Knight endgames are really Pawn endgames

Hello!

In this game, I gain a reasonable plus in the middlegame, only to squander much of my advantage by excessive liquidation. However, my opponent makes a questionable queen exchange that leads to a knight endgame in which I can easily win a pawn and make the position very uncomfortable for White. Later, my opponent blunders by allowing me to trade off the knights and reach a winning pawn endgame. Please help me learn how to keep the tension in the position.

Thanks,

chessman

Thanks for commenting! When I made the title, I had that quote in mind :) I had never heard that follow-up, though. Intriguing!

That is an interesting comment on move 29. Making somewhat committal pawn advances with limited material on board always risks weakening, but I'm not sure White is able to take advantage. However, I know knight endgames are often combinational wormholes, so it's likely there is someway of causing me to be tied to the defense of the a-pawn, but I also thought I could sacrifice a queenside pawn at the appropriate time to activate my king to attack his queenside pawns or, as you suggest, to support my kingside pawn majority, which would be liable to threats from the knight. I completely agree that these pawn advances need to be made with care.

I'll have to bite the bullet and jump into the wormhole :)

I heard from someone here the quote lies on the fact knights can't lose tempi without changing the position.

Are pawn endgames with knight making it more complicated.

That makes a lot of sense. For the side on the offensive in pawn endgames, a typical idea is to induce zugzwang by taking away the defender's waiting moves if a dominant king position has been reached. Usually, a similar principle holds in endgames with just one piece. Bishops can usually wait without revealing any new weaknesses, but knights immediately give up all the squares they protect with one move, a quality that can be easily abused.

Absolutely. It introduces the possibility of being able to limit king activity by forcing them to defend weaknesses. Knight endgames, like pawn endgames, usually lend themselves to heavy calculation due to many forcing maneuvers, but there are forking possibilities with knights that make it more complex, which is why I'm jumping into the wormhole against my will! :)

Yes.  I love the possibilities of zugzwangs in these endings.

shkrelis_nemesis wrote:

In the 90s I played GM Yermolinsky in a simul, and from his games in Chess Life I could see he played the Torre attack, which I countered with 3...h6 4.Bxf6 exf6.  After 25 moves I had reached a knight and pawn ending which I think should have been able to have been drawn -- I didn't have any serious weaknesses, and our king and knight positions were approximately equal.  I think I just blundered and let him fork a couple of pawns, or my king and a pawn, with his knight, so I lost.  A few minutes later when I was leaving I passed him and he gave me this look like no way you should have lost (or you're a big time patzer, LOL).  Wish I still had the game score, I played that under-appreciated defensive variation pretty well.

It's good that you made it that far! I would love to play in a simul.

@achja,

You are correct; it also weakens the b6 square, which would be another thing to think about.

You played a nice game and I really like your endgame, nicely played. 20...Be4?! is a little shaky.

Yeah, it was a move I made quickly. I don't think it carries that much risk, but there were probably better moves.

I was looking at it 20...Be4!?, it does seem too bad. 20...Be4!? 21.Rxd8 Rxd8 22.b4 attacking your knight but you have 22...f6! 23. Qxc5 Qxc5 24. fxe5 his pawn seem more weaker than yours, so maybe your move is not bad.

Yeah, I also think Qxe5 should end up fine for Black.

Isn't 27. b3 "forced" if White wants to keep the b-pawn? I think White will lose a pawn by force otherwise, though 27.b3 also loses a pawn as in the game.

Yes, that makes sense. Even though it wasn't a knight endgame, Fischer-Taimanov 1971 (game 4) showed how rook/knight pawns can overwhelm the opponent's knight. "There's not another side to a rook pawn!" so they say.

What if white plays 1.a3 in your position?

It can't possibly be that simple, could it? It's certainly true that queenside pawn majorities are often more significant than kingside pawn majorities in late middlegame/endgame struggles, but I don't think simply having a queenside pawn majority should constitute such a substantial advantage. I definitely don't think White can force a passed pawn on the queenside, if I play correctly. Of course, that's one big if. Nevertheless, I have seen games where a queenside pawn majority was allowed to fully blossom and led to a lost position for White, as in this game:

http://www.chessgames.com/perl/chessgame?gid=1094832

So, of course I'm intrigued and will search for other games and try to grasp how these games might shed light on what White could have done in my game.

Intriguing. Your only advantage was the queenside majority? I'd be interested in seeing a game from you that was, as you say, totally won with that being your only advantage. Which parts of your repertoire typically lead to such positions?

HueyWilliams wrote:

Why do people around here constantly ask us advice on games they won?  Show us games you lost!

I used to have a series that I've thought about continuing on my losses. If you're interested, check my archives. My losses here, however, are usually not reflective of my issues in chess understanding; they're usually stupid blunders because I have a short attention span on a computer, because I can have multiple tabs. Thanks for asking.

HueyWilliams wrote:

Right, nobody's losses ever reflect their true level of chess understanding. lol

That's not what I said. What I said is that they usually don't, because I end up making some stupid blunder that I don't need anyone to point out because I have a short attention span while playing chess and browsing Facebook. When I do lose and have no idea of what might have gone wrong or it's interesting for some other reason, I post it.