11 January 2017

Patterns: Some Evidence

Working through a lesson series called "Advanced Tactics" on Chess.com, I came across this problem.

Black to move

The diagram is upside down (Black on bottom).

This lesson series was created for Chess Mentor by Thomas Wolski. It contains many lessons from the games of Wilhelm Steinitz. In addition to enjoying the tactics practice, I am becoming impressed with Wolski's ability to extract lessons from Steinitz's play. I am beginning to think that a more sustained study of the first official World Champion's games might be in my future.

I spotted the first several moves of this combination in one second, that is, instantly. This instant recognition of most of the solution, and the confidence that the rest would be forthcoming stemmed from having seen essentially the same ideas in two other problems that I put in front of youth players in the past two weeks. See "Carlsen's Queen Sacrifice," problems 3 and 6.

This instant recognition strikes me as evidence of pattern recognition as an element in the development of chess skill (see "Patterns and Calculation").


  1. 1...Qxh3+ 2.Kxh3 Ne3+ 3.Kh4 and now two moves work:

    3...Ng2+ 4.Kg(h)5 Rf5+ 5.Kg4 Rf4+ 6.Kh5 (6.Kg5 h6+ and 7...Be8#) Be8+ 7.Kg5 h6#

    but some attractive mates also come from:

    3...g5+ and then
    (A) 4.Kxg5 Rf5+ 5.Kh4 (5.Kh6 Ng4#) 5...Ng2+ 6.Kg4 (6.Kh3 Rf2# is the prettiest, most economonical finish) 6...h5+ 7.Kh3 Rf2#
    (B) 4.Kh5 Bg4+ 5.Kh6 (5.Kxg5 Rf5+ and 6...Rh5#) 5...Rf6+ and ...Rf5-h5#

    1. I got as far as 3...Ng2+ before doing any calculation. Had I not created the other two problems referenced, I might have looked at 3...g5+.