Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: thought on Human Vs Programs

Author: Sandro Necchi

Date: 05:04:36 12/13/05

Go up one level in this thread


On December 13, 2005 at 07:54:18, Bernhard Bauer wrote:

>On December 13, 2005 at 07:03:46, Sandro Necchi wrote:
>
>>On December 13, 2005 at 03:50:29, Bernhard Bauer wrote:
>>
>>>On December 13, 2005 at 03:25:12, Russell wrote:
>>>
>>>>Wouldnt Programs get destroyed if they were not using any opening books or
>>>>database? From What I see even Top Programs doesnt play well without the book.
>>>
>>>As programs become stronger they will do with less book help.
>>>Long book-lines may be faulty.
>>
>>Sorry, but I do not agree...It depends how the book is made.
>
>Of course, it depends how the book is made. This answer is trivial.
>We may assume that a GM like M. Adams knows his lines very well, but in his
>first game against Hydra he met a new move 14.Rb1 instead of the usual 14.a4.
>14.Rb1 was found by Hydra.

Yes, but since he knows this line well it does not means he would play the best
move even if he knew it.
GMs tend to play moves that put in trouble the computers, so maybe this was his
plan.

>
>>
>>>Chrilly Donninger writes that Hydra's book is 150 KByte large and contains only
>>>solid main lines up to around move 10.
>>
>>If a book is really sound, it would be better than this solution. Of course if
>>the program taking over is weaker it would loose anyway.
>>
>>The "solution" of cutting the lines was made to avoid weak moves later in the
>>variations, and this can be done easily, but best would be to have a good book
>>without weak moves in the lines...it is possible.
>
>Yeah, a good book without weak moves would be best. Trivial too.
>If we had a *very* good book, we wouldn't need an engine.

Well, a good book cannot include all moves, but maybe the most important ones
that a chess engine may not be able to find or would play weaker moves.

>Solving the book problem sounds like solving chess.

Of course not. That would mean to include a huge amount of moves, but to find
the best moves is quite difficult and takes time or a very good system...

>
>You think it is possible? May be, but with a lot of work.

Yes.
This is a secret...SMK knows what I am talking about...

>
>>
>>>Kind regards
>>>bernhard
>>
>>Sandro
>Bernhard

Ciao
Sandro



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.