Wednesday, August 22, 2007

How to shuffle cards

(looks like cheating to me)

Once I explained my mother how to shuffle cards. I told her that a supple wrist was absolutely essential. We hadn't a deck at hand, so I could not show her. When I came home and took a deck of cards and shuffled them, I noticed that my wrists were stiff during shuffling. In fact it was virtually impossible to shuffle the cards with lose wrists.

On the internet there is no lack of good advice about chess. But even if somebody is a renowed chessplayer, that doesn't mean that his advice is good. Because what happened to me, can happen to the mere mortal side of any master. It is absolute necessary that the advisor has put his advice into practice and has used the feedback for finetuning his idea's.

When you want to improve at chess you have to be very carefull with your use of time. Since there is more good advice around than you can ever follow within a lifetime. And most of it is not put to the test.

Take for instance the famous tree of analysis from Kotov. It is impossible to follow the advice from Kotov in practice. Nunn critisizes this, but his own proposal can't stand the test either.

After studying according to the tree of analysis during a few weeks, a hierarchy in branches starts to unfold. The basis is the risk that is involved. If you sac a knight, the line must work. So what happens when the opponent doesn't take the knight is the last line to investigate. Further the mind doesn't treat serial lines and parallel lines the same. For that reason I consider it to be best to handle the long serial lines first. Then you work your way backwards to the beginning. I hope to formulate a more definite tree-handling-system within a few weeks. The fact that these obvious points aren't mentioned, signals that the authors have devised their idea's theoretically, but that they haven't put them to the test.

The same is true for the famous CCT-system. Checks, Captures and Threats. If you apply this in practice in a complex situation you will find that your list with candidate moves is way too long. A lot of the candidates are simply not worth to investigate. In general it works better from the end backwords in stead of starting with the candidate moves.

I must express myself precise. I don't say that the idea's of the tree of analysis and the CCT system are wrong, I say that they are not worked out in practice. Which means you have to invest a few months yourself to make these theoretical idea's usefull in practice.

And even that is not bad, but of course you always hope that someone has done the dirty work for you.


  1. all perfectly good. thank you.

    if you asked any significantly overweight person how to diet, most of these persons could provide you with lots of perfectly good information on how to do it.

    the issue isnt knowledge, but embodiment.

    or, as Gurjieff said, essence rather than knowledge.

    thanks, dk
    ...bedtime here...

  2. as to the CCT system, I imagine most of play with no formalized methodology of looking at each and every position. I imagine practically speaking it would be a vast improvement for the majority of players.

    Your post reminded me of an old American saying.

    Trust your mother but cut the cards yourself. 8)

  3. As I said in response to a similar comment at my recent post, the key is that you want to avoid overlooking important checks, captures, threats. One inefficient way to do this is to look at all CCTs. While perhaps not super efficient, it isn't all that bad. Most such moves can be quickly dismissed out of hand with a second or two of analysis. My problem isn't looking at too many CCTs, but overlooking those that end up biting me in the ass!

    Heisman addressed this concern in a response to a reader sometime, but for the life of me I couldn't bloody find it this morning.

  4. Here it is. Basically, you can make use of the fact that if you have been checking cct's the whole game, there aren't a whole lot of new ones to check each move. Evaluation inertia! That should help trim the tree.