Reply
 
LinkBack Thread Tools Display Modes
  #1   Report Post  
Old January 4th 04, 10:10 PM
Damir Ulovec
 
Posts: n/a
Default kbrkr & knrkr

Can somebody tell me what's the purpose of this tablebase files, because most
endings with this pieces are dead draw. Only cases for win is when weaker side
can't protect his only piece, rook.
And how to know how many major iterations will be when TBGen creating
tablebase? For some takes whole eternity...
Thanks in advance...
---
Try ChessMaster's The King vs ChessBase engines on two exactly the same
computers (two computer match), and see how swiftly CM wins...



  #2   Report Post  
Old January 5th 04, 08:56 AM
Anders Thulin
 
Posts: n/a
Default kbrkr & knrkr

Damir Ulovec wrote:

Can somebody tell me what's the purpose of this tablebase files, because most
endings with this pieces are dead draw.


Answer 1: It's a byproduct of how the endgames are created. In order to
produce a database for kbr-kbr, the subgames (such as kbr-kb and kbr-kr) must
be completely known, or the scores won't be analytically correct. So, they
have to be calculated, and if they are, they could just as well be included
with the rest. If you're convinced that the chess program will do the 100%
right thing without them, you just remove the corresponding files.

Answer 2: They main purpose is para-religious. They prompt you to question
their utility, and even their purpose for existing, and so will lead you on
to ask more advanced question along the same lines: What is the purpose of the draw?
Why can't I win? Is chess the right game for me? Is there any other game that
makes it more likely to win? and finally end up questing the purpose of your
own existence, since you don't seem to be likely to do better than draw, anyway.
Surely there must be a better game somewhere?

Answer 3: It's cheaper to have the files around than to implement the
corresponding knowledge in the chess program, and also less likely
to contain any errors. And the time saved from having to write special endgame
scoring functions, or vary scoring weights with game state,and debug and fine-tune
these functions, can be more profitably spent writing better 3D board display
functions, and interfaces to Internet chess ervers, since that's where the money
is, anyway.

A few misguided souls have some exaggerated notions of absolute truth in chess,
and particularly chess endgames, and so use these files as the best approximation
of truth this side of paradise. But they're the people who took the wrong turn out
of Answer 2, so they're probably lost, anyway.

--
Anders Thulin http://www.algonet.se/~ath

Reply
Thread Tools
Display Modes

Posting Rules

Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On



All times are GMT +1. The time now is 01:42 PM.

Powered by vBulletin® Copyright ©2000 - 2019, Jelsoft Enterprises Ltd.
Copyright 2004-2019 ChessBanter.
The comments are property of their posters.
 

About Us

"It's about Chess"

 

Copyright © 2017