Key squares (king + pawn vs. king endgames)

Hey chess folks,

as a chess beginner, the endgames' patterns and systems are yet nothing but mysterious riddles to me. Therefore, I dig in into the concepts of Opposition and Key squares. However, my understanding remains at best fragmentary. The Chess Mentor system, Rensch's videos as well as various blogs and wikis deal with these topics, however when it comes to key squares, the instructions become rather metaphoric. One has to "control", "go for", "reach" key squares. But: What does that precisely mean? Does my king have to physically be on one of the key squares to reach it or do I already control and reached the key square when I am one square next to it and hence prevent the opponent's king to step physically on that key square?

Additionally to the metaphoric use of language, it seems to me that the instructions hint towards key squares (plural) to be controlled and yet again point out that the occupation/control/reaching of a key square (singular) is enough. This leaves me with the question: Is it about all key squares which matter? Or is one all it needs?

So, in short: How do I have to approach the issue to distill clear, precise rules out of all this?

To illustrate one of the numerous endgames which ought to be solved by using some understanding about key squares, see attached one position found on http://amateur-chess.blogspot.in/2013/02/end-game-study_28.html:

 While I see after some trial-and-error that I have to reach b6 to win, I still don't get the big picture.

Kind regards

The first thing about K+P endings is the question: can I get my King in front of my pawn, if the opponent´s King is also in the vicinity? (If he´s far away it doesn´t matter). The answer here is "yes", so the next question is: If my pawn is not yet on the 5th rank, can I get my King two squares in front of my pawn, without endangering it? If the answer is again "yes", then you´ve won. If you only manage to get the King onto the square in front of the pawn before the opposing King blocks you with the opposition , it´s a draw. (If the pawn is already on the 5th rank and you can move your King in front of it and take the opposition by doing so, you´ve won).

The key square here is obviously c8; if you can control (ie protect) that with your King, your pawn can promote. So it´s principally not about trying to race your pawn forward; it´s about getting your King to b7 or d7.

K+P vs. K is not hard to learn, there are good videos, or better still, good endgame books (Silman for instance). Here are the basic principles for this position:

Step by step instruction for promoting the pawn, except the a or h pawn, from any square on the board.

1. Get your K in front of your pawn.

2. Get the opposition on the enemy K so that it is his turn to move.

3. If the enemy K backs up straight one square. Re-establish the opposition by moving your K forward one square. If the enemy K moves laterally one square left or right, move your K diagonally forward to control the square that the enemy K has just vacated. When the opposition has been established and it is your turn to move. Move the pawn forward one square. This will establish the position of the Ks in opposition with his turn to move.

4. After repeating this pattern about 3 to 4 times you K will have control of the key square on the 7th rank on either side of the pawn's promotion file. At that point the enemy K cannot stop the pawn from promoting.

The moves in the forced win line in your example position above with white to move:

1.Kf4 K(any legal move)

2.Kg3 K(any legal move except Kf1) if 2...Kf1 then Kf3 re-establishing the opposition with the enemy K to move. After 3.Kf3 if 3...Kg1 then4.Ke2; if 3...Ke1 then 4.Kg2

If in the example position you posted it is Black to move then it is a draw because after 1...Ke3 the White can no longer get in front of the White pawn.

I hope this helps. If you would like to know more or you have questions please let me know.

You described it better than I did Yaroslavl, but you seem to have got the board mirrored in your moves, they´re nowhere near f4 / g3!

@wasted_youth

Thanks, you are right.

Hello Wasted_youth and Yaroslavl,

thank you both for your valuable input! Those principles you enlisted seem good and promising; the shown position can be solved perfectly well with your list. And it turns out that while that position is taken from a tutorial for key squares (being here: b6, c6, d6), that whole concept seems at least here to be redundant (and can be solved by opposition and outflanking) - or do I go wrong somewhere? For example: Principle One you both mentioned (K in front of P) does not refer to the key squares, does it? In the given position, K on c5 does not yet occupy any key square and still K is in front of P.

Nevertheless, key squares as such seem to be a decisive concept in K + P endings and numerous resources attempt to explain and emphasize its capability to solve positions at once.  In a corresponding thread by mauerblume (http://www.chess.com/forum/view/endgames/knowing-the-opposition-rules-can-help) a similar topic is discussed. Amongst others, a position like follows is reached:

It is black's turn to move. But: When I apply the concept of key squares here, isn't this a winning position for white anyways? He got his K onto one of the key squares (namely c6).

Absorbing all what you said, I tried to solve a Chess Mentor lesson (King and Pawn versus King: The Opposition: http://www.chess.com/chessmentor/view_lesson?id=335) - and miserably failed. The position is:

The comments are the original Chess Mentor ones. My question: Is it wrong to approach the original position from a key squares-perspective (f7/f8, g7/g8/,h7/8)? Do I have any way to spot from the beginning on where to get and how to do it? Or do I truly have to calculate all this through and see that I have to play e.g. 5.g7? It just seems impossible to me to calculate that all and conclude from the original position onwards that 5. g7.

Thank you for your patience, all this has to look rather imprecise and goofy to you.

Best wishes

For the Chess Mentor problem, it's not too hard to evaluate. The key squares approach is one I don't really buy though, and for this a simple knowledge of K+P v K suffices. A thought process could go:

Firstly 1. Ke5 is forced (1. g7 Kf7) so 1. Ke5 Kf8 2. Kf6 opposition on the sixth rank, White wins. Then if 1. Ke5 Ke7 2. Kf5 (to stick to the pawn) and no matter what black does, opposition on the sixth and wins. Therefore White wins.

The key idea for this case is knowledge of K+P v K; you just have to know that K with opposition on the sixth rank wins no matter what (without the pawn in between of course :-P), and all caluclation past that point is superfluous.

Another simple puzzle for your amusement. Black to play and draw.

I don't think in this particular example that key squares are of huge importance as with every move there is only one move that doesn't lead to a draw so you only have to see a couple of moves ahead. In a position like this the opposition is a much more important concept.

I´d go absolutely with what Remellion and Checkmatealot say - the concept of key squares in simple K+P endings leads to confusion for an inexperienced player, since he's going to have to decide in advance what the key squares are and then act accordingly. There are however so many pitfalls for the attacker, ie draw by stalemate or repetition, that it´s much better 1. to learn the concepts of near and distant opposition, and 2. to study the basic techniques of K+P vs. K using these concepts, WITHOUT the concept of key squares.

I really would recommend getting a good book on endgames and working through it with a real chess board in front of you. I found Silman´s Complete Endgame Course great value for money - very well structured (he simply starts with the easiest mating / winning techniques, and progresses to the harder and much harder ones), very thorough, and very easy to read and understand.

The thing about endgames is that there really are cast-iron techniques for winning (and drawing when down in material); that´s not the case in openings and the middle game. And the techniques are really not hard to learn; you just have to find a good instructor who explains things in the simplest way possible. In this case, that would be forgetting about the key squares and concentrating on the opposition.

I disagree. Occupying the key squares is a win regardless, and they only change when the pawn is advanced to the 5th.

On the other hand taking the opposition does not always lead to a win and may in fact be a mistake.

Key squares are the most accurate and simplest way to win simple pawn endgames.

And I agree here. I personally don't like using key squares for the simpler pawn endings (most endings with 3 pawns or fewer, barring those named ones) as it complicates and confuses my thinking. But yes, it is a far more general/fundamental concept; the opposition and such are merely tools to get to the key squares.

Plus another lovely classic study, which illustrates the idea of key squares (w.r.t. a rook pawn.)

Hi,

now the pieces fall slowly together. Thank you all for your well-put explanations!

Going with Scottrf and roig_11, I still would like to analyze positions from the point-of-view of key squares and seeing thus opposition as a means to an end. After your explications I now understand why those key squares are two ranks in front of the pawn itself; it is about that magic buffer move to get opposition in case the opponent holds it.

To spice that key square concept up, i'll try, as suggested, to burn that squeeze-technique into my brain. If I understood it correctly, Remellion's formula "opposition on the sixth wins" does in fact work out so well due to that squeeze.

And: That book from Silman you mentioned, wasted_youth, seems to be outstanding - checked the preview on their webpage. Also other sources do recommend that book and the writing style seems light and easy. Thanks for the suggestion!

Nice. I'll let that all now sink in and revisit that respective Chess Mentor course.

Thank you all & have a Happy New Year.

Here is a really nice example, a tactic over 2600 with only 1 pawn each.

http://www.chess.com/tactics/?id=64691

Taking opposition on the second move loses. Changing the position of the key squares draws.

I found 'Dvoretsky's Endgame Manual' to be extremely good for getting my head around this. I don't think there are many better chess books on the market.

There is a specific section on key squares or squares of correspondence with some famous studies well explained. I recommend it!

Hi

thank you Simon for your book recommendation! For the moment though I think I'll stick to Silman's Endgame Course - that easy-going style of writing seems to suit me best. I'll put Dvoretsky's Endgame Manual on the back burner for now and certainly come back to it at a later stage.

Best regards