Author: Uri Blass
Date: 05:30:11 06/18/03
Go up one level in this thread
On June 18, 2003 at 08:01:58, Uri Blass wrote:
>On June 18, 2003 at 05:34:51, Thorsten Czub wrote:
>
>>they move without any strategic idea.
>
>I beated GM abir Har Aven when all the moves were of chess programs.
>I also beated the female world champion when almost all the moves were computer
>moves.
>
>I do not think that you can generalize about programs.
>If they search deep enough they can find often good moves during all the game.
>
>There are position that they are weak but you cannot be sure of getting them
>in games.
>
>Uri
I can add the game against har aven with some comments for people who are
interested and people who prefer direct pgn can get also pgn(with no comments).
I guess that I have pgn with comments but I am too lazy to search for it.
I guess that har aven tried to confuse the computer by playing anti chess but
this plan did not work:
Note that ICCF GM Abir Har Aven got the second place in the Israeli championship
with 5.5/8 so inspite of the impression from this game he is not a weak player.
(14) Abir Har aven - Uri Blass [A03]
[Blass,U]
33554424kB, f5book.ctg. PentiumIII800
1.g3 d5 2.f4 h5 (this move was
suggested by the tigers that I had at that time(tiger13 and gambit1) these
program planned 3.Nf3 h4) 3.Nf3 h4 4.Nxh4 e5 (Some programs like e5 and
other programs like Rxh4. I did not want to sacrifice the h file so I chose
e5) 5.Ng2 (I believe that Nf3 was better and I expected 5.Nf3) [5.Nf3 exf4
6.gxf4 Bg4 7.Ng5; 5.d3? exf4 6.Bxf4 g5 7.Be5 f6 8.Ng6 Rh6 9.Nxf8 fxe5]
5...Nf6 5...Nh6 was played in one game in my databases and I searched for
games with 5...Nf6 and did not find games with this move but my programs
prefered 5...Nf6 6.e3 [6.fxe5? Ne4] 6...Nc6 7.d3 white has problems also
after Nc3 and I remember that chess programs changed their mind from Nc3 to
d3 after searching for a long time 7...Bg4 8.Be2 Qd7 9.Nd2 0-0-0 (I expected
10.Nf3 and after 10.Nf1 I decided that this anti-chess cannot work so I gave
my Deep Fritz to search for more than 80 hours,it changed it's mind from
exf4 to Bd6 after more than 22 hours and I checked 10...Bd6 with my other
programs and found no problem with this move so I decided to play it) 10.Nf1
Bd6 11.Bd2? after this move it can be proved by only material evaluation
that white is losing the pawn advantage. I believe that white also lose more
material and the game. Deep Fritz prefers Bxg4 Nxg4 Qf3 that was probably
the best line(Deep fritz needs a lot of hours to change it's mind to Qf3)
11...exf4 12.gxf4 Rh3 13.c3 g5 Deep Fritz found it after more than 16
hours(Re8 is also good for black and Deep Fritz's line suggest that I get
one pawn advantage but it was not enough for me) 14.fxg5 Ne5 15.Nf4 Nf3+
16.Kf2 Bxf4 17.exf4 Nxd2 18.Ng3 Bxe2 19.Qxd2 Ng4+ 20.Kxe2 d4 21.Rag1 Nxh2
22.Kf2 Qg4 23.Ne4 Qh4+ even better than Qf3+ 24.Ng3 Re8 25.cxd4 f5 26.gxf6
Ng4+ I proved mate in 18 by chessmaster6000(chessmaster is the best program
to find mates) but told nothing to my opponent because I hoped for a mistake
that let me to win evenfaster 27.Kf1 Ne3+ 28.Qxe3? (28.ke1 was the slowest
way to lose the game) 28...Rxe3 29.Rxh3 my opponent resigned after the
following moves that were one condition 29...Qg4 30.Rh8+ Kd7 31.Rh7+ Kc6 0-1
[Event "correspondence game""]
[Site "Tel-Aviv"]
[Date "????.??.??"]
[Round "?"]
[White "Abir Har avan"]
[Black "Uri Blass"]
[Result "*"]
[PlyCount "62"]
[TimeControl "300+5"]
{1MB, Shredder7.ctg, URI-PC
} 1. g3 d5 2. f4 h5 3. Nf3 h4 4. Nxh4 e5 5. Ng2 Nf6
6. e3 Nc6 7. d3 Bg4 8. Be2 Qd7 9. Nd2 O-O-O 10. Nf1 Bd6 11. Bd2 exf4 12. gxf4
Rh3 13. c3 g5 14. fxg5 Ne5 15. Nf4 Nf3+ 16. Kf2 Bxf4 17. exf4 Nxd2 18. Ng3 Bxe2
19. Qxd2 Ng4+ 20. Kxe2 d4 21. Rag1 Nxh2 22. Kf2 Qg4 23. Ne4 Qh4+ 24. Ng3 Re8
25. cxd4 f5 26. gxf6 Ng4+ 27. Kf1 Ne3+ 28. Qxe3 Rxe3 29. Rxh3 Qg4 30. Rh8+ Kd7
31. Rh7+ Kc6 *
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.