Sunday, December 25, 2022

Now the dust starts to settle

 Now the dust starts to settle I can describe the way to acquire tactical skill in the most simple way:

Learn the 50 most frequent tactical motifs (mates, tactics and preliminary moves) by heart. No matter the method.

For instance:

  • use ChessTempo
  • select batches of max 30 problems per motif
  • repeat them until you know them by heart. Typically at least 20 times per motif
  • do it slow by using system 2. Speed is for testing system 1, not for learning. Say at least 3 minutes per repetition. Let chess logic (system 2) be your guide.
  • select the problems that are 5 ply deep
30 x 50 x 20 x 3 = 90,000 minutes = 1,500 hours is needed for a complete tactical overhaul, but you can focus on the most frequent motifs 

That will cost you 30 x 32 x 20 x 3 = 57,600 minutes = 960 hours which covers 80% of the most frequent tactics that occur in real games. We already knew that TINSTAAFL, but with 3 hours per day you can outplay anyone tactically within a year. So that is a work in progress.

When you plug the tactical holes in your bucket, the other holes start to leak more.

For instance in my opening repertoire. I plugged two of the three holes. The London Jobava system, and as back up the Colle Zukertort work very well with white, as does my recently adopted French defense with black. Apparently nobody below 2000 seems to have a clue when they play against the French with white.

I adopted the classical Dutch with black, for the very reason that it starts with 1.d4 e6, thus inviting white to the French defense, But if  they don't accept the challenge, I am forced to play the Dutch defense. It is a good defense, but everyone tries to avoid it with an obscure gambit with an early e4 or  g4. These gambits aren't great, but black needs a thorough theoretical preparation to parry them. And at this moment, I'm not willing to invest time in that.

So I'm looking for other ways to plug this hole with less effort. At the moment I'm looking at a repertoire based on the Bogo Indian by GM Elizabeth Paehtz which starts with 1. ... e6 2. ... Bb4+ (delaying Nf6). I hope that I can learn this without too much time investment in order to get rid of those pesky early gambits against the Dutch defense.

What I really want to invest my time and energy in is the middlegame. The Art of Attack in Chess is a great extension of my LoA landscape idea. The middlegame (and the opening!) is about building a suitable LoA landscape. Finally, I can bring my usual logic to the game!

Another hole is endgame strategy. I decided not to plug it for now, since it will cost me one to two years to do so. I want to absorb the middlegame patterns to the bone first. Since it is not sufficient to get a winning endgame on the board. You need enough time on the clock to play it. So matters don't begin with the endgame, as Capablanca said. They start with the middlegame.

Merry Christmas!


  1. 30 x 32 x 20 x 3 shows where you can go wrong:

    30 problems per theme: if you make the amount of problems in your set way too big, the knowledge dilutes (guilty)
    32 most frequent occurring themes in practice: if you choose the non frequent themes, you are wasting time (guilty)
    20 amount of repetitions per theme: if you are not patient enough, you stop the repetitions before you have absorbed it by heart (guilty)
    3 minutes per repetition: if you value speed too much, you don't take the time that is needed for absorption (guilty)

  2. A nice summary of the tactical learning process. If you can't understand and articulate a tactical motif, then you haven't learned it. And the conceptual foundation will last a lifetime, while rote memorization of patterns does not.

    Since you mentioned it, I have to say that I found The Art of Attack in Chess to be rather obtuse and not well explained by the author, with some concepts also inherently contradictory or fuzzy in terms of their logic.

    1. I bought the book at chessable, with 30 hours of video with explanation by gm Williams. That clarifies a lot.