Part 1 | Part 2

During my first forays into the realm of positional play, I learned that the two bishops can be a powerful force, controlling key diagonals and overpowering the bishop and knight or two knights that oppose them.

I can still recall first seeing wonderful examples of the bishops' combined power. These made a great impression on me. As a result, I often single-mindedly chased after the goal of attaining the "dynamic duo," only to find that they were frequently hemmed in by pawns (friend or foe), dominated by knights on central outposts, or readily exchanged by the opposition. What advantage?

With time came recognition that the two bishops is an important consideration in a position (what Silman, in HOW TO REASSESS YOUR CHESS, would call an “imbalance”). It is not, however, the magic elixir that guarantees an advantage, let alone a win, in any position. The win will go to the player who better understands and adapts to the unique needs of the position, with the two bishops as one component of the larger whole.

The following game highlights the struggle that revolves around the two bishops. White obtains the "cleric pair" and Black endeavors to bottle them up and make them less effective than Black's knight – the only cavalry left on the board. The resulting interplay, in a board-one final round game at a Grand Prix tournament, is instructive.

Randy Bauer (2200) - Kari Hoijorvi (2130)
Leavenworth (Kansas) Open, 5-15-94
King's Indian Attack

1.g3

What's the point of this, you say? Well, it's a flexible move that commits White to nothing except the fianchetto of the king's bishop. It's especially effective when, as in my case, the player also plays 1...g6. Even though the move has been played with some regularity by players the strength of Larsen, Benko (to beat Fischer, for example), and Suttles, it still causes a raised eyebrow from time to time from Black players.

1...d5 2.Nf3 Nf6 3.Bg2 e6 4.0-0 Be7 5.d3 0-0 6.Nbd2 c5 7.e4

We now have the ubiquitous King's Indian Attack, which can arise from a variety of openings, including 1.e4 e6 or 1.e4 c5 as well as with 1.Nf3 or 1.g3. I've often found this a useful weapon when less theoretically prepared or unsure of my opponent's openings predilections.

7...dxe4

I don't think that ceding the center is Black's best approach. My opponent didn't want to allow the stock kingside attacking schemes based on e4-e5 for White; Black's idea allows him to post his knight on the strong d5 square in response to e5.

8.dxe4 Nc6 9.Re1

This is a constructive waiting move by White, who posts the rook on a central file and removes problems associated with ...Ba6. White will likely play c3, move the queen, and then play Nc4 or Nf1. There are, of course, other ways to set up White's pieces.

9...e5

Black reasons that he can afford the tempo loss (e7-e6-e5) because White has placed his rook on e1. When Black has his pawn on e5 a typical strategy for White (or Black, in similar King's Indian Defense positions) would be to play for f2-f4, but that idea would be better with the rook still on f1.

However, Black's move creates a major structural imbalance: Black has allowed a serious hole on d5, which a knight can settle into, while Black's natural knight outpost, d4, can be contested by a White pawn on c3. White decides that this imbalance is important enough to center his play around.

10.c3 h6

Black would like to play 10...Be6, which prevents 11.Nc4, but 11.Ng5 is useful for White, since 11...Bg4 12.Bf3 offers to trade Black's good bishop for White's bad one (notice the White and Black pawn structure – a key for determining the effectiveness of a bishop. In this position, the e4-pawn hampers White's light squared bishop, while the c5 and e5 pawns limit Black's dark squared bishop), or 12.f3 Bd7 13.f4 gets White's kingside play started with tempo.

11.Nc4!

I think that Black thought this move unplayable, since 11...Qxd1 12.Rxd1 leaves the e4 pawn unprotected. In fact, this exchange is good for White, since 12...Nxe4 13.Nfxe5 Nxe5 14.Nxe5 gives White a clear lead in development and a strong bishop on the h1-a8 diagonal.

11...Qxd1 12.Rxd1 Bg4 13.h3?!

This "wins" White the two bishops, but it may not do so in the most efficient manner. I think better was 13.Re1. Why, you say, would White want to leave the open d-file for the blocked e-file? The primary reason is that White cannot easily hold the d-file, since ...Rfd8 will contest it at a time when connecting the rooks, with Be3, robs the Nc4 of its natural post.

White's key idea, after all, is to occupy d5 with his knight. The Re1 is actually well placed to help with this, since if Black responds ...Nxd5, White's exd5 will simultaneously expose the pawn on e5 to attack. This is a common theme in Black positions resulting from the King's Indian and Pirc Defenses. This also helps explain why it's often useful to study structures that can be utilized for both White and Black.

13...Bxf3 14.Bxf3

So White has won the two bishops, but his light squared bishop is, at the moment, not the greatest of pieces. White's plan must be to finish his development while improving the outlook for his bishops. At the same time, his Nc4 has a ready-made home on d5. Since Black has a knight controlling that square, White must always reckon with the resulting exchange: would the passed d5 pawn be a strength or weakness? That depends on the placement of the other pieces and pawns, but both sides must take it into consideration as they make their plans.

14...Rfd8

Black logically contests the d-file before White has a chance to smoothly connect his rooks. White's "logical" developing move, 15.Be3, takes away the White knight's best square. I envision putting the bishop on b2 or a3, keeping e3 clear. 15.Be3 b5 starts pushing White's pieces back.

15.Rxd8+

White isn't thrilled about giving up the open file, but he reasons that the bishops can control the squares on the file and allow White to eventually play Kf1-e2 and Rd1 if necessary. As noted above, 15.Re1 may be just as good, although in that case White is a tempo down on the note to move 13.

15...Rxd8 16.Ne3 b5

Black logically contests the c4-square and envisions c5-c4, which frees his dark squared bishop and restricts White's light squared bishop as well as artificially "isolating" a White pawn should an exchange on d5 occur. Black is playing actively, but his advanced queenside pawns can also become targets. That's the two-edged sword inherent in such advances, and, in the end, it helps White by giving his bishops something to shoot at.

17.Nd5

Is this advance good or bad? White believes that it is good because it dissolves the "ram" pawns (e4 vs. e5, a blocked type of structure that generally helps a knight in a battle with a bishop), which increases the scope of the  light squared bishop and, in this particular circumstance, forces the Black Nc6, after an exchange, to a less central location.

17...Nxd5 18.exd5 Na5!

Black recognizes that c4 is a critical square, and the knight keeps Black's grip on it while preventing a2-a4.

19.b3

White decides that he should deny the Black knight any possibility of effectively going to c4 while also keeping his pawns flexible. If Black plays c5-c4, White may choose to answer b3-b4 followed by a2-a4. With the pawn on b2, Black can capture en passant. White also wants to make it difficult for Black to play his knight to d6, where it would control all the important central light squares, making it more than a match for the White light squared bishop. This would be an example of a well-placed knight neutralizing the two bishops.

19...Bg5?

Black starts on a misguided plan. 19...Nb7 comes into consideration, although 20.Be3 calls Black back to the defense of his own weakened queenside pawns, and White will follow up with a4 with a queenside initiative.

Black's most combative plan would be to blunt the light squared bishop (and thus threaten the d5-pawn) with ...f5 and ...e4. The game gets very interesting after 19...f5!

Part 1 | Part 2