4-2 vs 3-3 (only pawn) endgames arising from Sicilian/Anti Sicilian

Question

I play Sicilian (mainly Najdorf) against e4. Sometimes (in anti-Sicilian or Sicilian openings) I end up with exchanging everything except 6 pawns.

So it becomes a 4-2 pawn (4 on the kingside) versus 3-3 end game. I am the side with 4-2 and most likely a6-b5 being the two.

I remember losing a few such games that I could have drawn or perhaps won.

Can you help me with the main strategies for black and and the opposing side white? Also in the specific game that I attached below, we reach to that position at move 28. But I appreciate any advice before move 28 as well.

[Event "Rated Blitz game"]
[Site "https://lichess.org/wVcYyus2"]
[Date "2020.09.02"]
[Result "1-0"]
[UTCDate "2020.09.02"]
[UTCTime "18:34:04"]
[WhiteElo "2174"]
[BlackElo "2033"]
[WhiteRatingDiff "+3"]
[BlackRatingDiff "-3"]
[Variant "Standard"]
[TimeControl "300+0"]
[ECO "B22"]
[Opening "Sicilian Defense: Alapin Variation, Barmen Defense"]
[Termination "Normal"]
[FEN ""]
[startply "54"]

1. e4 c5 2. c3 d5 3. exd5 Qxd5 4. d4 Nc6 5. Nf3 Bg4 6. Be2 e6 7. O-O Nf6 8. Be3 cxd4 9. Nxd4 Bxe2 10. Qxe2 Be7 11. Rd1 Nxd4 12. Bxd4 Qc6 13. Nd2 O-O 14. Nf3 Rfd8 15. h3 Rac8 16. Ne5 Qc7 17. a4 Nd7 18. a5 Nxe5 19. Qxe5 Qxe5 20. Bxe5 a6 21. Kf1 Kf8 22. Ke2 f6 23. Bd4 Ke8 24. Bb6 Rxd1 25. Rxd1 Bd8 26. Rxd8+ Rxd8 27. Bxd8 Kxd8 28. Kd3 Kd7 29. b4 Kc6 30. c4 g6 31. Ke4 b6 32. axb6 Kxb6 33. f4 h5 34. g4 hxg4 35. hxg4 Kc6 36. g5 f5+ 37. Ke5 Kd7 38. c5 Kc6 39. Kxe6 Kb5 40. Kf6 Kxb4 41. c6 a5 42. c7 a4 43. c8=Q 1-0

EDIT: Brian’s answer well addressed the case when my king’s position was inferior but I am actually more curious about the strategies in the case when they are the same or mine is superior. In the game below, I have a superior king position, but then I blunder. What should be the strategy of black in that case? Instead of pushing the a-pawn, I guess I should have done f6 to avoid g5. Then should I try pushing e5 to make it a passed pawn? But then wouldn’t it be an isolated, hence a vulnerable pawn? Again any comment before the start ply will be also useful.

To put it another way, considering the respective start plies in those games,in the first game, what to do to hold a draw, and in the second, what to do to win strategically?

EDIT 2: After Evargalo and David’s comments, I am more or less ok with the first game in the sense that there is no much more to say.

But as Evargalo pointed out in his comment, I don’t quite see why I had +0.9 in the very last position of the second game. To compare, I think both sides violated the second principle of Brian (don’t play in the minority side) and I think I had a better king position. My guess is that it could be due to losing the opposition or the g pawn blocking both f and h. But a deeper perspective could be better.

EDIT 3 After Brian’s great answers and my analysis with Stockfish, I seem to get an idea. But even then pawn endgames are very sensitive to mistakes that might be hard to spot. Thanks to all the contributors.

[Date "2020.09.02"]
[Result "1-0"]
[UTCDate "2018"]
[UTCTime ""]
[WhiteElo ""]
[BlackElo ""]
[WhiteRatingDiff ""]
[BlackRatingDiff ""]
[Variant "Standard"]
[TimeControl ""]
[ECO ""]
[Opening ""]
[Termination "Normal"]
[FEN ""]
[startply "70"]


1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. Bc4 e6 7. O-O b5 8. Bd3 Bb7 9. a3 Be7 10. Be3 O-O 11. f3 Nbd7 12. Qe1 d5 13. exd5 Nxd5 14. Nxd5 Bxd5 15. c3 Qc7 16. Rd1 Rad8 17. Qf2 Ne5 18. Be4 Qb7 19. Bxd5 Rxd5 20. Ne2 Rfd8 21. Rxd5 Rxd5 22. Nf4 Nd3 23. Nxd3 Rxd3 24. Bd4 Qd5 25. Qg3 g6 26. Re1 Bd6 27. Qh4 Bc5 28. Bxc5 Qxc5+ 29. Qf2 Qxf2+ 30. Kxf2 Kf8 31. Re3 Rd2+ 32. Re2 Rd5 33. f4 Ke7 34. Ke3 Kd6 35. Rd2 Rxd2 36. Kxd2 a5 37. g4 a4 38. g5 Kd5 1-0

Answer 1

King and pawn endgames are much easier than ones where each side also has a rook. The key principles are fairly simple. You want to try and queen a pawn. With pawns on both sides of the board and an imbalance as here there are two key points:

1. Getting your king to the middle of the board is very important.
2. The only pawns you are likely to queen are ones on the side where you have the majority. If the kings keep away then the further up your pawns get the easier it is to get a passed pawn and queen. So, don't move pawns on your minority side unless you absolutely have to! Doing so makes your opponent's task easier. Move pawns on the side where you have the majority.

In the position you show after move 27 you are already behind in the race to centralize your king. Whether that means you were already losing I doubt but it does make things difficult and it was your choice. Your king has taken 3 moves to go from g8 to d8. In the same number of moves you could have gone from g8 to d5 and that would have been a very different story.

Your transition into the endgame was very poor. Perhaps 20...a6 was already a mistake. You are breaking rul2 by advancing a pawn on your weak side. Better would have been f6 which as well as being a free move, because it kicks the bishop, frees f7, which is a much better square for your king than f8, and moves a pawn on your majority side.

After 24. Bb6 you are in too much of a rush to get all the pieces off with RxR and Bd8. You exchanged pieces in a way which was good for your opponent and bad for you. Bd8 was particularly bad. Your rook on c8 paralyzes white's queenside pawns.

Suppose instead it had gone something like this after move 20:

[fen "2rr2k1/pp2bppp/4p3/P3q3/3B4/2P4P/1P3PP1/R2R2K1 w - - 0 1"]

1. Bxe5 f6 2. Bd4 a6 3. Bb6 Rd6 4. Rxd6 Bxd6 5. Rd1 Rc6 6. Kf1 (6. b4 Bc7) Kf7 7. Ke2 e5

After 5...Rc6 you have a simple plan of bringing your king quickly to the center with Kf7, e5, Ke6. If white tries immediately to kick your rook away from the defence of the bishop by rushing forward the queenside pawns starting with b4 then Bc7 is a good response.

Answer 2

Without looking at those gamescores, I'm guessing that you're not taking charge in the center when it's there for the taking.

You know Sicilian players are supposed to win more of the endings, right? White wins the shorter, sharper games that get into magazines, and Black wins the longer ones after weathering the storm.

After weathering that storm, Black is left with a preponderance in the center (which you demonstrate in these endings), and then it's time to go back at them hard. If you don't have the tactical chops for that, cultivate them.

Answer 3

and in the second, what to do to win strategically?

OK, I've just seen your second edit with another position. Strictly speaking it would be better as a new question, but I'll add another answer looking specifically at this.

The first two things that strike me looking at the final position are that black has a much better king position and white has 2 "reserve moves" to black's none. Whether or not white's poor king position is immediately fatal is a question that requires calculation.

First let me explain "reserve moves" and why they are important in the endgame.

Remove the kings and look again at the position and ask yourself how many moves each side can make before disaster strikes. Black can't move. Any move will result in the loss of a pawn. White has two moves before having to lose a pawn, h2-h3 and h3-h4. What this means is that in any one-off mutual zugzwang position white is going to win thanks to one extra reserve move. Furthermore, if 2 zugzwangs are required for victory then again white is going to win.

In the actual position although black has an optical advantage the queenside foolishness means the game is drawn. As part of the optical illusion of black superiority it looks like white's next move, Kd3, is forced, but such is the tragedy of black's queenside pawns (and the lack of reserve moves) that I think white can even play Kc2. If black responds with the obvious Kc4 then black can even play b3+! and because of the extra reserve moves this just accelerates the development of a passed pawn.

The problem for each side is that pushing their minority side pawns has accelerated their opponent's development of a passed pawn. Let's have a look at a couple of lines. In each case I think that if one side had kept their minority side pawns at home they would likely have won.

[fen "8/5p1p/4p1p1/1p1k2P1/p4P2/P1P5/1P1K3P/8 w - - 0 1"]

1. Kd3 (1. Kc2 e5 (1...Kc4 2. b3+ axb3+ 3. Kb2 Kd5 4. Kxb3 Kc5 5. h3 Kd5 6. Kb4 Ke4 7. c4 bxc4 8. Kxc4 Kxf4 9. a4 e5 10. a5 e4 11. a6 e3 12. Kd3 Kf3 13. a7 Kf2 14. a8=Q) 2. fxe5 Kxe5 3. b4 f5 4. gxf6 Kxf6 5. Kd3 g5 6. c4 bxc4+ 7. Kxc4 Ke6 8. Kc5 h5 9. Kc6 h4 10. b5 g4 11. b6 g3 12. hxg3 hxg3 13. b7 g2 14. b8=Q g1=Q) e5 2. fxe5 Kxe5 3. c4 bxc4+ 4. Kxc4 f5 5. gxf6 Kxf6 6. Kb4 g5 7. Kxa4 h5 8. b4 h4 9. b5 Ke7 10. Ka5 g4 11. b6 Kd7 12. Ka6 g3 13. hxg3 hxg3 14. b7 g2 15. b8=Q g1=Q 16. a4 Qc5

And finally,

Instead of pushing the a-pawn, I guess I should have done f6 to avoid g5. Then should I try pushing e5 to make it a passed pawn? But then wouldn’t it be an isolated, hence a vulnerable pawn?

I think the idea of f6 is not to try and stop white pushing a pawn to g5, because that would be bad for white, but for black to try and try and push g5 to get rid of the white f4 pawn and leave the e pawn as a passer.

OK let's look at the position after 35 moves. White is going to recapture the rook. Then what are the plans for each side and how will the position look after they are implemented?

Both sides are going to try, and if they try hard enough they are going to succeed in creating a passed pawn on the side where they have a majority. So, white will have a queenside passed pawn and black will have a kingside passed pawn. What happens then?

Well, one side or the other is going to push their passed pawn and force the other side to undefend their passed pawn in order to take it. So white will end up with the king on the kingside and white's king will be on the queenside and the plans for both sides will be to take as many pawns on that side of the board as required to be able to queen a pawn, basically a race.

In this race the advantage will be with the side that has to move their king the fewest moves to achieve that together with having a pawn closer to queening. It looks like the starting gun for this race will be fired after black has taken a passed pawn on c4 and white has taken a passed pawn on e4 and black will be aiming to go on and take the white pawn on a3 and white is likely going to be aiming for a black pawn on h6.

It looks to me as if black is winning that race. If so then that is a scenario that white must try and avoid. If white does try avoid that I think that black can force that situation by first creating a passed pawn and then advancing the queenside pawns to paradoxically create a passed pawn for white.

Let's look at a few lines to see how it could pan out.

[fen "8/5p1p/p2kp1p1/1p6/5P2/P1P1K3/1P1r2PP/8 w - - 0 1"]

1. Kxd2 Kd5 2. Kd3 f6 3. b3 (3. h4 a5 4. b3 h5 5. c4+ bxc4+ 6. bxc4+ Kc5 7. Kc3 e5 8. fxe5 fxe5 9. Kd3 e4+ 10. Kxe4 Kxc4) 3... g5 4. fxg5 fxg5 5. c4+ (5. Kd2 a5 6. Kc2 e5 7. Kd3 e4+ 8. Kd2 a4 9. bxa4 bxa4 10. g3 Kc4 11. h4 gxh4 12. gxh4 Kb3 13. c4 Kxc4 14. Ke3 Kb3 15. Kxe4 Kxa3) 5... bxc4+ 6. bxc4+ Kc5 7. h3 a5 8. g3 a4 9. Kc3 e5 10. Kd3 e4+ 11. Kxe4 Kxc4 12. Kf5 h6 13. Kg6 Kb3 14. Kxh6 Kxa3 15. Kxg5 Kb3)

Caveat: I'm sure a good engine with tablebases will find lots of holes in this analysis but it gives a general idea of what black is aiming for in these kinds of positions. And I guess that these kind of endgames are what Sicilian players expect and are aiming for.