Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: u2600 Position of the Week

Author: blass uri

Date: 14:14:28 03/01/99

Go up one level in this thread



On March 01, 1999 at 15:46:14, Will Singleton wrote:

>
>On March 01, 1999 at 12:28:07, Bruce Moreland wrote:
>
>>
>>On March 01, 1999 at 05:50:44, Peter McKenzie wrote:
>>
>>>On March 01, 1999 at 04:48:28, Will Singleton wrote:
>>>
>>>>u2600 Position of the Week
>>>>--------------------------
>>>>
>>>>5R2/8/8/4rp2/3r4/5k2/2B1p2P/2K1R3 w - -
>>>>ArasanX Amateur, 0-1, 1999.02.27, fics
>>>>
>>>>In this position, White played 78. Rg8 and lost.  Could your
>>>>program avoid that move within 10 seconds or so?  The game might be
>>>>lost in any event (not sure); there is still some play left.
>>>
>>>Rxf5 Rxf5 Bxf5 Kf2 Rxe2 is a dead draw (even without the pawn on h2) - the white
>>>king is in the correct corner to easily draw this ending.  Just play the king to
>>>a1, and oscillate the bishop between b1 and a2 :-)
>>>
>>>It takes LambChop 28sec to lock onto Rxf5 (at depth 8) using a P133 searching
>>>about 20K nps.
>>>
>>>Peter
>>
>>Yes, R vs B knowledge helps with this.  1. Rxf5 is a draw.
>>
>>bruce
>
>
>Well, that's an unexpected result.  I ran this on Hiarcs, and after the
>continuation (as above), it still has white down well over a pawn.
>
>So that begs the question, does Lambchop or Ferret actually see that the
>position is drawn?  Can a tablebase tell you that?
>
>Will
I am sure tablbase of KR vs KBP can tell it
I expect good programs to see it only by evaluation
KR vs KBP is usually a draw.

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.