Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Chess Tiger Won and is the leader with Elo perf, 2744 (crosstables)

Author: Uri Blass

Date: 08:45:50 06/27/01

Go up one level in this thread


On June 27, 2001 at 10:44:51, Robert Hyatt wrote:

>On June 27, 2001 at 02:07:10, Uri Blass wrote:
>
>>On June 27, 2001 at 00:26:28, Robert Hyatt wrote:
>>
>>>On June 26, 2001 at 03:58:02, Uri Blass wrote:
>>>
>>>>On June 26, 2001 at 03:34:44, Adolfo Bormida wrote:
>>>>
>>>>>Win 3 draw 1 Croos Table
>>>>>
>>>>>                                      1 2 3 4 5 6 7 8 9 0 1 2
>>>>>
>>>>>  1   Slipak,Sergio         2448      * 1 1                 1   3.0/0  0.00
>>>>>  2   Matsuura,Everaldo     2467      1 *                 1 0   2.0/0  0.00
>>>>>  3   Limp,Eduardo          2465      1   *             1 1     3.0/0  0.00
>>>>>  4   Rodriguez,Andrés      2500            *         3 1 0     4.0/0  0.00
>>>>>  5   Andres,Miguel         2382              *     1 1 1       3.0/0  0.00
>>>>>  6   Hoffman,Alejandro     2453                * 1 1 1         3.0/0  0.00
>>>>>  7   Dorin,Mauricio        2410                1 * 1       0   2.0/0  0.00
>>>>>  8   Valerga,Diego         2468              1 1 1 *           3.0/0  0.00
>>>>>  9   Scarella,Enrique      2361            0 1 1     *         2.0/0  0.00
>>>>>  10  Ricardi,Pablo         2554          1 1 1         *       3.0/0  0.00
>>>>>  11  Panno,Oscar           2471        1 1 3             *     5.0/0  0.00
>>>>>  12  COMP Chess Tiger      2632*     1 3         3         *   7.0/0  0.00
>>>>>
>>>>>Standard crosstable
>>>>>
>>>>>IV Magistral República Argentina  2001
>>>>>
>>>>>                                     1 2 3 4 5 6 7 8 9 0 1 2
>>>>>  1   COMP Chess Tiger   2632  +102  *     ½           1 1     2.5/3
>>>>>  2   Panno,Oscar        2471  +131    * ½         1   ½       2.0/3
>>>>>  3   Limp,Eduardo       2465   +26    ½ * ½ ½                 1.5/3  2.50
>>>>>  4   Slipak,Sergio      2448   +73  ½   ½ *           ½       1.5/3  2.50
>>>>>  5   Ricardi,Pablo      2554  -105      ½   *   ½ ½           1.5/3  2.25
>>>>>  6   Valerga,Diego      2468   -53            * ½   ½   ½     1.5/3  2.00
>>>>>  7   Andres,Miguel      2382   +79          ½ ½ *         ½   1.5/3  2.00
>>>>>  8   Rodriguez,Andrés   2500   -38    0     ½     *       1   1.5/3  1.75
>>>>>  9   Hoffman,Alejandro  2453   -40            ½     *   ½ ½   1.5/3  1.75
>>>>>  10  Matsuura,Everaldo  2467   -75  0 ½   ½           *       1.0/3  1.75
>>>>>  11  Dorin,Mauricio     2410   -17  0         ½     ½   *     1.0/3  1.50
>>>>>  12  Scarella,Enrique   2361   -41              ½ 0 ½     *   1.0/3  1.50
>>>>>
>>>>>  Promedio elo: 2467 <=> Categoría: 9
>>>>>  gm = 7.00 m = 5.0 (las partidas con ChessTiger no están incluídas)
>>>>>  * corre en PIII 866 Mhz, 256 MbRAM. SSDF Rating al 13 de Junio del 2001
>>>>
>>>>I do not understand the big number of draws with the 3 points for a win system.
>>>>I expected the 3 points for a win system to encourage players to play for a win
>>>>but I do not see more wins than in normal tournaments and it seems that the
>>>>number of draws is even higher than in a normal tournament with 1 point for a
>>>>win and 1/2 point for a draw.
>>>>
>>>>Uri
>>>
>>>
>>>This may be a flawed system.  The thinking is "I get 3 for a win, 1 for a draw,
>>>I will go for the win if at all practical."  But it might end up "I get 1 for
>>>a draw, my opponent gets 3 for a win, so I will try for the draw rather than
>>>take a chance on giving him 3."
>>>
>>>IE it might actually increase the number of draws, not decrease it.
>>
>>simple logic say that if I have to choose between a forced draw
>>and between probability of 30% to win and 40% to lose I will choose the 30% to
>>win if there is 3 points for draw.
>>
>>If I get 3 wins 4 losses and 3 draws I get 3*3+4*1=13 points
>>
>>If my opponent win against me and get 9 draws in the other games he is going to
>>get only 3+9=12 points.
>>
>>Uri
>
>
>The math still works as I gave it however.  If you expect to win 30%, then
>3 of 10 times you will get +3 points.  If you expect to lose 40%, then 4 of
>10 times your opponent gets +3.  No matter how you look at it, he is going to
>win the tournament.

I see that I did a mistake in my math and +1 =9 gives exactly the same points as
+3 =3 -4 (my mistake is that I calculated +3 =3 -4 as 3*3+4 and not 3*3+3
but I still think that +3 =3 -4 is better than 10 draws or +1 =8 -1 by the place
and not only by points.


+3 =3 -4 result mean that you get 12 points and the opponents get +4 =3 -3
against you that means 15 points against you and the average number of points
thaey get against you is 1.5

+1=8 -1 means that you get 11 points and the opponents get also 11 points
against you so the average number of points that they get against you is 1.1

+3 =3 -4 improves your result relative to +1 =8 -1 by 1 point when
the opponents get average of 0.4 points.

It means that you get more improvement relative to your opponents so if you have
to choose between one strategy that gives you 10% chances to win and 10% chances
to lose and another strategy that gives you 30% chances to win and 40% chances
to lose then it is clear that the second strategy is better from ranking's point
of view.

Uri



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.