Chess theory

Sort:
eastyz

The other thing that has occurred to me is that your concept of a system is static as it relies solely (?) on pattern recognition whereas my system is intended to be more dynamic.  For example, I have a notion called string theory to cater for the fact that there are numerous permutations when there is attack and counterattack in a position.  I invented the idea to formulate some relatively simple rules to deal with such complexity.  The idea was to point the player in the right direction (the "probable" solution) rather than have the player wade through the variations aimlessly.  Strong players do this intuitively but they also miss things.

Sqod
eastyz wrote:

Can you elaborate on sou's or blocks?

 

Earth64 is calling units "blocks," and I'm not clear on how he has defined "forces." (https://www.chess.com/forum/view/general/how-to-analyze-position-scientifically-part-1)

Conceptually, I separated location (= squares) from units for the purpose of developing a very general chess notation that described both, although later I found such a notation already exists (called RAN), although I'll still need to modify it extensively. I had a long response written for you here, but since your question got me to rethinking my approach, I think I'll hold off until tomorrow before posting it.

 

najdorf96

Indeed.

Fascinating thread guys. Others may scoff, be dismissive of such an idea, to me I would rather think that it is a common endeavor every serious, competitive player goes about doing (unconsciously or intuitively as Robert noted) on a progressive trend. Through trial & error, accumulated knowledge experience or with some level of guidance.

Take the Polgars for example. Everyone knows how their parents trained them early on, whether they already had an innate talent cannot be known, nonetheless maybe in hindsight we can actually say that they must of had some degree of capacity to facilitate learning, improving and developing great skill through scientific means.

Essentially, as with the forums theme, creating their own personal system of playing as opposed to prodigies with natural talent (in-born system?)

Many people have related this about Bobby Fischer. That it seems as though he not only thinks in just moves, a string of variations, but in systems. Funny that

najdorf96

I don't know about anyone else, I've always approached my training parallel to Bruce Lee's JKD Philosophy (paraphrasing)

Absorb what is useful

Discard that which is useless

Add only what is uniquely your own.

eastyz

Thank you najdorf96.  Fischer was ahead of his time.  He did not give much away in terms of his approach.  Few great players do.  Capa never did.  He liked the aura of being a natural genius of the game.  Perhaps what we are trying to do here is a bit more scientific than the approach of those players, perhaps not.  We will never know.  All I know is that embarking on a system leads to great improvement if you are diligent enough.  I don't believe in the so-called intuition of strong players.  If it was all intuitive, why did they have to work so hard.  Every strong player I have ever met was a hard worker.  I have never met Wei Yi but he looks as if he could put chess away and get a good night's sleep for a change.  And then there is the Polgars.  Exactly my theory with one minor point of difference only.

eastyz

najdorf96, I agree with your approach.  You have to be openminded so that you don't miss out on something useful.  We know that from Steinitz's days.  He was scoffed at but proved everybody wrong.

RobbieCoull
eastyz wrote:

The human brain tends to do what the ancient Greeks (Plato, Aristotle) discovered and that is to recognise common traits and categorise responses, whether consciously or not.  After all, your input vectors in chess (and even in life) are rarely ever going to be identical.  What the brain can do is recognise that certain differences in the input vectors are likely (or in fact) immaterial.  

 

Which is why people recognise faces in, for example, wood grain or clouds while computers do not.

https://en.wikipedia.org/wiki/Pareidolia

RobbieCoull
eastyz wrote:

najdorf96, I agree with your approach.  You have to be openminded so that you don't miss out on something useful.  We know that from Steinitz's days.  He was scoffed at but proved everybody wrong.

 

That's how paradigm change works.  You can't effect change without people laughing at your ideas at first.  (Of course, having people laugh at your ideas does not guarantee that you are on the verge of achieving profound change....)

 

"First they ignore you, then they laugh at youthen they fight youthen you win." - Mahatma Gandhi

eastyz

RobbieCoull, it would be really something if a computer were to say:  "I recognise a computer face in the wood grain and/or clouds".  I would then pack my bags and go home to mother.

Ghostliner

Nice thread, haven't read through all of it but it looks like it would be well worth the time and effort.

Good man!

eastyz

Ghostliner, I am surprised you have looked at this thread.  I have been much maligned for starting it.  Anyway, my point is that if you want to improve your tactics, you cannot do it at random unless you have all the time in the world and even then, you are better off using some system.  Many have said that it is not possible.  But I disagree.  My rating peaked until I sat down to think about it.

zborg

Simplify through exchanges.  Win via pawn promotion.  That's the simplist way to cut through the confusion.

Indeed -- there's a large body of "chess knowledge," and thousands of exceptions along the way.  But it's not hard to conceptualize this problem, overall.

Just buy a good book by GM John Nunn, or others, and "get working."

End of Story.

eastyz

zborg, as always, we can come to you to get to 1200 on TT.  But some people are a bit more ambitious.

Ghostliner
eastyz wrote:

Ghostliner, I am surprised you have looked at this thread.  I have been much maligned for starting it.  Anyway, my point is that if you want to improve your tactics, you cannot do it at random unless you have all the time in the world and even then, you are better off using some system.  Many have said that it is not possible.  But I disagree.  My rating peaked until I sat down to think about it.

I don't understand, why have you been maligned for this?

eastyz

Ghostliner, best to read the posts and make your own judgment.  There are some very strong opinions about this.

Ghostliner

Ok, understood. I was intending to read through the whole thread anyway, form what little I've seen so far it looks completely fascinating.

Btw, I enrolled on an online chess course about two years ago and, of course, it includes modules on tactical training - it will be interesting to see if there's any crossover with the material here, I'm sure there will be.

Sqod

The way I understand it, eastyz has collected a number of patterns to watch for, and these (often geometrical) patterns (often with strange names, not surprisingly) are clues to which types of tactics/mates are likely to apply in such positions. This is only logical, and it's what we all should be doing as chess players anyway.

I've been collecting such patterns over the past year, though so far they have been mostly aimed at opening heuristics rather than tactics, but it would require little work (though a lot of time) for anyone to come up with their own system of mapping board position attributes to tactics to look for. For example, in my recent thread on a recurring but slightly uncommon type of mating position (https://www.chess.com/forum/view/general/have-you-seen-this-type-of-r-p-mate), the main patterns to watch for are: (1) opponent's king in the "tea cup" of your pawns; (2) your rook in a position to put the "lid" quickly on the "tea cup". This could be computerized relatively easily by defining the essential patterns in a general way, making boolean variables that become true if such a pattern exists on the board, and then conjuncting (= AND-ing) those patterns, and then firing a special procedure to focus on specifically finding that type of mate. For example:

BOOLEAN subpattern_01, subpattern_01, potential_gaiwan_mate, components of <opponent's king in tea cup>, components of <your rook in a position to move quickly to the lid rank>

subpattern_01 = <opponent's king in tea cup>

subpattern_02 = <your rook in a position to move quickly to the lid rank>

potential_gaiwan_mate = subpattern_01 AND subpattern_02

IF potential_gaiwan_mate THEN seek_gaiwan_combination

It sounds like eastyz has gone beyond this stage to a certain extent, and maybe has tried to combine such disparate patterns, or maybe has tried to map directly from specific pattern to specific combination without requiring the user to do calculations. Regardless, such possibilities are absolutely fascinating to me, and I regard eastyz as a true chess pioneer, which seem to be rare nowadays. One idea I had is that if enough members are interested in this idea (and believe in it!) then they could create their own club to collaborate on finding their own patterns. This would be easy to do, especially since on this site there exists a regular source of puzzles accessible to all (the daily puzzle), so everyone in the club could be looking for common threads and patterns behind the puzzles, documenting these, and coming up with a system to compete with a system like eastyz's system. Similar clubs could be created for finding opening heuristics, plans, and other chess essentials.

eastyz

Sqod, there are non-pattern recognition components.  But looking at what you have done, they could also be programmed.  I suppose you have been taken in by the pattern recognition mania, the latest in "science", and how others have slammed me here and on another thread.  It is one legitimate approach but it has limitations because otherwise every strong player (say strong IM and above) would never miss any tactics.  The obvious question is what happens if you don't see a familiar "pattern"?  Do you then have to resort to the random approach?  Where are those knockers?  (Actually, please stay away.)  The received wisdom used to be that if you were good at maths (always plural in Australia), then you would have an aptitude for chess.  Wrong.  So is it conceivable that, while science says that pattern recognition goes with strong players and tactics, it is not the only explanation or method?  It is a logical fallacy to think otherwise to my mind.  But who am I?

Sqod
eastyz wrote:

I suppose you have been taken in by the pattern recognition mania, the latest in "science", and how others have slammed me here and on another thread.  It is one legitimate approach but it has limitations because otherwise every strong player (say strong IM and above) would never miss any tactics.  The obvious question is what happens if you don't see a familiar "pattern"?  Do you then have to resort to the random approach?

Maybe you are referring to the Big Data / data mining / artificial neural network / Bayesian Nets statistical types of approaches. No, I'm certain those are not the road to AGI, if that's what you mean by being "taken in" by those. I was just trying to put together your clues into an overall understanding of what your system does, but none of us will know for sure until you publish it. The fact that such a system even exists blows me away, though, and makes me realize that wonderful possibilities exist that I (and probably many others) hadn't imagined.

HEL555

That great players have some sort of system to improve, that strikes me as obvious. Great players (or great anything) don't simply work and study at random - they work and study in a smart, efficient way.

Like when a student studies for an exam. You can go through all the book, from beginning to end, or you can study key concepts and extract the rest of the information from those key concepts by applying a bit of logic. Or, if this doesn't relate to you, imagine that you're studying for an exam in which the professsor always asks the same questions - what are you going to study, the whole book or are you going to focus on the questions you know he will ask?

How exactly great players do study and how they structure their thought I do not know.