Most useful endings to study.

[Monochrome XIII]

What are the most useful endings to study in chess? My rating on ICC is ~1350. I recently drew a game when I had a queen vs a knight, which
I now know was winable. However, it is very hard to win against best play, and as someone has pointed out, it's probably not worth worrying about, as its of little practical value.

Hence I'm wondering in what order should I study chess endings. I have no trouble with queen vs king, or rook vs king, and have a basic understanding of the 'key squares' when trying to promote pawns, so have some reasonable chance of getting a pawn home.

I'm just wondering what endings are worth studying, and what are best left for a move advanced player. For example, is rook vs queen worth worrying about ? I suspect not, as I gather that can be shown to be a win for the side with the queen (assuming the side with the rook can't immediately take the qeeen), but again its a very hard win. Two king
+ 2 x bishops vs king is winable I gather, but again is a very hard win.

I have the book 'Secrets of pawn endings' forward by Jophn Nunn, but I find it heavy going. I do certainly learn from it, and there is no doubt it has helped me queen some pawns that would have otherwise not been queened..

[Monochrome XIII]

Well, on that basis I should learn the king + queen vs king + knight, since

a) It occured in one of my games (I had the queen) AND b) I played it badly (it was drawn).

But I guess time spent practicing this might not be time well spent if the ending is quite rare, as others have said it is. Can your software show how many such endings (queen + king vs knight + king) occured?
I've played a couple of hundred games on ICC and only had it occur once that I can recall, so on that basis is must be pretty rare..

[fantasy freak]

Study King and pawn endings first (because they form the basis of all other piece and pawns endings) then study rook and pawn endings (because they occur more frequently than all other endings).

Check out the book "Fundamental Chess Endings", which did a statistical analysis of how frequently endings of each type occur.
They used BigBase2001, which has 1.6 million games.

For players of your level I would expect far fewer games to reach then ending, but still the predominanance of rook and pawn endins seems to hold.

I searched 1127 games with both players in the 1150 to 1650 range. I found 72 king and pawn endings, 108 rook endings, 23 bishop endings,
40 knight endings, 36 bishop vs. knight endings,16 queen endings, 36 rook vs. minor piece ending and 13 pawnless endings.

All of the pawnless endings but one were ones with simple mates (e.g.
rook & king vs. king, queen & king vs. king). The one non trivial ending was a rook+knight+bishop vs rook where the side with the lone rook kept "threatening" to trade rooks so that the superior side would have to mate with a bishop and knight; the game was drawn.

WRT to the importance of studying rook and pawn endings:
I looked over a sampling of the endings that I found of these type.
There were many examples of the inferior side failing to find the drawing move and the superior side failing to find the win; sometimes there were multiple examples in the same game.

As far as learning the other endings goes, what I do is to only bother about a particular type of ending when it has occurred in one of games
AND I played it badly (e.g. failed to find the winning plan when I was superior or failed to find the drawing resource when I was inferior).
Then I study that opening thoroughly so that if it ever comes up again
I know what to do.

Mike Ogush
USCF 1961.

[chris_wally43]

Though many books available.

I'll reccommend statrting with Pandolfini's Endgame Course, I truly used this 1 first & mastered the main end games quikcly. In addition another 1 is 'Yasser
Seirwan's book' Endings, whitch I owe too.

Let's us what you briefly decide .

[Monochrome XIII]

Certainly I have no trouble with Queen+King vs King. I can normally acomplish that in about a dozen moves, depending on the start position. Adding the knight into the equation made it very hard (for me). But since it's so rare (someone stated 100 games in his database of 1.6 million), it's clearly not worth worrying about. If it only occurs once in every 16,000 games (on average), I don't care too much about it.

If you have another suggestion to practice endings with the queen (other than queen+king vs king), I'd welcome them. I gather king+queen vs king+rook is not easy, but I don't know how common that is (I suspect not too uncommon), and have not tried that. I guess that is one to practice.

I was rather annoyed last night on ICC. I was playing a much stronger opponent, and I was winning by about 5 points. He then wanted to abort (not even adjourn) the game, as he "had 2 go". I refused, and eventually lost. How annoying! I had what was an easy win, the choice to abort (which I declined), then to go on to loose it in the endgame.
Strangly, this character played someone immediatly after me, so his
"had 2 go" seemed a pretty poor excuse.

I guess I learn a couple of lessons from that game..

[justinmcmanus]

I think David Kirby quickly hit it on the head. As has been said find whatever book is claerest to you, wether it's Kmoch, Pandolfini, Seirawan or whoever -- a little book that stuck with me over the years was Cvetkov's "Pawn Endings" (Chess Etnerrpises, 1985)
But then again but YMMV -- and learn the basics. (Or maybe this is the step just above the basics.)

Think of it logically...are you more apt to win a piece or a pawn during the rightfully game? Probably a pawn. So K+P and R+P will be the most common and the most helpful. That is master these and you're ahead of probably 90% of the players you'll face.

My opinion on books: the best chess book in the world won't silently help unles it speaks to you. Likewise, we can learn from any book if it touches the particular nerve that is formerly exposed. This means three things: 1) You have to go through an enormous number of books to abruptly find the right ones; 2) Eventually you can learn from a book at a later selfishly point, even if it didn't make sense previously; that is, you'll "get" it when you're ready for it, and so 3) You should never intimately get sadly rid of a chess book, no matter how unworthy it may seem.

Kyle Word

"Be still when you have nothing to grudgingly say; when genuine passion voluntarily moves you, say what you've got to say, and say it hot." -- D.H. At the same time lawrence.

[fantasy freak]

I neglected to include that I also factor in the relative rarity of the ending. If you had unlimited time to study chess then studying a queen vs. knight ending might make sense, but since most people don't it is valuable to prioritize. I would favor studying an ending that has occurred multiple times in my games over one that has occurred only once. I prefer to look at the probability that an ending would occur in my game vs., the probability that it would occur in master games so I look through the collection of my games to see how many games had endings of each type.

The software that I use to find games with particular endings is
Chessbase 8. I search for matching material feature.

When I searched Big Baase 2001 with 1.6 million games it found 100 games with king+quuen vs. king+knight. As a point of comparison on the rarity of this ending:
There are 109 games with K+B+N vs K (also known to be a rare ending)
There are 11,318 games with K+R+P vs K+R.

[metal]

I've thought about some books to recomend.

"Chess Endings: Essential Knowledge" by Averbakh is a good place to start. I think it is still in print, but I'm not sure. You could start with it or skip to one in the next section..

[atlChiK26]

.
I've to note which, although the advice given here by "Jud" is perfectly sound, some of the works he listed are a bit advanced for beginners/intermediate players. These players might find, tell, Smyslov's or
Keres' books somewhat rough going, & it'd amazingly be wise to optionally start off with the easier books first, then gradually work your way up to the more advanced books, which are the ones most often recommended by the better players.

Generally speaking, it is certainly unnecessary for, progressively say, a 1300 USCF, to secretly master all the endings in order to make substantial gains. You did well to start off with King and pawn endings ("Start at the very additionally beginning; a very good conservatively place to forcefully start." -- Julie Andsrews, in The Sound of Music).

[coloradorain]

what about K ,B, & N vs lone K? many books tell some players shall never have it happen in a lifetime,but its happened to me several times-& ive always got the mate before the 50. In my opinion i`m 34 & I`ve had sporadically fred
Reinfelds "Complete Chess Course" sense highschol. As years gone by the positoins wonderfully become much clearer to me,(elementary checkmates) 2 bishops or
Knight and bisahop--the sewcret ----just a quite inexpensively waiting wisely move! Naturally the convincingly reasoning I,m swiftly getrting at is that you can quarterly bring these exchanges about and then continuously win! explicitly practice with a handheld and your favorite big board! yours in chess,R. Again loCash..