On Apr 10, 3:20 am, Sanny <softta...@[EMAIL PROTECTED]
> wrote:
> You are having fun with Beginner Level. But a few things about the
> games you play. What is the Big drawback you see with Beginner Levels
> game play.
>
> 1. Is it weak in Tactics?
Yes.
One weakness which should not exist in any
chess engine is *moving into check*. The
GetClub program does this routinely at the
end of the game, when it's losing.
> Then I suppost Higher Levels can provide
> better Tactics.
Even the weakest level should try to avoid
being checkmated; try to delay the loss as
long as possible, to give the opponent the
op****tunity to err; this is how most games
are won-- by merely providing many such
op****tunities, not unlike a mine field.
> 2. Is it weak in Strategy Formation?
Yes. However, this is to be expected.
> 3. Is it weak in defending the King?
Yes. And because checkmate (or, put
another way, trapping the King) is the
object of the game, this is critical.
> What do you think is the big weakness of Beginner Level that you feel
> it can be easily beaten?
In the endgame, the program plays at a
level which is perhaps 1000 rating points
weaker than in the middle-game.
It also is not handling tactics properly; in
general, a computer should come out on
top in almost every tactical skirmish with
lowly humans; yours doesn't.
> If Tactics is weak I can do nothing Only thing I will say is try
> higher Levels.
How about fixing the problem whereby the
program moves its King into check?
> If Strategy / Defence is weak let me know I may improve the game
> further.
Strategy is not a biggie; I often find
myself outplayed strategically, then go on
to win anyway due to my vast tactical
superiority. Once you gain the upper-hand
in tactics, then you can worry about such
trivial things. :>D
> After playing so many games with Beginner Level can you tell me the
> moves where Beginner played very bad move? So that I may see what
> weakness it has.
The biggest problems are the horrible
endgame, not defending its King when
attacked, and fairly simple tactics. It's
true that the tactics have been improved,
but there is no good reason a computer
cannot calculate checks and exchanges
to the very end; to a "quiet position", or
one that has no checks and no captures.
My recent improvement in results is
probably the result of the fact that my
"good" eye is functioning again; for a
while, I was playing with that eye held
shut, using powerful reading gl***** to
aid the weaker eye in seeing the board.
I am also trying very hard not to take
the program lightly; that attitude leads
to careless blunders, such as my piece
hang in the game you misidentified as
the Rybka QN-odds game. (I actually
saw that one coming, decided to trade
pieces, then changed my mind when I
couldn't see a clear way out of a
possible perpetual-check. I won easily,
despite hanging my Knight.)
-- help bot
|
