🚀 go-pugleaf

>A question.  In trying to determine the weaknesses in my play, I have
>decided that an inability to compute variations quickly and accurately
>is the trademark of my numerous defeats.  Upon further investigation,
>it was discovered that I am completely unable to follow even the "main
>line" of a maneuver more than about 6 or 7 halfmoves.  After this I
>do not have a concrete grip on where the pieces are.
>
>How do people analyze variation so far out?  Does one keep a mental
>picture of the board in one's head, and make the necessary chganges
>on that board?  (I hope not).  How does this work?

There are exercises one can do to strengthen one's ability to
calculate variations. My favorite is to solve positions where
there is a forced mate in x moves, in my head. A useful book
for this purpose might be Reinfeld's 1001 Brilliant Ways to
Checkmate. (Not that I am a fan of Reinfeld by any means :-)
There are quite a few other such books that are useful. I
prefer working on positions where there is a forced mate to
other combinations because the solutions tend to be clear
once you find them, whereas sometimes non-mating combinations
are obscure and it is frustrating trying to find the idea
which may not be that obvious even after you peek at the
solution. (Also I missed a mate in 3 I had against John Grefe
about 10 years ago when I was an A player. I drew that game
but resolved never to miss a mate in 3 again.)

From: derek@uwvax.UUCP (Derek Zahn)
> How do people analyze variation so far out?  Does one keep a mental
> picture of the board in one's head, and make the necessary changes
> on that board?  (I hope not).  How does this work?

I could be mistaken, but I think much of it relates to better "chunking",
as was discussed recently in this newsgroup.

Take a simple example. Suppose you are White to move with this board fragment:

		  BLACK (lower-case)
    ---------------------------------
    |   |   |   |   |   | r | k |   |
    ---------------------------------
    |   |   |   |   |   | N | p | p |
    ---------------------------------
    |   |   |   |   |   |   |   |   |
    ---------------------------------
    |   |   |   | Q |   |   |   |   |

Now, any experienced player knows that's a win for White. They no
longer have to visualize the details (N-R6+ K-R1; Q-N8+ RxQ; N-B7 mate);
it's enough to be able to visualize GETTING to this position, since
this position is known to be a win.

Presumably, better players are that much more familiar with the
various ways in which the pieces combine as they relate from different
positions.

Dave Sherman
The Law Society of Upper Canada
Toronto
-- 
{ ihnp4!utzoo  pesnta  utcs  hcr  decvax!utcsri  } !lsuc!dave

I might also mention that playing a LOT of good-quality
speed-chess does wonders for your game. Find good
people to play with, preferably in a tournament setting so
everyone's trying hard (the Toronto Chess Club, which I was
president of many years ago, had a couple of speed-chess
tournaments a week, for example).

All that playing, along with having to think quickly,
really sharpens your mind for those tactical combinations -
and lets you, in effect, "see" further ahead.
-- 
{ ihnp4!utzoo  pesnta  utcs  hcr  decvax!utcsri  } !lsuc!dave

> From: derek@uwvax.UUCP (Derek Zahn)
> > How do people analyze variation so far out?  Does one keep a mental
> > picture of the board in one's head, and make the necessary changes
> > on that board?  (I hope not).  How does this work?
> 
> I could be mistaken, but I think much of it relates to better "chunking",
> as was discussed recently in this newsgroup.
> 
 . . .
> 
> Dave Sherman
> The Law Society of Upper Canada
> Toronto
> -- 
> { ihnp4!utzoo  pesnta  utcs  hcr  decvax!utcsri  } !lsuc!dave

The last issue of Chess Life has an article on blindfold chess which
also supports the 'chunking' theory. (An aside: Trying a few games of
blindfold chess might help remembering and projecting chess positions.)

I wonder if Nimzovich's 'overprotection' is in fact, when all the
mysticism is removed, really based on moving as many pieces as possible
into the same 'chunk' so that combinations could be arrived at more
easily?  Has anyone tried this, for example, as a computer chess algorithm
(doing exhaustive searching only on pieces within a given strategic
'chunk' instead of the entire board?)?