Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: KB*KP* -> some first promising results

Author: Uri Blass

Date: 06:10:17 02/03/05

Go up one level in this thread


On February 03, 2005 at 06:41:51, Thomas Mayer wrote:

>Hi Uri,
>
>On February 02, 2005 at 16:54:28, Uri Blass wrote:
>
>>On February 02, 2005 at 16:29:34, Thomas Mayer wrote:
>>
>>>Hi Uri,
>>>
>>>> I think that being good at chess is not a significant advantage.
>>>
>>>sure, but that was not what I mean. It's simple: I can not teach the engine
>>>things I do not know myself. So as long I have no idea what I should implement
>>>in the eval I can not do it... Do we agree here ? Once ? :)
>>>
>>>Greets, Thomas
>>
>>I can agree about it but I think that strong players also have no idea what to
>>implement in the evaluation.
>>
>>The main problem is that things that you know to be productive from your
>>experience can be counter productive when you test the program(because the
>>program can evaluate them indirectly by another term or because you have a bug
>>in your implementation or because they make some pruning less effective and it
>>is hard to know the exact problem).
>
>Well, that is clear -> evaluation is always a try out -> does it help ? Does it
>slow down too much ? Etc. -> But when you are weak in chess your problem is:
>What is missing ? What can I try ? That's my problem in middlegame...
>In the endgame it is different, there are a lot of books with clear rules and
>ideas -> it's not too complicate to program them and to try them and to throw
>them out if they do not help or if they are counterproductive.
>For most books about middlegames I am simply too stupid -> They explain games
>and gave lines why this and that is not as good... I usually can not create
>rules out of that. E.g. I did not understand the idea of good and bad bishop
>(except a totally blocked one) for a long time.

Movei also does not have a special knowledge about bad bishops and I think that
this knowledge is usually included in mobility of the boshop and bishop with bad
mobility is a bad bishop.

I do not know more than it as a player(or at least I am not aware of more
knowledge that I know about it as a player) and I do not think that I am a weak
player(my rating is near 2000 and I believe that it can be easily more than 2100
if I only learn to be better in tactics.

I thought about writing a program to teach me to be better in tactics and I have
an idea about a program that teach me by asking me questions and respond to them
but unfortunaly I had to give up because I do not know how to write an interface
so it seems that I will never be able to know if my idea is productive to teach
me to play better.

I could try to ask the same questions without board but I doubt if I can learn
to play blindfold and I do not like the idea of a program that give me
situations that I do not need to deal with in games so I want a program that
show me a diagram ask question and respond to my reply and I need an interface
for it.

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.