Author: Heiner Marxen
Date: 11:12:28 08/02/01
Go up one level in this thread
On July 09, 2001 at 14:06:01, Dieter Buerssner wrote:
>On July 09, 2001 at 11:14:35, Kurt Widmann wrote:
>
>>How deep do both have to search before dropping 9... Qc6?? for Bxe6! instead?
>>Fritz 6 [B02]
>>09.07.2001
>>
>>1.e4 Nf6 2.e5 Nd5 3.Nf3 d6 4.Bc4 Nb6 5.Bxf7+ Kxf7 6.Ng5+ Kg8 7.Qf3 Qe8 8.e6 g6
>>9.Nf7* Qc6 10.Nh6+ Kg7 11.Qf7+ Kxh6 12.d3+ Kh5 13.g4+ Kh4 14.Rg1 Qb5 15.Nd2 Qa5
>>16.Qf4 Qb4 17.Qxb4 Kg5 18.Qd4 Kh4 19.Nf3+ Kh3 20.Rg3# 1-0
>>
>>Do other programs have the same proplem?
>
>Yace on AMD K6-2 475 with 35 M hash needs 75 seconds to switch to Bxe6:
>
> nodes time score depth
> 4995317 26.566 1.84 9. 9...Qc6 10. Qg4 Qxc2 11. Na3 Qf5 12. Qxf5 gxf5
> 13. Nxh8 Kxh8 14. Nb5 {190}
> 12185259 1:03.2 1.44 10-- 9...Qc6 10. Nh6+ Kg7 11. Qf7+ Kxh6 12. d4+ Kh5
> 13. g4+ Kh4 14. Bg5+ Kxg5 15. h4+ Kxg4 16. Rg1+
> Kh5 17. Qf4 Qh1 18. Rxh1 g5 19. hxg5+ Kg6 {-41}
> 13822728 1:12.0 0.64 10t 9...Qc6 10. Nh6+ Kg7 11. Qf7+ Kxh6 12. d4+ Kh5
> 13. g4+ Kh4 14. Bg5+ Kh3 15. Qf4 Qxh1+ 16. Ke2
> Qe1+ 17. Kxe1 Kg2 18. Qf7 Nc6 {109}
> 14275398 1:15.1 0.65 10t+ 9...Bxe6 10. Nxh8 Kxh8 11. O-O
> 16798952 1:34.4 1.73 10t 9...Bxe6 10. Nxh8 Kxh8 11. Nc3 Nc6 12. O-O Qf7
> 13. Qe4 Bf5 14. Qf4 Bg7 {190}
> 17736647 1:42.0 1.73 10. 9...Bxe6 10. Nxh8 Kxh8 11. Nc3 Nc6 12. O-O Qf7
> 13. Qe4 Bf5 14. Qf4 Bg7 {190}
> 22714576 2:21.1 1.64 11t 9...Bxe6 10. Nxh8 Kxh8 11. O-O Nc6 12. Qe3 Qf7
> 13. Re1 Bxa2 14. b3 {HT} {270}
> 26133781 2:44.5 1.64 11. 9...Bxe6 10. Nxh8 Kxh8 11. O-O Nc6 12. Qe3 Qf7
> 13. Re1 Bxa2 14. b3 {HT} {270}
>
>Yace thinks (but cannot prove) that Black gets "only" mated in the 21st move,
>when analysing from back to front:
>
> 9406323 36.321 Mat12 12. 10. Nh6+ Kg7 11. Qf7+ Kxh6 12. d3+ Kh5 13. g4+
> Kh4 14. Rg1 Qd5 15. Nd2 Qe5+ 16. Kf1 Nc4 17.
> Nf3+ Kh3 18. Nxe5 Nd2+ 19. Bxd2 dxe5 20. Qf3+
> Kxh2 21. Qg3# {311}
>
>Perhaps Chest or some other programs can prove this?
Unfortunately, Chest cannot do this (fast enough).
After 50 minutes on a K7/600 it has just proven that there is no mate in 8.
And the effective branching factor around 20 prohibits larger depthes.
After 4 more full moves (which are equal in your PV and in the game given
by Kurt), i.e. after 13...Kh4 we have
[D]rnb2b1r/ppp1pQ1p/1nqpP1p1/8/6Pk/3P4/PPP2P1P/RNB1K2R w KQ -
where Chest verifies a mate in 8 by 14.Rg1, as Yace found it:
PV: Rg1 Qc5 Nd2 Qe5+ Kf1 Nc4 Nf3+ Kh3 Nxe5 Nd2+ Bxd2 h6 Qf3+ Kh4 Nxg6#
The black move allowing for the shorter mate in Kurt's line is 15...Qa5
Cheers,
Heiner
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.