A Chess forum. ChessBanter

If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below.

Go Back   Home » ChessBanter forum » Chess Newsgroups » rec.games.chess.computer (Computer Chess)
Site Map Home Register Authors List Search Today's Posts Mark Forums Read Web Partners

Castling bug in Chessmaster 9000 tablebases?



 
 
Thread Tools Display Modes
  #1  
Old August 28th 03, 05:47 AM
Peter Bereolos
external usenet poster
 
Posts: n/a
Default Castling bug in Chessmaster 9000 tablebases?

I think I may have found a bug in the CM9000 tablebases. I've patched CM9000
to version 1.0.2 and regenerated the tablebases with FEG 3.03c. I have found
problems in positions where White can castle like
W: Ke1 Ra1 Rh2 B: Ke8 .Here, every legal move for White will win,
but when I go under Mentor/Chess Coach, it only reports mate in # for king moves,
moves by Ra1 and 0-0-0. For moves by Rh2 there are no comments (except Rh8
"check", Re2 "check" and Rh7 "threatens mate") which from my understanding
implies that those positions are designated as not winning for White in the
tablebase. Replacing the Rh2 with other pieces doesn't help. The problem
persists in 5 and 6 piece endings, too. An extreme example is
W: Ra1 Ke1 Pa2 B: Ke8 Rb3 none of the moves including axb3! are given any
comment. However, take away the possibility of castling by putting the king
on d1 instead of e1 and it reports axb3 - White mates in 12.

Have I found a bug, or am I doing something wrong.

Pete
bereolos.tripod.com/chess.html
  #3  
Old August 28th 03, 11:46 PM
Peter Bereolos
external usenet poster
 
Posts: n/a
Default Castling bug in Chessmaster 9000 tablebases?

(John Merlino) wrote in message . com...
(Peter Bereolos) wrote in message . com...
I think I may have found a bug in the CM9000 tablebases. I've patched CM9000
to version 1.0.2 and regenerated the tablebases with FEG 3.03c. I have found
problems in positions where White can castle like
W: Ke1 Ra1 Rh2 B: Ke8 .Here, every legal move for White will win,
but when I go under Mentor/Chess Coach, it only reports mate in # for king moves,
moves by Ra1 and 0-0-0. For moves by Rh2 there are no comments (except Rh8
"check", Re2 "check" and Rh7 "threatens mate") which from my understanding
implies that those positions are designated as not winning for White in the
tablebase. Replacing the Rh2 with other pieces doesn't help. The problem
persists in 5 and 6 piece endings, too. An extreme example is
W: Ra1 Ke1 Pa2 B: Ke8 Rb3 none of the moves including axb3! are given any
comment. However, take away the possibility of castling by putting the king
on d1 instead of e1 and it reports axb3 - White mates in 12.

Have I found a bug, or am I doing something wrong.

Pete
bereolos.tripod.com/chess.html


This is not a bug, nor have you done something wrong.

Many tablebase formats intentionally do not support castling, because
it "never" happens in a real game when there are only 5 or 6 pieces
left on the board. This is also true of the de Koning format.

jm



I think you are misunderstanding my example, or I am not presenting it
correctly. The bug has nothing to do with supporting castling, it
just
occurs in positions with a White King on e1 and a White Rook in a
corner.

In the first example
W: Ke1 Ra1 Rh2 B: Ke8

Clearly any legal move by White wins. Moves by the king or Ra1
(including 0-0-0) are noted as White Mates in #. However, moves by
Rh2
do not have a "White Mates in" message, implying that White does not
win.

Pete
  #4  
Old August 29th 03, 04:37 AM
John Merlino
external usenet poster
 
Posts: n/a
Default Castling bug in Chessmaster 9000 tablebases?

(Peter Bereolos) wrote in message om...
(John Merlino) wrote in message . com...
(Peter Bereolos) wrote in message . com...
I think I may have found a bug in the CM9000 tablebases. I've patched CM9000
to version 1.0.2 and regenerated the tablebases with FEG 3.03c. I have found
problems in positions where White can castle like
W: Ke1 Ra1 Rh2 B: Ke8 .Here, every legal move for White will win,
but when I go under Mentor/Chess Coach, it only reports mate in # for king moves,
moves by Ra1 and 0-0-0. For moves by Rh2 there are no comments (except Rh8
"check", Re2 "check" and Rh7 "threatens mate") which from my understanding
implies that those positions are designated as not winning for White in the
tablebase. Replacing the Rh2 with other pieces doesn't help. The problem
persists in 5 and 6 piece endings, too. An extreme example is
W: Ra1 Ke1 Pa2 B: Ke8 Rb3 none of the moves including axb3! are given any
comment. However, take away the possibility of castling by putting the king
on d1 instead of e1 and it reports axb3 - White mates in 12.

Have I found a bug, or am I doing something wrong.

Pete
bereolos.tripod.com/chess.html


This is not a bug, nor have you done something wrong.

Many tablebase formats intentionally do not support castling, because
it "never" happens in a real game when there are only 5 or 6 pieces
left on the board. This is also true of the de Koning format.

jm



I think you are misunderstanding my example, or I am not presenting it
correctly. The bug has nothing to do with supporting castling, it
just
occurs in positions with a White King on e1 and a White Rook in a
corner.

In the first example
W: Ke1 Ra1 Rh2 B: Ke8

Clearly any legal move by White wins. Moves by the king or Ra1
(including 0-0-0) are noted as White Mates in #. However, moves by
Rh2
do not have a "White Mates in" message, implying that White does not
win.

Pete


Well, I guess I got the idea that the bug might have to do with
castling from your subject, which says "Castling bug...?" :-)

And, the problem DOES have to do with castling. Note that if you move
either the White King or the Rook on a1, then the problem goes away.
Basically, as I said before, any position in which castling is legal
will be IGNORED by the de Koning endgame databases.

So, for your initial position, all moves by the h2 Rook will NOT have
the "White Mates in" message because castling for White is still legal
after that rook moves. If any of the other pieces move, INCLUDING
castling to the queenside, then castling is no longer available, and
all such moves DO have the mate in N message.

The best way to show this is to take your original position, and make
the following moves:

1. Kd1 Kd8
2. Ke1 Ke8

Now, we're back to the original position, but castling is no longer
legal. You will see that all of White's legal moves now have the "mate
in n" message.

jm
 




Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT +1. The time now is 03:38 PM.


Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.Content Relevant URLs by vBSEO 2.4.0
Copyright 2004-2017 ChessBanter.
The comments are property of their posters.