Opening surveys in CBM 137

The old version of a modern variation: the 4...a5 Bogo-Indian

by Mihail Marin

After the first game of the Kramnik-Topalov match, the following anti-Catalan line became rather popular at the top level: 1.d4 Nf6 2.c4 e6 3.Nf3 d5 4.g3 dxc4 5.Bg2 Bb4+ 6.Bd2 a5, with the idea of answering the natural 7.Qc2 with 7...Bxd2+. Since then, I have commented on several games with this line for the respective issues of CBM, but it was only when working with the fourth game of the Sofia match between Anand-Topalov that I had a sudden revelation.

I happened to read a live comment by Shipov, mentioning that in the position above Smyslov used to play 7...Nc6. A whole new (or rather well forgotten) world (re-)opened in front of my eyes. I instantly remembered that as a teenager I was deeply impressed by the following Bogo-Indian line, which in the '80s enjoyed certain popularity thanks to the efforts of Taimanov, although Smyslov, Polugaevsky and Timman also employed it occasionally: 1.d4 Nf6 2.c4 e6 3.Nf3 Bb4+ 4.Bd2 a5 5.g3 (the most ambitious continuation, according to old and new theory) 5...d5 6.Bg2 dxc4 7.Qc2 Nc6 8.Qxc4 Qd5.

After White's seventh move the position is the same as in the modern anti-Catalan line, but the subsequent approach is entirely different. Instead of giving up the tension (to some minds, ...Bxd2+ may not be entirely logical, just one move after Black has defended his bishop), Black aims to get the best out of it. To me, all those years ago, this looked like a highly refined way of playing chess and, although I never had the Bogo-Indian in my repertoire, I enjoyed playing over the games with this line from Schakhmatnii Biuletin or Shakhmati v SSSR.

In the meanwhile, I have become a fan of the Catalan Opening, which made me very curious to find out how would I evaluate this line now. I started examining the possible lines with the purpose of proving the whole system to be playable for Black. The purpose of the present article is to share my impressions with the reader. I will cover all the significant continuations for White, but mainly the best ideas for Black.

In the position from the diagram above, White's queen is attacked and the d4-pawn is under pressure. Therefore, his choice is rather limited. We have A) 9.Qxd5 and B) 9.Qd3. The character of the play is quite different in these lines and we will deal with them separately.

A) 9.Qxd5 exd5

Releasing the tension may seem like a significant concession on White's part, but things are far from clear from the practical point of view. A rapid description of the position suggests a similarity with the Carlsbad structure, characteristic for the Exchange Variation of the Queen's Gambit Declined. There are several important elements that induce important differences, though.

The white light-squared bishop is not optimally placed for this structure (in the QG it usually goes to d3). This will cause White some problems starting the queenside minority attack, which tends to leave the c4-square weak.

On the other hand, Black's queen's knight is placed in front of its own c-pawn, preventing ...c6, which would consolidate the d5-pawn and reduce the g2-bishop to relative passivity. The way it is, the potential pressure against d5 should be taken into account permanently. Black's main trump is the territory gained on the queenside so far. Two of his pieces create a tense situation, while the a5-pawn plays an important part. Facing the strategic threat of ...a4-a3, White will have to weaken some square on this area of the board. For instance, a2-a4 would leave Black dominating on b4, a2-a3 would weaken b3 and c4 (after ...a4), while b2-b3 as an answer to ...a4-a3 would weaken both c3 and b4! Of course, these evolutions have an abstract character yet, but they have to be taken into account, too.

One important aspect refers to the kings' position. In most of the cases, White would castle short, but Black has a wider choice. In order to compensate for the temporary lack of structural flexibility, the king may go to d7, or (in the case of an exchange of the dark-squared bishops) even to d6, keeping c7 and d5 well defended. Short castling remains an important alternative, but ...0-0-0 should not be discarded either. In some cases, it would allow a rapid centralisation of the rooks (if followed by ...Rhe8).

White has a relatively wide choice on the next move. We will examine A1) 10.a3, A2) 10.Nc3 and A3) 10.0-0. 10.Bxb4 axb4 is likely to transpose to one of the sublines of A3) after 11.0-0 Bg4.

A1) 10.a3 Bxd2+ 11.Nbxd2 looks like a radical way of getting rid of the irritating piece tension and speeding up the queenside development.

However, the weakness of the b3-square usually offers Black a clear queenside plan. Besides, the bishop exchange tends to favour Black in the long term, ensuring him some stability on the dark squares. See Arambel,S - Soppe,G 0-1.

A2) 10.Nc3 looks like a natural developing move, but has the significant drawback of releasing the tension in a slightly passive way. In the next phase of the game, it will be only Black who puts pressure along the a5-e1 diagonal.

The game Browne,W - Smyslov,V 0-1 is a wonderful illustration of Black's main strategic ideas in this opening.

A3) By the method of elimination, 10.0-0 must be considered the most constructive continuation.

White maintains the status quo on the queenside by making a generally useful move. (We know already from the previous game that the white king is not too stable in the centre). Black has tried several moves in this position, but most of them tend to leave White a free choice for manoeuvring. Polugaevsky,L - Taimanov,M 1-0 is a good example of what can happen if White's concentration of forces gets out of control.

I believe that Black should act energetically, which makes 10...Bg4 the main candidate.

Black increases his pressure in the centre, threatening to win the d4-pawn. As we know, castling is not a high priority for Black. The king will choose the right place according to the circumstances.

White has two main ways of defending the pawn: A31) 11.e3 and A32) 11.Be3. The exchange 11.Bxb4 axb4 is a concession and does not have independent value. After 12.e3 Ne4 play would transpose to a subline of A31).

A31) 11.e3 looks like a solid way of restricting Black's piece activity.

On the negative side of this move, we can mention the fact that White's own dark-squared bishop's activity also is restricted. Black should build on the partial success with another active move, increasing the pressure.

Hence, 11...Ne4, when White has to take a decision already.

12.Bxb4 axb4 only improves Black's position, as in Vladimirov,B - Taimanov,M 0-1.

12.Bc3 is more consistent.

White does not threaten to continue with his development yet (Nbd2 would lose a pawn), but Black has no favourable way of releasing the tension. An exchange on c3 would strengthen White's centre after bxc3. I have made a few suggestions for Black in my notes to Arkell,K - Speelman,J ½-½, revealing also some of the hidden strategic dangers.

Generally speaking, 11.e3 is a good move, but Black should not feel too threatened.

A32) 11.Be3 is more cunning.

Admittedly, the bishop does not stand well in front of the e-pawn, but this is a temporary solution. The important thing is to keep the c1-h6 diagonal open, allowing a later Bf4 or Bg5.

Black has tried many moves here and I must confess that I faced a tough job making my choice.

I would mention that Black should not try harassing the bishop with 11...Ne7 (planning ...Nf5) because of 12.Bf4!, see Gaprindashvili,N - Taimanov,M 0-1.

The most active and probably best move is 11...Ne4.

Although the knight is not threatening anything concrete, it certainly makes White's further development quite difficult. Having exhausted the useful moves, Black will have to make up his mind in the near future. For 12.Rc1, see Kuligowski,A - Pokojowczyk,J 1-0 and for 12.h3 see Burger,K - Taimanov,M ½-½.

I would conclude that A) 9.Qxd5 leads to interesting play, but Black can count on equalising completely with accurate play.

B) 9.Qd3 is slightly more popular and presents Black with a bigger challenge.

Having mentioned Smyslov's name in the introduction, I will add that he repeatedly defended Black's point of view in the line 9...0-0 10.Nc3 Qh5.

However, the games Akopian,V - Smyslov,V ½-½ and Tukmakov,V - Smyslov,V ½-½ are not really appealing for Black.

The variation that had impressed me as a teenager goes 9...Qe4.

The strong threat ...Qxd3 forces White to exchange queens with 10.Qxe4. (It should be mentioned that after 10.Qc4? Black can do a lot better than agree to a draw by repetition, by playing 10...b5! as in Tatai,S - Taimanov,M 0-1). After 10...Nxe4 a first critical moment of the variation is reached.

If compared with the A) line, Black has less control in the centre (his pawn is on e6 rather than on d5), but his king's knight is more active (some would add: also more unstable!). Black's declared short term aim is to get the bishop pair with ...Nxd2. In the long term, he will advance his a-pawn, for the purpose of causing the same kind of weaknesses as in line A).

White's problem is that he cannot castle because the d4-pawn would be hanging after general exchanges on d2. His main continuations are B1) 11.a3 and B2) 11.Bxb4. For less common moves see Meyer,P - Cladouras,P 0-1.

B1) The main tabyia of the variation arises after 11.a3 Nxd2 12.Nbxd2 Be7 (Black has tried other continuations on the 12th move, too, but it is the position below that had captured my attention as a teenager).

The game Kouatly,B - Polugaevsky,L 0-1 had had a strong theoretical impact in that era. In my comments to it I have revealed some of the essential aspects of the position. White tried improving over this game in Portisch,L - Timman,J ½-½ and Torre,E - Maninang,R 1-0, but Black managed to prove the correctness of his general strategy.

So far, things look encouraging. My actual perception is similar to that from a quarter of a century ago. Black has chances to outplay his opponent in a heavy strategic battle without queens on the board. Still, I had an unexpected blow when I became aware of the less popular:

B2) 11.Bxb4

After having examined the branch of lines grouped under A), we are used to the fact that this is a strategic concession. However, things are different with the long diagonal wide open.

The most desirable reaction is 11...axb4.

However, as revealed by the game Biriukov,O - Faibisovich,V 1-0, the b4-pawn is more of a weakness than strength.

Black does better playing 11...Nxb4.

Black clears the path for the c-pawn with gain of time, which represents a small achievement. The king will be well placed in the centre, contributing to the defence of the queenside, while his rival will have to get "out of play" with 0-0.

The overcautious 12.Na3 is inoffensive, see Sosonko,G - Hartmann,W ½-½.

The more ambitious 12.0-0! is a better try for an advantage.

Black has no immediate path to equality, but he can build up a solid defensive position, based on the presence of the king in the centre. In Beliavsky,A - Ljubojevic,L 0-1 things looked a bit messy for Black at some point, but he nevertheless emerged well out of the complications and eventually won on time. In Vaulin,A - Faibisovich,V ½-½ events had a slower character, but Black managed to hold his own all the same.

Conclusion: Although I have a marked feeling of disappointment provoked by the fact that in line B2) 11...axb4 is not recommendable, I believe that this whole old-fashioned variation of the Bogo-Indian remains playable for Black. I suspect that the engines' overall scepticism sometimes is caused by the fact that the black king is in the centre (actually, from a human point of view, this is an advantage!). Players with good endgame technique will feel very much at home with Black. Although, as mentioned above, there rarely is an immediate way to clear equality for Black, White cannot give play a one-sided course. His queenside weaknesses tend to cause him long term problems, with an increasing degree of risk in the event of simplifications.