Author: leonid
Date: 15:00:40 02/06/01
Go up one level in this thread
On February 06, 2001 at 16:58:23, Heiner Marxen wrote: >On February 06, 2001 at 16:29:06, leonid wrote: > >>On February 06, 2001 at 08:59:10, Heiner Marxen wrote: >> >>>On February 06, 2001 at 06:21:32, leonid wrote: >>> >>>>Hi! >>>> >>>>If you like to solve forced mate... >>>> >>>>[D]BKB4R/R1R5/6r1/B3bbr1/4bbb1/1R1bbbb1/3b4/6k1 b - - >>>> >>>>Please say your result. >>>>Leonid. >>> >>>Huh, for this one Chest finds 6 different solutions in 7 moves: >>> >>>Bxa5 Rb1+ Bxb1 Rh1+ Bxh1 Bc6 Bexc7+ Ka8 Bxc6+ Rb7 Bxb7+ Bxb7 Rg8# >>>Bxa8 Rab7 Bxc7+ Bxc7 Baxb7 Rb1+ Bxb1 Rh1+ Bxh1 Bxb7 Bxc7+ Ka8 Rg8# >>>Bxc7+ Bxc7 Bxa7+ Kxa7 Bde3+ Rb6 Rxb6 Bc6 Rxc6+ Bb6 Bb8+ Ka8 Rxc8# >>>Bxc8 Bc6 Bxc7+ Bxc7 Bxa7+ Ka8 Bxc6+ Kxa7 Bde3+ Rb6 Ra5+ Kb8 Bxc7# >>>Bxh8 Bab7 Bxc7+ Bxc7 Bxc7+ Ka8 Bxb7+ Rbxb7 B5e4 Bf5 Bxb7+ Rxb7 Ra6# >>>Rg8 Bb6 Bxc7+ Bxc7 Rxh8 Ra1+ Bde1 Rxe1+ Bxe1 Bb7 Bxc7+ Ka8 Rxc8# >>> >>>7.7 minutes K7/600 60MB hash. >>>In the last depth there was sort of an explosion: >>># 1 0.00 0.87 1- 0 >>># 2 0.00 1.00 1- 0 >>># 3 0.01 0.97 57- 0 >>># 4 0.10 1.07 599- 0 >>># 5 0.78 1.19 4747- 0 >>># 6 9.84 1.33 58492- 0 >>># 7 456.38 1.67 2705767- 1136381 >>>(depth seconds speed nodes in nodes out) >>> >>>I have not yet tried with side to move switched ;-) >>> >>>Heiner >> >>Probably this position is the champion by the number of possible solutions. >> >>Mine can't take this position by selective in 7 moves. Only brute force do the >>work. In 7, first solution come in 1 min 12 sec. It is nice that you find in 7. >>Without this was not even sure if search by brute force have some sense. Mine >>first solved position by default selective in 13 moves. Too far away! Time, 0.33 >>sec. >> >>Cheers, >>Leonid. > >I do not quite understand... when your program finds a forced mate in 0.33 >seconds, isn't that great? This "shortest mate" business as done by Chest >is interesting, but not a big deal for a chess playing program, I think. >For a playing program it is important to quickly find a forced mate, >if there is one, but it need not at all be the shortest. It shall be quick. >Do you disagree? > >Heiner Here are few "buts" and this is why I ask from time to time to put here (if somebody really have) few positions from real chess championship. Game played by human. Final response of how effective can be mate solver (for now I see its effectiveness only inside of chess game) can be founds by statistics for positions arrived from real games. This last position can give very wrong impression about my program capacity. It could be so quickly solvable by my program because I compose it using my selective search on the first place. How many my selective search missed existing mates, before I reached this last position, nobody knows. The fact that selective found mate in six moves from shortest, make me feel miserable. Probably in the future I must try to improve capacity of my selective search more that I ever expected. This is why those few positions that I put here and found some varied responses to their solutions are so valuable. Alone I could hardly obtain data that I badly need. Leonid. Leonid.
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.