Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: 1.BELCT2001: discussion about the games, part 1 : zarkov4.5T - gt14.6

Author: Uri Blass

Date: 13:05:40 10/05/01

Go up one level in this thread


<snipped>
>(Blass, Tel-aviv 05.10.2001)
>Zarkov4.5T - Gambit-Tiger14.6
>[D]5rk1/ppp3pp/1b1np3/4q3/B7/2P1P2P/PP1P2P1/R1B1Q1K1 b - - 0 1
>
>Analysis by Deep Fritz:
>
>21...Qxe3+--
>  +-  (5.72)   Depth: 1/3   00:00:00
>21...Qxe3+-- 22.dxe3
>  +-  (8.44)   Depth: 1/4   00:00:00
>21...Bxe3+!
>  +-  (4.91)   Depth: 1/4   00:00:00
>21...Bxe3+! 22.Qxe3
>  +-  (2.62)   Depth: 1/5   00:00:00
>21...c6!
>  +-  (2.53)   Depth: 1/5   00:00:00
>21...c6!
>  =  (-0.09)   Depth: 1/5   00:00:00
>21...c6--
>  ²  (0.44)   Depth: 2/4   00:00:00
>21...c6--
>  ²  (0.44)   Depth: 2/4   00:00:00
>21...Ne4!
>  ²  (0.41)   Depth: 2/4   00:00:00
>21...Ne4! 22.d4
>  ²  (0.37)   Depth: 2/4   00:00:00
>21...Qa5!
>  ²  (0.34)   Depth: 2/5   00:00:00
>21...Qa5 22.Qh4 c6
>  =  (0.22)   Depth: 3/8   00:00:00
>21...Qa5--
>  ²  (0.53)   Depth: 4/9   00:00:00  2kN
>21...Qa5-- 22.Bb3 Qf5 23.d4
>  ±  (0.72)   Depth: 4/9   00:00:00  2kN
>21...Ne4!
>  ²  (0.69)   Depth: 4/11   00:00:00  2kN
>21...Ne4! 22.d3 Qa5 23.Bd7
>  ²  (0.47)   Depth: 4/12   00:00:00  2kN
>21...Qf6!
>  ²  (0.44)   Depth: 4/12   00:00:00  3kN
>21...Qf6 22.d4 c6 23.Bc2 Bc7
>  ²  (0.50)   Depth: 5/12   00:00:00  6kN
>21...c6!
>  ²  (0.47)   Depth: 5/14   00:00:00  7kN
>21...c6! 22.d4 Qe4 23.Bb3
>  ²  (0.37)   Depth: 5/14   00:00:00  8kN
>21...c6 22.d4 Qe4 23.Bb3 Bc7 24.Qe2
>  ²  (0.37)   Depth: 6/15   00:00:00  15kN
>21...c6 22.Bc2 Ne4 23.d4 Qf5 24.b4 Bc7
>  ²  (0.47)   Depth: 7/20   00:00:00  47kN
>21...Qe4!
>  ²  (0.44)   Depth: 7/20   00:00:00  69kN
>21...c5!
>  ²  (0.41)   Depth: 7/20   00:00:00  114kN
>21...c5! 22.Bc2 c4 23.d4 cxd3 24.Bxd3
>  ²  (0.28)   Depth: 7/21   00:00:00  149kN
>21...c5 22.Bc2 c4 23.b3 Ne4 24.Ba3 Rd8 25.Bxe4 Qxe4 26.bxc4
>  ²  (0.34)   Depth: 8/21   00:00:00  237kN
>21...c5 22.Bc2 Nf5 23.Qf2 g6 24.Qf4 Qxf4 25.exf4 c4+ 26.Kh2 Kf7
>  =  (0.25)   Depth: 9/22   00:00:00  454kN
>21...c5!
>  =  (-0.06)   Depth: 10/25   00:00:02  1382kN
>21...c5 22.Qh4 Ne4 23.d4 cxd4 24.cxd4 Bxd4 25.Bc2 Rc8 26.Bxe4 Rxc1+ 27.Rxc1
>  =  (0.06)   Depth: 11/28   00:00:07  4419kN
>21...c5 22.Qh4 Qf5 23.Qe1 c4 24.b3 Bc5 25.bxc4 Qd3 26.Bd7 Rf6 27.Qg3
>  =  (0.03)   Depth: 12/29   00:00:18  10819kN
>21...c5 22.Qh4 Ne4 23.d4 cxd4 24.cxd4 Qa5 25.Bc2 Bd8 26.Bd2 Nxd2 27.Qxh7+
>  =  (-0.03)   Depth: 13/35   00:00:44  26179kN
>21...c5 22.Qh4 Ne4 23.d4 cxd4 24.cxd4 Qa5 25.Bc2 Bd8 26.Bd2 Nxd2 27.Qxh7+
>  =  (-0.13)   Depth: 14/37   00:01:35  57526kN
>21...c5 22.Qh4 g5 23.Qg4 Ne4 24.d4 Qf6 25.Qe2 cxd4 26.Bc2 dxe3 27.Bxe3
>  =  (-0.13)   Depth: 15/41   00:05:04  186195kN
>
>(Blass, Tel-aviv 05.10.2001)
>
>Uri
Analysis continued

21...c5!
  ³  (-0.44)   Depth: 16/40   00:11:35  413867kN
21...c5! 22.Qh4 Bc7 23.d4 cxd4 24.cxd4 Qa5 25.Bc2 Nf5 26.Bxf5 Rxf5 27.e4
  µ  (-0.72)   Depth: 16/42   00:19:18  692120kN

(Blass, Tel-aviv 05.10.2001)


It seems that I did not give Deep Fritz enough time for the last move
c5 failed high for Deep Fritz at depth 16 so maybe Bb3 is still the losing
error.

I thought that c3 is better because it opens the diagnol for the bishop at a4.
If Nxe6 is really losing that it is going to be a very hard problem to see it
from the root position.

It is possible for top programs to see cloe to draw evaluation for Nxe6 but the
question is if they evaluate the alternative Ne2 as better and it seems to me a
question of evaluation because proving the material gain(if there is a material
gain) seems to be possible only some plies after c3 and Deep Fritz seems to need
19 plies after c3 and it means 20 plies before c3

It may be interesting to give Deep Fritz at least 20 plies to analyze the
position at move 19 of white but I leave it to people with better hardware if
they are interested.

Deep Fritz used p800(64 Mbyte hash tables) in all the analysis

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.