Author: Dieter Buerssner
Date: 07:24:06 08/23/03
Go up one level in this thread
On August 22, 2003 at 14:35:51, John Merlino wrote:
>>8/3p1P2/3Pp3/2p1P2K/p1p3P1/PpP3p1/1P3pP1/6k1 w - - 0 1
>>
>>1. f8=R f1=Q 2. Rxf1+ Kxf1 3. g5 Kxg2 4. g6 Kh2 5. g7 g2 6. g8=Q g1=Q 7. Qxg1+
>>Kxg1 8. Kg6 Kf2 9. Kf6 Ke2 10. Ke7 Kd2 11. Kxd7 Kc2 12. Kxe6 Kxb2 13. d7 Kxa3
>>
>>and at this juncture:
>>
>>[D]8/3P4/4K3/2p1P3/p1p5/kpP5/8/8 w - - 0 14
>>
>>white has to underpromote again in order to win...
>>
>>Mihaly
>
>Fair enough.
>
>However, I think that asking an engine to find an underpromotion 27 ply away
>from the current position is a bit much. AND it looks like finding the
>underpromotion from THIS position requires at least a dozen ply search. So,
>you're asking the engine to search 40 plies from the initial position in order
>for it to give a correct answer.
>
>Can you tell me an engine that can do this? I looked at the other replies so far
>and none of them have done it.
Yace shows it after a long search for a while:
4218848324 5:07:50 0.34 27. 1. f8=R f1=Q 2. Rxf1+ Kxf1 3. g5 Kxg2 4. g6 Kf1
5. g7 g2 6. g8=Q g1=Q 7. Qxg1+ Kxg1 8. Kg5 Kf2
9. Kf6 Ke1 10. Ke7 Kd2 11. Kxd7 Kc2 12. Kxe6
Kxb2 13. d7 Kxa3 14. d8=Q b2 15. Qd5 b1=Q 16.
Qxc4 Qb3 17. Kd5 Kb2 18. e6 a3 {HT} {0}
780537524 5:27:51 0.74 28++ 1. f8=R f1=Q 2. Rxf1+ Kxf1 3. g5 Kxg2 4. g6 Kf1
5. g7 g2 6. g8=Q g1=Q 7. Qxg1+ Kxg1 8. Kg5 Kf2
9. Kf6 Ke1 10. Ke7 Kd2 11. Kxd7 Kc2 12. Kxe6
Kxb2 13. d7 Kxa3 14. d8=R Kb2 15. Kd5 Kc2 16.
e6 b2 17. Rb8 a3 18. e7 b1=R 19. e8=Q Rxb8 {341}
1960815916 5:54:59 1.59 28t 1. f8=R Kxg2 2. Kh4 f1=Q 3. Rxf1 Kxf1 4. Kxg3
Ke2 5. g5 Kd3 6. g6 Kc2 7. g7 Kxb2 8. g8=Q Ka1
9. Qc8 b2 10. Qxd7 b1=Q 11. Qxe6 Qd3+ 12. Kg4
Qxc3 13. Qd5 Qe3 14. d7 Qg1+ 15. Kf5 c3 16. e6
c2 17. Qe5+ Ka2 18. Qe2 Kb3 {HT} {0}
Then it switches to 1...Kxg2 as the black defense, which should mean, that the
2nd underpromotion is understood. Before, the score decreased from iteration to
iteration.
Regards,
Dieter
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.