Please turn on javascript in your browser to play chess.
Only Chess Forum

Only Chess Forum

  1. 24 Apr '07 10:39
    Please help to clarify the following:
    1) If I WIN AFTER LONG BATTLE , WOULD I BE AWARDED FEWER POINTS? HOW MUCH?
    2) iF I WIN AGAINST RATED PLAYERS BETTER THAN ME, WOULD I BE AWARDED GOOG POINTS? HOW MUCH?
    3) NOTHING LIKE ABOVE AND ALL GAMES ARE AWARDED WITH SAME WAITAGE?!
  2. Standard member onyx2006
    onyx2007
    24 Apr '07 10:41
    deja vu
  3. Standard member wormwood
    If Theres Hell Below
    24 Apr '07 10:55
    Originally posted by venkatesh murali
    Please help to clarify the following:
    1) If I WIN AFTER LONG BATTLE , WOULD I BE AWARDED FEWER POINTS? HOW MUCH?
    2) iF I WIN AGAINST RATED PLAYERS BETTER THAN ME, WOULD I BE AWARDED GOOG POINTS? HOW MUCH?
    3) NOTHING LIKE ABOVE AND ALL GAMES ARE AWARDED WITH SAME WAITAGE?!
    FAQ
  4. Standard member onyx2006
    onyx2007
    24 Apr '07 10:58
    FAQ off
  5. 24 Apr '07 17:33
    Chess ratings aren't so much a score as they are a predictor of how well you'll tend to do in any given game, as evidenced by past histories of both players. The rating number is therefore a representative guess as to how strong a player you are in general.

    The methods are fairly standard for any chess microcosm (and contained in the FAQ)
  6. Standard member Phlabibit
    Mystic Meg
    24 Apr '07 17:44
    Originally posted by onyx2006
    FAQ off
    I was about to checkmate a user, and they resigned! Now I got far less points because they quit.
  7. 24 Apr '07 18:18
    Originally posted by Phlabibit
    I was about to checkmate a user, and they resigned! Now I got far less points because they quit.
    Check mate or resignation, even time out, gives the same effect on rating. The method of winning a game, nor the length or material differences has anything to do with the rating calculation.
  8. Standard member Phlabibit
    Mystic Meg
    24 Apr '07 19:23
    Originally posted by FabianFnas
    Check mate or resignation, even time out, gives the same effect on rating. The method of winning a game, nor the length or material differences has anything to do with the rating calculation.
    Yeah, this is a debate I had with a player in help section.

    "Well, it only stands to reason a checkmate would be worth more... Think about it."

    I got a laugh out of that, and the more I thought about it the more I laughed.

    P-
  9. 24 Apr '07 20:16
    Originally posted by Phlabibit
    Yeah, this is a debate I had with a player in help section.

    "Well, it only stands to reason a checkmate would be worth more... Think about it."

    I got a laugh out of that, and the more I thought about it the more I laughed.

    P-
    haha, you are making me laugh now
  10. Standard member Dutch Defense
    Stealer of Souls
    25 Apr '07 00:44 / 2 edits
    The post that was quoted here has been removed
    I fail to see how writing in capital letters is offensive. Maybe it's just you.
  11. 25 Apr '07 00:51
    Originally posted by Dutch Defense
    I fail to see how writing in capital letters is offensive. Maybe it's just you.
    Bold is sooooo last week
  12. 25 Apr '07 01:21
    Just in case it still wasn't clear: The answer is 3.
  13. 25 Apr '07 06:44 / 1 edit
    To provide a bit more depth to the answer, the only factors which affect rating change.

    1) The difference in ratings between you and your opponent (using the top ratings held during thecoiurse of the game, I believe).

    2) The result (win, loss, draw)

    3) How established you and your opponent's ratings are (Unrated, Provisional, Established)

    The theory is as follows. Players are assumed to play at a certain strength of play, although this strength cannot be determined exactly. It must be estimated by looking at past game history, and adjusting this estimate after every game.

    Given 2 players and their ratings, there is a formula for expected result, with a value from 0 to 1. This formula is based entirely on difference in ratings.

    Rating adjustments are made by comparing the actual result to the expected result and shifting the ratings to reflect this new data, from the underachiever to the overachiever. Past game history doesn't directly affect this shift, save for accounting for new and provisional players.

    This makes the process simpler by eliminating extraneous information from consideration, as well as allowing easier adjustments for those cases when a player may genuinely have grown stronger in their play (say after a week long intensive chess class from which they gains a lot of insight).

    Established players don't tend to shift their level of play as significantly as provisional players, however, so the most their score will adjust will be 32 points, whereas provisional and especially unrated players allow for much higher shifts, because the system lacks enough data to firmly and reliably establish their level of play.