Saturday, November 11, 2006

The best defence against e4 ?

with some more analysis i found a strong line for black in the
closed Ruy Lopez, namely in the Zaitsev variation.
After 9.h3 Bb7 (Zaitsev) 10. d4 Re8! 11. Nbd2 Na5 12. Bc2 black
usually plays c5 (transpositions in black move order are possible)
but what next ? For example, after 13.Ndf1 (or d5) black often
plays 13 .. Bf8, a good move, which after .. g6 (to prevent the
white Nf1g3 and then Ng3f5) makes the fianchetto Bf8g7 possible.

But black has an even stronger move, namely 13 .. exd4 !
Now after 14. cxd4, black can bring his knight on a5 back
to c6, and also later still can play Bf8, making it very hard
maybe even impossible for white to achieve advantage.

In fact, because of this defence, it might even be better
to play 10 d3. for white, or 9 d4 (Boguljubow) but
in these cases also no structural advantage is achieved.

What does this mean ?
Well, having analysed many white openings, it appeared
that in most cases with e4 white could achieve a slight
advantage. But not anymore in above line.

S.. is that black move .. exd4! a novelty ?
Well, not exactly, in the high level -Chessbase sponsored-
'freestyle' (man+computer) tournament in *this* year,
2006, Rybka, the topengine played this move against
the engine Goldbar and the game ended in a draw.

For my personal research, and the E-book better
chess opening play, it means that now other opening moves
as 1.d4, or Nf3, or even 1.c4 are fundamentally just
as good as 1.e4. A major change ?

Well not really, for beginners i still recommend
1.e4, even although you might encounter a defence
like the one give above, or difficult Sicilian lines.
Medium/tactical players probably could better
switch to a gambit repertoire, which even would
be sharper than a general 1.e4 repertoire
more about that later (i'm thinking about
listing a general gambit repertoire, when always
starting with e4, eg. 1.e4 d5 2.d4!? etc.

And indeed, advanced players can maybe better
use a more positional repertoire with 1.d4, but..
also with this move i believe not fundamental
advantage can be achieved for white, which
means chess indeed is a draw; as was expected
already many years by those knowledgeable in
the subject (and top-GM's like Anand and Karpov).

Monday, November 06, 2006

Back to the Rossolimo

With white i got some problems against the accelerated fianchetto
in the Sicilian, even when going for the Maroczy variation,
e4 c5 Nf3 Nc6 d4 cxd4 Nxd4 g6!! c4! etc.
So i tried to improve my Rossolimo variations with Bb5 instead
of d4 in the line above, and i now again achieve advantage for white.

So with these improvements my computer rating is going up again,
now with standard nr 5 on ICC.
To a large extent still due to the improved performance of Rybka
on my new AMD X2, with Windows X64 installed,
but the opening book used must also be good of course.

In fact, the higher the ratings of the opponents playing me,
the more important the quality of the book becomes,
as i've clearly seen. Small 'errors' , ie deviations from the
best book move can have big consequences, a well known
fact in chess; at least with respect to tactical errors.

Monday, October 16, 2006

New bullet record

well, as could be expected with the strong Rybka engine,
now i've also achieved a new 'bullet' (lightning/speedchess)
rating on the Internet Chess Club with the Bookbuilder account.

And upgraded my computer again, now running
with the multiprocessor version of Rybka 2.1
with a AMD X2 dual core (4200+ type).

New goal: increasing the standard rating again..
And new idea: uploading my largest Bookbuilder database
(more then 13 million positions) via the Internet
to a site like; only for registered
Bookbuilder users though; as the support otherwise will
take me too much time of course.

For those interested in downloading Bookbuilder:
Get it from CNET!

14 days trial
And don't forget to have fun with your chess

Friday, September 15, 2006

improvements with Rybka analysis

Recent months i've checked the opening variations in my
E-book (better chess opening play) with the top engine
Rybka, now version 2.1-o UCI.
Initially it appeared to become more difficult to obtain
a structural advantage for white as result of the
less optimistich Rybka analysis, eg. when compared
with Shredder. But while continuing the analysis
new variations were found !

This new analysis is not completed yet, i'm working
on an update of the book, and testing new variations
on ICC (latest standard rating record 2851
on August this year).

Some new findings:
- against the Caro Kann the advance variation again now is preferred
again, instead of the 'English' variations with c4.
- for the Sicilian it is difficult, if not impossible
to find the 'best' defence variation for black; it appears
that different strategies are possible; currently i'm still
experimenting, eg. with Pelikan, Taimanov, and Najdorf
- also after .. e5, when white plays Bb5, the Ruy Lopez,
many different defence variations seem to be equally good,
e.g. Berlin, Cordell, closed, etc.

Yet after renewed analysis and gradually improving the
best lines for white, certain patterns in middle game
planning appear, whereby the advantage for white
with 1.e4 still remains clear in all cases.

Against 1.d4, the Nimzovitch and Queens Indian
defences remain the 'best' defences in practice,
as result of a complete minimax of my large database.
Having recently acquired part 2 of Shereshky's
excellent book Endgame strategy, closed games,
it appeared that i've discovered something which
some Russian top chessanalysts apparently already knew,
namely that the above mentioned 'light square' (bish0p)
strategy is better for black than the black square strategy
like Kings Indian, Slav (with c6), etc. And if black would choose
the QGD it has to be played very carefully, i.e. with moves
like Bb4, in order to adhere to the black square strategy.

It certainly is encouraging that i've achieved similar
conclusions with my own computer analysis, which
means that for th 1.e4 variations we are not far away
of almost 'perfect' results. NB for such results i've
have to extend the analysis as done by Shereshevky
in his Endgame strategy part I book (open games)
to deeper lines, and the results will soon be puslished
in the latest version of the E-book 'better chess opening
play' on

NB believe it or not, the second hand version of
the excellent, highly original, and apparenlty quite
rare book part I for open games by Shereshvky
now is selling at for a price
of more than 1000 dollars.

So buying my E-book book, including the new version
-when it becomes available in a few weeks- certainly
won't be a bad deal, as it is much more up to date
compared with the rather oldfashioned variations
and games as published by S., and..
last but not least you get the computer
program Bookbuilder included..

Tuesday, April 04, 2006

ICC standard rating record

not suprising, with the new engine Rybka 1.1 achieved a new
rating record on ICC (2843), at about 10th place. With faster
hardware i believe an even higher rating could easily be achieved.
No human GM is capable of achieving such ratings on ICC..

All games played with self made book, own 1.e4! repertoire,
sometimes 1.a3 as suprise, and the Arena chess interface.

The eval of this super engine Rybka clearly is different
from Shredder, which we still use in FICS, so some lines
eg. in the Ruy Lopez are played a bit differently.

As a result next action will be to update our 'best'
variants again on our site
And yes, thereafter the goal is to achieve
yet another new rating record..

Bookbuilder program/download link

Apparently at the site the download link for my chess opening program Bookbuilder disappeared Here it is: Bookbuilder demo ...