Number One (on ICC)

(mainly) as result of the strong Houdini 1.5 engine(*), but
also due to the recent improvements in my Chess Partner book,
got a new 'standard' rating record of about 3050 on the
Internet Chess Club (ICC) with my 'bookbuilder' account,
on 28 December 2010, and nr 1 rating
for standard (rapid and slower_ play) !

On a 'simple' Quad comp, while there are some 6 core
comps around (with often an intel icore extreme), so
using a good book, does it matter ?

Yes, indeed it matters, especially at higher levels
of play (i.e. slower time controls, like in standard),
this now simply is proven in actual play,
and it makes sense, a chain is as weak as its weakest
part, and for a chess program this means everything has
to be without weaknesses, opening book, middlegame
(engine) practical endgame (engine) and theoretical
endgame (tablebases); have a lousy book, and against
a similar or even slightly weaker engine the
game can be lost, it makes a difference +/- 100
Elo or so, is my experience..but above a certain
level you can't increase improvement ofcourse
anymore, some booklines are perfect, others
are weak, and to some extent you can tune it
for the engine, playing style blitz/standard,
but thereafter its up to the engine (and the
hardware ofcourse); some concrete examples later
as i'm currently testing my CP book against
other CP books offline (preliminary results:
about 63 % with my recent book, vs 37% with
the older , general CP book, with the
same engine. How much Elo difference that is,
you may calculate..
:)

until next time,

jef

PS (*) quite an improvement now with Houdini,
and it plays especially well in the endgame, so with
a good book, achieving a slight advantage after the
opening, a reasonable (but not extremely) fast comp,
maintaining the advantage, Houdini often converts
such a slight advantage to a win in the endgame,
especially in slow (standard) games.