Author: José de Jesús García Ruvalcaba
Date: 14:28:19 01/05/00
Go up one level in this thread
On January 05, 2000 at 14:54:51, Dann Corbit wrote: >On January 05, 2000 at 14:18:48, José de Jesús García Ruvalcaba wrote: >>Many years ago I composed the following study: >>8/8/3k3P/b4K2/8/8/8/5N2 w - - 0 1 >> According to my analysis, it is a win for white. My main line was 1. Ne3 Ke7 2. >>Kg6 Kf8 3. Nd5 Kg8 4. h7+ Kh8 5. Ne7 and mate in three more moves (this last >>move was the motiv I was trying to exploit, it is the longest forced mate with >>this material, the black bishop on dark squares the white king in g6, the white >>pawn in h7 and the black king in h8). Unfortunately, 4. Nf6+ Kh8 5. Ne4 is also >>a win for white, for this reason I did not submit this position for publication. >> I do not remember exactly why other first moves do not win, and how does white >>win after 1... Bc3. Could somebody with tablebases please confirm the following? >>-It is a win for white. >>-The only winning move is 1. Ne3 >>-The main line I posted is correct. >> Thanks in advance. Any supplementary analysis is also appreciated. > >I suggest Heiner Marxen's fabulous Chest program. It can solve any 8 move >checkmate that I have tried overnight (no matter how cluttered the board is), >and some of them in just a few moments. It will also solve almost all 9 move >checkmates overnight, but some will take a weekend. It does more than a chess >program ever could, in that it will write a detailed refutation table if you ask >it to (though for 8 moves, you have better trim it down some by starting at a >much later point or you'll be sorry). The version on Heiner's web site is >already UNIX ready, so your Solaris port should be a snap. > >http://www.drb.insel.de/~heiner/Chess/chest.html > I have saved the link. I will try to compile it next week. If you wish to keep a Solaris binary of Chest in your ftp site, just say so and I will gladly send it to you when I have it. >Heiner's tool does not use tablebase files, but that is on the todo list. > >By the way, Ne3!! is definitely the winning move, but that position takes at >least ten moves for checkmate, doesn't it? > It takes a lot longer for mate, if black is trying to delay mate. When I composed this position tablebases were unknown for most people. Delaying mate was not considered always the best defence, mainly becuase it was not known which move delayed the mate the most. That is why I considered 1... Ke7 as main line, as a human over the board it was more difficult to find a win against it. Several of my friends tried to solve this position, but only one found the correct solution in one week (at that time computers were relatively weak chess players), the same guy found the dual solution on the fourth move of my main line. >I have the 4 piece tablebase files only, but here is what Crafty says so far: [analysis snipped] Thanks, I have saved your other post with the longer analysis. At least I know now how was the win against 1... Bc3. > >Thanks for that game, by the way. It will take me at least 3 or 4 days to enjoy >it. I am having a hard time in the other games, at least with this draw I will not be shooted out. I will send them to you as I finish them.
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.