Sign in with
Sign up | Sign in
Your question

Endgame analysis beyond tablebases

Last response: in Video Games
Share
Anonymous
June 10, 2005 6:35:56 PM

Archived from groups: rec.games.chess.computer,rec.games.chess.analysis (More info?)

[ Followup-To: rec.games.chess.computer ]

There's an interesting article at Chess Cafe on computer analysis beyond
what is currently possible with tablebases.

http://www.chesscafe.com/mueller/mueller.htm

(which will soon be consigned to the archives at

http://www.chesscafe.com/text/mueller50.pdf

-- why does Chess Cafe archive in PDF instead of HTML?)

The idea is that a human annotates a position with commets such as `White
loses if the black king gets to the sixth rank'. The the computer can
then do the sort of retrograde analysis that is used to construct
tablebases (using existing tablebases where available) and determine the
result of the position. Of course, the results might not be correct if
the annotation are wrong. In particular, if the annotations say that
White's king must stay within a particular area to hold the draw, the
system will miss any possible wins where the white king is more
adventurous.

This doesn't seem to be much use for improving engines' endgame play but
looks like it could be handy for analysis purposes. More technical
details can be found at

http://www.minet.uni-jena.de/preprints/bleicher_04/FREE...

The software's homepage is at

http://www.shredderchess.com/freezer/index.php?topic=ho...

Alas, you have to pay for it.


Dave.

--
David Richerby Old-Fashioned Perforated Gnome (TM):
www.chiark.greenend.org.uk/~davidr/ it's like a smiling garden ornament
but it's full of holes and perfect
for your grandparents!
June 19, 2005 9:56:53 PM

Archived from groups: rec.games.chess.computer (More info?)

Hi David,
Thanks for the link to the article. sorry if I have no comments as of yet. I
need to understand the theory first.

Rubix

"David Richerby" <davidr@chiark.greenend.org.uk> wrote in message
news:yah*pxMQq@news.chiark.greenend.org.uk...
>[ Followup-To: rec.games.chess.computer ]
>
> There's an interesting article at Chess Cafe on computer analysis beyond
> what is currently possible with tablebases.
>
> http://www.chesscafe.com/mueller/mueller.htm
>
> (which will soon be consigned to the archives at
>
> http://www.chesscafe.com/text/mueller50.pdf
>
> -- why does Chess Cafe archive in PDF instead of HTML?)
>
> The idea is that a human annotates a position with commets such as `White
> loses if the black king gets to the sixth rank'. The the computer can
> then do the sort of retrograde analysis that is used to construct
> tablebases (using existing tablebases where available) and determine the
> result of the position. Of course, the results might not be correct if
> the annotation are wrong. In particular, if the annotations say that
> White's king must stay within a particular area to hold the draw, the
> system will miss any possible wins where the white king is more
> adventurous.
>
> This doesn't seem to be much use for improving engines' endgame play but
> looks like it could be handy for analysis purposes. More technical
> details can be found at
>
> http://www.minet.uni-jena.de/preprints/bleicher_04/FREE...
>
> The software's homepage is at
>
> http://www.shredderchess.com/freezer/index.php?topic=ho...
>
> Alas, you have to pay for it.
>
>
> Dave.
>
> --
> David Richerby Old-Fashioned Perforated Gnome
> (TM):
> www.chiark.greenend.org.uk/~davidr/ it's like a smiling garden
> ornament
> but it's full of holes and
> perfect
> for your grandparents!
!