Please turn on javascript in your browser to play chess.
Site Ideas Forum

Site Ideas Forum

  1. Standard member flexmore
    Quack Quack Quack !
    07 Feb '05 11:59 / 1 edit
    Game 810605 earned 0 rating points for Ironman31.
    Deepfault was rated 1740.
    you may say ironman took no risk .... but:

    ... Game 502661
    hungtn was rated a measly p1702.

    the ratings should not be an integer, they should be stored as a number with 1 decimal place accuracy, and ironman should gain a portion of a point from his win against Deepfault.
    Ironman should only gain a little, perhaps only 0.1, but he should gain something.

    (the obvious flipside is that ironman does not deserve the full 1 point he gets from beating an 1800 player ... he should only get part of a point, perhaps 0.5 )
  2. 08 Feb '05 01:15
    Originally posted by flexmore
    Game 810605 earned 0 rating points for Ironman31.
    Deepfault was rated 1740.
    you may say ironman took no risk .... but:

    ... Game 502661
    hungtn was rated a measly p1702.

    the ratings should not be an integer, they should be stored as a number with 1 decimal place accuracy, and ironman should gain a portion of a point from his win against ...[text shortened]... point he gets from beating an 1800 player ... he should only get part of a point, perhaps 0.5 )
    Anybody who plays someone rated 700 points lower than himself takes the same risk--it's not just Ironman. I think it's fair because it discourages people from gathering easy points against much weaker players.
  3. Standard member Crowley
    Not Aleister
    08 Feb '05 09:27
    Originally posted by Svin1
    Anybody who plays someone rated 700 points lower than himself takes the same risk--it's not just Ironman. I think it's fair because it discourages people from gathering easy points against much weaker players.
    The problem for him will be that (if) he gets to a stage where everyone is rated 300+ less than him and his rating will stagnate. Then he'll only be able to lose, but gain nothing...
  4. Standard member flexmore
    Quack Quack Quack !
    08 Feb '05 10:44 / 4 edits
    Originally posted by Svin1
    Anybody who plays someone rated 700 points lower than himself takes the same risk--it's not just Ironman. I think it's fair because it discourages people from gathering easy points against much weaker players.
    this is not quite true .... ironman has a different rating rule from the rest of us .... 2100+ players also have a different rule.

    http://www.redhotpawn.com/help/index.php?help=faq
    "
    How is my rating calculated?
    When you join, you are given a rating of 1200.

    If you enter into a rated game, your rating will be recalculated on its completion so long as at least two moves have been made by each player.

    Note: Until you have completed 20 games your rating may go down even when you win a game!

    Until you complete 20 games, you will have a provisional rating. This rating is calculated as the average of all the games you have played. Each game is scored as being your opponent's rating (for a draw), your opponent's rating plus 400 (for a win), or your opponent's rating minus 400 (for a loss). If your opponent is also on a provisional rating, then the 400s above become 200s. If your opponent has played fewer than five games then their rating is treated as 1200 when calculating your rating.

    Players are rated using the following formula:

    New Rating = Old Rating + K * (Score - Win Expectancy)

    K is a constant (32 for 0-2099, 24 for 2100-2399, 16 for 2400 and above)

    Score is 1 for a win, 0.5 for a draw and 0 for a loss.

    The Win Expectancy is calculated using the following formula :

    Win Expectancy = 1 / (10^((OpponentRating-YourRating)/400)+1)

    The Win Expectancy is used in the rating calculation but is interesting on its own. For example, the calculation below is for a rating difference of 200. This shows that if you have a rating 200 points higher than another player, you can expect to beat them, on average, three times for each four games played.

    Win Expectancy = 1 / (10^(-200/400)+1) = 0.76

    Note: ^ = "to the power of", e.g. 2^3=8.

    If you have a non-provisional rating and you play a provisional-rated player, then you receive (or lose) only half the number of rating points you would normally. If the provisionally-rated player has played fewer than five games, then their rating is treated as 1200 when calculating your rating.
    "

    i do not understand the need for this "k", it seems unnecessary.

    i suspect it is only here to stop players like ironman embarassingly getting a rating higher than kasparov.
    rhp ratings lose legitimacy when players have a rating higher than kasparov.
    but here we are a correpondence site and so the ratings gap must be greater than o.t.b.
    here we go from casual unskilled players moving very fast ... up to skilled players googling opening books and regularly taking many days to calculate important moves.

    i say get rid of the double standards ... make k=32 for everyone.
  5. Standard member Toe
    08 Feb '05 12:17
    Originally posted by Crowley
    The problem for him will be that (if) he gets to a stage where everyone is rated 300+ less than him and his rating will stagnate. Then he'll only be able to lose, but gain nothing...
    Only in a ratings sense.

    I suspect he plays 'cos he wants to. It is a occupational hazzard of being shockingly good that there are few people on site who can compete with you. Fortunately there are a lot of folk who will happly be beaten sensless though.

    If it were all about ratings, why not ask if the lower rated player takes an unreasonable risk of loosing a point against him? Same answer: 'cos they like to play, and what's an RHP rating point worth anyway?
  6. Donation Quirine
    Tovenaar
    08 Feb '05 14:10
    Originally posted by flexmore

    i say get rid of the double standards ... make k=32 for everyone.[/b]
    I do not agree.
    I think the system is fine as it is. It is comparible to the rating system used by FIDE and ICCF.
    Only a minor point as far as I am concerned: I like the k-factor to be 8 or 12 for 2600+ rated players
  7. Standard member Deepfault
    Devil's Advocate
    08 Feb '05 18:40
    Surely at his level ratings are meaningless as there is no real comparison.

    I would guess maintaining his record of loosing only one game has more meaning.

    If he has a rating of 2400 or 3500 and he gets a point no matter it means little then to say he has won another X number of games.

    I must admit playing against him is daunting and I wonder if he actually enjoyed my games I know I found then painful in the same way as a rider in the charge of the light brigade must have felt that painful..on to glory and death certain, but to leave this mortal coil with some honour if only in the stupidity of accepting the order to charge.

    I thank the clan system and my esteemed superior officer for issueing the order that gave me the oppertunity to play him.



  8. 09 Feb '05 11:06 / 2 edits
    The K-factor makes things a little harder for Ironman, but the same problem do arise for lower rated players.

    1800 vs. 1000:
    ---
    Win expectancy: 1 / (10^(1000/1800)+1) = 0.99
    New rating = 1800 + 32 * (1 - 0.99) = 1800 + 0.32 =~ 1800.
  9. 09 Feb '05 13:19
    Originally posted by flexmore

    i do not understand the need for this "k", it seems unnecessary.

    i suspect it is only here to stop players like ironman embarassingly getting a rating higher than kasparov.
    [b]rhp ratings lose legitimacy when players have a rating higher than kasparov.

    but here we are a correpondence site and so the ratings gap must be greater than o.t.b.
    here ...[text shortened]... calculate important moves.

    i say get rid of the double standards ... make k=32 for everyone.[/b]
    I'm pretty sure, that the K-factors has been taken from the USCF rating algorithm.
  10. 09 Feb '05 17:38
    the K factor is the maximum number of rating points you can win or lose from a particular game,

    the rest of the formula attenuates this number according to the players rating. for example if a low rated player beats a high rated player then they can never get more than the value of k, but the bigger the rating difference the closer to k this gain will be
  11. Standard member flexmore
    Quack Quack Quack !
    10 Feb '05 04:55
    Originally posted by Quirine
    I do not agree.
    I think the system is fine as it is. It is comparible to the rating system used by FIDE and ICCF.
    Only a minor point as far as I am concerned: I like the k-factor to be 8 or 12 for 2600+ rated players
    why?
  12. Donation Quirine
    Tovenaar
    10 Feb '05 05:37
    Originally posted by flexmore
    why?
    A lower K-factor at the top of the ratings prevents that ratings go to incredible high levels. This works pretty well for FIDE-rating where the top players have a K-factor of 5 I think (may well be 10) where as 2000-rated players have a K-factor of 20 (may well be 30).
    Ten to fifteen years ago Kasparov had a rating around 2800, now he has 2820 or so. There's a slight increase in rating but not much.
    So it is easier to compare players of different eras (still difficult though).
    Here at RHP the top had less than 2000 some 2 years ago, around 2200 1 year ago and 2400 now. It would be nice if this enormous increase every year came to a halt..
  13. Standard member flexmore
    Quack Quack Quack !
    10 Feb '05 05:57 / 4 edits
    my expectation of our ratings is that they reflect the quality of moves that we play when moving our rhp chess pieces.

    Tebb and Ironman, using opening books, and spending several days on a move will presumably play as well as kasparov does in 2 minutes.

    Their rating should be massively above that of some unskilled player, unaware of enpassant, logging into rhp for a few very quick moves followed by timeouts in all of their games.

    The standard of play difference is much greater on rhp than at fide.

    the entry players are set at 1200 ... this is our only fixed point.
    I believe it is natural for the top rated players at rhp to eventually have a rating of 1600 above the entry level players (2800+).

    i think this "k" barrier just slows rhp ratings from rapidly getting to where they will eventually be.

    and the worst problem ... it may discourage the very top players from feeling free to play us all.

    of course it is nice to see some link with the fide ratings ... but it is illusory and disjointed to do it this way.
  14. 10 Feb '05 07:27
    Just a thought give these massively high rated players, since they can't get points the normal way because of mathematics) a point per game in the tournaments they win, or some such, it gives them an incentive to keep playing all us mortals and who nows one of us mortals may draw them some day and then they'll be back to the normal way of gettin rating points
  15. Standard member flexmore
    Quack Quack Quack !
    10 Feb '05 07:50 / 1 edit
    Originally posted by Quirine
    A lower K-factor at the top of the ratings prevents that ratings go to incredible high levels. This works pretty well for FIDE-rating where the top players have a K-factor of 5 I think (may well be 10) where as 2000-rated players have a K-f ...[text shortened]... ould be nice if this enormous increase every year came to a halt..
    K is the development coefficient.
    K = 25 for a player new to the rating list until he has completed events with a total of at least 30 games.
    K = 15 as long as a player`s rating remains under 2400.
    K = 10 once a player`s published rating has reached 2400, and he has also completed events with a total of at least 30 games. Thereafter it remains permanently at 10.

    http://www.fide.com/official/handbook.asp?level=B0210

    so fide only uses a 30% adjustment in k for high rated playters, rhp is using 50%