Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: IQ debugging and some tough positions ...

Author: John Merlino

Date: 12:56:58 07/29/02

Go up one level in this thread


On July 29, 2002 at 15:30:55, Jim Monaghan wrote:

>Hi everyone,
>
>I've also been working a little more with the really tough positions. I thought
>I'd post a couple with the solutions to see if anyone out there can solve them.
>A good test position is always interesting. If there's interest I'll post
>others. On a Cel 1.3 Ghz/256 with 32MB hash after 5 minutes Shredder misses
>these:
>
>[D]4k3/pp2pp2/4n1p1/3RP1P1/2r1N3/P1P5/1P2q3/1K5Q w - -
>
>4k3/pp2pp2/4n1p1/3RP1P1/2r1N3/P1P5/1P2q3/1K5Q w - - bm Nc5; id "IQ.1081";
>Dzhindzhikhashvili-Tukmakov, Leningrad 1971
>1. Nc5!! Rf4 (1... Rxc5 2. Qh8+ Nf8 3. Rxc5) 2. Qh8+ Nf8 3. Rd8+! Kxd8 4. Qf8+
>Kc7 5. Qxe7+ Kc6 6. Qxb7+ (1-0)

Chessmaster 9000, for the first 1:05 (on a PIII-600), never considers anything
other than Nc5. However, it feels that 1...Qh5 is Black's best reply. For a long
time after that, it feels that 1.Qh8+ is best. Then, at 11:07, it switches back
to 1.Nc5. Perhaps 1...Qh5 holds for Black?

Time	Depth	Score	Positions	Moves
0:00	1/3	0.04	3833		1.Nc5 Qh5 2.Qxh5 gxh5 3.Nxb7
0:00	1/4	0.00	10867		1.Nc5 Qh5 2.Qxh5 gxh5 3.Nxb7 h4
0:00	1/5	0.18	33081		1.Nc5 Qh5 2.Qxh5 gxh5 3.Nxb7 Re4
					4.Kc2
0:01	1/6	0.10	92341		1.Nc5 Qh5 2.Qxh5 gxh5 3.Nxe6 fxe6
					4.Rb5 b6 5.Kc2
0:02	1/7	0.08	194494		1.Nc5 Qh5 2.Qxh5 gxh5 3.g6 Rg4
					4.gxf7+ Kxf7 5.Nxb7
0:06	1/8	0.03	481071		1.Nc5 Qh5 2.Qxh5 gxh5 3.g6 Rg4
					4.gxf7+ Kxf7 5.Nxb7 Rg1+ 6.Kc2 h4
0:13	1/9	0.07	1220164		1.Nc5 Qh5 2.Qxh5 gxh5 3.g6 Rg4
					4.gxf7+ Kxf7 5.Nxb7 h4 6.Rd7 Rg1+
					7.Ka2
0:30	2/10	-0.05	2876505		1.Nc5 Qh5 2.Qxh5 gxh5 3.g6 Rg4
					4.gxf7+ Kxf7 5.Nxb7 Rg1+ 6.Kc2
					h4 7.Rd3
1:05	2/10	0.00	6209020		1.Qh8+ Nf8 2.Nf6+ exf6 3.Rd8+ Kxd8
					4.Qxf8+ Kc7 5.Qd6+ Kc8 6.Qf8+ Kc7
2:46	3/11	0.00	16278497	1.Qh8+ Nf8 2.Nf6+ exf6 3.Rd8+ Kxd8
					4.Qxf8+ Kc7 5.Qd6+ Kc8 6.Qf8+ Kc7
9:27	4/12	0.00	54626657	1.Qh8+ Nf8 2.Nf6+ exf6 3.Rd8+ Kxd8
					4.Qxf8+ Kc7 5.Qd6+ Kc8 6.Qf8+ Kc7
11:07	4/12	0.06	64298251	1.Nc5 Qh5 2.Qxh5 gxh5 3.g6 b6 4.Nxe6
					fxe6 5.Rd3 Rg4 6.Rh3 Rxg6 7.Rxh5
					Rg1+ 8.Kc2 Rg2+ 9.Kb3

>[D]4k3/pp2pp2/4n1p1/3RP1P1/2r1N3/P1P5/1P2q3/1K5Q w - -

Here's another one that Chessmaster 9000 waffles on. From 2 through 7 seconds,
it picks 1.Nxf5. Then from 0:08 through 1:29, it wants 1.Nxd5. After 1:29, it
goes back to (and sticks with) 1.Nxf5:

Time	Depth	Score	Positions	Moves
0:00	1/5	-1.19	1949		1.Ng2 Nc4 2.Nf4 Nd2 3.Ke3
0:00	2/6	-1.14	5137		1.Ng2 Nc4 2.Nf4 Nd2 3.Ke3 Ne4
0:00	3/7	-1.21	10676		1.Ng2 Nc4 2.Nf4 Nd2 3.g6 fxg6 4.Nxe6

0:00	4/8	-1.16	24050		1.Ng2 Nc4 2.Nf4 Kc6 3.g6 fxg6 4.Nxe6
					Nd2 5.Nf4
0:01	5/9	-1.24	88354		1.Ng2 Nd7 2.Nf4 Nb8 3.Nd3 Nc6+
					4.Kc3 d4+ 5.Kb3 Na5+ 6.Ka2 Nc4
					7.Kb3
0:02	5/9	-0.17	157901		1.Nxf5 exf5 2.e6 fxe6 3.g6 e5+
					4.Kxe5 Nd7+ 5.Kxf5 hxg6+ 6.Kxg6
					Nf8+ 7.Kg7 Ne6+ 8.Kg8 Ng5
0:04	6/10	0.93	380795		1.Nxf5 Nc4 2.Nd6+ Nxd6 3.exd6 Kc6
					4.Ke5 Kd7 5.c3 Kc6 6.g6 fxg6 7.Kxe6
0:07	7/11	0.93	637042		1.Nxf5 Nc4 2.Nd6+ Nxd6 3.exd6 Kc6
					4.Ke5 Kd7 5.c3 Kc6 6.g6 fxg6 7.Kxe6
0:08	7/11	1.47	769227		1.Nxd5 exd5 2.e6 fxe6 3.g6 e5+
					4.Kd3 hxg6 5.h7 e4+ 6.Ke2 f4 7.h8=Q
					f3+ 8.Kf2 Nc4
0:12	8/12	1.34	1097278		1.Nxd5 Nxd5 2.c4+ Kc6 3.cxd5+ Kd7
					4.d6 Kc6 5.Kd3 Kd7 6.Ke3 Kc6 7.Kf3
					Kd7 8.Kf4
0:23	9/13	1.42	2112221		1.Nxd5 Nxd5 2.c4+ Kc6 3.cxd5+ Kd7
					4.d6 Kc6 5.Kd3 Kd7 6.Ke3 Ke8 7.Kf3
					Kd8 8.Kf4 Kd7
0:52	10/14	1.34	4797279		1.Nxd5 Nxd5 2.c4+ Kc6 3.cxd5+ Kd7
					4.d6 Kc6 5.Kd3 Kd7 6.Ke3 Ke8 7.Kf3
					Kd8 8.Kg3 Kd7 9.Kf4
1:29	10/14	1.38	8248855		1.Nxf5 Nd7 2.Nd6+ Kc6 3.Nxf7 Nf8
					4.c4 dxc4 5.Kxc4 Ng6 6.Kd4 Ne7
					7.Nd6 Kd7 8.Ke4 Kc6
2:44	11/15	1.42	15062972	1.Nxf5 Nd7 2.Nd6+ Kc6 3.Nxf7 Nf8
					4.Nd6 Ng6 5.Ne8 Nh4 6.Kc3 d4+ 7.Kd3
					Nf3 8.Nf6 Nxg5 9.Kxd4 Nf3+ 10.Ke3
5:54	12/16	1.46	32232027	1.Nxf5 Nd7 2.Nd6+ Kc6 3.Nxf7 Nf8
					4.Nd6 Kd7 5.c4 dxc4 6.Ne4 Ke7 7.Kxc4
					Kf7 8.Kd4 Ng6 9.Nd6+ Ke7

jm



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.