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

Only Chess Forum

  1. Subscriber no1marauder
    Humble and Kind
    12 Sep '08 01:59
    I've been wondering about some of the high OTB ratings claimed by British players here in comparison to their American counterparts. I finally found a site giving distribution figures of Brit ratings here:

    http://www.bluehorizonweb.com/blog/2008/03/grading/ecf-grading-statistical-distribution/

    The summary is:

    25th percentile - 82 (1660 Elo*)
    50th percentile - 110 (1800 Elo)
    75th percentile - 140 (1950 Elo)
    90th percentile - 168 (2090 Elo)
    95th percentile - 185 (2175 Elo)
    99th percentile - 228 (2424 Elo)


    Now I compared this to the last data available from the USCF here:http://www.uschess.org/ratings/ratedist.html

    Focusing on the column for non-scholastic members (scholastic includes many children who play few games) the differences are striking. A 25th percentile in the USCF would be about 1150 over 500 points below the Brit ELO. The breakdown in the same way as the above would be (approximate):

    25th percentile: 1150
    50th percentile: 1475
    75th percentile: 1750
    90th percentile: 1990
    95th percentile: 2115
    99th percentile: 2365

    Though the numbers get closer at the top, it seems that a USCF rating of people at similar percentiles is hundreds of points below what the same level of achievement would get you in an ECF rating at the below Expert level. When comparing USCF OTB ratings to Brit ratings this should be kept in mind.

    Comments?
  2. 12 Sep '08 02:22
    I'm confused.Your rating is calculated by the rating of your opponents and the results you scored against them.Right?Then what are these charts for?

    Also,how does this distribution work?Does the USCF,and their counterparts,simply get to choose what rating they attribute to a certain percentage scoring?
  3. Standard member wormwood
    If Theres Hell Below
    12 Sep '08 02:43 / 1 edit
    finnish national SELO ratings:

    25th 1391
    50th 1565
    75th 1778
    90th 2004
    95th 2153
    99th 2419


    looks closer to ECF than USCF on the lower levels... I've noticed the big number of low rated american players on CTS as well, I don't think what's visible here is just a coincidence.

    I'm guessing it has to do with national chess culture. there's no school chess here, but as I understand in u.s. it's quite common. so it would make sense to me that there is a lot of beginning players in u.s. who stop the second there's no chess in school.
  4. Subscriber no1marauder
    Humble and Kind
    12 Sep '08 12:29
    Originally posted by Katastroof
    I'm confused.Your rating is calculated by the rating of your opponents and the results you scored against them.Right?Then what are these charts for?

    Also,how does this distribution work?Does the USCF,and their counterparts,simply get to choose what rating they attribute to a certain percentage scoring?
    The charts show the number of players at each rating level based on their performance.

    I have no idea what your last paragraph means.
  5. Subscriber no1marauder
    Humble and Kind
    12 Sep '08 12:37
    Originally posted by wormwood
    finnish national SELO ratings:

    25th 1391
    50th 1565
    75th 1778
    90th 2004
    95th 2153
    99th 2419


    looks closer to ECF than USCF on the lower levels... I've noticed the big number of low rated american players on CTS as well, I don't think what's visible here is just a coincidence.

    I'm guessing it has to do with national chess culture. there's no sc ...[text shortened]... ere is a lot of beginning players in u.s. who stop the second there's no chess in school.
    I used the non-scholastic figures which obviously omits beginning players in school.

    It seems clear that the differences in formulas for the rating systems lead to highly different results. These are closed systems; it seems highly unlikely that the average Brit tournament player is truly much stronger than the average American one. Therefore, the numbers should be adjusted accordingly when you attempt to compare an American player's rating to a Brit one.
  6. Standard member wormwood
    If Theres Hell Below
    12 Sep '08 12:42
    Originally posted by no1marauder
    I used the non-scholastic figures which obviously omits beginning players in school.

    It seems clear that the differences in formulas for the rating systems lead to highly different results. These are closed systems; it seems highly unlikely that the average Brit tournament player is truly much stronger than the average American one. Ther ...[text shortened]... be adjusted accordingly when you attempt to compare an American player's rating to a Brit one.
    the recent ECF adjustment might also be the cause. in the recent thread about it many of the lower rated ECF players seemed to feel that their new ratings are undeservedly high.
  7. Subscriber no1marauder
    Humble and Kind
    12 Sep '08 12:56
    Originally posted by wormwood
    the recent ECF adjustment might also be the cause. in the recent thread about it many of the lower rated ECF players seemed to feel that their new ratings are undeservedly high.
    This is the post by Squelchbelch a while back that got me thinking about this:

    Nonsense!

    1500 is below average club player:
    Basic opening knowledge, spotting routine tactics for both sides, some prophylaxis, some strategy, basic endgame knowledge, sound thought process etc...

    2000+ (OTB or CC) is like a different language. Congrats - you're in the top 0.5 % of the chess playing population:
    Strong club player - advanced combinative/positional play, extensive opening repertoire, depth of knowledge of middlegame strategy & endgames...

    http://www.timeforchess.com/board/showthread.php?threadid=89500&page=4


    I thought the comment about 1500s was nonsense based on what I know, the 1500s I've played and the USCF rating distribution chart. But Squelchy was certainly right if you're discussing a Brit 1500; he'd be in the lowest 10% or so of rated players. It seems we were comparing apples and oranges.
  8. 12 Sep '08 13:08
    Originally posted by no1marauder
    The charts show the number of players at each rating level based on their performance.

    I have no idea what your last paragraph means.
    Ah,think I get it now.I thought the chart was used to determine ratings.
  9. Standard member wormwood
    If Theres Hell Below
    12 Sep '08 13:35
    Originally posted by no1marauder
    This is the post by Squelchbelch a while back that got me thinking about this:

    Nonsense!

    1500 is below average club player:
    Basic opening knowledge, spotting routine tactics for both sides, some prophylaxis, some strategy, basic endgame knowledge, sound thought process etc...

    2000+ (OTB or CC) is like a different language. C ...[text shortened]... be in the lowest 10% or so of rated players. It seems we were comparing apples and oranges.
    oh, right. that might explain it. -although his explanation of the skills of a 1500 also fit my CC perfectly. I haven't seen any kind of categorical differences so far, there's no "different language", my openings are not very good, endings are crap, middlegame strategy ad hoc, and practically all of my moves are based on simple 2-move 'analysis'. most of my errors are still due to simple evaluation errors after a single move. not usually dropping stuff, but underestimating simple moves which create big problems.

    then again, maybe that's also why I haven't been able to break 2000, and generally get stomped by the higher rated players.