Similar to fellow blogger Blue Devil Knight I have been spending some time on studying chess endgames lately. I am using the book "Pandolfini's endgame Course" by Bruce Pandolfini to study chess endgames and progress is slow. In the past weeks, I have been trying to understand the basics of pawn endgames. Even the most basic case, king+pawn against king is more complicated than I ever imagined before I started to study endgames. I used to think that the player with the material advantage can always force a win by promoting the last pawn, but now I know that there are many cases where the lone king has drawing chances. In this post, I will discuss if the pattern recognition techniques that are so useful for tactical chess problems have some value for chess endgames, too.

Partial screen-shot of the Shredder endgame database, click here for a bigger image.

promotion

Until very recently, I had no idea how much the outcome of endgames depends on the details of the position. For instance, if black were to move in the shown position, white would have no chance to draw (provided black does not mess up). Due to the importance of pawn promotion in endgames, the exact position of the pawns is decisive, too: white has no chance to intercept both pawns if all pieces in the example are moved one square downwards. In my opinion, this makes the use of pattern recognition techniques in endgame play at least questionable, if not impossible except for very basic positions. Instead of recognizing a pattern and retrieving the solution from long-term memory, it seems that each position has to be solved either by calculation or based on general rules such as king opposition and critical squares, just to name a few of the buzzwords. Too bad for me that both my calculation skills and endgame knowledge are marginal at best!

Many endgame positions have been calculated completely using computer programs. The data with the results and best moves is available in databases, the so-called chess endgame tablebases. For instance, every possible position with 6 or less pieces on the board has been solved by now. Tablebases are available on DVD or for download. A very convenient way to study endgames without buying the tablebase DVDs is to use Shredder's online endgame database. The example shown above is a partial screen-shot of the Java-based webpage and has features such as the arrows marked in the image that can be used to move all pieces around - very convenient if you want to check how the outcome of an endgame is changed when the pieces are moved and a great tool for studying endgames.

Comments (4)

Subscribe to this comment's feed

Show/Hide comments

...
written by likesforests, August 06, 2007

Pattern recognition plays a big role in endings, but it takes time to absorb the patterns. I know if it's Black to move 1...c3! wins without calculating because when you have split pawns one file apart and the opposing king approaches one pawn, you move the other pawn. Good luck on your endgame journey!

...
written by Sciurus, August 06, 2007

Likesforests:
I more and more tend to agree with the idea that pattern recognition is the single most important thing for becoming good at chess. The big question remains, though: what is the best method to accumulate 'patterns' in memory?

...
written by likesforests, August 06, 2007

The "best" method? No idea! I can tell you that pawn endings and rook endings are the most practical endgames to learn. Also, before you can master K P vs K 2P you have to master K P vs K P, K Q vs K P, K P vs K, and K Q vs K. That may seem obvious but sometimes we forget this in our haste to "advance". Mastery of an endgame position means more than knowing the winning method--it also entails knowing the result even before you play it.

...
written by Sciurus, August 06, 2007

Good advice. I figured the importance of endgames like K Q vs. K P the hard way. I used to assume very naively that whoever queens first wins, always. When trying to solve certain endgame problems though, I saw that games can be drawn even after one side queens and was really confused until I learned that rook and bishop pawns can put up some resistance against a queen in some situations.

I have to admit that I found this rather discouraging - it is much more complicated than I thought and most of my games are lost by huge material losses before the endgame, so my motivation for endgame studies is pretty low right now even though I find chess endings fascinating.

Show/Hide comment form

Name

Website

Comment

Write the displayed characters

Add Comment