Monday, January 21, 2008

About calculation

The idea of narratives is so flexible that you can use it for almost anything. Even the first problem of my new problemset unveils an abundance of new information when pried open with narratives. Some information on important elements of calculation is revealed. What is it what you do exactly when you calculate? If you know that than you can devise a method how to learn it.

So be patient, the position itself is not specific important or interesting, the implications are, though.

White to move.

White is an exchange and a pawn down. Black's kingposition is unsafe though.
If you have been on a diet of 6 queensacrifices before breakfeast during the past years, like me, then you will see a red flashing "Bxh7" appear in neon-letters above the board.

I'm sure there are a lot of chessplayers who will just play the move by intuition and "look what happens". If you are less lazy, you look if there is a forced mate somewhere. Since you are a full rook down after the bishop sac, you are not interested in or worrying about material compensation, unless black must give his queen to prevent mate.

So what are you trying to calculate? You try to drive the king into a place where he is with his back against the wall. Into a matingnet so to speak. The fabric of the matingnet can consist of 3 types of material:
  • Squares covered by your pieces
  • Hostile pieces
  • The rim of the board
We always talk about patternrecognition. But there are no pins or skewers or forks here. So what is actually the nature of the patterns we talk about? I will show you.

To see the rim of the board will cause you no problem. Nor will the opponents pieces, since you physically see them. But the squares that you cover, that radiate from your pieces, that is much more difficult to see. When you drive the king to his demise, you must know where that is. This is one of the aspects of calculation. I call this chess vision in it's broadest sense. If you are training calculation, it must cover the aspect of chess vision at the least. There is more to it, of course, but this must be learned in any case.

There are 3 main lines in this position, each with increasing difficulty. More difficult to calculate.

The first variation goes as follows:
1.Bxh7+ Kxh7 2.Qh3+ Kg6 3.Qg4+ Kh7 4. Qh4+ Kg8 5.Qh8#
So you drive his king simply back to g8. Look how inert the black pieces seem to be. The same for most of your own pieces. They just form the rigid background of the action.

If you enter this solution Renko will not count that has a correct answer. His line is
1.Bxh7+ Kxh7 2.Qh3+ Kg6 3.Qh6+ Kf5 4.g4+ Kxg4 5.Qg5+ Kf3 6.Qg3+ Ke2 7.Qd3#

It's not clear why this is preferred by Renko. It has to do with the amount of useless sacrifices black can do to delay the mate, I belief. For the essence of the story this is irrelevant.There are 3 flavours in this line with the pawn sacrifice on 4.g4+ with the black king going mate on e2, e4 or h3.

If you look at the moves white make they are all checks, captures or threats. That are the only 3 tools you need. When you move your pieces around, you must be well aware of the radiation coming from your opponents pieces, so you will not place a piece en prise.

Rybka comes up with a quiet move. A tool usually too difficult to use for mortals. That is hard to calculate for a human, since it gives black actually a tempo for free and you must check all candidates for their harmlessity. Which is a hard thing to do for a human. That line goes as follows:
1.Bxh7+ Kxh7 2.Qh3+ Kg6 3.Qh6+ Kf5 4.Re1

So how does your chess vision improve? When you know you can work on it consciously.

1 comment:

  1. I call this "Cover the squares technique" and has helped my very poor tactical game improve to reasonably good over the past couple of years.

    In your example I saw them in the exact same order but on the second set was about to stumble by hanging my queen with a check on g5after the queen check on h6. But then I did see that it failed as the black king could simply capture the white knight.

    So after taking stock saw the deflection move of g4+. Without the "cover the squares visualization technique", I wouldn't have noticed these three sets of attacks a couple of years ago.

    Perhaps you are right, I am learning. Old dogs do learn new tricks---but very slowly I fear.