Sicilian Endgames: Three Pawns for the Piece

  • GM BryanSmith
  • | May 13, 2014

In various lines of the Sicilian, Black plays ...a6 and ...b5, which sometimes gives White the chance to sacrifice a minor piece on b5. The knight that ends up on b5 subsequently captures on d6, the queens are traded, and a typical endgame arises. Through this endgame, we can learn the themes of not just one specific alley of the Sicilian Defense, but also begin to attain a greater understanding of endings where pieces battle pawns in general.

To illustrate how this sacrifice occurs, let's see the first known example of this archetypal Sicilian sacrifice. This was the game Rauzer-Makogonov, Leningrad 1934:

The famous slayer of the Sicilian Defense, Vsevolod Rauzer, now continued with 10.Bxb5! axb5 11.Nxb5

Due to the position of the black queen on c7, he picks up the pawn on d6 as well, and will get three pawns for the piece. It is likely that the queens will also be traded. How do we evaluate the resulting endgame and those like it?

To get our bearings, let us first look at a game which took place twenty years later where Black decided to immediately contest this ending.

Now we have reached our thematic ending. What is going on here?

White has three connected passed pawns. This is a formidable force. However, they must be carefully advanced as they will be opposed by Black's piece play. Black will try to put pressure on the pawns frontally, poke holes, and occupy those holes with minor pieces. In some cases, Black can look for counterplay on the kingside.

The bishops are of opposite colors. This is typically the case, since White usually sacrifices the light-squared bishop on b5, and Black has to trade off the dark-squared bishop when he takes the knight on d6. This in principle favors Black as it is harder to advance passed pawns when there are opposite-colored bishops.

However, the specifics of the position often mean that the black bishop is sorely restricted, especially by the e4/f3/g2 pawn chain. White generally should aim to keep the queenside pawns on the light squares to avoid a blockade and to further restrict Black's bishop.

A major factor is the position of the black king. In the Bronstein-Najdorf game, Black decided to castle kingside, hoping that by spiriting the king away, he would be able to more effectively pressure the pawns with his rooks. This turned out to be a mistake, and in most cases the black king should play a role in opposing the pawns.

Miguel Najdorf | Image from the Dutch National Archives & Spaarnestad Photo / Wikipedia

Now let's see how these factors played out in the rest of the game:

While in principle the exchange of queens after the "sacrifice" of the bishop for three pawns favors White, the evaluation of this endgame requires some subtlety. A not-insignificant factor is the position of White's kingside pawns. In many variations of the Sicilian, moves like g2-g4 and h2-h4 can be played before this endgame is reached. While these moves are great when you are trying to checkmate your opponent's king in the middlegame, once the situation changes, you might come to regret them.

In the ending we are discussing, these moves will usually work in Black's favor. For one thing, they will leave the f3 pawn behind, making it more vulnerable to attack from Black's pieces. Another important idea is that the black king might invade via e5 and f4. And finally, Black has more chances to break up the pawns by moves such as ...h5 or ..f5.

In the following game, the position is similar to the Bronstein-Najdorf game, but here it turned in Black's favor. Why? At first I was confused. But several major factors made the difference. First of all, White did not manage to retain the dark-squared bishop as Bronstein did. Second, the white kingside pawns were somewhat weakened. And third, GM Gilberto Milos had a much better place for his king on e7.

In examining these contrasting games, one can hope to achieve a better understanding of this typical and thematic endgame. As an overarching principle, I would suggest the following: The side with the pawns seeks stability in the position, hoping to slowly improve and advance their pawns. The side with the piece hopes to create more chaos and counterplay to upset the stability of the position.



  • 2 years ago


    Thank you for increasing my insight into Najdorf endgames. I play the Najdorf regularly but have not seen this sac exchange. I will read part 1 and 2 of this series in preparation for an up coming otb tournament.

  • 2 years ago


    Thank you for raising the issue with both sides presented.  Some of the moves in the second game were pretty rad for black

  • 2 years ago


    the najdorf game reminds me of a game i played gainst a 2000+ player where i traded a rook for the abc pawns which were passed and assisted by my king and the rook couldnt stop them cuz his king was displaced on the ks and it was sicilian like the najdorf game and while the king moved to the qs i improved on my position and won. i was very proud of  that game cuz it was correspondent and no inaccurcies were made on his part. it took 11 moves for me reap the reward but it was worth it.

  • 2 years ago


    Outstanding article! Thank you

  • 2 years ago


    Perfectly chosen game Bronstein-Najdorf, white's play is clear and instructive :)

  • 2 years ago


    ChicagoDave1968, cool ! Thanks for sharing.

    I saw the game before, but it still looks impressive.

    And I thought I read he was only 13 years old when the game was played, making it even more impressive to me.


  • 2 years ago


    Another very instructive example is Wei Yi - Shirov, World Cup 2013 --

  • 2 years ago



    41.Kxe5 in the last game is not a legal move. The e pawn is covered by the knight on d7 (even though that knight is pinned).


  • 2 years ago


    Great article!    Why not 41. KxE5 ??? He could have gotten both of those pawns.  No?

  • 2 years ago


    Great !!!

  • 2 years ago


    Fantastic article ! Thanks a lot ! Laughing

  • 2 years ago


    This is the single most useful article I've read on - I play peresypkin's sacrifice in the sveshnikov sicilian frequently! Thanks!

  • 2 years ago


    Thank you for such a didactic article. Loved it!

Back to Top

Post your reply: