Author: Uri Blass
Date: 07:51:16 03/15/02
Go up one level in this thread
On March 15, 2002 at 10:40:12, Uri Blass wrote: >On March 15, 2002 at 01:23:43, Ed Schröder wrote: > >>On March 14, 2002 at 20:19:58, Art Basham wrote: >> >>>Someone here posted this problem about 2 months ago, >>>and I just "fell in love" with it..! >>> >>>White plays 1. Ne3-g4+ and mates in 30...WITHOUT a Queen on the board! >>> >>>Be sure to play this out to the end, and see the beauty found in this >>>unique chess masterpiece! >>> >>>[D]5n2/B3K3/2p2Np1/4k3/7P/3bN1P1/2Prn1P1/1q6 w - - >>> >>>My Crafty engine ann. "mate in 30" in only 25 sec...! >>> >>>Have fun with it! >>> >>>Why not post your favorite chess problem here, >>>Or, It may just be your favorite move in chess... >>>the move that really surprised you, >>>or "knocked your socks off" so to speak..! >>> >>>(The man who taught me how to play chess was my Latin and math teacher >>>in H. S....If he were living today, he would be just thrilled to see >>>what chess computors can do_! >> >> >>It's a known position posted many times here, there are several chess programs >>which solves this instantly, pick any Rebel version it will announce the mate in >>less than a 1/10 of a second. >> >>Ed > >I have a special version of my program that can also do it but I am not going to >use it in games. > >This special version needs only 4476 nodes to find the mate by extending 2 plies >for every single reply that happens in the first 60 plies but it is a bad idea >to use it in games because there are position when >it can in theory crush at depth 1 because the attacker may have 10 possible >moves to force single reply in every move and in 60 plies it may need 10^30 >nodes. > >It is possible to decide to use that special version only for 10000 nodes >and continue with the original version if that version does not find a mate. > >There will not be significant difference when you do not consider blitz >but I do not like that solution. > >Here is the winboard.debug file of my special program: > >61058 >first : black >61058 >first : time 30000 >otim 30000 >61058 >first : white >go >61068 <first : Error (unknown command): black >61068 <first : Error (unknown command): white >61128 <first : 1 9941 6 4476 e3g4 e5f5 g4h6 f5e5 h6f7 e5f5 f7d6 f5e5 f6g4 e5d5 >g4e3 d5e5 d6f7 e5e4 f7g5 e4e5 g5f3 e5e4 f3d2 e4e5 d2f3 e5e4 f3g5 e4e5 g5f7 e5e4 >f7d6 e4e5 e3g4 e5d5 g4f6 d5e5 d6f7 e5f5 f7h6 f5e5 h6g4 e5f5 g4e3 f5e5 f6g4 e5e4 >g4f2 e4e5 f2d3 e5e4 d3f2 e4e5 f2g4 e5e4 g4f6 e4e5 e3c4 e5f5 c4d6 f5e5 f6g4 e5d5 >c2c4 >61138 <first : move e3g4 >GameEnds(0, (null), 2) >Interrupting first >70071 >first : force >70071 >first : quit > >Uri Note that things are easier when you consider only single reply to check extension because in that case there are not many options to force a single reply but I also believe that exrtending every single reply to check by 2 plies is going to do the program weaker and I do not like solutions of extending only when all the moves are single reply to check because it means that one quiet move in the end of the line may prevent the program to find the solution. Single reply to check extension by 2 plies also does not help to solve the mate in 19 that chessmaster could solve in an hour. I believe that I may try different ideas but I am not going to do my program weaker. Uri
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.