Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: how to evaluate KRP vs KR (Diagram) Draw

Author: Kurt Utzinger

Date: 23:41:25 06/05/04

Go up one level in this thread


On June 05, 2004 at 15:04:22, Uri Blass wrote:

>On June 05, 2004 at 14:24:17, Kurt Utzinger wrote:
>
>>On June 05, 2004 at 12:14:03, Uri Blass wrote:
>>
>>>On June 05, 2004 at 10:59:22, Kurt Utzinger wrote:
>>>
>>>>On June 05, 2004 at 10:53:14, Kurt Utzinger wrote:
>>>>
>>>>>On June 05, 2004 at 10:30:48, Daniel Shawul wrote:
>>>>>
>>>>>>Hi
>>>>>>
>>>>>>My engine lost against Gothmog after playing Rg3?? in this position
>>>>>>
>>>>>[D]8/8/8/4k1K1/6R1/8/5rp1/8 w - - 0 1
>>>>>>
>>>>>>The position is in the 5 man TB but i don't have it yet.
>>>>>>DanChess searched to a depth of 18 and still can't solve this position.
>>>>>>What is the general guideline to evaluate this position? Should rook
>>>>>>pawn endgames(with more than one pawn in general) be evaluated differently?
>>>>>>
>>>>>>best
>>>>>>daniel
>>>>
>>>>     There is no need to have tablebases here. Some endgame
>>>>     knowledge would do the same. It's of course dead draw
>>>>     as black's king will never be able to support its
>>>>     pawn. The only thing for White is to keep its rook
>>>>     where it is and to play Kg6/Kg5 and so on.
>>>>     Kurt
>>>
>>>
>>>Some search can also do the same without lot of knowledge
>>>
>>>Here is some analysis of latest movei
>>>
>>>13 -176 4205 10834368 g4g3 f2f8 g5h6 f8h8 h6g7 h8h2 g3g4 e5f5 g4g3 f5f4 g3g6 f4
>>>f3 g7f6
>>>14 -228 7728 19816754 g4g3 f2f8 g5h6 f8h8 h6g7 h8h2 g3g4 e5f5 g4g3 f5f4 g3g6 f4
>>>e3 g6e6 e3f3 e6e1 f3f2
>>>14 -227 10356 26656758 g5g6
>>>14 -198 29157 71570225 g5g6
>>>14 -175 52458 125280528 g5g6 f2e2 g6g5 e2c2 g4g3 c2c8 g5h6 c8h8 h6g7 h8h2 g3g4
>>>e5f5 g4g6 f5f4 g7f6
>>>14 -175 52507 125405519 g5g6 f2e2 g6g5 e2c2 g4g3 c2c8 g5h6 c8h8 h6g7 h8h2 g3g4
>>>e5f5 g4g6 f5f4 g7f6
>>>
>>>
>>>
>>>Uri
>>
>>
>>     Hi Uri
>>     movei has also no idea of this endgame. The variation
>>     given by the engine is a complete loss for White
>>     14 -175 52507 125405519 g5g6 f2e2 g6g5 e2c2 g4g3
>>     this is a terrible and losing mistake
>>     c2c8 g5h6 c8h8 h6g7 h8h2 g3g4 e5f5 g4g6 f5f4 g7f6
>>     and Black has an easily won rook ending.
>>     Kurt
>
>I did not claim that it knows but only that search is enough for it to avoid the
>blunder.
>
>The blunder is in the pv later but at that point the win is too deep for it to
>see.
>
>Uri


       In that case: would movei play this rook ending
       correctly to a draw?
       Kurt



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.