Main
Date: 10 May 2005 12:11:27
From: Peter Billam
Subject: types of move you fail to predict
Greetings.

I hadn't previously seen this suggestion from (sorry bout this) the Go
faq www.faqs.org/faqs/games/go-faq/ about how to improve your game...

Another way of trying to diagnose "mistakes you did not know you were
making" is to play through a lot (say 10 or more) of professional games
trying to predict the next move. See if you can reach some sort of a
conclusion on which types of move you are consistently failing to predict

That's interesting, "types of move"... In a chess context that might
be things like pawn-advances-when-challenged, bishop-moves-backwards,
unite-the-rooks, regroup-rather-than-exchange and hundreds of other
"types". Is the concept of "types of move" systematised in chess ?

(L.Polgar divided his book of endgames into chapters
which could be considered endgamish "types of move".)

Is the identification of "types of move you are consistently
failing to predict" part of some teaching tradition in chess ?

--

Regards, Peter

Peter Billam, DPIWE/CIT/Servers, hbt/lnd/l8, 6233 3061




 
Date: 10 May 2005 13:37:30
From: Mark Houlsby
Subject: Re: types of move you fail to predict
Hallo Oliver

>Such testbooks are also available for endgames - i found here that my
play
in pawn endgames is better than
in N-endgames (however i found after some analysis that also FMs and
GMs
overlook things in these N-endgames).
And also in endgames king + 1 pawn vs. king + 1 pawn sometimes even
experienced GMs make mistakes.

I have "Rate Your Endgame" by Ed Mednis and Colin Crouch. Could you
please tell me of any other "HGIYC" endgame books which you may know?

Many thanks

k



  
Date: 12 May 2005 00:37:35
From: Oliver Maas
Subject: Re: types of move you fail to predict
Hi k

i have "Testbuch der Endspieltechnik" (testbook of endgame technique) by FM
J. Konikowski and P, Schulenburg, which appeared in the Beyer Verlag. I dont
know whether theres an English version. Its part of a series of testbooks
(editor B. Feustel, Bamberg, Germany).

friendly greetings,

Oliver





 
Date: 10 May 2005 22:29:59
From: Oliver Maas
Subject: Re: types of move you fail to predict
Hi Peter

i agree that the method "playing through GM games trying to predict the next
move"
is quite good. In fact this is the method GM Danny King is using in some of
his books.
("How good is your chess?") This training can be some fun too.

I personally think that its also important to know the types of positions in
which one plays better
than in other positions. Looking at several GM games, i found for example
that i predicted
Blacks moves in the game Magerramov - Kasparov (1977, 0-1) mostly correct
(58 out of 70 points),
(consequent attack against the white king, queens on board) while i was
quite planless on the White side
in Seirawan - Karpov (1990, 1-0) (middlegame without queens).

Such testbooks are also available for endgames - i found here that my play
in pawn endgames is better than
in N-endgames (however i found after some analysis that also FMs and GMs
overlook things in these N-endgames).
And also in endgames king + 1 pawn vs. king + 1 pawn sometimes even
experienced GMs make mistakes.

Types of moves, yes, interesting concept, remembering some of my last games
i found that i overlooked more than once
an exchange sacrifice by my opponent. :)

regards

Oliver




 
Date: 10 May 2005 02:40:04
From: Neil Schemenauer
Subject: Re: types of move you fail to predict
Peter Billam <[email protected] > wrote:
> Is the identification of "types of move you are consistently
> failing to predict" part of some teaching tradition in chess ?

I'm not sure but I think the training strategy I'm using qualifies.
What I do is blunder check my games using a chess engine. All of
the positions in which I make a major mistake (e.g. >= 1.5
evaluation drop) go into a database along with a list of acceptable
moves.

I'm using "Chess Position Trainer" to maintain the database. I
don't like the training interface though so I wrote my own (it reads
the CPT database). Right now I have about 300 positions in the DB.
I make a lot of errors. :-)

My idea is that I have certain blind spots. Usually I make okay
moves but sometimes I completely miss what move is necessary in a
position.

Neil