Author: blass uri
Date: 11:26:47 12/26/99
Go up one level in this thread
On December 26, 1999 at 14:14:28, Christophe Theron wrote: >On December 26, 1999 at 11:47:10, Frederic Friedel wrote: > >>On December 26, 1999 at 09:56:46, Enrique Irazoqui wrote: >> >>>>Very Very interesting, i adore this kind of stuff, i am beggining to try to >>>>solve this now, my english is regular, i like to confirn: >>>> >>>> >>>>1. e4 xx >>>>2. xx xx >>>>3. xx xx >>>>4. xx xx >>>>5. NxR++ >>>> >>>>correct? >>> >>>It could also be 5...-NxR++. It may be more likely, given the way Frederic wrote >>>it. >> >>Enrique is exactly right. It _could_ be 5...NxR#, which gives you an extra ply. >>On the other hand he will confirm that, knowing me, it might just be 5.NxR#. >> >>BTW I posted this problem on http://slashdot.org (my favourite site) last night >>at 3.00 a.m. Twelve hours later I have over 100 emails from SlashDot visitors. >>Suddenly I realise what I should really be doing. There is a huge amount of >>interest for chess out there, in the real world. >> >>No correct solutions so far from SlashDot, just clever ones, like "Knight takes >>Rook, mate!" (i.e. buddy). One correct email solution from here (Alex, don't >>post it!). And a number of correct solutions from east German GMs after the >>puzzle appeared in ChessBase Magazine 73. Apparently it was known there. > > >After all this problem is only 9 or 10 plies deep. Many programs should be able >to solve it very quickly. > >For example Fritz and Rebel-Tiger can go to ply depth 10 in only several >seconds. > >Maybe I could add a line in my program to test, in case of checkmate, if the >last move was NxR, so I would quickly get the solution. > >Maybe there are even several solutions to this problem. > >Oops... I realize it would maybe not work. Probably the solution lays in a part >of the tree that is pruned early by alphabeta. I believe that a computer program can solve it but not a chess program because this is not a chess problem because both sides have the same target. The solution in the worst case(if black mate in the 5th move) is 9 plies but you do not have to generate all the legal moves for the last plies and you can identify lot before the final position by distance between the knights and the king that there is no solution. I believe that a good program can find all the solutions in a short time. 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.