.. and points dont have any bearing on the game result.. it is just a basic framework to help beginners understand the value of different pieces
Also, the king is worth the whole game
Answering my own question, it looks like ratings change based on where you stand at the moment the game ends. That seems right from what I've observed on here.
So when you first sign on for a game, let's say the Vegas odds line reads "Win: +42 Lose: - 112." Then, during the course of play, another game of yours ends and you win it. So now your rating is higher. So if you then win the game that promised you 42 points, you'll now only get like +15 because your rating improved during the game, altering the original contract.
( ? )
(Just a little confusion I had over how correspondance chess is rated.)
Something I note about ratings that doesn't seem right: you can get an inflated rating by beating lesser players without challenging yourself to play better ones.
I was rated 1380 at the start of the Ruy tournament, so I was put in the 1200-1400 bracket. Being 2nd-seeded in my pool, I've pulled a bunch of wins against sub-1300 players. My rating has gone to 1480. Not that I mind, but really, beating 6 straight 1200ish players does not make me better than a 1450 player, because I've consistently struggled to beat those 1450 players lately. Seems like a gap in the system. Not sure how to correct for it.
I do challenge myself to play better players, but during the tournament I've got this batch of 1200ish players, so its throwing my rating off. Personally, my goal is for my rating to be less than the average rating of my opponents - that's how I'll know when my rating is meaningful. Because if my rating is higher than the average rating of my opponents, that means I haven't been taking enough challenges. Perhaps there's a way to factor average opponent rating into your rating. Because if you keep getting +10 when you beat a 1200 player, could you get to 2000 that way?
I really dont understand how the rating system came together either
Apart from the standard Elo system, has there been any other systems tried to give players a rating system?
It seems the system will self-correct because of human nature -- most people will want to improve their game and not simply languish playing lower rated players exclusively and continuing to wrack up points for their own sake. If I understand correctly, if someone follows their best instincts to always improve and play equal or higher rated players, some of them would defeat you and your own rating would level out to a rating that more accurately reflects your skills. The system would seem to only "fail" people who fail themselves by not necessarily wanting to play better chess or improve, but only really wanting to play that sad, machismo-ladened game of "my rating is bigger than your rating" -- also part of human nature, but... hopefully not prevalant.
i have a question. If you download a PGN-file can you the replay it on your chess-simulator (for example Fritz) and in that case how? Please help!!..
I would like Chess.com to consider two suggestions
1. Chess.com says that players rating consists of two component - Rating itself and a varaible that says how accurate is the rating. Then why we see only the rating for the players? We also should be able to see the accuracy of the rating. Isn't that possible?
2. In case of timeouts player should have option to start another game from the same position when timeout happened(if both the player agree) as timeout can happen for any cases like network issues, the players computer got malfunctioned, the player got ill or he went for vacation and could not come back on time for any reason (like he or his family member got ill, he could not catch his flight/train on time) and his vacation time expired.
Like they said before. Of course they have to start you some where, and it won't be 0.
lol mznor, everyone starts at 1200: average elo rating...
i still don't get the varible thingi with elo ratings swinging aggresively...
elo rating of 1200 when it is ur first time