On Nov 13, 2:01 pm, SBD wrote:
On Nov 13, 12:59 pm, help bot wrote:

of the opposition and direct calculation will do the job. In
fact, I believe the mate with BB or with BN comes up more
often than corresponding squares.

That strikes me as incredibly implausible, do you have any proof?

Yes. I proved this as my doctoral thesis at Harvard,
in 1952. Before then, it was mistakenly assumed
that corresponding squares were the way to world
peace -- but I showed that it was, in reality, the mate
with Bishop and Knight that would end disease and
poverty once and for all. Of course, now that Al Gore
has a Nobel Peace prize too, I no longer brag of mine.


-- doc bot

On Nov 13, 2:20 pm, help bot wrote:

That strikes me as incredibly implausible, do you have any proof?

Yes. I proved this as my doctoral thesis at Harvard,
in 1952. Before then, it was mistakenly assumed
that corresponding squares were the way to world
peace -- but I showed that it was, in reality, the mate
with Bishop and Knight that would end disease and
poverty once and for all. Of course, now that Al Gore
has a Nobel Peace prize too, I no longer brag of mine.

OK, so much for the wit parade......

Seriously, I've only ever seen one B+N versus king (never played one),
but as I noted, I've used corresponding squares quite a bit - maybe
every 30 or 40 games.

I thought maybe you had used Chess Query Language in Fritz or one of
the big databases to show this.

On Nov 13, 4:09 pm, SBD wrote:

Seriously, I've only ever seen one B+N versus king (never played one),
but as I noted, I've used corresponding squares quite a bit - maybe
every 30 or 40 games.

Many players carefully avoid the possibility, fearing
an embarrassing post-mortem where some patzer
comes over to show how "easy" it is.

It has occurred in some of my games, and I have
even had the ending of KBB vs. KNN on deck, though
my opponent somehow rejected that edge in favor of
something with *zero* winning prospects. As for the
corresponding squares theory, I cannot say how often
it may have come up, but not once have I ever been
shown after a game that I could have won or drawn,
if only I have moved to this key square; perhaps this
is because nobody else has studied the idea either?

OTOH, *countless times* I have been shown that I
made a boo boo in the opening or middle game --
even by players I would expect to beat at Knight odds.
And in my computer analysis of completed games, I
find the same thing to hold true -- no matter who the
players were, or of what strength.


I thought maybe you had used Chess Query Language in Fritz or one of
the big databases to show this.

I am not familiar with CQL in Fritz, but that may be
because I have Fritz 5.32, not the more recent versions.
As far as I can determine, the chess programs I have
used do not "know" corresponding squares theory, but
instead they try to calculate -- just like me. But they
are familiar with (to put it nicely) the theory of direct
opposition. One of these days, I need to install some
tablebases... .


-- help bot

I'm the author of the original post. I have won dozens of games with
KBN vs K, and KQ vs KR. But the technique in these endings is
straight forward. Let's get back to the subject.

I'm beginning to understand how opposition is meaningless on some
lines, how this status is transfered to a major line, and once my
opponent leaves the major line outflanking can occur following the
standard algorithm used for orthadox opposition. The best position I
can think of that illustrates the problem that I do not understand is
found in a game Gulko vs Short, Riga, 1995.

FEN 5k2/8/5p2/7p/8/4PK2/8/8

Elementary knowledge suggests taking distant opposition with Kf4 or
Kf2, but according to the theory of corresponding squares the only
correct move is Kg2! I just don't get it. Too, som GM call this
hexadox opposition. I can't find this term anywhere on the net or in
any other text by anyone othe then Max Euwe.

Can anyone suggest any book or articles on the subject?

SBD wrote:
David Richerby wrote:
I only said it was `more common' in problems and correspondence. I
mentioned correspondence chess because, there, unlike OTB chess,
you're allowed to make notes.

I'm sort of starting to see.... is it a depth of analysis you are
referring to?

To some extent, yes. And the fact that corresponding squares, in all
but the simplest cases, requires an amount of book-keeping that is
hard (though not, of course, impossible) to do in one's head.[1] I
certainly don't dispute your claim that you've used corresponding
squares in an OTB game, for example.


Dave.

[1] I'm aware that this argument is close to circular, since there's a
great temptation to define the simplest cases as `those that can be
done in one's head'

--
David Richerby Disgusting Unholy Postman (TM): it's
www.chiark.greenend.org.uk/~davidr/ like a man who delivers the mail but
it's also a crime against nature and
it'll turn your stomach!

On Nov 14, 1:41 am, Petrovitch wrote:

I'm the author of the original post. I have won dozens of games with
KBN vs K, and KQ vs KR. But the technique in these endings is
straight forward.

Riiiight. Just stay there and I will be right back with Fritz,
to test you in Q vs. R. :D


Let's get back to the subject.

I'm beginning to understand how opposition is meaningless on some
lines, how this status is transfered to a major line, and once my
opponent leaves the major line outflanking can occur following the
standard algorithm used for orthadox opposition. The best position I
can think of that illustrates the problem that I do not understand is
found in a game Gulko vs Short, Riga, 1995.

FEN 5k2/8/5p2/7p/8/4PK2/8/8

Elementary knowledge suggests taking distant opposition with Kf4 or
Kf2, but according to the theory of corresponding squares the only
correct move is Kg2! I just don't get it.

If you "don't get it", then how do you know that
the correct move is supposed to be K-g2?



Too, som GM call this
hexadox opposition. I can't find this term anywhere on the net or in
any other text by anyone othe then Max Euwe.

One problem is the way most endgame books are
organized; they divide matters up by which men
remain on the board, and leave nothing for concepts
like corresponding squares, preferring instead to
give typical examples from grandmaster play, then
show where the loser went wrong.


Can anyone suggest any book or articles on the subject?

I went to the biggest chess book review site on
the Web and did a search for "corresponding
squares"; it went numb, froze, then gave me a
blank stare. But when I search for say, Sicilian
Dragon Bodarevsky Yugoslav hyper-attack
gambit with a lemon twist, I get on the order of
two thousand hits! Go figure.


-- help bot

On Nov 14, 12:41 am, Petrovitch wrote:
I'm the author of the original post. I have won dozens of games with
KBN vs K, and KQ vs KR. But the technique in these endings is
straight forward. Let's get back to the subject.

I would say of course to Q vs R, and unfortunately, say that, unless
you play an average of say, 30 games a day, the probablility of you
having "dozens" of BN games is remote. But that is for another
day.....


Can anyone suggest any book or articles on the subject?

As always, I think Cheron is the best book for this. If there is a
book on corresponding squares, it sounds like something Grondijs or
another author may have done as a work of love.

Cheron has 30 detailed pages on this subject, an outgrowth of an
article he wrote in 1944. Most Americans don't know Cheron, probably
why they can't play endgames.

Why not write GM Benko and ask him if he might be interested in
covering this topic in CL?

On Nov 14, 3:25 pm, SBD wrote:

As always, I think Cheron is the best book for this. If there is a
book on corresponding squares, it sounds like something Grondijs or
another author may have done as a work of love.

Cheron has 30 detailed pages on this subject, an outgrowth of an
article he wrote in 1944. Most Americans don't know Cheron, probably
why they can't play endgames.

Actually, most chess players from that era are now
either dead or else they no longer need to study such
things.

More recently (read: in the modern era), American
(and perhaps British) writers have simply "lifted" work
from others and translated it into English, taking the
credit themselves. I would be very much surprised
if, for example, Bruce Pandolfini has not covered
this somewhere, in books written mainly for kids.

The finest book I have ever read on the endgame
was written by GM Averbakh; it was targeted at
stronger club players, did not overwhelm with fluff,
and had no analytical errors as far as I could see.

But far more popular at the time was the work by
GM Fine -- a massive compilation of what I must
assume were other people's analysis, since he
made so very many errors. This error-riddled
work was widely promoted for decades, and was
supposedly the (fool's)gold standard until fairly
recently.

By comparison, multi-volume works by authors
like Cherov and Averbakh were either difficult to
find in English, or else prohibitively expensive -- if
not both. One should also consider the fact that
it is of little use to be a master of the endgame, if
every endgame reached is dead lost due to a
weakness in the opening and/or middle game.

It has been my experience that Americans have
a decided tendency for overkill in the opening
phase (relatively speaking), are somewhat weak
at tactics, but truly horrible only in the endgame.
But as I wrote before, it is usually sufficient to
know "opposition" and be able to calculate
accurately, as even world-class players can err
in these endings. While some may leap at the
opportunity to ad hominize me, they would no
doubt be embarrassed were I to post an example
of their favorite GM botching this stuff in a
famous game. :D


-- help bot

SBD wrote:
On Nov 14, 12:41 am, Petrovitch wrote:
I'm the author of the original post. I have won dozens of games with
KBN vs K, and KQ vs KR. But the technique in these endings is
straight forward. Let's get back to the subject.

I would say of course to Q vs R, and unfortunately, say that, unless
you play an average of say, 30 games a day, the probablility of you
having "dozens" of BN games is remote. But that is for another
day.....\\
Not so.

It's common on ICC for opponents to continue to play on in KPPk endings.
My solution is to promote the PP to N and B and then mate.

I get lots of KBNk endings that way.

--
Kenneth Sloan
Computer and Information Sciences +1-205-932-2213
University of Alabama at Birmingham FAX +1-205-934-5473
Birmingham, AL 35294-1170 http://KennethRSloan.com/

On Nov 14, 5:50 pm, Kenneth Sloan wrote:
SBD wrote:
On Nov 14, 12:41 am, Petrovitch wrote:
I'm the author of the original post. I have won dozens of games with
KBN vs K, and KQ vs KR. But the technique in these endings is
straight forward. Let's get back to the subject.

I would say of course to Q vs R, and unfortunately, say that, unless
you play an average of say, 30 games a day, the probablility of you
having "dozens" of BN games is remote. But that is for another
day.....\\

Not so.

It's common on ICC for opponents to continue to play on in KPPk endings.
My solution is to promote the PP to N and B and then mate.

I get lots of KBNk endings that way.


Are those blitz games, or (heaven forbid) does this
take up much valuable time? I doubt I could put up
a serious fight in blitz, but then, the time pressure
is mainly on my opponent, since he is the one
trying to prove he can mate. Also, at ICC, do you
have to claim a 50-moves draw, or does the site's
interface keep track for you?

I've never tried B&N at GetClub, though I have
many times underpromoted for various reasons.


-- help bot