Monday, July 25, 2022

It works

 Today I finished the first tournament of 7 rounds long OTB games. The result might look a bit meager. My start rating was 1691. My TPR of this tournament is 1716. Normally I don't hide myself behind excuses since they obscure an objective result. But in this case, the result might obscure what has actually happened.

Generally, I try a new training method for a few months. I use to start euphoric, but after a few months reality sets in and I abandon the method. In this case, it is different. I trained this method during 80 days, for a few hours per day. A few of my wins can clearly be attributed to the method. At the same time, the losses can be attributed to other reasons. Like abandoning all opening preparation for my new repertoire in favor of the new method, my poor physical and mental condition due to a long sick bed, my being not used to games longer than 30 moves and so on. During my penultimate game I had trouble to focus my eyes on the board at move 78 after 5 hours, and during my last game I noticed to my surprise that had lost a piece without even detecting it. I'm due for a rest day.

No worries, Tuesday the next tournament starts with 9 long games. So though the method has no actual proof that it works in a way that it is reflected in my rating, I know that it worked, and I will share the details. I will try to be as precise as possible.

What was not paramount:

  • piece visualization
  • move visualization
  • geometrical patterns of the pieces
  • auras of the pieces
  • phantom auras of the pieces
  • conceptualization

They all play a role, mind you, but none of them were paramount. So what is paramount?

Familiarization of the basic elements of a combination.

You might think that you are familiar with all these elements, I know I thought so. But I'm not. The word thought expresses what the problem is. Thinking is a system 2 activity. But you need those patterns to be imprinted in system 1, in order to make it a skill. First of all, how do I know that this works?

I use approximately 750 puzzles. About half of them are mates. The past weeks, I saw 10 hook mates on a regular basis. I studied them thoroughly. In a game during the tournament, at move 53, I all of a sudden recognized a potential hook mate. My opponent didn't recognize the pattern, but he smelled a rat. So he started to calculate. In the end he sacrificed the exchange to prevent the mate. But in stead he created another mate pattern. So I mated him. I wasn't familiar with the hook mate before, and I don't remember I have ever used it.

The difference was, I saw the mate (system 1) while he had to calculate it (system 2). Thinking is by its very nature slow, energy consuming and error prone. A grandmaster doesn't calculate much better than mere mortal people. But since his seeing is so much better, the end result is much better too. I clearly saw now how it works in my own mind. I didn't need to calculate the hook mate at all.

What is important is:

  • the amount of patterns
  • the importance of those patterns
  • the frequency of occurrence
  • how deep your familiarity with the pattern is

The database I use, consists of the more well known patterns. It consists of the basic tactical elements and the mates. Most people know these patterns. And knowing happens to be a system 2 activity. This means, they can calculate those patterns, but they can't see them easily. So their calculation can't be very deep, is error prone and slow.

The quality of your calculation and how deep you can calculate is directly related to your familiarity with the patterns that are needed for that specific calculation.

MDLM stumbled upon this by accident, but he failed to analyze it properly. He took a lot of time for his first circles himself, where he built his familiarity. But in his book he stressed on the last circles, especially the speed of them. Which is completely counter productive and defies the goal of gaining familiarity.

The first step is to master the ABC of tactics. I will give an example how to do that exactly later on. Maybe after the next tournament, since I might need my rest day.

You cannot expect that this immediately rockets your rating trough the roof. Since everybody knows most of these patterns well. But they are not familiar with them. They can simulate a result by calculation. But since that is done with the aid of system 2, their results will be slow, energy consuming and error prone. That is where they will drop points to you.

After mastering the tactical ABC, you need to specialize in certain areas. You can't gain points in an area that everybody knows well alone. You need to enter areas where your opponent is totally blind. When I saw the hook mate, my opponent was totally blind for it. He smelled a rat, but he couldn't find the two moves that could have saved him by calculation alone.

Most people know approximately two third of the type of mates. So you can enter the area where they are blind, by being familiar to them all. The endgame is a clear area where most people have little knowledge of. Since time trouble is lurking around the corner, it can be especially fruitful to increase your familiarity there.

It is of the utmost importance, to keep an eye on the frequency of occurrence. If a pattern doesn't occur frequently, you cannot derive points from it, no matter how familiar you are with it. Luckily, most people are addicted to opening theory, thus neglecting much frequently occurring patterns. Positional moves are well known, but they are again not familiar to most people. You can specialize in exchanges, for instance.

In general, it is a good idea to focus on patterns that happen in every game. Mate, tactics, exchanges, targeting pawn weaknesses, positional moves et cetera. In this digital time, there is an abundance of good training material. But you have to choose wise, young padawan.

5 comments:

  1. In a previous post, you stated:

    A Fide master said that he knew only 20 patterns of the 32 checkmate patterns. What does this tell us? It is not the amount of patterns that counts. What is important, is the frequency of occurrence of the patterns you know and how deep your familiarity with the pattern is.

    The requirement for playing at a higher level of chess is KNOW-HOW (SKILL), not KNOW-WHAT (KNOWLEDGE). Your Fide master may only “know” approximately 2/3 of the most common mate patterns, but I guarantee you that he has the SKILL to identify the potential pattern in ANY position, and that he can then “see” the entire process in one fell swoop (chunk?).

    I found the following confirmation of your current hypothesis in GM Jonathan Tisdall’s excellent book Improve Your Chess NOW, Section 2: Tactical Patterns (pg 90):



    The power to guess the unseen from the seen, to trace the implications of things, to judge the whole piece by the pattern, … this cluster of gifts may almost be said to constitute experience.” - Henry James, The Art of Fiction.

    Our first task in this section is to summarize well-know, fundamental, tactical patterns. Due to their forced nature, these themes are easily catalogued, and have been so in many earlier works. In order not to reproduce too much existing material, I will now refer the less experienced reader to Appendix 1 (mating patterns) and Appendix 2 (tactical patterns), where many of these themes are covered in a relatively pure form. In Appendix 2 I have essentially devoted one example to each common motif [actually THEME, in the Lasker sense], and tried to avoid any examples where there are a combination of motifs involved in a tactical finish.

    One disadvantage of the popular method of presenting a large selection of combinations purporting to illustrate a motifs that in practice several themes tend to blend together. While problem solving is always beneficial, it is not the most efficient way to communicate the essence of the motif. I was extremely pleased to have an agreeable conversation with training guru Mark Dvoretsky on this topic. He feels very strongly that there is no need to be flooded with examples. Illustrate the point and then proceed to USE [KNOW-HOW = SKILL] this knowledge constructively. (I raise this issue for those who might feel that a breezy catalogue in the back of the book is somehow skirting the issue.) These appendices can also serve as a quick review and reference for those who have seen most of this before.

    Once one has encountered recurrent tactical patterns, the nest job is to ROOT THEM DEEPLY IN ONE’S CHESS SUBCONSCIOUS. They should become virtually a reflex, freeing one to puzzle out the more original problems to be solved at the board. In this way, even if the theme is not readily obvious, its proximity to a current position should make its presence felt. It is this kind of learning, ABSORPTION, that is the basis of intuition.



    The problem most of us have is that we misconstrue “familiarity” to be equivalent to SKILL. Sadly, that’s just not true.

    ReplyDelete
    Replies
    1. Maybe absorption is a better word than familiarity. I can live with that.

      Delete
  2. I use a database of 750 problems. Every problem contains 3 elements or motifs or themes or whatever. That are 2250 elements. I wouldn't be surprised if the study of one seventh of the amount of problems would be enough. But "over the top" is my middle name.

    ReplyDelete
  3. Effectively, all boils down to absorbing the tree of scenarios in a way that you immediately see what to do. Being able to reconstruct the tree of scenarios when you need it is not enough. Since you need it every move, and reconstructing is a system 2 activity. You must immediately see that a defender can be eliminated. The method is not difficult at all. In order to learn to see, you need to look. Which is a system 1 activity.

    ReplyDelete
  4. That the method works is evident for me after the next 4 games. Of course that is still subjective, but for me that is enough. I see the proof in how the games went.

    I invested a roughly estimated 240 hours in the new method. To complete a full absorption of all 2250 elements hidden in de 750 problems, I expect to need another 700 hours. With 3 hours per day we talk about 230 days or so.

    We found the method. I found the proof, albeit it being subjective, for the moment. The next step is about optimizing the method. 700 hours of mainly staring at a position in a meditative manner is probably not very efficient. Tomorrow I have a rest day. Maybe I can find some time and energy to write some guidance to optimize the staring process a bit.

    ReplyDelete