Sunday, June 05, 2005

Pattern recognition in action

In awaiting of my new endgamebook I have time to surf around the web, looking for articles and opinions about pawnendings. I stumbled on this diagram in an article of Karsten Muller, who wrote it for Chesscafe. I recognized the pattern immediate so the solution was obvious.
Since I just started with endgames, I could remember that I had seen the same pattern about two weeks ago in the book of John Nunn. I have published both diagrams, to make it possible to compare the recognized pattern with the original.

Diagram 1, the recognized pattern
White to move draws, black to move wins.

This was the original position:

Diagram 2, the original.
White to move draws, black to move wins.

The interesting point though is that the patterns that triggered the recognition are actual invisible. I studied diagram 2 for a couple of hours a few weeks ago. The study revealed the importance of the f-line. Thats the line black is forbidden to pass. So that's the borderline white has to defend, using the opposition.
The 2nd rank is imporant too, since any bypass operation of black will have that rank as an axis.

So the recognized pattern exists mainly of the invisible cross, formed by the f-line and 2nd rank. Invisible on the board that is, but not invisible in the mind. So during the study I made the patterns myself by imagineing them in the mind.

Now I all of a sudden realize that it's the same with tactical problems. The invisible patterns are more important than the visible ones.
I remember well doing Polgars 5333+1 a few hundreds exercises mate in #1.
I was very enthousiast because I saw the shift from looking at the pieces to looking at the fields they cover.
Maybe we have here the reason why this study is so time consuming. It takes time to generate the important invisible patterns.

A few Knights talked about a kind off video that automatic plays tru all problems and their solution. Though the idea is very good, I can now predict that that is not gonna work.
Because it shows only the visible patterns. Maybe it could work with a computerprogram that generate those invisible patterns in a position. How is the quality of the patterns that are generated by Convekta, the so called arrows of shame and fields of ignorance?
Does that help to see what's important or are they confusing?

This reveals the whiff of magic you feel when you see a grandmaster in action. I showed these positions to a friend who didn't have these invisible patterns stored. Time after time he failed with both colours, unable to see why I reached my goal everytime, no matter what he tried.

1 comment:

  1. I had a simalar experience when I did the Polgar mates in one. Initially I did them a few months before starting my MDLM program. The second time I did it after reading King of the Spills blog where he talked about force fields and had charts of the force fields for each piece.

    I found it to be helpful in that I started thinking very clearly of what squares were locked down by my pieces (the king cannot move to) and what squares are open
    ( free for the king to move to).I felt this was a turning point in that I quickly started to see the position in terms of these squares and what type of piece could effectively lock down these remaining squares.

    As for the "video" of the solutions I do believe this would still be helpful. I would use them in addition to the traditional circles. This would be used for enforcement. This would be especially helpful to go through a video at staged intervals after one completed the 7 circles as well.

    I would also like to have a Polgar brick size book that would show a few thousand pictures of mating positions.(not the problems just the pictures) I think that this would help mating patterns pop out while playing a game. Has anyone seen a website or book that has this ?