Author: leonid
Date: 07:30:00 09/21/02
Go up one level in this thread
On September 21, 2002 at 09:53:05, Paul wrote: >On September 21, 2002 at 09:37:32, leonid wrote: > >>Hi, >> >>I tried to find some old mate that represent completely "normal position". I was >>able to find only this one that was sufficently deep to presest here. All others >>are between 3 or 5 moves deep. Later, when my mate solver had no apparent >>mistakes in logic, it was essentially "heavy positions" that were created to >>verify new speed for new logic variation. >> >>[D]rnbqk3/ppp3P1/4p3/3pPp2/3P1N2/6p1/PPP5/RNBQKB1r w - - > >Hi Leonid, Hi, Paul! I see that Your program take those positions very easily. Good sign! >This one is much too easy for chess playing programs (I think) ... too many >checks! ... here's Pretz (p3/1000): > >00:00 WM8 05 g8=Q+ Ke7 Qg7+ Ke8 Qh5+ Rxh5 Qg8+ Ke7 Ng6+ Kd7 Bb5+ Nc6 Qf7+ Qe7 >Qxe7# Yes, this is position that take no time (00 sec) to be solved by selective. My initial positions were almost all done for the sake of creating quick selective search. Only later I wrote brute force search. Selective search is practical solution for finding mate in real chess game in very short time >>And one more position from my "usual" but very easy one. >> >>[D]4Rrnq/1pQ1qNkN/1P1QQnqr/q3bbRq/5Q1Q/1B4q1/qB1q3Q/4Q1QK w - - > >I like this one better :) ... > >00:10 WM9 06 Rxg6+ Rxg6 Qexf6+ Nxf6 Qhxf6+ Bxf6 Qxf6+ Rxf6 Qexg3+ Bg4 Qgxg4+ >Qag5 Qgxg5+ Kxh7 Qgxh5+ Rh6 Qhxh6# The above position is 9 moves deep. It was solved by default selective only in 10 moves (12 sec). Brute force solved it in 9 moves in 5 min and 17 sec. My usual setting: Celeron 600M, No hash. How long can take first mate in 8 by brute force I don't know. I looked it only 7 moves deep to verify that mate is really in 8. Seven moves took more that 3 minutes to say "no". Cheers, Leonid. >>Please indicate Your result. >> >>Thanks, >>Leonid. > >groetjes, >Paul
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.