Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Enrique games-problem with Junior5 and adjudicated games

Author: Matthew Herman

Date: 05:08:34 01/06/99

Go up one level in this thread


On January 06, 1999 at 02:35:22, blass uri wrote:

>
>On January 05, 1999 at 21:02:35, Matthew Herman wrote:
>
>>On January 05, 1999 at 17:33:37, Enrique Irazoqui wrote:
>>
>>>On January 05, 1999 at 16:34:27, blass uri wrote:
>>>
>>>>This is the position from game 13 of enrique's games:
>>>>8/8/2k3P1/p5r1/K7/8/2B5/8 w - - 0 1
>>>>
>>>>black last move was kc5-c6
>>>>
>>>>and my Junior5 prefers 79.Bd3 at depth 20 and at bigger depthes
>>>>
>>>>In enrique's games Junior5 chose 79.Bb3 with evaluation -1.12
>>>>and my Junior5 gives 79.Bd3 with evaluation -0.xx at depth 20 and at bigger
>>>>depthes
>>>>
>>>>What is the reason for this
>>>>
>>>>Is it a bug in the Autoplayer or another reason?
>>>
>>>Amir can answer better than me. I don't know why a given program plays a given
>>>move at a given moment. Right now I can't check this out because I would have to
>>>stop playing the remaining matches. Once the tournament will end, in about 2
>>>weeks, I'll be more than glad to analize this position with J5.
>>>
>>>>Another problem is that games are adjudicated too early
>>>>
>>>>game 146 between tiger11.7.5 and Rebel10 was adjudicated as a draw in the
>>>>following position
>>>>
>>>>8/7k/3R4/7P/6K1/4b3/8/8 w - - 0 1
>>>>
>>>>Is it a draw by tablebases?
>>>
>>>Tiger and Rebel don't use tablebases. I adjudicated this game because no
>>>progress was being made. Again, I can replay from this position when the
>>>tournament will be finished.
>>>
>>>Enrique
>>>
>>>>I think that even if the result is positive the programs should play it because
>>>>black may do a mistake.
>>>>
>>>>The fact that the programs see no plan to win does not prove that it will be a
>>>>draw.
>>>>
>>>>I saw cases when programs succeed in losing dead draw positions.
>>>>
>>>>Uri
>>
>>
>>The position 8/8/2k3P1/p5r1/K7/8/2B5/8 w - - 0 1 is also a win after 1. Bb3,
>>Rxg6 as given by Ken Thompson's tablebases (for instance 2. Kxa5 ?? leads to
>>mate in 15 for black (and is also found as a strong position (+5.00 after a few
>>seconds from Fritz5.32)))
>>
>>After that black is either mating or wins the bishop in 30 moves ..
>
>Bb3 Rxg6 were the game moves but the question is if black wins after Bd3(The
>move of my Junior5).
>I think that using crafty with tablebases may help.
>
>Bb3 was a mistake because after Rxg6 there is no practical chance for white.
>
>Uri


Actually Bb3 isn't totally bad as KR+ Rookpawn vs KB is notoriously hard to win
if you dont have the tablebase..



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.