Wednesday, September 17, 2008


by IM Igor Khmelnitsky

IM Igor Khmelnitsky

Igor is a winner of many national and international tournaments in Europe and the United States. At various points during his career, he has won individual encounters with many of the game’s best players. He has been a participant in the Ukrainian National Championship as well as a three-time contestant in the US National Championship.

The common theme in the feedback I am getting from those who took my evaluation Exam is how shocked they are with their rating estimate in the Calculation subcategory. Well, with those who submitted their results for my personal review and also had a phone training session with me, I was able to investigate further. I've discovered many different issues, some specific to a person and some general enough that can be applied to everyone.

One of the common issues that I discovered I'd like to share with you today. Humans can't compete with pure calculators, whether mathematical or chess, without certain strategies, rules, principles, standard ideas that can help the "mechanics" of their calculation. I am calling it a personal "knowledge base." The bigger it is the easier is for you to calculate - as you will:

  • discover moves candidates much faster,

  • eliminate the less promising ones more easily, and

  • assess the position at the end of each variation more accurately.

I recently came across this position:

White to move

What is your thought process? Take 15-20 minutes then compare your ideas with mine.


Let's take a look together now:

  • Quick assessment - White is dominating materially, but can't stop Black's e-pawn.

  • Ideas - attack against the Black King while also seeking ways to catch the e-pawn.

  • Move candidates - Nxe6, Rc8+. Let's start calculating…

Well, if these were your initial step (precisely in the order shown) - excellent - I would say Expert-Master level. Although it is not likely lead you to a solution of this clever study. One can easily get lost in the lines with multiple ideas of White and Black. Example: good for White is 1.Nxe6 Kg8 2. Rc8+ Kf7 3.Ng5+ Ke7 4.Nf3!! exf 5.Rc5 e1Q 6.Re5+ winning; and not so good is 1.Nxe6 h6 2.Rc8 Kh7 3.Rc7 Kh8 =, but wait, 2. Nc5 e1Q 3.Nxe4 - how to assess this position…Oh, my head is spinning...

What is missing? Finding the idea of getting the White King involved and letting Black to get the Queen and then even make some Queen moves.

This is based on general knowledge that except for maybe one "sure thing" position - (Knight + Rook vs. King) there are no other mating setups to use as a spring board in calculation in this particular example. And that last one can't be reached without Black's cooperation!

When I studied this position, 1. Ke5 was one of my original moves-candidates and the first one I studied. White is planning to meet 1…e1Q with 2.Kf6. The problem - 2... Qxf2+ breaking the setup, and there is no way, even at the expense of the Knight (3. Nf3 Qxf3+) to stop it. Oops… at the expense of Knight.... Hmmm. So I don't really need the Knight, as long as I can play Kf6. How about 1.Nf3 - not just trying to give up the Knight, but rather trying to block the f-file.

The rest is an easy calculation: 1…exf 2.Ke5 e1Q (else 3.Re4) 3.Kf6 Qa1+ (3…h6 4.Kg6) 4.d4! (4.c3? h6 5.Kg6 Qb1+) h6 5.Kg6.

This is a study by the Platov brothers (Vasily and Mikhail) published in 1924. Also, it is probably busted because of a possible second solution (a no-no for studies!) as after 1.Nxe6 h6 2.Nc5 and 3.Nxe4 White might have a practical winning chances.

A very difficult study! Probably impossible to solve, unless you search for ideas first.

No comments: