Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Fritz 9 - the first human-like chess engine in the world

Author: Günther Simon

Date: 15:18:33 10/28/05

Go up one level in this thread


On October 28, 2005 at 18:00:22, Drexel,Michael wrote:

>"The program has become more human in a different way. It now has extensive
>chess knowledge and a human evaluation structure, ensuring that it finds good
>moves and interesting ideas where there are no direct tactics to guide it. You
>will appreciate the value of this when you analyze with the program. Fritz 9
>thinks and plans like a human, while retaining the tactical ferocity of the
>world’s top chess engine" www.chessbase.com
>
>You don´t believe this? You think that´s a load of crap?
>
>You are wrong, please look at the following examples and decide yourself:
>
>Example 1:
>
>[D]rnbqkb1r/ppp2ppp/3p1n2/4P3/4P3/2N5/PPP2PPP/R1BQKBNR b KQkq - 0 4
>
>Analysis by Fritz 9:
>
>4...Sfd7 5.exd6 Lxd6 6.Sb5 Lb4+ 7.c3 Lc5 8.Lg5 f6 9.Dh5+
>  ±  (0.91)   Depth: 7/21   00:00:00  53kN
>4...Sfd7 5.Sd5 dxe5 6.Lg5 f6 7.Lh6 Sc6 8.Dh5+ g6
>  ±  (0.76)   Depth: 8/27   00:00:00  255kN
>4...Sfd7 5.exd6 Lxd6 6.Sb5 Le7 7.Lf4 Sa6 8.Dd5 Lb4+ 9.c3 Lc5
>  ±  (0.94)   Depth: 9/27   00:00:00  504kN
>4...Sfd7 5.exd6 Lxd6 6.Sb5 Lb4+ 7.c3 La5 8.Dd5 0-0 9.Lg5
>  ±  (1.07)   Depth: 10/29   00:00:01  1063kN
>4...Sfd7 5.exd6 Lxd6 6.Sb5 Sf6 7.Lg5 0-0 8.Sf3 Sc6 9.Sxd6 cxd6 10.c3 De8 11.Lxf6
>Dxe4+
>  ±  (1.14)   Depth: 11/32   00:00:03  3319kN
>4...Sfd7 5.exd6 Lxd6 6.Sb5 De7 7.Sxd6+ cxd6 8.Ld3 0-0 9.Se2 Sc5 10.0-0 Sxe4
>11.Te1
>  ±  (1.02)   Depth: 12/37   00:00:10  9001kN
>4...Sfd7 5.exd6 Lxd6 6.Le3 0-0 7.Dd2 De7 8.0-0-0 Td8 9.Lf4 Se5 10.Sd5 De8
>  ±  (1.16)   Depth: 13/34   00:00:24  23816kN
>4...Sfd7 5.exd6 Lxd6 6.Le3 0-0 7.Dd2 De7 8.0-0-0 Sc6 9.Lg5 De6 10.Sf3 Sde5
>11.Sd4 Dg4
>  ±  (1.08)   Depth: 14/41   00:01:00  60691kN
>4...Sfd7 5.exd6 Lxd6 6.Le3 De7 7.Dd2 Sf6 8.Ld3 0-0 9.Sf3 Sc6
>  ±  (1.13)   Depth: 15/39   00:02:29  156351kN
>
>4...Nfd7!
>A Master would not even think about the materialistic 4...dxe5? which is the
>typical computer move in this position.
>Instead true masters of the game willingly give up pawns in exchange for a
>better position.
>Unlike other engines Fritz 9 understands the importance of development and King
>safety.

I have not yet looked at the other examples but the first one at least is
not convincing. Can you tell me why F9 gives already nearly winning score
of +1.13, when it is so happy being 'Human' like?
I can tell you why it does not play dxe5 here and I doubt this is a computer
typical move at all. F9 obviously just has a bigger malus for uncastled
kings in the centre(as after dxe5 Qxd8+ Kxd8 etc...).

Guenther



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.