Monday, March 19, 2012

Devising a training
















.
.
.
.
What training is covered by CT?
If you have a close look at the grand scheme of tactics of the previous post, you will find that CT already covers a training for the following elements which play a role in the grand scheme:
  • Pin
  • Double attack
  • Discovered attack
  • Skewer
  • Rontgen attack
  • Trap
  • Annihilation of the defender
  • Clearance
All these can be simply trained if you are a member of CT and you define your own filters. Personally I like speedtraining, but any regimen is good, ofcourse.

Munich insists, and I'm inclined to agree, that he has found a winning and proven method here. I'm going to give it a try. Maybe he is willing to repeat it one more time here and describe exactly what he is doing and in what order and at what speed. For those who didn't follow/are confused by the comments and just to be perfectly sure that we are talking about the same.


What training is not covered by CT?
What is in the grand scheme of a tactic not covered by CT:
  • Identification of the attackers.
  • Identification of the attacking squares.
  • Identification of the problems on the road from attacker to attacking square.
  • Solving of the problems on the road from attacker to attacking square.
  • Identification of the targetsquares.
  • Identification of the targets.
  • Identification of the problems on the road from target to targetsquare.
  • Solving the problems on the road from target to tatgetsquare.
There are two approaches:
  • Start with the attackers and work forward.
  • Start with the targets and work backwards.
I belief that the first approach is the most natural and easiest to implement.

I have experimented in the past with target scans here, here and here for instance. But since I didn't exactly know what I was looking for, the attempts lacked the necessary precision.

So let me first formulate what I want to accomplish.
If I'm trying to solve a high rated problem at CT, and after initial probing by means of trial and error I am still clueless,  I need to guide my thoughts in the right direction. What thoughtprocess guides my thoughts in the right direction?

Identification of the attackers.
There are always quite a few pieces and pawns of which you immediately know: this piece is not going to be involved in an attack of any kind. These pieces can be eliminated beforehand. All other pieces of which you are not sure must be investigated for if they can act as an attacker.

Identification of the attacking squares.
This is a bit more complicated. What are the properties of an attacking square?
  • The attacker is not yet on the attacking square.
  • The attacker must be in direct contact with the attacking square OR
  • The attacker must be able to reach the attacking square with tempo.
  •  An attacking square is in direct contact to the target squares.
In general every square an attacker  reach with tempo is a potential attacking square. This must be investigated for every potential attacker.

Identification of the problems on the road from attacker to attacking square.
The following problems can occur underway:
  • The attacking square is in the hands of the enemy.
  • The road towards the attacking square is in the hands of the enemy.
  • The road towards the attacking square is blocked by an enemy piece.
  • The road towards the attacking square is blocked by a piece of your own.
Solving of the problems on the road from attacker to attacking square.
It is useless to think further about attacking squares if you can't reach them.
These are the methods to solve the problems:
  • Exchange the defender.
  • Annihilate the defender.
  • Chase the defender away.
  • Decoy the defender.
  • Block the defender (with tempo!).
  • Clearance by your own piece (with tempo!)
If you can't solve the problems for an attacking square, you can prune it from further thoughts.

Identification of the targetsquares.
Targetsquares come into two flavours: with a target on them and without a target on them.
It's way easier when the target is already on the targetsquare.
There is a geometrical relationship between the attacking square and the targetsquares which depends on which piece is attacking. If the attacker is a knight, the targetsquares are at a knightforks distance. If the attacker is a bishop, the targetsquares are on the same diagonal as the attacking square etc..

Identification of the problems on the road from target to targetsquare.
  • When the target is not yet on the targetsquare, it has to be forced onto it.
  • The target is of lower or equal value than the attacker AND the target is defended.

Solving the problems on the road from target to tatgetsquare.
If the target is not on the targetsquare, you have to force it there. There are to methods to do so:
  • If there is a piece on the targetsquare of low value which is protected by a piece of high value, you can exchange the lower valued piece, so that the high value piece has to step onto the targetsquare.
  • If the target is almost trapped yet has just one escape route: the target square, you can chase the piece onto the targetsquare.
  • Get rid of the defender. See methods above.
Coming weeks I'm going to try this out in practice. Adjustments will follow, I'm sure.

11 comments:

  1. I think there is a problem with causality with your ideas about a new training:
    A problem has an/some attacker but a attacker dont have problems, meaning you cant detect attackers without already knowing the pattern. ( same with targetsquare and so on )

    Of couse the attacker is "active" somehow, so you can say (sometimes) wich piece might not be involved in a tactic, but the attacker has to be able to adress a special weakness of the opponent.

    I am more with Heisman: its necessary to see the "roots of tactical destruction". A fork f.e does usualy work, because there are 2 pieces/squares wich are not efficient protected

    There is a need of a weakness at the opposite side, and there is the need of the Ability to use this weakness.

    The three Trainings of Fritz: Check, Attack and Defence are adressing this form of "boardvision" for weaknesses.

    ReplyDelete
  2. When you look at a difficult problem, there comes a moment that you "run out of patterns". Not because you haven't them in stock, but because your mind have set a certain context which limits the popping up of them. In that case you have to guide your thoughts in a promising new direction. To reset the context.

    That is the purpose of this idea.

    The seeds of tactical destruction you take allready care of by exercising at CT and doing Fritz-exercises. There is no need to stop with that. But this plan A only brings you so far. When you are clueless in a position it means that plan A doesn't work, so you need a plan B. This is plan B.

    I identified what is NOT trained bij plan A but is still an element of the tactic scheme.

    The scheme is based on the high rated problems I have done the past months. Every problem fits in this scheme. Except for promotion problems. The reason for this is that I only accept pieces as targets, not squares. Maybe I should make an exception for the promotion squares.

    ReplyDelete
  3. For those who are interested in estimates: my estimated FIDE blitz-rating is 1798.

    ReplyDelete
  4. Hello Temposchlucker,
    I just sent you an email with my training scheme in a nutshell.
    Make sure my email does not get into your junk folder due to my strange name: ex_monster

    greetings

    Munich

    ReplyDelete
  5. If you run out of pattern, then your set of pattern was not big or relevant enough. You need to start calculating or searching ( you may attribute the searching pattern =: "guided" ;)

    I dont see how a "attacker-", "targetsquare-" training can work. How do you see an attacker without knowing the pattern?

    If you are using high rated problems.. why are you not using compositions. They are always difficult.

    ReplyDelete
  6. @Aox,
    you don't seem to grasp the concept of guidance. Maybe I haven't explained it well enough. Or do you reject the concept?

    ReplyDelete
  7. f you are using high rated problems.. why are you not using compositions. They are always difficult.

    Why would I use irrelevant patterns? To my knowledge CT hasn't any either.

    ReplyDelete
  8. menvsk xpeopyou are using high rated problems.. why are you not using compositions. They are always difficult.

    Why would I use irrelevant patterns?


    My point ;-) High rated problems = less common = less relevant.

    Where are differences between high rated problems and compositions?


    you don't seem to grasp the concept of guidance. Maybe I haven't explained it well enough. Or do you reject the concept?


    I think i understand you concept. But i think you will get your guidance AFTER you already solved the problem.

    How can you find a targetsquare without already knowing/estimating the tactic.
    Maybe you can give an example?

    by the way, now i type the 5tht time these silly letters, i cant read this captcha

    ReplyDelete
  9. @Aox,
    I think i understand you concept. But i think you will get your guidance AFTER you already solved the problem.

    A strange thing to say. So I assume that this means you reject the idea. In that case I understand that you stick with plan A.

    My point ;-) High rated problems = less common = less relevant.

    For plan A relevance is key. For plan B the patterns of A are irrelevant.

    How can you find a targetsquare without already knowing/estimating the tactic.

    by finding new patterns that are relevant to plan B. Patterns that you learn with plan A won't help you here. So far I have only identified what is needed to implement the tactical scheme and what is not covered by training via CT or Fritz.(The WHAT).
    The next question is to find the HOW.

    ReplyDelete
  10. No doubt: There is something like guidance, i use it in a different way. My personal experience is, that i can "understand" a position. I see the function of every piece, the neccessitys of positions the strengths and weaknesses and come by this way to the solution of a puzzle. Its a lot of "If" "Then", some "Cant" and "but" a few "can only", "would like to", and so on. Its an analytical process, thats A form of guidance, you are looking for a different form of guidance. That there might be something is no problem with me. But the points you named ( targetsquare and so on ), this! seems to me not helpful.

    Of couse, if you know targetsquare attacker and so on, you can construct the tactic "easy", but i think you cant detect these tactical elements without already knowing the concrete pattern.

    To me, it sounds like if you want to build a house and you want to know, where every stone will be BEFORE its build up. Stone 512 : will bee Kitchen outside wall top left.

    Maybe you can give an "concrete" example how that might work?

    ReplyDelete
  11. Many patterns we "know". The meaning of "to know" is here - you heard about it, and if you think long enough for a minute or so, you will solve it.
    But real knowledge of a pattern is, if you see it quickly. In almost no time. Like you read these words.

    I give you an example of a pattern you know quickly:
    http://chesstempo.com/chess-problems/64208

    It least you should find it within 10 seconds, maximum 12 seconds. Not longer. If you needed 36 seconds like the average user did, then you dont know the pattern well. I reckon I would solve this tactic including all 5 moves in about 8 seconds.

    Now imagine you are in a real game and the situation of the puzzle is only a future possibility. We talk of the position before white took the Bg7 with 1.Qd4xBg7??

    What was blacks previous move?
    I looked it actually up, and my guess was right. It was 0.Nf6-g4!
    To move this move, you need to see that your bishop Bg7 is hanging and the queen could take it. But if she takes that Bg7, then you have a smother check mate.
    However, this is only a possibility.
    As black you also need to see what white else could do.
    Hm. 1.Qb6 looks agressive and still prevents the check mate, but then you have 1...Be4! attacking the pawn h2 and a guardian of the square b8, so you can maybe move your rook here to attack the queen while she needs to stay on the diagonal to prevent the check mate.

    No matter what happens, you will find, that Nf6-g4 is quite winning. But you should not just hope for bad moves of white, but you need to consider Qb6 as well.
    You see, a 1500 rated player would only be able to find the check mate and needs on average 36 seconds. However, a strong player looks at Nf6-g4, sees the Bg7 is hanging, but also sees within 5 seconds that if the queen takes, he delivers a nice check mate. You would never ever consider Nf6-g4, if you cant find this check mate within 10 seconds. Because you would look at it see that it is hanging, then try to see if the check is any good, but since you dont know the pattern inside out you simply see, that you dont fork anything with the check. So you discard the move Nf6-g4 and are worried instead if you can survive the situation, because you wont be able to castle anymore. You are in deep trouble.

    So to "know" a pattern within 36 seconds is not good enough.
    If you look at the patterns you really know, you will see, you dont know much. All you can say is "yeah, I think I have seen this before", like you heard about it, which is not good enough.

    And honestly, I believe I would have seen Nf6 in a Blitz game, maybe even in a bullet game. I really know this pattern inside out, and since Nf6-g4 is a discovery, I would look at it and see if I could use the Ng4 for an attack. I would not miss it, I am sure about it.

    ReplyDelete