ForumPosition SearchText SearchMy ProfileRepertoireSite InfoGuests InfoHelpLinks
In last month's update I looked at the line 1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.Nf3 0-0 6.Be2 and now 6...Nbd7 followed by 7...e5, finding it to be interesting and quite flexible for Black. Assuming this is a good weapon, can we not then use an Old Indian move order with 1.d4 Nf6 2.c4 d6 3.Nc3 e5 (or 3...Nbd7 4.e4 e5 5.Nf3 g6) 4.Nf3 Nbd7 5.e4 g6 6.Be2 Bg7 7.0-0 0-0, avoiding a lot of White's other systems? This approach was popular amongst East German players including Rainer Knaak, Lutz Espig and Thomas Casper, now having strong exponents such as Robert Markus and Sergei Movsesian.
These move order considerations can get very messy, so let's simplify the goals of this update. What are the pros and cons of 3...e5 and 3...Nbd7 in reaching a Classical King's Indian with ...Nbd7 included? The sections below address what I think are the key issues:

Download PGN of April ’24 KID games

>> Previous Update >>


Old Indian Variation 3...e5 4.Nf3 Nbd7 5.Bg5 [A54]

White's 5.Bg5 is only a problem with the 1.d4 Nf6 2.c4 d6 3.Nc3 e5 move order, after 1.d4 Nf6 2.c4 d6 3.Nc3 Nbd7 4.Nf3 Black can play 4...c6 if he wants to play the Old Indian instead of the King's Indian (putting his bishop on e7 instead of g7), which is usually included in an Old Indian set-up. After 4...c6 5.Bg5 he can then play 5...h6 6.Bh4 g5 7.Bg3 Nh5, which hunts down the dark square bishop with a tempo save over similar positions in which Black plays ...g7-g6 and only later ...g7-g5. If he wants to play a King's Indian he can delay committing to 4...e5 with 4...g6 instead. After 1.d4 Nf6 2.c4 d6 3.Nc3 Nbd7 4.Nf3 g6 5.Bg5 Bg7 we reach a Smyslov System against the King's Indian, which is not considered to be a critical test. On the other hand both these move order finesses commit Black to the early ...Nbd7, which might not be Black's choice against other White set-ups such as 1.d4 Nf6 2.c4 d6 3.Nc3 Nbd7 4.e4 e5 5.d5 (see below).

Black has quite a few ways to play against 5.Bg5 and the games of two leading exponents, Robert Markus and Dmitri Andreikin, were the first thing that caught my attention. When faced with 5.Bg5, Markus has played the rare 5...exd4 6.Nxd4 h6 (or 6...g6 immediately) 7.Bh4 g6 (Carnic, S - Markus, R ):

Andreikin meanwhile has favoured 5...Be7 6.e3 0-0:

usually followed soon by ...exd4 (Gupta, A - Andreikin, D ).

Yet as I investigated this line I felt it was probably better to play the mainstream approach, of ...c6 followed by some move of the f6 knight to exchange dark square bishops. Klukin, K - Gareev, T is a good example of this plan, with Black outplaying his opponent in a nice game.

On the other hand Joerg Hickl's immediate exchange of bishops with 6...Ng8 has not brought him much joy, despite this being recommended in his book and Chessable course, 'Play 1...d6 Against Everything' (Zude and Hickl).

Here only Hickl's draw with Dautov gets mentioned, not his losses. Black's position is solid but rather passive, with White winning an attractive game in Gabriel, C - Hickl, J. Finally it's worth noting that after 3...e5 4.Nf3 Nbd7 5.Bg5, Black cannot play in King's Indian style with 5...g6 because of 6.dxe5 dxe5 7.Nxe5, winning a pawn.

Other strong players have wheeled out various 5.Bg5 counters in online blitz, but if they do not also use them in more serious games they probably lack confidence in the objective value of their approach.


Old Indian Variation 3...e5 4.d5 [A53]

In playing this, White wants to set up a Saemisch style pawn structure with e2-e4 and f2-f3, thereby ensuring a long-term space advantage. Black certainly has more options after 3...e5 4.d5 (rather than 3...Nbd7 4.e4 e5 5.d5) and I wanted to make 4...Bf5 work, trying to prevent e2-e4 altogether. Unfortunately the engine did not like anything I came up with so I looked at the flexible 4...g6 and found that the obscure internet rapidplay game, Nishant, M - Balabaev, T, appealed to me.

If Black wants another alternative he might choose to emulate Akopian's play in Repka, C - Akopian, V, and here the queen's knight goes to a6.


Old Indian Variation 3...e5 4.dxe5 [A53]

4.dxe5 has been a major option since David Bronstein and Isaac Boleslavsky introduced the King's Indian with 3...e5 in the 1940s. After 4...dxe5 5.Qxd8+ Kxd8 White's best is 6.Nf3. At one time 6...Nfd7 was considered compulsory, but it turns out that Black has several options including 6...Nbd7, 6...Bd6 and 6...Nc6. It's this latter move that is the engine favourite, claiming fully equality for Black after 7.Bg5 Be6:

Anton Demchenko has played a lot of blitz games this way, usually reaching it via 1.d4 Nf6 2.c4 d6 3.Nc3 e5 4.Nf3 Nc6!? 5.dxe5 dxe5 (5...Nxe5 is also playable) 6.Qxd8+ Kxd8. I intend to revisit his 4...Nc6!? in a future update, and it goes without saying that 5.d5 is the critical response. Meanwhile I've chosen Aczel, G - Sedlak, N to illustrate the endgame.


Old Indian Variation 3...Nbd7 4.e4 e5 5.d5 [A53]

This appears to be the main problem with a 3...Nbd7 move order. After 4.e4 e5 5.d5 followed by 6.f3, White will get a long-term space advantage:

Can Black get adequate counterplay? In Rapport, R - Markus, R White gave a model demonstration of how to gradually use White's space advantage, though I think that Markus would have shaken the grip of a lesser player.


There are cons to both 3...e5 and 3...Nbd7, my main concern about 3...e5 being 4.Nf3 Nbd7 5.Bg5, and the key issue with 3...Nbd7 being 4.e4 e5 5.d5. At the same time I'd say that both moves are quite playable and lead to interesting and relatively fresh positions. Together with last month's coverage of the 6...Nbd7 Classical this forms the basis of an economical King's Indian repertoire, though we also need to consider lines in which Black's queen's knight is committed to d7 and White plays g2-g3.



See you next month! Nigel

>> Previous Update >>

Don't hesitate to share your thoughts and suggestions. Any queries or comments to the KID Forum, or to me directly at support@chesspublishing.com (subscribers only) would be welcome.