Chess theory

Sqod, I used to knock about with a lot of strong players and I was a strong player here where I live when I was playing competition.  All players have different approaches but they all go back through their games and learn from their mistakes if they want to improve.  And they all study endgames and puzzles to help them improve.   Chess is supposedly about logic.  Therefore, having some approach to solving the puzzles rather than approaching them randomly is always going to be an advantage. Getting a TT rating above 2000 does not take much. It is when you hit a wall at something 2500 that you have to have another look at your approach and refine it.  In one sense, it is no different to a competition player wanting to improve their rating.  A random approach will always lead to slow progress and an early peak. You have to understand what you are doing to improve it.  That is normally the role of a chess coach but if you don't have one, it is purely up to you.

Robbie, I think Magnus is very much doing what you say and keeping it a secret.  I have suspected as much for a long time.  He would not be the only strong player.  The difference between Magnus and other players is that his system is superior.  In terms of pure talent, somebody like Nakamura strikes me as having more talent (although I know others might disagree).  But Magnus is a better thinker ie he has a better idea of what he is doing because of his superior system.  When I say system in the case of Magnus, I don't necessarily mean a system he developed through research etc.  It could have been something developed organically.

I've also been wondering if GMs have "secret systems." One quote I read from Bobby Fischer's trainer went something like "I won't reveal, nor would Bobby want me to reveal, what his-prematch preparations are." I always simply envisioned something like practice in visualization, the equivalent to a weight lifter lifting weights or a musician practicing scales, but maybe there's more to it. In recent years, however, I've never gotten any clue from anything I've read that GMs have secret systems. If they did, you'd think they would reveal it before they died, or publish it once they were out of competition, or outright publish parts of it while they're still competing (I'm thinking of Nigel Short's books and all their recommendations), or give hints or pieces of it in their books (I'm thinking of Fischer's "My 60 Memorable Games"), but I've never seen any such hints of the existence of such a secret system. It sounds like, ultimately, they're still basically struggling like the rest of us, just with a lot more (emphasis on "a lot"!) knowledge, experience, and capability in memory and visualization. Sadly, it sounds like they never try to document their knowledge, as is standard in our society today (I'm thinking of all the places I've worked, where virtually every company must train every new employee from scratch, and must let them make the same mistakes that every employee before him/her has had to learn through painful and sometimes very costly trial-and-error). What a waste. How inefficient.

Great players have an invested interest in keeping their training systems etc secret so that the myth of their talent survives their death.  Have you ever thought of that?

I guess I haven't considered that someone would have such a big ego that they'd want to perpetuate a false impression even after their death. What do they think is going to happen to the public's opinion of them after the public finds out? Public opinion will then flip-flop to the other extreme: "He lied to us, and fooled us!" I'm reminded of the violinist Paganini, whom some people thought had made a pact with the devil partly because the sequences of notes he played were almost impossible to reach on a violin at high speed, but it turned out that he just secretly retuned his violin in unusual ways when the audience wasn't watching (http://www.paganini.com/nicolo/nicindex.htm).

Sqod, I don't know if there is any false impression. If a great player discovers some system and keeps it to himself, he is no less a great player.  What Paganini did is no different to retuning the guitar to my mind (I used to play the guitar; I never learnt the violin.)

Still, I've never heard of even the slightest indication of the existence any secret system (until this thread, that is). Wouldn't some GM who made a discovery of some great method want to be remembered as the inventor, especially since somebody else will eventually rediscover it and get the sole credit? Or wouldn't their ego make them brag about their secret? Wouldn't some book somewhere at least mention the existence of a system that's unlike anything that most players are using? (I did find *something* like this in one book, basically the poker points system but applied to position evaluation in chess, but it wasn't secret and wasn't that profound.)

Hmm. Maybe there exists some secret chess society that preserves this secret knowledge throughout the ages? (http://elitedaily.com/life/culture/10-elite-secret-societies-history/) 

Botvinnik made a living after his competition days from his chess school teaching his "secrets".  If there are no secrets, then every chess coach would be as good as another but that is not the case.  The approach to training and practice matters a lot.

I suspect 'secret system' is perhaps open to too sinister an interpretation.  Like easytyz, I think people who develop their methods over time are unable to easily explain what exactly they are doing, and I'd be surprised if many players over time have sat down and worked out such a system in 'longhand'.  

There have been advanced recently in our understanding of memory and thinking processes, as well as our understanding of codified process structures, and it can take some time for someone to come along who thinks differently enough, and has the skills in other area sufficient to explain and teach a new system.

The ATLS system in emergency medicine that structures how everyone deals with major cases of trauma today revolutionised trauma care in the 1980s and 1990s when it rapidly disseminated worldwide.  Like all such systems, it took what had been an heuristic approach, taught in an apprentice system, and codified it into a much more efficient system that sped up care and reduced errors and was taught by instructors in standardised courses.    

However, it only came about because a physician had a crash and was horrified by the poor care he and his family received.  He went about creating the ATLS system and writing it down and teaching it.  I was lucky to be one of the first medical students to be taught it in the UK, and I remember the resistance we met from older colleagues when we tried to use it in the field.  

By secret system, I meant any system which was not divulged to the general public.  I had a soviet coach once and he showed me a concept that he himself had invented as part of his system (I was paying him for the lessons).

How does your system prevent hanging a piece in time pressure?

I use several methods to analyse positions-

1. Algebric
2. Algorithomic
3. Schematic
4. Descriptive

1.Algebric: The method is effective for basic mate and tactical positions.

2. Algorithomic: It is mainly Query based event seeking. It is the most effective method.

3.Schematic: It evolves with specific idea based search.One of these is- 4S schemes [the basic 4 aims of chess - Positonal improvement, tactical aim, promotion, mate threat ].

4. Descriptive: It is purpose and intention based. It is mainly used for annotation and describing games publicly without revealing true schemes & search ideas.

eastyz,

Thanks for your disclosure of some of the components of your system, which is much more than you needed to provide. Your system becomes even more believable to me now because all those components you mention are similar to existing concepts in chess, like center and extended center, the trick of counting attackers vs defenders (or even counting their values), the importance of having open lines toward the king, the limited mobility that comes from being near an edge or corner, and so on, all of which I've read in conjunction with tactics. It also fits with what I've been thinking about, that there exist patterns that we detect subconsciously through much experience, patterns that might start making us look more closely for a winning combination because it looks like there is imbalance. I believe you're right: since most players are merely trying to boost their ratings, just being able to detect the presence of patterns is sufficient for them to play better, without having to figure out exactly what those patterns are. I happen to be highly analytical with a scientific bent so such vagaries are not sufficiently satisfying to me.

Yes, people differ in how they use those patterns they recognize subconsciously. Some players just rely on the patterns for practical reasons and can't verbalize them, others can verbalize them well and become good coaches because they are capable of conveying them, even if their chessplaying isn't top notch. All these general phenomena occur in other fields, too, especially in music. I've always been astonished at how some relatively young adult musicians could show so much sophistication in songwriting when they're too young to have gotten a degree in music, and when the chord progressions they used weren't the type they could discover on their own. They didn't seem to be prodigies, especially since in later years it often seemed that they lost their songwriting skills. I decided that they just hung out with more knowledgeable musicians who taught them what they needed to know without having to learn it on their own through lengthy trial-and-error. That's the efficient way to do it.

In recent decades, with the preponderance of computers and push for efficiency and many old experts retiring from their jobs, there has emerged a field called knowledge engineering. The idea is to interview experts at certain jobs to collect their knowledge to put into a computer. The result is called an expert system, which is considered a subfield of practical AI. An expert system is time-consuming and tricky to write since computer structures and computer logic are so different from the way humans reason and store information. However, that is the way everything is headed how: automation of human reasoning, at least as far as we can push it. Since AI is my field, I've always assumed that managers and employees were aware of this important trend and would be steering in that direction, but the reality is that few managers are aware of it, and most employees consider their private knowledge of the job to be "job security" since if they ever get fired it will cost the company a lot of money and years of time to train somebody to match the same level of skill. That fits with your example of Botvinnik making money as a coach because knowledge becomes financially valuable in areas where people badly want that knowledge when it isn't publicly available.

As for a new language of chess, I'm rapidly developing my own terminology and I'll be posting some of that terminology in my thread on determination of default moves, probably today or tomorrow. I've already borrowed Earth64's term "force" so feel free to borrow my terminology, too, if you find it useful. Although I'm currently using it only for openings, I suspect it will have good applicability to tactics, as well.

Easytyz,

Agreed, thanks for sharing, and makes a lot of sense.

I tried using some of your ideas in some tactics today and I found even the little you provided quite helpful.

Yes, the best manager I ever hired actively sought out such employees: the one who don't want to share their knowledge and expertise (deliberately holding back to make themselves harder to replace, I guess) and made them share for the good of the group.

Your area I suspect has many similarities to the work being done in expert thinking around diagnosis in medicine.  

Ziyrab, thank you for your sarcasm.  However, as I said elsewhere, the system is not for blitz.

Earth64, can you say more about your algorithmic method?  Or is that a secret?  I would understand if you say it is a secret.

Tactics trainer problems are timed. Full credit requires rapid solving. Do not confuse skepticism for sarcasm.

Ziyrab, did it occur to you that I got quicker even though I did not always get full credit each time for speed?  Also, it is much easier and quicker solving a puzzle, knowing that there is a solution, than playing blitz where you have additional thinking processes.  Many of the puzzles have a common theme.  It was so bad at one stage that a certain person would cynically write "Not another queen sac".  Things have improved somewhat since but there is still the issue that you approach a puzzle thinking that you have seen this type of puzzle before, if not the very puzzle.  The hardest puzzles were the endgame studies, in my opinion, but only because we don't normally study endgame studies.  So I started looking at typical endgame studies and sure enough there are a number of common themes and I started to get the knack.  There is a language of its own for composed problems, one which I don't understand.  But that tells you that you can evolve your own understanding of these things to facilitate problem solving.  Practice makes perfect.  But intelligent practice will always be better still.  I should also say that I don't have the know how to use an engine for assistance.  I would not know how you would begin to do that and solve the puzzle in time.  I have had that accusation thrown at me a number of times but I would like to know how you could do it.  Do you know because it is a total mystery to me?

I saw an explanation in another forum. I don't know how it works.

Your method seems plausible. But if it doesn't translate to better play in blitz and OTB, then something important is missing from the TT selections.