There are many of us on here striving for improvement and i have recently been looking at ways of improving my general undersanding of positions and calculative techniques.
One technique i've come across is recomended by GM J Rowson, who advocates looking for games in databses, finding the key turning point and then as an exercise work out a better line or lines for the losing player.( all without your AI)
whilst looking for candidate games i came across the game Speelman-Kasparov Madrid 1989.
WOW!
It has made me realise just how little i really understand about this game. It seems to blow everything we talk about regarding general positional concepts completely out of the water. Kasparov sacs 2 pawns for what appears to be absolutely zero compensation then goes on to destroy his opponent in the most brilliant fashion.
I sort of get the first pawn sac. he obviously feels opening up lines for the DS Bishop to be worth it but from then on the game is just completely unfathomable!
After 25..Qg5+ fritz would suggest Kas has lost his marbles with a strong advantage to white. But just watch how he handles the position. This is a game i could spend a fortnight with. seriously great stuff.
I present for your pleasure



Jonathan Speelman - Garry Kasparov [E92]
Madrid Madrid (2), 1989

1.c4 g6 2.e4 Bg7 3.d4 d6 4.Nc3 Nf6 5.Nf3 0-0 6.Be2 e5 7.Be3 Ng4 8.Bg5 f6 9.Bh4 g5 10.Bg3 Nh6 11.h3 Nc6 12.d5 Nd4 13.Nxd4 exd4 14.Qxd4 f5 15.Qd2 f4 16.Bh2 Nf7 17.h4 h6 18.hxg5 hxg5 19.g3 f3 20.Bxf3 Ne5 21.Be2 g4 22.Bg1 c5 23.dxc6 bxc6 24.0-0-0 Be6 25.Qxd6 Qg5+ 26.Kb1 Qg6 27.Ka1 Rab8 28.Rh5 Qxh5 29.Qxe6+ Kh8 30.Qe7 Nf3 31.Bxf3 Rxf3 32.Na4 Re8 33.Qxa7 Qe5 34.Qb6 Qxe4 35.Qb4 Qc2 36.Rb1 Rd3 37.Qc5 Re2 38.Qh5+ Kg8 39.Qxg4 Qxb1+ 0-1

Quite astounding.

And apparently it was a rapid game!

Very cool game - I need to go through it today. I love games exchanging material for initiative.

btw, my both my database and copy of Stohl's 'Garry Kasparov's Greatest Chess Games' have the game in 1988, not '89.

Thanks for sharing! Quite amazing... he does everything "wrong" and win at the end... when you feel like you understand something about chess a game like that just show how inocent you are (well, me in this case).

There is a LOT that I don't understand about that game, but I'm most vexed by the rook for bishop trade by white. ???

Originally posted by monteirof
Thanks for sharing! Quite amazing... he does everything "wrong" and win at the end... when you feel like you understand something about chess a game like that just show how inocent you are (well, me in this case).

Originally posted by MrHand
There is a LOT that I don't understand about that game, but I'm most vexed by the rook for bishop trade by white. ???

The reason a few of you cannot fathom out this game is because
it's not a very good game to study. It's entertaining and has
a tactical trick or two but it is a blitz game.

The players are playing by instinct and not analysis.
It's a scrappy affair with some clever tactics by Kasparov.
Nothing more and useless from an instructive point of view
because it is flawed - too many unanswered questions.

If you want to study how to play blitz and steal the initiatve
this will suffice but hardly ideal. Some of Korch's blitz games
would serve the purpose better.

I doubt Kasparov would rate it amongst his best 500 games.

I think you got the Jonathan Rowson idea about lifting games
from databases and studying them from Zebras.

I reviewed Zebras:

http://textualities.net/writers/non-fiction-reviews/chandlerg02.php

My one gripe was this piece of advice without giving examples of what
games to choose.
.
I knew if weaker players just selected games and they chose the wrong one,
they would simply not understand it. The case in point.

Sorry if I ruffle a few delicate feathers but this hack and slash in no
way compares with any of Kasparov's real masterpieces.

(I do hope another strong player comes on and agrees with me,
else I'm in for a pasting.)

Here is a game that studying will repay.
I know because I ripped this game apart when I was getting good.
All I had to work with was the bare score.

A very instrucitve attacking game.
Watch those Black Knights and the Queen work together.
Oh. Black's 12th move is not a typing error

And to show you what I learned from that game I posted above.
Where do you think I got the idea of my 1oth move from?

Rapid game or not, there are some great trade-offs made in this game. Kasparov's first pawn sac essentially traps White's dark square bishop, who's fate is sealed by 19. g3. 19. f3 was required to fix that f-pawn before trying to free the bishop. Each pawn sac gains tempi, positional advantage and activity. Another reminder for me to think about to not just count up material when evaluating tactics. That bishop never did make it out of its little corner prison.

Agreed the pawn sacs for activity is a blitz ploy.

And f3 was practically a necessity but a timed f4 and suddenly the Bishop
is out - not studied it any length - sorry but I find it an ugly game.

Cannot see anything crushing for Black after 28.f4 instead
of the dodgy Rook sac. Then suddenly the Bishop is free and
it's going to d4 - White's OK

I stand by what I said - this is not a good game to study in any
great detail. You only have to read the previous posts by players
who I would not class as duffers, they are truly baffled by the game.
So would I be if I did not see it for it is.

It's a Blitz game where Black, gambled, created threats,
planted seeds of doubt and hoodwinked his opponent into making
a few rash mistakes. It's certainly not a masterpiece.

How were you able to transfer the game into the pgn viewer from the text in the first post?

Highlight the game and copy it , then pasted it in between
those pgn markers. (pgn)game here (/pgn)
Thus:

(pgn)
1.c4 g6 2.e4 Bg7 3.d4 d6 4.Nc3 Nf6 5.Nf3 0-0 6.Be2 e5 7.Be3 Ng4 8.Bg5 f6 9.Bh4 g5 10.Bg3 Nh6 11.h3 Nc6 12.d5 Nd4 13.Nxd4 exd4 14.Qxd4 f5 15.Qd2 f4 16.Bh2 Nf7 17.h4 h6 18.hxg5 hxg5 19.g3 f3 20.Bxf3 Ne5 21.Be2 g4 22.Bg1 c5 23.dxc6 bxc6 24.0-0-0 Be6 25.Qxd6 Qg5+ 26.Kb1 Qg6 27.Ka1 Rab8 28.Rh5 Qxh5 29.Qxe6+ Kh8 30.Qe7 Nf3 31.Bxf3 Rxf3 32.Na4 Re8 33.Qxa7 Qe5 34.Qb6 Qxe4 35.Qb4 Qc2 36.Rb1 Rd3 37.Qc5 Re2 38.Qh5+ Kg8 39.Qxg4 Qxb1(/pgn)

To get it to work changed the () round brackets to [] square brackets.

Very nice! Kasparov was very daring.