On Sep 9, 12:18 am, Sanny wrote:

I was thinking the GetClub Chess will win a lot of games now, But I am
seeing that the results are not comming as predicted.

Winners Go through the Recorded Games and tell me what mistake the
GetClub Chess is doing. Why it is loosing to strong players?

What weakness are you seeing in the Games played at GetClub Chess. Why
it is unable to beat the strong players? Is there any strategy mistake
or it is still making tactical mistake.


In spite of comments which strongly imply a certain
depth of search, I find that the program is not even
aware of itself being checkmated until one or perhaps
two moves beforehand. This is good evidence that the
program is weak tactically, and until it can see simple
tactics like mate-in-two there is no reason to waste
time thinking about strategy or trying to punch in more
opening lines for it to play by rote.

Other Java applets play just fine, so I don't quite
understand why it is that with GetClub the applet
takes so long to load and then runs so s-l-o-w-l-y
and why there must be frequent disconnects, why
the Web site is so messed up, etc. A lot of time
has been spent to get the site where it is today (no
matter how sad that may be), so I understand the
reluctance to try another approach -- one where you
follow programming guidelines and test software for
bugs and so forth.

Let me suggest an experiment, whereby you can
keep your buggy Web site and your present chess
programs, and simply add a side project then. Set
up a new "level"; call it Fish Level and write the code
so the program only does two things:

1) generates a list of all legal moves in a given
position

and

2) use simple arithmetic to add up the values of
the men on the board, returning for instance the
value "0" for the starting position and "-1" where it
has lost a pawn.

Now, using only this simple arithmetic, analyze
the possible (legal) variations until a fixed depth
has been reached, and then have the program play
the move which nets the largest value in its own
favor. In other words, if it can grab the King, have
it grab the King; if it can take the enemy Queen,
have it do that, etc. Just look a few moves deep
and then play the best move it has found.

My prediction is that unless even this approach if
screwed up somehow, it will play better than what
you have now because it will *always* get the simple
tactics part of chess right.


-- help bot