@ReneDescartes

I really want to comment on many of the good points you have made in this thread, but I'm a little short of time now.

I will just mention that Avni's "The Grandmaster's Mind" has 12 interviews with example games and positions, 9 of which are with top Israeli GMs on their playing style and thinking processes. The interviews are not overly structured; Avni was "looking for insights, not for proofs". 

Then follows a 40-page discussion comparing the findings from the interviews with the standard conclusions from de Groot, Simon & Chase, Gobet, Saariluoma and Charness. Among Avni's criticisms of standard chess psychology:

- The "thinking aloud" methodology seriously interferes with thinking and is a poor simulation of actual thought
- Problems used in research are often too easy to differentiate GMs from mere masters or experts
- Many studies call players as low as 2100 "experts" and proceed without any titled subjects, while pretending that they are studying "expertise". Therefore, while chess thinking has been studied quite a bit, the thinking of GMs is still poorly understood.

For me the most important insights of the book were:

- How prophylactic thinking is absolutely fundamental to all these players (and presumably to all GMs), fully vindicating Dvoretsky (and Nimzowitsch before him) in his "obsession" with this concept.
- How ruthlessly practical GMs are; very conscious of the clock as a factor in the game and unwilling to spend lots of time on a tempting or fascinating line that is unlikely to succeed.
- The room for difference that still exist between these players: Intuitive vs. constantly calculating, defensive vs attacking, etc.

The other Avni book I like is "Danger in Chess", which has similarities to Heisman's "Looking for Trouble", but on a higher level.

@Jupp53

Very good point about fitting the study material to playing strength. I've also seen Heisman write a lot about that in his ChessCafe column. Many people (particularly adult beginners) make the mistake of thinking they can skip the "easy" tactical drills, for example.

@Jupp53

It seems we agree about a lot--especially Lichtenberg. Thanks for reminding me of him.

ReneDescartes wrote on 06/24/10 at 23:31:51:


What was the chunking concept introduced for? (For someone who wants to read it by the creators the article: "The magical #7 +-2" is a recommendation. But without psychological education it's a lot of work.) In the research about short term memory were a lot of contradicting results in terms of "pure" information theory measures. It became evident that the memory system uses non-physical-world units with a base in further systems. (Or in other words: What we already knew was more and wrong. ) [/quote]

Quote:

True. But why but? 

Quote:

You can't blame the empirical findings of DeGroot or the chunking theory for what some people make with it. As Lichtenberg once said: A book is a mirror. If an ape looks into it you can't expect an apostel to look back.
Quote:
My opinion too about the prospective research. But I see other consequences and the "congenial, experienced and successful trainer" will not contradict the chunking theory. There is some information in it useful (b.e. forget to think about "talent" as long as you're able to work over the material and some more if you know about memory reseach). 

For the chess player without empirical psychological education I'd prefer Dan Heisman as literature till 1800 and Dworetzki from 1800 to give a hint about the idea.

@ReneDescartes

Well, I'm not necessarily against using the results of psychological research in practice. After all, if it's not useful for anything, what's the point?

I think the "psychology of chess" as it now stands has many useful conclusions that should be taken into account by writers of "chess training" books and vice versa, but the problem is the two "traditions" tend to ignore each other!

Modern chess authors that I think have insights the psychologists could make use of include Rowson, Dvoretsky, Aagaard and especially Avni (who is ignored by the chess psychologists even though he is a psychologist!). His brilliant "The Grandmaster's Mind" should be required reading for anyone doing research in chess psychology.

Btw. Fernand Gobet besides being a psychologist is an IM and former Swiss national player, and I've seen him quote Kotov, Dvoretsky, Karpov and Shirov in his papers. So to give him some credit, he's more of a bridge-builder than most of his colleagues in psychology.

You're right about De la Maza. His method is basically sound and many amateur players do need to work hard on their tactics to improve. But if his program is all the chess training someone does, their enjoyment of the game may be reduced and burn-out becomes a possibility. So it's not just pleasant but also wise to study some strategy, great GM games or sublte endgames in-between all the tactics.

TicklyTim wrote on 06/25/10 at 16:03:13:

I think you quoted some such material yourself, namely the books by Baburin, Vukovic, Grooten and Gelfer? Nothing wrong with either of those. 

Or if you want it even more comprehensive/systematic there's Polgár's brick "Chess Middlegames" with 5000+ positions and Aagaard's two-volume "Attacking Manual". On endgames there's Dvoretsky's Endgame Manual, Averbakh's 5-volume "Comprehensive Chess Endings" and Flear's "Practical Endgame Play: Beyond the Basics".

A lot of those ex-Soviet ring-binders are now sold as computer training programs by Convekta; this should be a more effective way of assimilating material than books. For example they have the aforementioned Comprehensive Chess Endings on an updated DVD.

So the material is there on paper or on disk (whichever you prefer), that's not the problem. The real problem is investing lots of time in learning and repetition while keeping it interesting enough not to quit the training from boredom. One good way to cut down the workload (at least initially) is to follow Gobet and Jansen's advice and focus on those patterns that are likely to appear from your chosen openings.

ReneDescartes wrote on 06/24/10 at 23:31:51:

Didn't I just say the same thing in the previous post? 

I like your wording, though - chunking does seem to be more "descriptive rather than prescriptive". That's a good way to describe it.

I do somewhat disagree with your assessment of Michael de la Maza, though. I don't think he has contempt for master play so much as just lack of understanding. He found a "quick fix" to get from 1400 to 1800 by over-addressing the biggest weakness in his game (tactics), and he got frustrated trying to figure out how to reach the next level after that. 

But I do think his original (in)famous "400 Points in 400 Days" is a worthwhile read for class players, just to emphasize the importance of tactics to players who routinely make tactical oversights and could learn from his study methods. I don't prescribe following his method exactly, but doing lots of tactics puzzles, along with other forms of Chess study, is important.

Easy answer:

1. Play lots of Chess.
2. Study lots of master games.
3. Do other Chess study.

Eventually, you'll see a lot of the same stuff over and over, and your brain will be quicker about recognizing that stuff during your own games. That's your chunking.

A few mentions have been made that stronger players use chess 'chunking' or 'templating' when assessing a position and deciding on their move. It is said that through their experience (or study) they have an idea of what to do already as a position will be in parts familiar. To some extent, they have trained their intuition.

If this is one of the main differences between Masters and club players then aquiring these chunks/templates knowledges seems like a good step towards improving.

So how can the club player try and attempt learning this chunking/templates?
I think I heard that a youthful Bobby Fischer set about the task of trying to assimilate every possible position (don't quote me, I'm not sure where I heard that from).

It seems to me that studying differing pawn structures might be a start. The book "Pawn Structures in Chess" by Soltis covers pawn structures based on various openings. Baburin's "Winning Pawn Structures" concentrates more generally on isolated and hanging pawn structures.

Structures would be chunking/templating on a large scale. These pawn structures cover the whole board. Is it possible to do some macro-chunking? That is to study little motifs in an isolated area of the board.

I'm sure many people have studied the options of the 'classic bishop sacrifiace' probably through reading Vukovic's "Art of Attack". This systematic breakdown of tactics (the attack vs Dragon also comes to mind) seems easily attainable, but I feel the chunking/templating that is required is more positional. Do people agree with that?
It seems more difficult to break down positional ideas. Grooten's "Chess Strategies for the Club Player" seems to do this to some extent. I also vaguely recall "Positional Handbook" by Gelfer, which takes typical middlegame themes and tries to demonstrate them through repeated examples.

Does chunking/templating need to be hooked onto a specific pawn structure or is macro-chunking (localised motifs) possible.
What should people set out to assimiliate for this chucking/templating?
Pawn structures seems obvious, but is anything else possible?

Is there more to this than just reading a good middlegame book? I hope so.
People tend to learn maths by repeating multiple exercises on the same theme. Maybe to aquire this chunking/templating, we need to do something similar for chess positions. Maybe the difficult task is sourcing the information in a form that is readily learnt. 
Do we need a middlegame book that is more like a school textbook, in a similar way to the endgame books are laid out?

Once deciding on what to learn to build up chunking/templating, the next question is what differnent techniques are there for learning them.
Surely they need reviewing every so often just as we review our opening to keep them fresh in our mind?

So three questions come to mind:
- What does the chunking/templating actually consist of?
- What sources of material can anyone recommend for this?
- Are there any good learning techiniques that might be more appropriate for building this knowledge base?


- Tickly.