"BigKhat" wrote in message
om...
Who is Kasimdzhanov? And how many of you know Khalifman before he
became World Champion?

I knew him. Investigated some arrangements for him at the NY Open, and also
varieties of chess analysis, esp. on the Armenian French in 1997-1999.

Excuse me for top-poting, but I see what you write below - maybe there is an
element of truth in it. However, most people are motivated to play other
people, and motivation has a lot to do with performance.

Even technically 'playing computers' is unlike playing human beings. The
computer or program usually has access to an opening book, for example, and
can trot out the best analysis for the past 30 years in 3 seconds. Even if
you provide the person with a stack of opening manuals and a data-base the
factor of time to research all possibilities as a sort of librarian is,
well... its boring! Therefore in playing a computer you are playing an
enigma!

The chess engine cannot actually calculate the moves it is playing (!) from
the opening book** - otherwise why use the opening book at all? Then, the
opponent changes character when the opening book is exhausted, or if you
deviate from known theory, and the chess engine begins to make its own moves
based on its evaluation rubric. There are various estimates of strength of
chess-engines with book=off, from 1900 to 2400. How interesting a challenge
can this be to a GM?

Recently we sponsored a series of game annotation at www.chessville.com in a
column called Lessons Learned. In Najer - Bezgodov, Alexei B points out the
danger of using computers, and incomplete databases for analysis, and
criticises his own preparation for not over-ruling his computer's analysis.

Phil Innes

**In an interesting exchange with Bob Hyatt of Crafty fame, he said that as
White his program would never play the third move in the Ruy! It would never
play 3. Bb5. It does not evaluate that move as strongly as some others. This
creates something of a conundrum when the otherwise universally recognised
strongest move on the board is rejected. If his program actually plays that
move, therefore, it is not as a result of understanding the position but of
plonking out a move from an opening book, possibly because of a statistical
chance of that move having better success than others.



We only get a steady diet of Kasparov, Anand, Kramnik, + the other
round robin GMs (a few of which lost when playing Kasimdzhanov).
These RR GMs are soft. In most games they only play a few moves and
take quick draws. It is not surprising to me that an unknown with
more fire in his belly wins because these softies aren't used to
playing against quality players who want to fight.

This is exactly the reason why these top GMs can't play computers.
They get unnerved because the computer will play on in any position
unlike these chicken (gotta keep my rating high so I can belong to the
RR club) GMs. I say no more agreed draws until after move 60. Then
we'll see more real games of chess.

The chicken GMs will respond, but the position at move 20 really was
drawn. Fine. Try playing a computer and see if you can really draw
it. Bet most of them would be too chicken or get unnerved and lose.
It is hard to imagine that all positions that are agreed drawn at move
20 really would be drawn if played out to move 60. Don't these
chicken GMs realize that chess players make mistakes and blunder. A
position with most of the material on the board at move 20 is not dead
drawn.

The other reason chicken GMs won't like this is a lot of chicken GMs
won't play on in unclear positions. Yielding to their mutual
chickeness, they'll both agree to a draw. Don't give these chicken
GMs this option!

Whew! I'm glad to get that off my chest.