Saturday, December 13, 2008

First time exposure

Lately I found myself writing the following as comment to a post of BDK:

Now that you are rubbing that old wound again. . . What if it has nothing to do with circles at all? If I look back I made my greatest progress when I entered a new area. When I started with tactics I didn't repeat them. Yet I gained 250 points.

I entered the (to me) new area of positional play last year. I gained another 70 points. If I count my latest winning streak which isn't processed by the rating committee yet, I can add another 80 points. Virtually I'm above 1900 now.
What if you simply can't avoid to memorize patterns when you enter a new area? What if plateauing simply means that you ran out of new area's? Not because they aren't there but you simply don't know how to get access?

For me, there is no proof that repetition is the key to improvement. I didn't improve when I repeated problems. I improved when doing tactical problems only once for the first time. It seems that it is the mere fact of exposure to the practical application of new knowledge that did the job. Looking at the other knights, they usually weren't all that much exposed to tactics before they did the circles. Accidently, unlike me, their first time of serious exposure to tactics coincided with doing repetitions. So it was easy to attribute their progress to the circles while in fact it might be caused by first time exposure.

First of all I have to make clear that I use here a very "Temposchluckeresque" interpretation of the word area. That is, I abuse the meaning of the word to make it fit into my current biassess. Take for instance the two fighting methods I showed you in my previous post. That specific knowledge I call an area. There are zillions of them.

It is by no means simple to get access to a new area. Look for instance at the struggle of BDK with the area of colorcomplexes. Without the feedback of a coach, or a helping blogcommunity, you are having a hard time to find your way in the jungle of chessknowledge.

It all begins with tactics. Without tactics you are nowhere. I remember only all too well how I learned about the power of the bishoppair. I quite wreckaged my position game after game to obtain that powerful asset, only to lose it the very next move by a simple trade.

But those tactics lead to a concrete variation bias which helps you to think forward, while for all other area's of chess you need to think backwards. Seperated by a gestalt-switch, as BDK called it. Only when you sufficiently master tactics, you are ready to enter other area's.

In order to enter the area of endgames, it took me two months only to find out where to start (pawnendings, I thought at that moment). After that I studied endgames for almost two years, only to find out that it was the wrong starting point. Lately I have found a more logical approach and I started with endgame strategy. Only now I have the feeling that I factually entered the area of endgames. Such entrance shows itself by that I get the positions I'm reading about on the board now, and that from the practical application of new knowledge arise new questions of which I am able to find the answers to. A little detour along the crenels and merlons rose a lot of questions which I have managed to answer now, for instance.

I like the idea of the gestalt-switch. When you have mastered a new area, all of a sudden you see it everywhere. While before the patterns associated with the area remained hidden. To me this indicates the road to go to master new patterns.

Based on my own experience I deem the progress by mastering the area of tactics alone at about 250 ratingpoints. High time to enter other area's.


  1. There are really two questions. First, how helpful is it to memorize problems (and of course their solutions). Second, is repetition the best way to memorize? It seems the answer to the second question is yes. The answer to the first question is not as clear.

    It probably depends on the type of study whether memorization is good. E.g., it seems most likely for opening it would be helpful. I would guess for elementary tactics too. I think pure rote memorization is rare, anyway: our brain may initially memorize a back rank mate using a problem in which the rook is on the a file, but it will likely generalize and see the same pattern when the rook is on different files.

    However, I'm starting to think you might be right, that it isn't the memorization as much as the exposure and struggle.

  2. I advocated for some time that our idea of what a pattern actually is might be wrong (rabbits and clouds, maybe you remember).

    For instance, I experimented to conduct my games according the metaphor of the merlons, the crenels and the walking pawnwall. At a certain moment I was beaten by a lower rated player when I had pushed the pawnwall to much forward. My position was overstretched and collapsed.

    That is the pattern of overstretching a position. Now I have seen the pattern once, consciously, the amazing human skill of pattern recognition kicks in. I can't help that, it all works by itself. Even in quite different positions I am able to see the faintest similarity with an overstretched position.

    The problem lies in identifying the relevant patterns. The rest follows by itself. There is no need for repetition. In this post, you might even replace the word area with pattern.

    We might not need so much patterns. Since our ability to see a pattern everywhere once stored is so well developed. Prof. de Groot put us on the wrong foot with his 50,000 - 100,000 patterns.

    The storage is not the problem. One lost game with the pattern, one aha-erlebnis during a study, and our emotions reinforce the storage of the pattern. So memorization is not the problem.

    The problem lies in finding the relevant patterns in the wilderness out there. You probably read about colorcomplexes. That hints towards a valuable pattern that needs to be hunted down. But in doing so you have to get rid of a lot of misunderstanding and of a lot of good advice. But there will come a moment that the coin falls.

  3. The thing with tactics is they don't stick after once.

    I still leave pieces en prise sometimes.

  4. I don't think that's a valid argument. Blundering is due to bad playing habits, not to lack of the ability to recognize a combination.

  5. I think you would be going too extreme in the opposite direction if you think a single exposure to an idea is enough. What we need are skills--I could give a friend a chess book and he could read it and understand the game very well (e.g., outposts, the concept of a fork, etc). But I will destroy him in a game.

    If it isn't repetition of problems that is key to mastery, it is repetition of something. It could be repetition of patterns in practice instead of individual positions (e.g., instead of doing the same back rank mate 10 times, do ten subtly different back rank mate problems). This would seem a very reasonable thing.

    Some kids are probably better at one-trial learning than adults, a more active seed-planter in their brains. :)

  6. Is it possible the first time through you did better, but it took repetition to help the abilities stick?

  7. Well,it is not the first time of exposure to new knowledge, that does the job. And it is not the first time that you try to apply the knowledge (that's usually a big disappointment). But it is the first time that you get an aha-erlebnis with emotional impact when applying knowledge that brings it home. Usually there elapse a few years before first obtaining knowledge and the final impact. But of the latter, you need it only to happen once.

    I said it before that I always had the feeling that my factual improvement in tactics took place in only 6 weeks, when all lights were green. But I can't really back up that feeling with proof.

    A lot of processing of the initial knowledge is needed to convert it into understanding. The fact that we tend forget to apply the knowledge after the first disappointment doesn't help either.

    And then again, I might be wrong alltogether. But what the heck, I going to try it.

  8. There is a problem inherent in developing a new chess skill. Basically, you don’t know what you don’t know. Chess coaches ( a luxury item for some of us) may help in that matter if you find the right one. The rest tend to struggle for a while until we hit the wall.

    Brute force methods like MDLM’s circle methods is a good bridge but should not be the sole method. At the best, it fills the vacancy of knowledge rather unselectively in great quantity.

    The problem with pure memorization of tactics, positional patterns, endgame patterns is that it creates static ( generic) memory markers. Making the recall that much more cumbersome. Pure volumes of puzzles will eventually have some of this seep into the deep abyss of long term recall …when you most need it.

    Back in the days of active MDLM posts and debates, I recall much was said about creating labels and narratives. I thought this was a nice “stepped” approach to adding color to the memory markers because now you are engaged a little more. I believe less volumes of patterns are required with this approach to achieve the same goal, recall when you need it of the appropriate pattern.

    I also believe ( as an evolutionary step of chess improvement) a refinement process could be a great improvement in augmenting chess memory markers for efficient recall. Why waste memory retaining patterns you never encounter? Rather, seek out patterns that would most probabilistically pop up in your games. This one is tricky because you have to first of all know your play well enough to know what patterns to study.

    This requires looking closer at your openings, the type of middle games you end up with and end games you will most likely run into. Other than cursory scans of databases in repertoire for miniatures, the best way to improve here is to create a training positions of the games I played where either I delivered a tactical shot, missed one, or received one. Now, the memory marker that I created has an emotional tag along with the narrative and pattern.

    As if this evolutionary progression of tactical acumen isn’t enough and should be developed ( in stages) by anyone serious enough on the chess improvement ladder, you need to also develop positional sense. It is through strategy that one improves the probability of a tactic being delivered.

    Here, I go through another evolutionary progression. Studying as many master games as possible is a quick way to get a baseline of positional concepts I find. I was back again at the stage of not knowing what I didn’t know. But, I was a little more selective this time. I do a narrative approach to improve the memory markers ( view working through the historic chess tournaments on my own before using the book or a chess engine). I also decided to do a progressive journey intent on seeking evolution of style. I felt this was the best approach. It seems to be effective for me as I do tend to recall certain stratagems quite efficiently.

    I see an event horizon approaching as the next stage ( perhaps next year) where I weed out the selection to a collection of specific games with more depth and meaning to my style of play and endgames I am most confident in.

    I’ve babbled enough. Thanks for hearing me out.

  9. BP,
    great description! I will build my next post on that.