Author: Christophe Theron
Date: 15:01:20 04/21/02
Go up one level in this thread
On April 21, 2002 at 01:50:34, Uri Blass wrote:
>On April 21, 2002 at 00:12:02, Christophe Theron wrote:
>
>>On April 20, 2002 at 12:35:04, Uri Blass wrote:
>>
>>>On April 20, 2002 at 11:53:46, Christophe Theron wrote:
>>>
>>>>On April 20, 2002 at 03:01:01, Peter Berger wrote:
>>>>
>>>>>On April 20, 2002 at 00:29:10, Uri Blass wrote:
>>>>>
>>>>>>The fact that tiger did no progress in the last moves is not a good reason to
>>>>>>afjudicate it as a draw.
>>>>>>
>>>>>>Tiger has KRN vs K and the game should continue.
>>>>>>
>>>>>>Maybe tiger does not know to win KRN vs K but the 50 move rule are going to
>>>>>>force it to sacrifice the knight and get KR vs K that it knows to win.
>>>>>>
>>>>>>I do not know about the knowledge of palm tiger but it is one possibility.
>>>>>>
>>>>>>Uri
>>>>>
>>>>>You are completely right and I came to the same conclusion. It was partly a
>>>>>technical decision as I was really too tired to watch this game any longer. And
>>>>>I expect exactly the same as you ( the Tiger sacrifices the knight because of
>>>>>the 50 move rule and then will know how to mate). I think I will let them replay
>>>>>this endgame (sigh).
>>>>>
>>>>>Peter
>>>>
>>>>
>>>>
>>>>I think that you are right and that the 50 moves rule is going to save Tiger
>>>>here.
>>>>
>>>>As I told you in email, there is a bug in version 14.9. Tiger thinks this KRN/K
>>>>endgame is actually a KRN/KR. So it uses the wrong knowledge here because of
>>>>this bug and is not trying to push the king to the border to mate it.
>>>
>>>I do not understand it because even in KRN vs KR it is better for the stronger
>>>side to reduce the distance between the kings and palm tiger did not do an
>>>effort to do it(even if you use only piece square table evaluation then
>>>centralizing the king should give better moves than the moves that I saw).
>>>
>>>Uri
>>
>>
>>
>>Tiger just consider KRN/KR as very drawish, and has no specific knowledge apart
>>from that.
>>
>>I wonder if you can win KRN/KR in comp vs comp games, in the general case. I
>>don't think so, so I did not spend much time on it.
>
>It is usually a draw but the same rules that are good for KRB vs KR or KQ vs KR
>are also good for KRN vs KR so I do not think there is a need for a special
>code.
Well... I have no special code for KRB/KR or KQ/KR either!
Christophe
>Usually there is not a win but in positions that there is a win the winner
>usually put the king in a better square.
>
>If I remember correctly then
>I rememeber a case when gandalf4.32 lost KR vs KRN agaionst Fritz6 when the
>sides did not use tablebases so knowing nothing is not good there.
>
>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.