Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: ECE #9

Author: José de Jesús García Ruvalcaba

Date: 15:07:46 03/18/99

Go up one level in this thread


On March 18, 1999 at 16:54:12, Dann Corbit wrote:

>On March 17, 1999 at 06:55:58, Steffen Jakob wrote:
>
>>I also dont understand ECE #9. It doesn't matter what black plays. He is lost...
>>ECE doesn't seem to be very usefull for automatic endgame testing. Any comments?
>
>You are right.  That is a typical test suite bust.  Here is the C.A.P. data:
>
>8/2P5/1K6/3r4/8/8/8/k7 b - -; acd 61; acn 17901; ce -32717; pv Rd6+ Kb5 Rd5+; bm
>Kb2; pm Rd6+; id ECE.0009;

	This is the famous Saavedra position. Black is lost, and I do not think anybody
intended this position to be a test, I think this is a very interesting and
revealing position and for that reason it is famous.
	There is a way to make it a test position, place white's pawn in c6 instead of
c7 and make it white's turn. It should be then:

8/8/1KP5/3r4/8/8/8/k7 w - -; bm c7
(sorry if it is wrong, I typed the epd string).
	It is interesting not only that a program finds 1. c7, but also the score and
the PV.
	By the way, the Saavedra solution would be 1. c7 Rd6+ 2. Kb5 Rd5+ 3. Kb4 Rd4+
4. Kb3 Rd3+ 5. Kc2 Rd4 6. c8=R Ra4 7. Kb3
	It seems that 1... Rd6+ is not the move that delays mate the most, but for
practical purposes it is better than Kb2. Well, who knows, the ending KQ vs KR
can be difficult to win against a computer with tablebases, but at Saavedra's
time there were no tablebases.




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.