Monday, April 25, 2011

The supertrick

















.
.
.
To give everybody a chance to participate I copied a few comments of Anonymous from 2 posts back (maybe you can nickname yourself?).

I did read all the heisman nooks (about a year ago), I saved 29 of them from cca 100. Just before I have written this I revisited "the seeds of tactical destruction" which looks suspiciously similar to idea I tried to convey in my 2nd post. (btw, what if we want to prearrange with a plan/opening to reach them as a piece configuration and obstruct the opponent to achieve as many as possible before we launch the attack?) I remembered I read something similar from him. So, I should make myself clearer here too, because I came down as if I think Heisman is an idiot, but he isn't of course. Still CCT isn't my cake.
I agree with your conclusion about visualization. It is a strong must.
To be a bit funny, it is so important you concluded it more than once throughout the years.
Actually I was very excited because we came to prod this calc/vis problem seriously about the same time, (February 16, 2011 Scandrills -which is a brilliant post) I started to think about it, how to dissect it (I call that decompression in my own lingo), how to practice it etc. I recognized that even if I can visualize the position in my head, I miss crucial features: "takes, takes, takes, not good it won't win anything." Then the engine says: there was a pin there after the exchanges.Ups. You described the same thing when you played that game recently where you DID recognize the future factor in visualization/calculation.


Tempo said: About the kids ..They use a trick. And I'm busy to find out what that trick is and how I can learn my brains to do the same. And that, my friend, is pretty 2D.

You will kill yourself with this idea. Now I will provide factors for your careful consideration.
You are smarter than these kids.
You are smarter than some of them will ever be. (ok, I can't be this confident but you are smarter than most adults, and we concluded rightly that chess != intelligence)
You are after this "supertrick" for years. (see I remember the lingo)

I did some research on the polgar sisters (they are hungarian just like me) and now some on Karjakin, and I read the Waitzkin book in 2010. They are probably good examples. Polgars learned chess every day for 6-8 hours according to a predeveloped program, Karjakin played chess from 5, then at 6 he joined a club, Waitzkin also had an IM coach renowned Bruce Pandolfini,and parental support, as his childhood rival, who's father forced the kid to play chess 8-10 hrs daily. This supertrick looks like play and study positions til nosebleed with expert guidance. Children are very single minded if they like something they put enormous time and mental effort into it without the distractions of the adult life.

To think about the supertrick further let's go back to the topic of my 2nd post, recognizing critical factors. It is a thin ice theorising here and there but maybe you will give it a thought.
We train children's mind to recognize and identify things. We show them cars, birds, dogs and they look at these things with awe. Most of them develop the unnerving custom to tell the adult about their discoveries with loud voice: DOG ! Yea. CAR ! Yea. ...
Because all the things are new, their emotional feedback constantly helps their memory by nudging them to think about their subject of awe frequently. They woke up, and their first thought is x and before sleep, they wanna talk about x. They wanna discuss x in the afternoon also at lunch.
See where this is going?

As a side question, do you know how much time was wasted with things you tried but they did not work? If you could go back, you'd tell yourself: don't bother with that, and bumm you saved months with one sentence. That isn't supertrick, just the fact that being own's coach and player and theoretican is a very time consuming business. Kids skip all this. They aren't explorers, they are a simple copy machine with unlimited paper and tint.

So, why there are super talents? I think now that super talent's brain works differently in the subject area and most often only in the subject area. You can't copy that and most of the time the ST don't know how and why he/she is different. But I don't think it is an issue for us, even with super talent they don't show more than a few percent edge over the rest.

Maybe they have a constant verbal connection or something, an instant "narrative" they employ.
Maybe they are very intent to find details. What we have to do with effort they do it customarily. Somewhere in their childhood they locked in on that type of reasoning.

Look at my brother again with the problem of recognizing critical factors and their consequences to make necessary decisions.
Should i tell him about the million little thing he misses? Yes.
Will he be dramatically better because of this? I doubt that. It is so intricate but it is so simple in the same time. We can talk it over but there are too many situations and quirky details. I chunk the position, he isn't.
He must practice the same critical situations with my help, (and I know which are the critical situations because of my experience) talk it over, and practice it again and again until he can hold all the details in his mind compressed into a single unit and manipulate the elements of it fast enough with the learned algorithm. In fact, he will decide under a second. His ability will grow according to the time he puts into it because I can dissect and identify the elements, and I am familiar with necessary practice for the ability they require.
Clearly, his supertrick is me.

Ok, let's see if we can put this conversation at work.

The moment you wrote that letter in februari you could actually summarize my blog of the past 6 years with one sentence:

"Get a coach!"

And if it wasn't for my discoveries the past few months that would be it.
I will try to explain why I say that there is a trick involved.
There are quite a few things that everybody can do:
  • Getting a coach
  • Read chessbooks
  • Study openings
  • Study endgames
  • Play long games
  • Etc.
If two intelligent persons decide to do this and they put the same amount of time and effort in it, I expect them to improve in about the same way. All these you can summarize under the noumer "hard work". The only difference that can be made in this area is by working harder, longer, more different books etc..

Compare this to the following:
A 13 year old girl which hasn't have a coach, who is dislectic and can't spell "Nimzowitch", who is out of book at move 3, who's IQ is 30 points lower hacks me off the board in 23 moves.
Now that is what I call a trick. This is something that NOT everybody could do, no matter the time invested or the intelligence or discrimination involved.

My moves can be positional subtle as Karpov's, as long as I don't see the pin after ply 8 it is not going to bring me anywhere.

I notice that I have just debunked the theory of deliberate practice as being the sole cause of expertship. Besides an enormous amount of practice a mental trick (a very specific skill) is needed.

In the chess world it are the children who master this trick that are invited to have a coach and to get involved in deliberate practice. From their point of view deliberate practice is the only thing that matters. In the mean time they haven't the slightest idea why these old guys like you and me play so badly.

5 comments:

  1. Your wish has been fulfilled. I also sent one of my notes file, because I was deeply moved by this:

    "I notice that I have just debunked the theory of deliberate practice as being the sole cause of expertship."

    Let's be up to date with the notes before we discuss further this expert motive. I have a feeling you will be interested. If you want I can send you the source too (almost a 1000 page book.

    On the attacking configurations:

    [Event "standard match"]
    [Site "ICC"]
    [Date ""]
    [Round "?"]
    [White "joel j"]
    [Black "nn"]
    [Result "1-0"]
    [ECO "A56"]
    [WhiteElo "2200"]
    [BlackElo "2000"]
    [Annotator "z"]
    [PlyCount "43"]
    [EventDate "2011.??.??"]

    1. e4 c5 2. Nc3 Nc6 3. f4 g6 4. Nf3 Bg7 5. Bc4 e6 6. f5 Nge7 7. fxe6 dxe6 8. d3
    O-O 9. O-O Ne5 10. Qe1 Nxc4 11. dxc4 Nc6 12. Bg5 Qa5 13. Qh4 Nd4 14. Bh6 Nxc2
    15. Ng5 {This represents one of my favorite methods of attacking a fianchetto
    bishop formation. White's last move, Ng5 takes aim of Black's weak h7 sqr.
    Only the black king defends that sqr, and now white has two attackers on that
    (indirectly the queen). By delaying the exchange of bishops, white preventing
    black from adequately defending h7. Trading bishops prior Ng5 would give black
    additional defensive options of h8,h5, h6 etc.} Bd4+ 16. Kh1 Qd8 17. Bxf8 h5
    18. Rxf7 Nxa1 19. Qf4 Bxc3 20. Rg7+ Bxg7 21. Qf7+ Kh8 22. Qxg7# 1-0

    This is from Joel Johnson's book "Formation attacks". Look at how this game flows, as he deliberately aims for this configuration from the start. Basically he can move automatically because he knows his aims (config) in the given situation(s). Interestingly but not surprisingly he mentions alternatives for conduction/continuing the attack if some of the factors are changed.

    [Event "ICC"]
    [Site "ICC"]
    [Date ""]
    [Round "?"]
    [White "Joel J"]
    [Black "NN"]
    [Result "1-0"]
    [WhiteElo "2200"]
    [BlackElo "2320"]
    [PlyCount "49"]
    [EventDate "2011.??.??"]

    1. e4 c5 2. Nc3 Nc6 3. f4 g6 4. Nf3 Bg7 5. Bc4 e6 6. f5 Nge7 7. fxe6 dxe6 8. d3
    O-O 9. O-O a6 10. a4 b6 11. Qe1 Nd4 12. Qh4 Nxc2 13. Bh6 Nc6 14. Ng5 Qd4+ 15.
    Kh1 Nxa1 {Here the idea for Black is to defend the bishop with his queen. But
    white does no thave to exchange bishops to make progress.} 16. Nxh7 Bxh6 $2 (
    16... Kxh7 17. Be3+) (16... Rd8) 17. Qxh6 Qg7 18. Nf6+ Qxf6 19. Rxf6 Ne5 20. h3
    Nc2 21. Nd5 {The idea behind this move is to prevent Black's queenside pieces
    from coming to the aid of Black's King by pinning the f7 pawn. An added
    benefit that it hinders g6 and prevents black evaquating his king.} exd5 22.
    Bxd5 Ra7 23. Rf4 g5 24. Qxg5+ Ng6 25. Qxg6+ 1-0

    and so on.

    ReplyDelete
  2. Compare this to the following:
    A 13 year old girl which hasn't have a coach, who is dislectic and can't spell "Nimzowitch", who is out of book at move 3, who's IQ is 30 points lower hacks me off the board in 23 moves.
    Now that is what I call a trick. This is something that NOT everybody could do, no matter the time invested or the intelligence or discrimination involved.

    I call this: superior calculation a visualization ability. I bet you didn't lose to fine positional moves. She was single minded and honed her only (and very relevantly positioned) skill constantly. DLM did something similar, as he also pointed out the importance of calculation in chess. Most probably she will not pass a certain level though. If I remember correctly Sancho and Blunderprone profited more from the circles meaning they got this ability? Btw I am between 1600 and 1700 on Fics (I dunno how strong is that really) without one circle ever completed and I didn't play more than a 100 games til that point. I lose most of my games by miscalc. Next time you meet those kids dont forget to interview them about their chess related activities.

    ReplyDelete
  3. In the mean time they haven't the slightest idea why these old guys like you and me play so badly.

    old guys are slow, the results in attack-training are worse as older the player.
    Old guys stick to their errors! Old guys have many wrong ideas and even if you tell them, explain them... no chnage. Kids have no "ideas", they see: works, works again, must do more often.

    Old guys knows a lot but they dont "see" it. For example: You know what a fork is, you know different types of forks, you know famous games with forks but you dont see the chance for a fork in 2 moves in your game.
    Kids learn from the result of their action ( they "play" ), adults try to prove ideas (they fight).
    But thats not all. I "think" kids still develop (brain) "hardware", while old guys only work/learn with software (mainly).

    ReplyDelete