Database Organizing Strategies
GREG SMITH
organizing strategy for a chess database. Roughly, his idea is that postal
players might want to organize their data by year or era.
I have been contemplating reorganizing (or perhaps more honestly, organizing
for the first time) my ChessBase data, but I am really stumped as to what the
best way to go about it would be. Who out there has an organization strategy
that you find effective and useful? Could you share it with us? How would you
change it if you had the time to do it all over again?
On a related note, I looked into upgrading ChessBase 4.0 to the new Windows
version, which here in the States costs $80 (fair price, I think). But what
is unfair, it seems to me, is that that $80 only buys the program. If
you want the 200,000 games that goes with it, the price is $200 (a mere $95
less than what the program-plus-games costs the newcomer). This seems a bit
steep considering most of us have already spent $295 on ChessBase 4.0. In
short, ChessBase is asking 4.0 users to pay $500 in total for a program
newcomers are paying $300 for.
Is it worth it, those of you who have made the switch?
Jeff Otto
>In this month's _Chess Life_ Alex Dunne devotes his postal column to an
>organizing strategy for a chess database. Roughly, his idea is that postal
>players might want to organize their data by year or era.
>
>I have been contemplating reorganizing (or perhaps more honestly, organizing
>for the first time) my ChessBase data, but I am really stumped as to what the
>best way to go about it would be. Who out there has an organization strategy
>that you find effective and useful? Could you share it with us? How would you
>change it if you had the time to do it all over again?
I am very happy with the way that my data are managed, and I wouldn't change it
if I started all over. First, I use two separate database programs, because I
feel that they serve different functions. My games database program is
ChessAssistant. Here I have just two databases: one of my games, and one with
everything else. I use ChessAssistant this way because if I want to search for
a particular middle game pawn structure or endgame material or whatever, it
doesn't matter to me whether the game was played in 1925 and started as a
King's Gambit or was played last year as something else. Consequently for the
sake of completion I see no reason to separate the games based on date or
opening. My other database program is Bookup. I use this database for all my
openings. Because Bookup has a far superior "tree" mode to CA, and because it
is much more user friendly in terms of adding/changing/deleting analysis, I use
it primarily for opening study. With Bookup, I have multiple databases, one
for each opening I play, one for my own games (which differs from the CA
database in that I add analysis here), and one for each endgame (K+P, Rook,
minors etc).
As I get new data, I import it into a temporary CA database. I search this new
database for games played in my openings. These are printed out, and I enter
them into bookup by hand. The new CA database is then joined with my total
database, and the new database deleted.
Jeff
Jay McKeen
I too was disappointed that my upgrade fee did not include the database
that new buyers were getting. We spent just as much on the original
program as the newcomers, didn't we?
Tim Mirabile
>
> In this month's _Chess Life_ Alex Dunne devotes his postal column to an
> organizing strategy for a chess database. Roughly, his idea is that postal
> players might want to organize their data by year or era.
>
> I have been contemplating reorganizing (or perhaps more honestly, organizing
> for the first time) my ChessBase data, but I am really stumped as to what the
> best way to go about it would be. Who out there has an organization strategy
> that you find effective and useful? Could you share it with us? How would you
> change it if you had the time to do it all over again?
I have about 180,000 games in my database. The main database
is broken up into 5 pieces: A,B,C,D,E after the ECO codes.
You can use search mask and export or the ECO utility to
do this. I also have databases for my games, active postal,
completed postal, endgames and combinations (these are games
that start from positions other than the starting position),
games from my local club's bulletins, etc.
Breaking up the large database makes it easier to kill
doubles, and also limits the search process a bit. i.e.
If I have white against someone, since I only play 1.e4,
I only need to search for him as black in B+C.
The same applies to doing a sh+F6 (compare or novelty search).
> On a related note, I looked into upgrading ChessBase 4.0 to the new Windows
> version, which here in the States costs $80 (fair price, I think). But what
> is unfair, it seems to me, is that that $80 only buys the program. If
> you want the 200,000 games that goes with it, the price is $200 (a mere $95
> less than what the program-plus-games costs the newcomer). This seems a bit
> steep considering most of us have already spent $295 on ChessBase 4.0. In
> short, ChessBase is asking 4.0 users to pay $500 in total for a program
> newcomers are paying $300 for.
I don't think you get the 200,000 games for the $295,
I thought that was extra.
> Is it worth it, those of you who have made the switch?
Yes, I made the switch, and I like CBWIN. I have a 17"
monitor, so I work in 1024x768 in windows, and CBWIN takes
good advantage of this. Unlike CB4, you can see the board
and the analysis at the same time. You can select a bunch
of games from the list, hit enter, and CBWIN will create
a single game tree out of them. You can search several
databases at once, and have several game windows open at
once. You can drag and drop one game into another.
You can do position searches, material searches, and motif
searches. There is also a set of TT fonts for printing
games with informant style notes and diagrams.
I'd say it was worth the $80, especially if you
did not already buy all of the extra programs like
CBTREE, FINALE, MOTIV, fonts, etc. needed to do this
in CB4.
DR. ROY SCHMIDT
[...]
:it primarily for opening study. With Bookup, I have multiple databases, one
:for each opening I play, one for my own games (which differs from the CA
:database in that I add analysis here), and one for each endgame (K+P, Rook,
:minors etc).
YEs, Bookup is the place for analysis, because the analysis is stored with
the *position* not with the game. So if you transpose into a position
already in the database, ...
:As I get new data, I import it into a temporary CA database. I search this new
:database for games played in my openings. These are printed out, and I enter
:them into bookup by hand. The new CA database is then joined with my total
:database, and the new database deleted.
^^^
Ouch! I think I know what you mean, but sounds strange.
Why do you enter the games into Bookup by hand???
--
Roy Schmidt sch...@uxmail.ust.hk
Information & Systems Management Dept, School of Business and Management
The University of Science and Technology
Clearwater Bay, Sai Kung, HONG KONG
Anders Thulin
>I have been contemplating reorganizing (or perhaps more honestly, organizing
>for the first time) my ChessBase data, but I am really stumped as to what the
>best way to go about it would be.
The best way will necessarily begin with your own requirements. A good
database organization is one that matches your questions.
List all your questions you want to have answered by the
database. Perhaps something like:
List all Kizeritzky gambits played in year X.
Find all games in which position Y occurred.
What openings have player P played recently?
And so on.
Once you have all these question, and have a decent idea of the
relative importance of each, it will very probably be possible to
design an organization that suits you.
If your most important requirements don't say '... in year X' an
organization based on years may be not be optimal: it seems as you're
looking for information that is not related to years. If you still
decide to organize your database after years, you'll probably going to
need some kind of support to make searches over a set of years.
If your requirements say '... in opening Y' or '... by player P'
often, it may be that you should organize your database after openings
or players instead.
If your requirements say '... in tournament T', perhaps you'll need
to use tournaments as the main basis for organization.
The question is very illuminating. It shows very clearly that for a
database to be effective for a particular user, it will need to be
flexible enough to support any organization that that user needs,
rather than some fixed preconceived organization.
--
Anders Thulin a...@linkoping.trab.se 013-23 55 32
Telia Research AB, Teknikringen 2B, S-583 30 Linkoping, Sweden
Stephan Busemann
>I have been contemplating reorganizing (or perhaps more honestly, organizing
>for the first time) my ChessBase data, but I am really stumped as to what the
>that you find effective and useful? Could you share it with us? How would you
>change it if you had the time to do it all over again?
I am using ChessBase (4.0) together with Alpha 2.0 that allows me to
construct additional keys (White, Black, Tournament/Year).
From within ChessBase I can select the key I need, e.g. for openings or
players. (Other keys are possible, too (endgames etc), but I don't use
them regularly.) I have quick access to all data I need: if I want to
export data, I can get all of a certain tournament in a certain year
(e.g. Linares 1993); if I want to prepare for a CC game, I look up
all games of my opponent with the color he has in our game; if i want to
prepare an opening variation, I use the opening key.
I have about six different databases, each containing between 60 and 120
thousand games. Runtime decreases on my 486-50 if DBs exceed the 100.000
size (this requires huge opening keys, which makes access slower).
So I have to split things up.
I keep corr games separate from the rest, and NIC style
name notation separate from ChessBase style notation. the latter distinction
is useful sinde Alpha operates on strings and would not be able to find out that
Karpov,Anatoly (NicBase) and Karpov,An (ChesBase) are the same player.
I would get different keys for the two anyway.
The rest is organized according to years so that I roughly know
where to look first, if I am in search of a certain game. One DB has
everything up to 1945, one up to 1979, another one for the Eighties,
and a third one for the Nineties.
Data I bought is again separate for copyright reasons (I should not
give away copies of that stuff).
I am reasonably happy with this organization. The biggest drawback
is that I have to search for openings six times (in all DBs). That's
time consuming. Upgrading to CBWin would certainly allow me to handle
this much easier, and I probably will do so sooner or later.
For me, a consistent notational framework (for names, locations, tournament
modes etc) is most important. Allthough there are at least two suggestions
(ChessBase, NicBase), neither has become THE standard, or even A standard.
People write "Moskau Olympiade" or "Moscow ol" or "Olymp.'95 Moscow" etc.
Or you see "Kortschnoi" "Korchnoi" "Korchnoj" and the christian name
before or after, separated by comma or not, abbreviated in one way or
another or not...
A program like Alpha doesn't have a fair chance to find what you're looking for
unless you tell it what variations might exist in your data. But you
don't know!
I try to stick to the ChessBase notational framework because most of my data
already commits to it quite well (the NicBase one is more explicit and
better, though). PGN conversion is important for me to edit 'awkward' data
more comfortably than it is possible with the ChessBase DBEdit tool
(unfortunately ChessBase do not sell their editing tools).
-------------------------------------------------------------------
Stephan Busemann, DFKI, Stuhlsatzenhausweg 3, D-66123 Saarbruecken
Phone: +49 681 302-5286; Fax: -5341; email: buse...@dfki.uni-sb.de
Chuckchess
gbpa...@online.dct.com
>If you use Chess Assistant these questions become redundant.
The sorting routines are so fast and easy to use, that you can extract
subsets of data in most any fashion in seconds. (My database is
421,000 games -- I can find most anything in 5 minutes or less -- the actual
search taking about 30 seconds. WHEW!