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

Only Chess Forum

  1. 06 Jun '08 11:11
    Do you find that ratings are accurate on this site?
  2. 06 Jun '08 11:22
    Originally posted by moteutsch
    Do you find that ratings are accurate on this site?
    As a rough estimate, I would say that 1/3 are too high, 1/3 are too low and the rest are about right.
  3. Subscriber AttilaTheHorn
    Erro Ergo Sum
    06 Jun '08 11:29
    Originally posted by David Tebb
    As a rough estimate, I would say that 1/3 are too high, 1/3 are too low and the rest are about right.
    I would tend to agree with that assessment.
  4. Subscriber Ponderable On Vacation
    chemist
    06 Jun '08 11:37
    Originally posted by moteutsch
    Do you find that ratings are accurate on this site?
    Actually there is no such thing as accurate rating.

    But I agree that you can get a rough estimate of a players strength after their rating settled.

    Of course in looking at a lot of players ratings you see artifacts.
  5. 06 Jun '08 12:28
    Surely the rating is always accurate and the question is: - does it necessarily represent the players true strength?Aand the answer to this is no.

    I say no because people can spend more or less time on their moves and this has a considerable baring on results. For example I think there are quite a few sub 1500 players with big game loads who could easily earn higher ratings if they took more time and so on.

    There are many anomalies that come about for a variety of reasons.

    Overall I'd say the rating is a reasonable guide for players who have a consistent method of play and that over the course of time it can serve as a useful measure of personal improvement.
  6. 06 Jun '08 13:04
    Originally posted by Ponderable
    Actually there is no such thing as accurate rating.

    But I agree that you can get a rough estimate of a players strength after their rating settled.

    Of course in looking at a lot of players ratings you see artifacts.
    I meant in comparison to real-world ratings...
  7. 06 Jun '08 13:13 / 1 edit
    Originally posted by moteutsch
    Do you find that ratings are accurate on this site?
    Ratings are only tool to compare your progress within that certain group. So outside of this website your rating of course is not accurate, because it is a different group, but within this website they are accurate. Remember after all, ratings where invented as a promotional tool and to that effect it still is. I'm not a fan of the rating system and think they should just get rid of it.
  8. 06 Jun '08 13:46
    Originally posted by Restless Soul
    Ratings are only tool to compare your progress within that certain group. So outside of this website your rating of course is not accurate, because it is a different group, but within this website they are accurate. Remember after all, ratings where invented as a promotional tool and to that effect it still is. I'm not a fan of the rating system and think they should just get rid of it.
    If they are taken away you wouldn't have an idea how well you're doing; since also serves as a goal to reach so people would improve their game
  9. Standard member wormwood
    If Theres Hell Below
    06 Jun '08 13:54
    Originally posted by moteutsch
    Do you find that ratings are accurate on this site?
    elo rating system:

    http://en.wikipedia.org/wiki/Elo_rating_system
  10. 06 Jun '08 14:14
    Originally posted by moteutsch
    I meant in comparison to real-world ratings...
    OK then in that case the answer is yes for some people (their rhp rating reasonably accurately reflects their rating in another group) and then no for others. There are so many variables that there can be no consistent comparison for everyone.

    There may well be more consistency the higher up the food chain you go but for a more accurate comparison you would require a player to be equally active in both arenas and even then there are many other elements that can skew the result.

    I'd say my rating on here is a much more accurate reflection of my playing strength than my ECF rating because my RHP grade is always up to date.

    My ECF grade is based on a handful of games I played in the 05/06 & 06 /07 club league seasons....and I've played a lot and studied a lot in the time since then. Even those OTB games were skewed by being out graded in most of the games and also from being beaten by strong ungraded players (joining OTB from internet chess!) thus placing further downward pressure on the ECF grade. Further more the ECF (English Chess Federation formally BCF British Chess federation but had to change their name because they refused to share the funding with Scotland!)...now has two separate conversion codes to create an ELO depending on what you're grade is. Very confusing...
  11. Standard member varoadstter
    woodpusher
    06 Jun '08 14:31
    Originally posted by moteutsch
    I meant in comparison to real-world ratings...
    What constitutes "real-world ratings"? There are numerous rating systems in use by various entities. Some of them use calculations that are close enough that the ratings would corrrelate fairly closely and some do not. Since every entity has their own take on this, it's really not particularly useful to try to compare them, however.

    One analogy: If I tell you it's 32 degrees outside you may think that's cold (if you think in degrees Farenheit) or you may say that it's hot (if you thing in degrees Celcius). The numbers don't match but they both move up and down in the same way when conditions change.

    At the end of the day, it depends on what you are comparing RHP ratings to to know whether or not there is a good correlation.
  12. 06 Jun '08 14:42
    However I think it's significant that this is a fairly large community. What I don't know about is how the playing strength of the community affects our ratings. If we had a large number of stronger players join the site would this have a downward pressure on everyones rating....if all the top players departed would all the ratings shoot up...maybe a mathematician or statistician could throw some light on this.
  13. 06 Jun '08 14:56
    Originally posted by moteutsch
    Do you find that ratings are accurate on this site?
    The way one approaches the game is different in CC vs OTB, OTB rapid chess, vs blitz on ICC. This makes it tough to measure "overall chess ability", no matter the rating algorithm/method used. You could say that one's chess ability is the relative ability to beat others... but then the question becomes, at what game? Fast/slow? open book take home exam (like CC) or closed book timed exam like OTB? When I was on the ICC, my rating difference across the various time controls was +/- 300 pts, and furthermore the Top 10 rating lists across the different time controls was different as well, so it wasn't just me. Same pool of players, different game.

    So as already mentioned, a person's rating here only indicates how good they are at CC, period (even then it only applies to RHP, not CC in general). Move to a different playing environment and that changes. time outs in CC is also an issue too but i won't get into that here.

    In my humble opinion, I think OTB with standard time controls, say 90+ minutes per side, is real chess. Everything else is a variation of it. My reasoning is that only in OTB are the conditions equal. Here in CC, you have some player's who take advantage of books & databases (which is ok), while others choose not to for whatever reason. In internet blitz, dexterity (a non-chess skill requirement) is a factor so slow non-computer savvy people are disadvantaged.
  14. 06 Jun '08 14:57
    RHP is an isolated island in the ocean of chess around of the world. In this island we have our own rating system as isolated as the island.

    When we begin to exchange rating numbers with the ELO players, national ratings etc, then we are not isolated anymore. But it takes many games to equal out the difference between the systems to achieve an equibrilium. After that, not before, are the rating comparable betwen RHP and the rest of the chess world.

    But we are an isolated island, and I think we should continue with this.

    Are we suffering from inflation (or deflation) of ratings? Can we trust that a 2100 player today is as good as an 2100 player from 5 years ago? I don't think so. We can only compare players within the same time. In order to have an absolute value that will be the same as time goes we have to invoke a inflatory compensating rating points.

    I say we should have an average of all active players at 1200 points. If we don't have this average we should compensate with inforced rating point loss, or win to regain the average of 1200 points.

    This is an idea to stop inflation of the ratings, but I can live witohut it too. The effects without a 1200 average rating is that the rating will flow with time, upwards or downwoards and further defferentiate our rating compared with the outside world.
  15. Standard member slappy115
    Slappy slap slap
    06 Jun '08 14:59
    Originally posted by AcapaYespada
    If they are taken away you wouldn't have an idea how well you're doing; since also serves as a goal to reach so people would improve their game
    I would look at my win/lose ratio if there wasn't any ratings.