Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: How are the Nalimov Tablebases different from the Thompson Endgame Datab

Author: Ernst A. Heinz

Date: 11:55:39 01/24/00

Go up one level in this thread


Hi Bill,

>How are the Nalimov Tablebases different from the Thompson Endgame Databases?

Nalimov:   distance-to-mate/mated for *all* positions

Thompson:  distance-to-conversion for *selected* positions

In order to clarify the kind of "selected positions" covered by
Thompson's endgame databases, I cite from Chapter 5 of my new book
"Scalable Search in Computer Chess" which contains more detailed
information about the subject:

``Thompson's databases only host positions where the materially weaker and
  potentially losing side enjoys the right to move. His databases contain
  exact results for these positions if the weaker side actually loses. Yet
  draws and possible wins of the weaker side remain unresolved because the
  databases do not discriminate between them. Consequently, far more than
  half of the endgame positions that occur during searches and are suitable
  for database recognition at all do not apply to Thompson's files on
  average because either the materially stronger side is to move or it does
  not win. In theory this gets accounted for by 1-ply searches starting at
  the positions with the wrong side to move.''

Hope this helps,

=Ernst=

P.S.
"Scalable Search in Computer Chess" now at Amazon.de and MKP online.
http://www.amazon.de/exec/obidos/ASIN/3528057327/
http://www.mkp.com/books_catalog/3-52805-732-7.asp  :-)

Please visit http://supertech.lcs.mit.edu/~heinz/node1.html for more
information about the book.



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.