Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Cadaques Tournament - Some Statistics

Author: Chessfun

Date: 12:07:52 01/15/00

Go up one level in this thread


On January 15, 2000 at 15:03:26, Frank Schubert wrote:

>Hi everybody,
>
>40 games played by each program so far. Time to make some first statistics
>of Enriques superb tournament:
>
><pre>
>
>General Statistics:
>
>Games        : 140
>
>White Wins   : 49 (35.0 %)
>Black Wins   : 30 (21.4 %)
>Draws        : 61 (43.6 %)
>
>White Perf.  : 56.8 %
>Black Perf.  : 43.2 %
>
>ECO A =     14 Games (10.0 %)
>ECO B =     35 Games (25.0 %)
>ECO C =     17 Games (12.1 %)
>ECO D =     41 Games (29.3 %)
>ECO E =     33 Games (23.6 %)
>
>
>Rating List:
>
>
>    Program                Elo    +   -   Games   Score   Av.Op.  Draws
>
>  1 Junior 6.0           : 2665  101  74    40    60.0 %   2594   45.0 %
>  2 Fritz 6a             : 2641   96  86    40    63.7 %   2542   37.5 %
>  3 Hiarcs 7.32          : 2615  111  75    40    52.5 %   2597   40.0 %
>  4 Rebel-Tiger 12.0e    : 2598  106  59    40    56.2 %   2554   57.5 %
>  5 Shredder 4           : 2597   80 106    40    43.8 %   2640   37.5 %
>  6 Nimzo 7.32           : 2592   76 102    40    41.2 %   2653   42.5 %
>  7 Century              : 2493   78  92    40    32.5 %   2619   45.0 %
>
></pre>
>
>Shredder and Nimzo had the strongest opponents so far. Most of the draws
>were played by Tiger (proportion 57.5 %).
>
>Elo calculation was done by iteration algorithm.
>Margins of error roughly between 60 and 110 Elo's (given with 95 % confidence).
>Distance between Junior and Nimzo is only 73 points. That means nearly
>everything can happen during the next rounds. Only Century seems to be
>"out of race", but compare the upper margin of Century (2493+78 = 2571) with
>the lower margin of Fritz (2641-86 = 2555) and you see that Century can
>still hope to reach a better rank.
>Also consider the fact that for halving the error margins four times as much
>games are necessary. Therefore I expect the margins to be slightly above
>50 Elo's at the end of the tournament (120 games for each program).
>
>Finally the single statistics:
>
><pre>
>
>(1) Junior 6.0                :  40 (+ 15,= 18,-  7), 60.0 %
>
>Shredder 4                    :  20 (+  6,= 10,-  4), 55.0 %
>Nimzo 7.32                    :  20 (+  9,=  8,-  3), 65.0 %
>
>
>(2) Fritz 6a                  :  40 (+ 18,= 15,-  7), 63.8 %
>
>Nimzo 7.32                    :  20 (+  6,=  9,-  5), 52.5 %
>Century                       :  20 (+ 12,=  6,-  2), 75.0 %
>
>
>(3) Hiarcs 7.32               :  40 (+ 13,= 16,- 11), 52.5 %
>
>Shredder 4                    :  20 (+  9,=  5,-  6), 57.5 %
>Rebel-Tiger 12.0e             :  20 (+  4,= 11,-  5), 47.5 %
>
>
>(4) Rebel-Tiger 12.0e         :  40 (+ 11,= 23,-  6), 56.2 %
>
>Century                       :  20 (+  6,= 12,-  2), 60.0 %
>Hiarcs 7.32                   :  20 (+  5,= 11,-  4), 52.5 %
>
>
>(5) Shredder 4                :  40 (+ 10,= 15,- 15), 43.8 %
>
>Junior 6.0                    :  20 (+  4,= 10,-  6), 45.0 %
>Hiarcs 7.32                   :  20 (+  6,=  5,-  9), 42.5 %
>
>
>(4) Nimzo 7.32                :  40 (+  8,= 17,- 15), 41.2 %
>
>Junior 6.0                    :  20 (+  3,=  8,-  9), 35.0 %
>Fritz 6a                      :  20 (+  5,=  9,-  6), 47.5 %
>
>
>(6) Century                   :  40 (+  4,= 18,- 18), 32.5 %
>
>Fritz 6a                      :  20 (+  2,=  6,- 12), 25.0 %
>Rebel-Tiger 12.0e             :  20 (+  2,= 12,-  6), 40.0 %
>
></pre>
>
>
>Source: ELOStat 1.0
>
>
>If you found the results interesting I will post the new table after each round.
>
>Frank

Very interesting, please do so Frank.
Thanks.




This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.