I have recently researched the O'Kelly variation in the Sicilian which goes 1. e4 c5 2. Nf3 a6! and found that it's a good surprise weapon.
As beginners, we’re taught that backward pawns are generally weak because they can’t be protected by another pawn from behind (otherwise, it wouldn’t be backward). This is especially a problem if the enemy forces are in a position to exert pressure on the backward pawn. If the pawn can’t be won by brute force, then it’s often the case that the defender becomes tied down in its defense.
Conceding a Backward Pawn
But what if the backward pawn is hard to attack? Is it reasonable to concede what might be a long-term weakness (or a violation of a principle) for a dynamic middle game? Absolutely, and I’ll show you an example from the black side of a Sicilian.
The line goes:
- e4 c5
- Nf3 a6 (you’d be surprised how often this move comes as a surprise)
- d4 cxd4
- Nxd4 Nf6
- Nc3 e5 conceding a backward pawn, but with tempo.
- Nb3 and Bb4 with a threat already to win the e-pawn.
- White parries that threat with Bd3.
- According to the database on chess.com, by far the most popular response for black is to break free with d5, thus ridding black’s backward pawn. Typical play can go as follows: exd5 Nxd5
- Bd2 Nxc3
- bxc3 Be7 and I suppose black is not mad about the opening outcome because the position is more or less equal and black has the better structure (although that counts for little with the queens still on.) This line may offer black equality, but it clarifies the position too much for my taste.
Instead of the approved 7. d5, I recommend 7. d6 simply supporting the e-pawn and opening the door for the light-squared bishop.
What If White Castles?
So what happens if white simply castles on his next move? Black meets 8. O-O with Bxc3 making use of 6. Bb4 and saddling white with doubled isolated c-pawns. After 9. bxc3 O-O, let’s take stock of the position. Black has a backward d-pawn and has given away the bishop pair (it was given but it wasn’t a gift!).
At first glance, it might seem risky, especially if white plays 10. Ba3. Isn’t white close to winning the d-pawn or at least making black’s pieces passive? The answer is a resounding no! In order to win the d-pawn, white has to move his bishop on d3 so white’s queen can partner with the dark-squared bishop. But, currently, that’s inadvisable because the e-pawn would drop.
For instance, 11. Bc4 would be met with Nxe4 12. Qd3 Bf5 and the tactics work in black’s favor. I4. Bd5 is met by the surprisingly simple Nf6 and 15. Qxf5 is countered with Nxd5. 14. Rae1 is a better try, but then b5 comes as a nice shot. White can win a pawn back after 15. Bd5 Nf6 16. Bxf7 Rxf7 17. Qxf5 but it’s clear white is not coordinated and black is much better. Even swapping queens off with Qd7 would be fine:
18. Qxd7 Rxd7
19. Rd1 a5
20. Rxd6 Rxd6
21. Bxd6 Ne4!
What If White Prepares a Bishop Move With Re1?
So if immediately moving the bishop is flat-out bad, what if white prepares it with 11. Re1? 11. Re1 is met with Qc7 and after 12. Bf1 black is in time to defend with Rd8. But more importantly, what is white doing? Both bishops are unimpressive and the only break is f4, but the king’s rook is on e1.
What If White Adds Pressure?
But here comes another question: What if white adds pressure with Qd2 intending Rad1? It’s too slow, and black is able to demonstrate the dynamism of the position. 13. Qd2 and a5 looks good, but also Be6 meeting 14. Rad1 with d5! The move we purposefully withheld until further development and a greater increase in tension. 15. exd5 Nxd5 16. c4 Nab4 and what’s not to like? But even if 14. d5 wasn’t possible, black has the positional Ne8 just keeping everything under control. The backward pawn isn’t weak.
A Backward Pawn Does Not Violate a Principle
To improve in chess, the beginner has to see past the superficial and understand that what seems to violate a principle (a backward pawn) is actually not. 7. d5 was a perfectly fine move and there’s a reason it’s the mainline, but I like to think of playing 14. d5 as the difference between pulling a rubber band back a little versus pulling a rubber band back a lot. When you pull back a rubber band a lot, there’s a greater amount of tension. In other words, the release of energy is amplified and therefore more dangerous.
Continuing the Line
Now, let’s go back so we can go forward. 8. O-O made black’s next move the only move because allowing Nd5 would be too much because the bishop would have to give itself up on c5. White would have the bishop pair for next to nothing. But what if white didn’t want to allow black to wreck his structure? Say with 8. Bd2? Then what?
I suggest Be6 increasing central control and not allowing Nd5. 9. a3 Bxc3 10. Bxc3 really illustrates my point. It’s true that black conceded the bishop pair and this time without inflicting white with doubled isolated pawns, but the position is closed enough for white’s bishops not to wreak havoc. Normal play might continue O-O 11. O-O Nbd7 12. h3 Qb6 and black has finished development and ready to play d5 if he wants.
What If White Makes a More Aggressive Play?
What if white plays something more aggressive after 11. Nbd7 like 12. f4? Then Bxb3 comes strongly. The point is after 13. Bxc3 Nc5 does it all. I even have a potential endgame after:
14. Qe2 Qb6
15. Kh1 Nxd3
16. Qxd3 Qxb3
17. Qxd6 Nxe4
18. Qxe5 Nxc3
19. Qxc3 Qxc3
20. Bxc3 Rac8
21. Rab1 b5 and black is playing for only two results.
A Seeming Weakness Can Be a Strong Play
Now that we’ve gone through a few different lines, it’s important to remember what this article is about. Hopefully I’ve shown that it’s strong play to take on what seems like a weakness if it means you’re able to attain a dynamic, strategically rich middle game.
Yes. There are instances when a backward pawn is simply a liability, but in the above-mentioned lines, black compensated for it with active pieces and almost always having the option to play d5 at the right moment.