Questions about opposition in the Van Nyevelt position.

I am just beginning to learn king and pawn endings so please bare with me for what might seem like stupid questions. In the position below, White to move can win, but black to move can defend. My questions are, if it's white to move, there are two options to take the opposition, the direct distant opposition of a4, or the virtual opposition of a2. Does it matter which opposition is taken? And if he goes to A4, will he then have to take the virtual opposition at some point anyway? Say, 1. A4 e5 2 a3? Assuming best play by black with white to move first, can someone explain what the best play from white would be move by move to reach the key squares and why? Would the white king have to give up opposition at some point to take the key squares?

Would the white king have to give up opposition at some point to take the key squares?

Yes. When you're defending using the opposition, you want to keep it. When you're attacking using the opposition, you usually have to give it up to attain your objectives (a bypass manouver).

1. Ka2? throws away the win after 1... Kd4!

Now White has a choice: keep the opposition, but stay on the second rank or bypass and give up the opposition.

Staying on the second rank gets nowhere: 2. Kb2 Ke4 3. Kc2 Kd4 4. Kd2 Ke4 5. Ke2 Kd4 6. Kf2 Ke4.

Bypassing doesn't help either: 5. Kc3 Ke5 6. Kc4 (6.Kd3 Kd5) Ke4 7. Kc5 Ke5 8. Kc6 Ke6 etc.

White needs to play 1. Ka4! and bypass further up the board.

I'm usually the one saying it's very good to study endgames, but in this case I'd say the problem is too advanced for someone just beginning to learn king and pawn ending. King+pawn vs King for example is much more imortant and then some simple King+Pawn vs King+pawn. The problem above should come much later when studying endgames and it wouldn't surprise me if players around my strength would misevaluate or maybe even misplay this position.

Thanks everyone. @ CaveatCanis. I've heard that diagonal opposition doesn't work for the defender, this is borne out in one of above lines  where the white king is marching right along the fourth rank and the black king is continuosly marching right along the sixth rank as well taking the diagonal opposition each time, but it does him no good. Though, in this same starting position were it black to move, I believe it could be appropriate for black to take the diagonal opposition by moving to d4 if the white king moved to b2? Am I wrong about this, and if not, does this mean that diagonal opposition actually does work for the defender depending on what key squares there are and the kings' relationship to them? Also, are you absolutely sure 1 A2 taking the virtual opposition loses for white?

@ QuantumKnight, It seems like the tablebases would just be giving the answers without me actually having to learn them. I'm just trying to learn the concepts behind the endgame positions and in particular here trying to expand my knowledge of opposition, I don't know how much tablebases would actually help me in this regard.

@Martin0 I've already mastered the basic king+pawn vs King endgames and many of the king+pawn vs king+pawn positions. I don't think this above position is all that complicated and I think gaining understanding of it is the logical next step for me. This position is actually given near the beginning of Karsten Mueller's Endgames volume 1 chessbase DVDs where he begins with how to mate with a queen and works his way up to more advanced endgame concepts, in the DVD this position is only 6 steps removed from the basic mating patterns. As well as being in one of the beginning chapters of "Secrets of Pawn Endings" by Mueller and Lamprecht where they gave the problem only a "medium" difficulty, though that was with black to move first in this position and draw. White to move first was not given as a problem in this case because it probably should be obvious, I think Caveat's explanation of using the simple "bypass" technique once you have the opposition to advance to key squares cleared it up for me though.

@AdorableMogwai

There's often "corridor" that's 3 files (or 3 ranks) wide that the attacker wants to use. The defender draws by preventing the attacker from advancing. This requires the use of lateral (or distant) opposition in the corridor. The attacker wins by seizing the opposition on the central file (or rank). When the defender moves to one side, the attacker advances on the other side (the bypass). The defender can then get the diagonal opposition, but that won't prevent the attacker's further advance.

Here the corridor is the 4th, 5th and 6th ranks. (I think it actually stretches as far the 8th rank, but 3 ranks wide is enough to win.)

Outside the corridor, the opposition doesn't matter, unless it affects the chances of getting opposition within the corridor itself. For example, in the first position, the diagonal opposition determines the outcome. In the second position, the opposition doesn't matter; it's drawn regardless of who has the move:

CaveatCanis, after moving to a2, couldn't white maintain prospects for moving off the second rank by taking the opposition this way? If so, it would be a loss of time, but they could still win.

No, because 3. Ka2 just repeats the same position we had after 1. Ka2? Black draws by repeating the same defence: 3... Kd4.

Thanks for the answer, I'm sure after studying all the info here over the next day or so I'll start to get it.

It seems that way.

Before you make snarky comments you might want to work on your own king and pawn knowledge first, as I see from your recent thread here http://www.chess.com/blog/GSHAPIROY/king-and-pawn-endgame-support-squares that you don't even know the very basics.

In my hasty analysis I did fail to notice it was a reptition, a mistake I made in part from lack of sleep and also because before now I always assumed that taking any form of opposition (in this case the virtual opposition) was automatically good. Now I've learned that that's not always the case. So you keep making your comments that don't contribute anything and I'll keep learning king and pawn endings.

I disagree. For practical purposes this may not be as useful as, say ideas on R+P endgames (Lucena, Philidor) or the general ideas behind all endgames (force weaknesses, invade with king, outside passed pawns or what-have-you.) But there are other important benefits.

This sort of position is solvable by intense calculation, which is useful no matter what. And it also reinforces the more basic concepts of opposition and reduction to other endgames (e.g. the reason 1. h5 draws although white can win the pawn, or how black's 1...Kf3 fails) which you have to see while calculating. It also reminds that endgames are tricky and precise calculation is always a must.

Pure K+P endgames are always good to learn, since they exemplify raw calculation, always good in endgames.

I agree it's good to solve positions like this. My main point was that you should start with the most basic things first which in this case include understanding opposition and be able to tell when K+pawn vs K is won or drawn and be able to play those positions correctly. I regret having said this should come much later though. It's not wrong to study positions like this before you even start studying rook endgames.

I selected this position to study because I knew I didn't fully grasp opposition and that if I came to understand all the details of this position (with both white to move and black to move versions) I would begin to fully grasp it.

Here's what I've learned, if I am wrong in any of this, please correct me.

1. Opposition is not a goal in itself, but only a tool to reach a goal of obtaining control of key squares. Before one carelessly takes any form of opposition, they must first identify the key squares and what their ultimate goal is.

2.  In order to use opposition to gain control of key squares, one must first take direct opposition on what's called a "main file or rank". This is something I had overlooked in my Secrets of Pawn Endings book as there was a very short paragraph at the beginning about it. In the above position, the key squares are f4, e4, d4, d5, d6, d7, e7, f7, and g7. The main ranks and files would be determined by the central key squares in the line of key sqaures. Here the central key squares in the line are d5 and d6, thus one would have to obtain direct opposition on either the 5th or 6th rank before moving right towards the key squares with the "bypass" manuever. Going around towards the bottom and obtaining direct opposition on the e file doesn't work because the black pawn on g4 hinders the bypass manuever by preventing white from moving to the f3 square.

Believe it or not, this is a relatively easy position to understand compared to the other hundreds of positions given in Secrets of Pawn Endings, the ones given in the later chapters would probably take the average chess player hours to solve. After looking at the K+P problems in that book the tactical puzzles I do on Chesstempo seem simple by comparison.

Believe it or not, this is a relatively easy position to understand compared to the other hundreds of positions given in Secrets of Pawn Endings.

That's an excellent book by the way. When you get to the chapter on Corresponding Squares, it helps to know that normal opposition is just one (very common) system of corresponding squares.

Roughly speaking, it's the one you get when pawns don't get in the way of the King's manouvers. When you're close to pawns (particularly in composed positions) the normal rules of opposition often don't apply and you have to look for other patterns instead.

I was looking through that chapter earlier this morning where the positions given with a lot of pawns and a number on almost every unoccupied square of the board seemed incredibly complex. I think it will at least take me six months to get to that chapter as I'm trying to go through the book very systematically and fully understand every last detail of the example positions and exercises before I move on. I'm not even on the fourth chapter yet, where the above van Nyevelt position is given, and I only saw it because I was watching Karsten Mueller's Endgames Vol 1. Chessbase DVD that's just about endgame basics and not strictly K+P endings. This position is one that Karsten Mueller seems to think is important for people to learn as he selected it for that DVD.

It seems crazy to think that K+P endings are such a small part of overall chess yet someone can spend years studying them. I wonder if I should start studying rook endgames at the same time (which the Secrets of Pawn Endings book says are more common) or if I should go through the whole book and only then move on to rook endgames. Learning the K+P endings requires enough time and effort as it is, and I think if I added rook endings on top of it it might be overwhelming. I do know the very basics of rook endings like the Lucena position already.

One thing I can add is that I believe it's good to study more endgames than you actually need in most otb games. When you start to master more advanced stuff the easier things will seem much easier and you probably won't struggle as much with the "easy" endings when you need them otb.

Some people disregard endgame study while others think the most basic is enough (for those rated <2000). I think the basic is enough, but that it's generally good to dig deeper to get a deeper understanding. Aiming for "minimum required" is generally not a good idea.