Author: Robert Hyatt
Date: 21:19:57 12/01/01
Go up one level in this thread
On December 01, 2001 at 23:53:30, Peter McKenzie wrote:
>I am working on an endgame test suite, and finding it surprisingly difficult to
>get good test positions. It is interesting that many of the positions found in
>chess books either have a dual solution, or (less common) the solution has a
>mistake.
>
>Take the following position for example:
>
>[D]8/3n1pkp/6p1/1P1N4/3K2P1/7P/8/8 w - -
>
>Chigorin - Marshall, Karlsbad 1907
>
>Chigorin played the thematic 1.g5!, which clearly wins as it prevents any black
>counter-play.
>
>But I think that most computers will initially prefer 1.b6, and it is unclear if
>this is winning or not. I think tablebases could be quite useful here, so
>perhaps someone could run crafty/yace/fritz/tiger etc on it for a decent amount
>of time.
>
>One possible attempted defence is 1.b6 f5 2.gf gf 3.b7, it might be worth
>starting a search from there.
>
>cheers,
>Peter
Crafty grabs onto g5 immediately and sticks with it....Until the move
fails low at a fairly deep search... Perhaps g5 isn't the right plan,
as the EGTBs might be seeing a subtle drawing plan that is not
obvious. Here is the output
from my PIII/750 laptop: (all 5 piece EGTBs used)
nss depth time score variation (1)
1 0.03 2.48 1. b6
1-> 0.04 2.48 1. b66
2 0.04 -- 1. b6
2 0.05 1.66 1. b6 Nxb6 2. Nxb6
2 0.06 1.83 1. g5 h6
2-> 0.07 1.83 1. g5 h6
3 0.08 1.92 1. g5 h6 2. Nf6
3-> 0.09 1.92 1. g5 h6 2. Nf6
4 0.09 1.75 1. g5 h6 2. h4 hxg5 3. hxg5
4-> 0.11 1.75 1. g5 h6 2. h4 hxg5 3. hxg5
5 0.11 1.82 1. g5 h6 2. h4 hxg5 3. hxg5 Kh7
5-> 0.13 1.82 1. g5 h6 2. h4 hxg5 3. hxg5 Kh7
6 0.14 1.87 1. g5 h6 2. h4 Nf8 3. Ke5 Ne6
(2) 6-> 0.15 1.87 1. g5 h6 2. h4 Nf8 3. Ke5 Ne6
7 0.16 1.82 1. g5 h6 2. h4 Nf8 3. gxh6+ Kxh6 4.
b6 Ne6+ 5. Ke5 Kg7
(4) 7-> 0.22 1.82 1. g5 h6 2. h4 Nf8 3. gxh6+ Kxh6 4.
b6 Ne6+ 5. Ke5 Kg7
(3) 8 0.24 1.73 1. g5 h6 2. h4 hxg5 3. hxg5 Nf8 4.
Nc7 Nh7 5. Ne8+ Kf8
(2) 8 0.37 1.84 1. Ke4 h6 2. b6 Nxb6 3. Nxb6 Kf6 4.
Nd5+ Kg5 5. Ke5
8-> 0.40 1.84 1. Ke4 h6 2. b6 Nxb6 3. Nxb6 Kf6 4.
Nd5+ Kg5 5. Ke5
9 0.63 1.89 1. Ke4 h6 2. b6 Nxb6 3. Nxb6 h5 4.
gxh5 gxh5 5. Nd5 Kg6
9 0.68 2.02 1. g5 h6 2. h4 hxg5 3. hxg5 Kf8 4.
b6 Nxb6 5. Nxb6 Kg7
(2) 9-> 0.73 2.02 1. g5 h6 2. h4 hxg5 3. hxg5 Kf8 4.
b6 Nxb6 5. Nxb6 Kg7
10 1.00 2.30 1. g5 h6 2. h4 h5 3. Ke3 Nc5 4. b6
Nb7 5. Kd4 Kh7
10-> 1.45 2.30 1. g5 h6 2. h4 h5 3. Ke3 Nc5 4. b6
Nb7 5. Kd4 Kh7
11 1.95 2.40 1. g5 h6 2. h4 h5 3. Ke3 Nc5 4. b6
Nb7 5. Kd4 Kh7 6. Ke5
11-> 3.25 2.40 1. g5 h6 2. h4 h5 3. Ke3 Nc5 4. b6
Nb7 5. Kd4 Kh7 6. Ke5
12 4.38 2.33 1. g5 h6 2. h4 h5 3. Ke3 Nc5 4. b6
Nb7 5. Kd4 Kf8 6. Ke5 Kg7
(3) 12-> 7.71 2.33 1. g5 h6 2. h4 h5 3. Ke3 Nc5 4. b6
Nb7 5. Kd4 Kf8 6. Ke5 Kg7
(2) 13 10.30 2.36 1. g5 h6 2. h4 h5 3. Kc4 Kf8 4. b6
Nxb6+ 5. Nxb6 Ke7 6. Nd5+ Ke6 7. Kd4
Kf5 8. Ke3
(2) 13-> 23.07 2.36 1. g5 h6 2. h4 h5 3. Kc4 Kf8 4. b6
Nxb6+ 5. Nxb6 Ke7 6. Nd5+ Ke6 7. Kd4
Kf5 8. Ke3
14 29.76 2.29 1. g5 h6 2. h4 h5 3. Kc4 Kf8 4. b6
Nxb6+ 5. Nxb6 Ke7 6. Nd5+ Ke6 7. Kd4
Kf5 8. Ke3 Kg4
(4) 14-> 1:00 2.29 1. g5 h6 2. h4 h5 3. Kc4 Kf8 4. b6
Nxb6+ 5. Nxb6 Ke7 6. Nd5+ Ke6 7. Kd4
Kf5 8. Ke3 Kg4
15 1:39 -- 1. g5
(3) 15 2:03 1.85 1. g5 f5 2. gxf6+ Kf7 3. Ke4 Ke6 4.
b6 g5 5. f7 Kxf7 6. Kf5 h6 7. b7 Nb8
8. Ke5 Na6 9. Nf6 Nb8
(4) 15-> 5:57 1.85 1. g5 f5 2. gxf6+ Kf7 3. Ke4 Ke6 4.
b6 g5 5. f7 Kxf7 6. Kf5 h6 7. b7 Nb8
8. Ke5 Na6 9. Nf6 Nb8
(3) 16 6:51 2.07 1. g5 f5 2. b6 Nxb6 3. Nxb6 h6 4. h4
hxg5 5. hxg5 Kf7 6. Ke5 Kg7 7. Nc8
<HT>
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.