Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Revised endgame suite

Author: Carsten Kossendey

Date: 15:47:02 01/29/98

Go up one level in this thread


On January 29, 1998 at 18:18:09, Howard Exner wrote:

>Any owners of the newer commercial programs want to try this one?
>Harald or Kai ? I've got some non-commercial results and am trying to
>determine if some of the positions are either too easy or too hard.
>Thanks
>
>Run for 10 minutes or until you know for sure that the program will keep
>the move. Is this the correct EPD notation?

No.

>I normally use
>Rebel's format but have been corrected by some posters not to include
>the = sign.
>
>Q6n/8/8/8/8/6K1/pk6/8 w - - id  "XEND01";bm not Qxh8

If some move should NOT be played that's "am" (avoid move) instead of
"bm" (best move). Also I think each operator must be preceded by a
space.

>8/8/8/1P2kp2/P2p2p1/3R2P1/1r6/4K3 w - - id "XEND02";bm Ra3
>8/8/6p1/6K1/R2k3P/8/p7/r7 b - - id "XEND03";bm Kc5
>8/Bpk5/8/P2K4/8/8/8/8 w - - id "bm"XEND04";bm Kd4
>8/1pPK3b/8/8/8/6k1/8/8 w - - id "bm"XEND05";bm Kc8

You got the id operators broken here.

>8/8/4K3/2k5/p6p/2P5/1P4P1/8 w - - id "XEND06";bm Kd7
>8/5p2/8/5R2/6P1/8/5rk1/r2RK3 b - - id "XEND07";bm Rf1
>8/1Q5n/8/8/8/2p5/7K/2k5 w - - id "bm"XEND08";bm not Qxh7
>8/7p/R7/7k/8/p5K1/8/r7 b - - id "bm"XEND09";bm Kg5 (a2?)

This is a tough one. I generally think it is okay to have both "bm" and
"am" operators in one position (if somewhat senseless tho since "bm"
already suggests all possible solutions are given), but it will cause
some programs to freak out (Crafty for sure).

>8/6p1/8/3kP2P/6K1/8/8/2b5 w - - id "XEND10";bm h6
>2b5/2k5/p1p3p1/Pp1p1p1p/1P1P3P/5PP1/2PNK3/8 w - - id "XEND11";bm Ke3
>8/5p2/p3p1p1/1pbkP2p/5P1P/1P4P1/P3KB2/8 w - - id "XEND12";bm Be1 or
>Be3(Bxc5?)

When there are multiple solutions you just list them, separated by
spaces

>1r6/5k2/p4p1K/5R2/7P/8/6P1/8 w - - id "XEND13";bm Kh7
>8/6k1/p4p2/P3q2p/7P/5Q2/5PK1/8 w - - id "XEND14";bm Qg3
>8/6p1/1p5p/1R1rk3/4p3/1P2P3/1K4PP/8 w - - id "XEND15";bm Rxd5
>8/8/6p1/3PkP2/8/2K5/6P1/n7 b - - id "XEND16";bm Kxf5
>8/8/3k4/3p2n1/b3p1p1/1N2PpP1/2BK1P2/8 w - - id "XEND17";bm Nd4
>8/r4pK1/2k1p1p1/p3P3/R4P1P/8/8/8 w - - id "XEND18";bm f5
>8/8/pp6/3b2K1/1P6/PN2k3/8/8 w - - id "XEND19";bm Na5
>8/8/2p5/3p1k2/1P1P2pp/1P3P2/1P3KP1/8 b - - id "XEND20";bm Kf4
>r5r1/8/bRP1pk1p/3pNp2/5P2/7P/PR4P1/6K1 w - - id "XEND21";bm Rxa6
>8/4bBp1/4P3/6pP/3k1PK1/8/8/8 w - - id "XEND22";bm Kf5
>
>in #3 Kd5 is ok but not quite as accurate as Kc5. Kc3? throws the win
>away.
>#1 and #8 test knowledge but can now be solved with raw search on the
>faster hardware.
>
>The toughest ones so far seem to be 3,9,12,14,16,18,19 and 22.

Here it goes again, corrected:

Q6n/8/8/8/8/6K1/pk6/8 w - - id "XEND01"; am Qxh8
8/8/8/1P2kp2/P2p2p1/3R2P1/1r6/4K3 w - - id "XEND02"; bm Ra3
8/8/6p1/6K1/R2k3P/8/p7/r7 b - - id "XEND03"; bm Kc5
8/Bpk5/8/P2K4/8/8/8/8 w - - id "XEND04"; bm Kd4
8/1pPK3b/8/8/8/6k1/8/8 w - - id "XEND05"; bm Kc8
8/8/4K3/2k5/p6p/2P5/1P4P1/8 w - - id "XEND06"; bm Kd7
8/5p2/8/5R2/6P1/8/5rk1/r2RK3 b - - id "XEND07"; bm Rf1
8/1Q5n/8/8/8/2p5/7K/2k5 w - - id "XEND08"; am Qxh7
8/7p/R7/7k/8/p5K1/8/r7 b - - id "XEND09"; bm Kg5
8/6p1/8/3kP2P/6K1/8/8/2b5 w - - id "XEND10"; bm h6
2b5/2k5/p1p3p1/Pp1p1p1p/1P1P3P/5PP1/2PNK3/8 w - - id "XEND11"; bm Ke3
8/5p2/p3p1p1/1pbkP2p/5P1P/1P4P1/P3KB2/8 w - - id "XEND12"; bm Be1 Be3
1r6/5k2/p4p1K/5R2/7P/8/6P1/8 w - - id "XEND13"; bm Kh7
8/6k1/p4p2/P3q2p/7P/5Q2/5PK1/8 w - - id "XEND14"; bm Qg3
8/6p1/1p5p/1R1rk3/4p3/1P2P3/1K4PP/8 w - - id "XEND15"; bm Rxd5
8/8/6p1/3PkP2/8/2K5/6P1/n7 b - - id "XEND16"; bm Kxf5
8/8/3k4/3p2n1/b3p1p1/1N2PpP1/2BK1P2/8 w - - id "XEND17"; bm Nd4
8/r4pK1/2k1p1p1/p3P3/R4P1P/8/8/8 w - - id "XEND18"; bm f5
8/8/pp6/3b2K1/1P6/PN2k3/8/8 w - - id "XEND19"; bm Na5
8/8/2p5/3p1k2/1P1P2pp/1P3P2/1P3KP1/8 b - - id "XEND20"; bm Kf4
r5r1/8/bRP1pk1p/3pNp2/5P2/7P/PR4P1/6K1 w - - id "XEND21"; bm Rxa6
8/4bBp1/4P3/6pP/3k1PK1/8/8/8 w - - id "XEND22"; bm Kf5



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.