Main
Date: 02 Nov 2004 05:07:55
From: Kevin Houston
Subject: find duplicate postions
I am developing a chess opening tree for myself. For each unique
position I want to respond with one and only one unique move. Since
there is more than one way to reach a particular position I'd like to
be able to search my tree for duplicate positions and insure that I am
responding the same way each time. Does anyone know of a chess
database program that will allow me to create an opening tree and
search that tree for dupicate positions? I've been using ChessBase
Light to create the tree but I have not been able to figure out how
to search the tree for duplicate positions. I think it allows me to
search for particular positions but not dupicate positions. I am open
to trying other database programs if this feature is available.




 
Date: 02 Nov 2004 15:13:41
From: CeeBee
Subject: Re: find duplicate postions
[email protected] (Kevin Houston) wrote in rec.games.chess.analysis:

> I am developing a chess opening tree for myself. For each unique
> position I want to respond with one and only one unique move. Since
> there is more than one way to reach a particular position I'd like to
> be able to search my tree for duplicate positions and insure that I am
> responding the same way each time. Does anyone know of a chess
> database program that will allow me to create an opening tree and
> search that tree for dupicate positions? I've been using ChessBase
> Light to create the tree but I have not been able to figure out how
> to search the tree for duplicate positions. I think it allows me to
> search for particular positions but not dupicate positions. I am open
> to trying other database programs if this feature is available.


Both CB and Chess Assistant create a tree with unique positions. They
simply add to every unique position how often it was played, and what
the result was for that game with that branche. There are no "duplicate
positions".

If you create a tree with those programs you will see all answers to a
particular move. If you want to have a tree which gives you one response
to every move made by your opponent, you can simply delete those other
moves (and their sub branche) from the tree.

Maybe a better way is to create an opening repertoire, say like in
ChessBase, both for your white and black games, and after building that
repertoire from single opening moves and variations you did input you
can create an opening tree for both black and white with the desired
result.


--
CeeBee


"I don't know half of you
half as well as I should like;
and I like less than half of you
half as well as you deserve."


  
Date: 02 Nov 2004 18:00:03
From: Mike Ogush
Subject: Re: find duplicate postions
On 2 Nov 2004 15:13:41 GMT, CeeBee <[email protected] > wrote:

>[email protected] (Kevin Houston) wrote in rec.games.chess.analysis:
>
>> I am developing a chess opening tree for myself. For each unique
>> position I want to respond with one and only one unique move. Since
>> there is more than one way to reach a particular position I'd like to
>> be able to search my tree for duplicate positions and insure that I am
>> responding the same way each time. Does anyone know of a chess
>> database program that will allow me to create an opening tree and
>> search that tree for dupicate positions? I've been using ChessBase
>> Light to create the tree but I have not been able to figure out how
>> to search the tree for duplicate positions. I think it allows me to
>> search for particular positions but not dupicate positions. I am open
>> to trying other database programs if this feature is available.
>
>
>Both CB and Chess Assistant create a tree with unique positions. They
>simply add to every unique position how often it was played, and what
>the result was for that game with that branche. There are no "duplicate
>positions".
>
>If you create a tree with those programs you will see all answers to a
>particular move. If you want to have a tree which gives you one response
>to every move made by your opponent, you can simply delete those other
>moves (and their sub branche) from the tree.
>
>Maybe a better way is to create an opening repertoire, say like in
>ChessBase, both for your white and black games, and after building that
>repertoire from single opening moves and variations you did input you
>can create an opening tree for both black and white with the desired
>result.
>
>
>--
>CeeBee
>
>
>"I don't know half of you
>half as well as I should like;
>and I like less than half of you
>half as well as you deserve."

Kevin,

As CeeBee points out, positions are stored only once in the tree.
What you really want is not to eliminate duplicate positions as much
as to know when you are entering the moves into the tree if the
position is already in the tree via some other move order.

As you are entering the moves and reach a position as long as you
don't see a suggested move for the position, you can be assured that
the position was not reached by some other move order previously
entered. If there is already a suggested move for the position then
you need to compare it with the move you where about to enter. If the
moves are the same, just have Chessbase make the move already in the
tree and continue entering your repertoire. If the moves are
different, you will need to investigate which move is the correct one
and make sure that the correct move is the only move in the tree for
the given position.

BTW, if you are interested in having the database drill you on the
repertoire after you have entered, there are databases whose priy
purpose to help users train on their opening repertoire: BookBuilder,
BookUp, or Chess Position Trainer. The first two are commercial
products, but all three have some version (with either miits on
functionality or on time that you can use it for the commerical
programs) that can be downloaded for you to try out.

Mike Ogush


   
Date: 03 Nov 2004 06:28:52
From: Kevin Houston
Subject: Re: find duplicate postions
[email protected] (Mike Ogush) wrote in message news:<[email protected] >...
> On 2 Nov 2004 15:13:41 GMT, CeeBee <[email protected]> wrote:
>
> >[email protected] (Kevin Houston) wrote in rec.games.chess.analysis:
> >
> >> I am developing a chess opening tree for myself. For each unique
> >> position I want to respond with one and only one unique move. Since
> >> there is more than one way to reach a particular position I'd like to
> >> be able to search my tree for duplicate positions and insure that I am
> >> responding the same way each time. Does anyone know of a chess
> >> database program that will allow me to create an opening tree and
> >> search that tree for dupicate positions? I've been using ChessBase
> >> Light to create the tree but I have not been able to figure out how
> >> to search the tree for duplicate positions. I think it allows me to
> >> search for particular positions but not dupicate positions. I am open
> >> to trying other database programs if this feature is available.
> >
> >
> >Both CB and Chess Assistant create a tree with unique positions. They
> >simply add to every unique position how often it was played, and what
> >the result was for that game with that branche. There are no "duplicate
> >positions".
> >
> >If you create a tree with those programs you will see all answers to a
> >particular move. If you want to have a tree which gives you one response
> >to every move made by your opponent, you can simply delete those other
> >moves (and their sub branche) from the tree.
> >
> >Maybe a better way is to create an opening repertoire, say like in
> >ChessBase, both for your white and black games, and after building that
> >repertoire from single opening moves and variations you did input you
> >can create an opening tree for both black and white with the desired
> >result.
> >
> >
> >--
> >CeeBee
> >
> >
> >"I don't know half of you
> >half as well as I should like;
> >and I like less than half of you
> >half as well as you deserve."
>
> Kevin,
>
> As CeeBee points out, positions are stored only once in the tree.
> What you really want is not to eliminate duplicate positions as much
> as to know when you are entering the moves into the tree if the
> position is already in the tree via some other move order.
>
> As you are entering the moves and reach a position as long as you
> don't see a suggested move for the position, you can be assured that
> the position was not reached by some other move order previously
> entered. If there is already a suggested move for the position then
> you need to compare it with the move you where about to enter. If the
> moves are the same, just have Chessbase make the move already in the
> tree and continue entering your repertoire. If the moves are
> different, you will need to investigate which move is the correct one
> and make sure that the correct move is the only move in the tree for
> the given position.
>
> BTW, if you are interested in having the database drill you on the
> repertoire after you have entered, there are databases whose priy
> purpose to help users train on their opening repertoire: BookBuilder,
> BookUp, or Chess Position Trainer. The first two are commercial
> products, but all three have some version (with either miits on
> functionality or on time that you can use it for the commerical
> programs) that can be downloaded for you to try out.
>
> Mike Ogush

Mike,

You are correct when you say that what I "want is not to eliminate
duplicate positions" but to see if a "position is already in the tree
via some other move order". You say that if the position already
exists in the tree then the program should suggest a move. While this
is not as satisfying to me as having the program tell me where all the
duplicate positions are at least it serves the same purpose. As I
said I am using chessbase light. The only way that I know of to have
it suggest a move is to click the button that looks like a IC chip to
start the fritz analysis mode. But the suggested moves it gives have
nothing to do with what is already in the tree but on chess engine
analysis. Is there some other way to have it suggest a move. Also I
intentionally setup a duplicate position to see how it would respond
but still the only way to get it to suggest a move was to click the IC
chip.

What would really be cool is a program that allows you to create a
tree that "folds in on itself". So that there could be no duplicate
positions. So if different branches reach the same position then
those branches would connect to the same "position node". There would
have to be some sort of "on the fly" move numbering depending the path
it takes to get to the position since the same position may be reached
by a different number of moves, but I think that could be worked out
by the programmer. Maybe I'll write something like this myself if it
does not already exist.


    
Date: 08 Nov 2004 19:15:56
From: Michael Leahy
Subject: Re: find duplicate postions

"Kevin Houston" <[email protected] > wrote in message
news:[email protected]...

> What would really be cool is a program that allows you to create a
> tree that "folds in on itself". So that there could be no duplicate
> positions. So if different branches reach the same position then
> those branches would connect to the same "position node". There would
> have to be some sort of "on the fly" move numbering depending the path
> it takes to get to the position since the same position may be reached
> by a different number of moves, but I think that could be worked out
> by the programmer. Maybe I'll write something like this myself if it
> does not already exist.

This definitely exists in Bookup. Download the free Bookup 2000 Express to
try it out. For a trial period it includes instant editing of the tree.

When Bookup arrives at a position that is reached by other move orders, it
displays a red 'T' for transposition.

Mike Leahy
"The Database Man!"
www.bookup.com




    
Date: 03 Nov 2004 18:26:11
From: Mike Ogush
Subject: Re: find duplicate postions
On 3 Nov 2004 06:28:52 -0800, [email protected] (Kevin Houston) wrote:

>[email protected] (Mike Ogush) wrote in message news:<[email protected]>...
>>
>> SNIP
>>

>> Kevin,
>>
>> As CeeBee points out, positions are stored only once in the tree.
>> What you really want is not to eliminate duplicate positions as much
>> as to know when you are entering the moves into the tree if the
>> position is already in the tree via some other move order.
>>
>> As you are entering the moves and reach a position as long as you
>> don't see a suggested move for the position, you can be assured that
>> the position was not reached by some other move order previously
>> entered. If there is already a suggested move for the position then
>> you need to compare it with the move you where about to enter. If the
>> moves are the same, just have Chessbase make the move already in the
>> tree and continue entering your repertoire. If the moves are
>> different, you will need to investigate which move is the correct one
>> and make sure that the correct move is the only move in the tree for
>> the given position.
>>
>> BTW, if you are interested in having the database drill you on the
>> repertoire after you have entered, there are databases whose priy
>> purpose to help users train on their opening repertoire: BookBuilder,
>> BookUp, or Chess Position Trainer. The first two are commercial
>> products, but all three have some version (with either miits on
>> functionality or on time that you can use it for the commerical
>> programs) that can be downloaded for you to try out.
>>
>> Mike Ogush
>
>Mike,
>
>You are correct when you say that what I "want is not to eliminate
>duplicate positions" but to see if a "position is already in the tree
>via some other move order". You say that if the position already
>exists in the tree then the program should suggest a move. While this
>is not as satisfying to me as having the program tell me where all the
>duplicate positions are at least it serves the same purpose. As I
>said I am using chessbase light. The only way that I know of to have
>it suggest a move is to click the button that looks like a IC chip to
>start the fritz analysis mode. But the suggested moves it gives have
>nothing to do with what is already in the tree but on chess engine
>analysis. Is there some other way to have it suggest a move. Also I
>intentionally setup a duplicate position to see how it would respond
>but still the only way to get it to suggest a move was to click the IC
>chip.
>
>What would really be cool is a program that allows you to create a
>tree that "folds in on itself". So that there could be no duplicate
>positions. So if different branches reach the same position then
>those branches would connect to the same "position node". There would
>have to be some sort of "on the fly" move numbering depending the path
>it takes to get to the position since the same position may be reached
>by a different number of moves, but I think that could be worked out
>by the programmer. Maybe I'll write something like this myself if it
>does not already exist.

Kevin,

WRT to the confusion about using Chessbase and to know when you are
entering moves that you have reached a duplicate position: I have had
the full (non-light) version of Chessbase for several years and
forgot that some of the features in the full-featured version aren't
available in the light version. In particular the way that I was
thinking of entering the tree is only avialbale in the full-version.
The method I was thinking of is to enter the tree into a Chessbase
opening book, where at each postition the program displays the
candidate moves that it knows about from the position. Sorry about
that!

Based on what you are tring to accomplish with your opening repertoire
I suggest you download Chess Position Trainer
(http://www.chesspositiontrainer.com/ ).It will do the right thing
with regard to handling duplicate positions reached by different move
orders. Almost the first half of the 49 page manual that comes with
the program is a tutorial for the program. So it's pretty easy to
come up to speed on it.

There are some features that Chess Position Trainer does not have that
other competitive products do. You did not mention whether you would
want the features or not so I will list them below:

1. ability to assign evaluations to leaf nodes and to
backsolve, that is to propagate the evaluations upward to
the root node.
2. the ability to import PGN games and use the game result to
supply a default evaluation for the leafn node
representing the end position of the game.
3. the ability to invoke a chess engine and copy the engine's

evaluation of the position into the database's evaluation
of the position.

BookUp has all three features, BookBuilder has features 1 and 3, but
not 2.

You can find information about BookUp at http://www.bookup.com/ and
about BookBuilder at http://www.superchess.com/. Both are commmercial
programs, but have a free trial version that you can download to try
the program out.






     
Date: 04 Nov 2004 07:37:38
From: Kevin Houston
Subject: Re: find duplicate postions
snip
>
> Kevin,
>
> WRT to the confusion about using Chessbase and to know when you are
> entering moves that you have reached a duplicate position: I have had
> the full (non-light) version of Chessbase for several years and
> forgot that some of the features in the full-featured version aren't
> available in the light version. In particular the way that I was
> thinking of entering the tree is only avialbale in the full-version.
> The method I was thinking of is to enter the tree into a Chessbase
> opening book, where at each postition the program displays the
> candidate moves that it knows about from the position. Sorry about
> that!
>
> Based on what you are tring to accomplish with your opening repertoire
> I suggest you download Chess Position Trainer
> (http://www.chesspositiontrainer.com/ ).It will do the right thing
> with regard to handling duplicate positions reached by different move
> orders. Almost the first half of the 49 page manual that comes with
> the program is a tutorial for the program. So it's pretty easy to
> come up to speed on it.
>
> There are some features that Chess Position Trainer does not have that
> other competitive products do. You did not mention whether you would
> want the features or not so I will list them below:
>
> 1. ability to assign evaluations to leaf nodes and to
> backsolve, that is to propagate the evaluations upward to
> the root node.
> 2. the ability to import PGN games and use the game result to
> supply a default evaluation for the leafn node
> representing the end position of the game.
> 3. the ability to invoke a chess engine and copy the engine's
>
> evaluation of the position into the database's evaluation
> of the position.
>
> BookUp has all three features, BookBuilder has features 1 and 3, but
> not 2.
>
> You can find information about BookUp at http://www.bookup.com/ and
> about BookBuilder at http://www.superchess.com/. Both are commmercial
> programs, but have a free trial version that you can download to try
> the program out.

I downloaded Chess Position Trainer and have used for a couple of days
now. So far it seems to be exactly what I wanted! Would you happen
to know what the best newsgroup to post to would be if I want to ask a
question or make a comment about that specific program. I have a
couple of questions about using it but I will probably wait until I
finish reading the manual before posting them.


      
Date: 04 Nov 2004 19:26:15
From: Mike Ogush
Subject: Re: find duplicate postions
On 4 Nov 2004 07:37:38 -0800, [email protected] (Kevin Houston) wrote:

>snip
>>
>> Kevin,
>>
>> WRT to the confusion about using Chessbase and to know when you are
>> entering moves that you have reached a duplicate position: I have had
>> the full (non-light) version of Chessbase for several years and
>> forgot that some of the features in the full-featured version aren't
>> available in the light version. In particular the way that I was
>> thinking of entering the tree is only avialbale in the full-version.
>> The method I was thinking of is to enter the tree into a Chessbase
>> opening book, where at each postition the program displays the
>> candidate moves that it knows about from the position. Sorry about
>> that!
>>
>> Based on what you are tring to accomplish with your opening repertoire
>> I suggest you download Chess Position Trainer
>> (http://www.chesspositiontrainer.com/ ).It will do the right thing
>> with regard to handling duplicate positions reached by different move
>> orders. Almost the first half of the 49 page manual that comes with
>> the program is a tutorial for the program. So it's pretty easy to
>> come up to speed on it.
>>
>> There are some features that Chess Position Trainer does not have that
>> other competitive products do. You did not mention whether you would
>> want the features or not so I will list them below:
>>
>> 1. ability to assign evaluations to leaf nodes and to
>> backsolve, that is to propagate the evaluations upward to
>> the root node.
>> 2. the ability to import PGN games and use the game result to
>> supply a default evaluation for the leafn node
>> representing the end position of the game.
>> 3. the ability to invoke a chess engine and copy the engine's
>>
>> evaluation of the position into the database's evaluation
>> of the position.
>>
>> BookUp has all three features, BookBuilder has features 1 and 3, but
>> not 2.
>>
>> You can find information about BookUp at http://www.bookup.com/ and
>> about BookBuilder at http://www.superchess.com/. Both are commmercial
>> programs, but have a free trial version that you can download to try
>> the program out.
>
>I downloaded Chess Position Trainer and have used for a couple of days
>now. So far it seems to be exactly what I wanted! Would you happen
>to know what the best newsgroup to post to would be if I want to ask a
>question or make a comment about that specific program. I have a
>couple of questions about using it but I will probably wait until I
>finish reading the manual before posting them.

You could post to rec.games.chess.computers, but it would probably be
more efficient to just send an e-mail to the author of the program
directly.

Mike Ogush