By Joel Benjamin

In the past three installments we dealt with the most popular Sicilian second move, 2...d6.  While the check line presents itself as an obvious anti-Najdorf or -Dragon line, 2...e6 often has “non-openers” in a quandary. Beginning this month we will consider the possibilities that stem from a wonderfully flexible move, 3.Nc3. White preserves the option of an open or closed Sicilian approach, which forces the Black player to carefully regard his move order.

This month we will start with the reply 3...a6, a virtually obligatory move for proponents of the Paulsen or Kan Variation (1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 a6). I recommend 3.g3, which I have used with good success on a number of occasions. I’ll start by showing a very good grandmaster demolished by this surprisingly dangerous system.

Joel Benjamin - Alex Stripunsky [B40]

Manhattan Club Championship 2001

1.e4 c5 2.Nf3 e6 3.Nc3 a6 4.g3 b5

Black counters white’s fianchetto with one of his own. In the past I would often transpose into an open Sicilian, but I’ve gone away from that after an unpleasant experience in this line: 5.d4 cxd4 6.Nxd4 b4 7.Na4 Bb7 8.Bg2 Nf6 9.Bg5 Be7! (Kasparov had earlier played the inferior 9...Qa5, lulling me into a false sense of security) 10.0–0 Qa5 11.Bxf6 Bxf6 12.Nb3? Qxa4 13.Nc5 Qc6 14.Nxb7 Be5! 15.Nd6+ Qxd6 16.Qxd6 Bxd6 17.e5 Bxe5 18.Bxa8 Bxb2 19.Rab1 Bc3 (Benjamin-Serper, Salt Lake City 1999).  With two pawns for the Exchange and a massive structural advantage, Black is already winning and ultimately registered the point. Furthermore, the whole game had been played before, and White was no more successful in Bistric-Sax, Sarajevo 1983. Readers will understand why this game left me a trifle bitter!

This game is by no means the end of the discussion. White could for instance sacrifice the e4-pawn with 9.0-0, inviting great complications. A recent novelty may have important theoretical implications: 9.0-0 Bxe4 10.Bxe4 Nxe4 11.Re1 d5 (11...Nc5 or 11...Nf6 may be better) 12.c4! Bxc3 13.Nxc3 Nxc3 and now 14.Qh5!! gave White a winning attack in Guthrie-Humphrey, Mount Buller 2004/5.

5.Bg2 Bb7 6.d3 d6 7.0–0 Nd7 8.Ng5!?

We see a recurrence of the same plan from the …Nxd7 lines of the Moscow Variation. It is surprising how well this seemingly time-wasting maneuver works.

8...Be7 9.Nh3 Ngf6 10.f4 b4 11.Ne2 c4?!

Black’s attempt at counterplay in the center actually just strengthens the tide of onrushing White pawns.

12.Nf2 cxd3 13.cxd3 Qb6 14.d4 0–0 15.g4 Rfc8 16.g5 Ne8 17.Ng3 Bd8 18.Be3 a5 19.Ng4 Nf8 20.f5 Qc7 21.Rf2 a4 22.h4 a3 23.b3 Qd7 24.fxe6 fxe6 25.Qf3 Qe7 26.Raf1

Twins are great, but triplets are even better!

26…Nd7 27.d5 Nc7 28.dxe6 Nxe6 29.Nf5, 1–0.

Joel Benjamin-Ilya Smirin [B40]

World Open 2001

1.e4 c5 2.Nf3 e6 3.Nc3 a6 4.g3 b5 5.Bg2 Bb7 6.d3 d6 7.Ng5

Here I trotted out my favorite maneuver a move earlier. The argument for haste is provided by the possibility of 7.0-0 Be7, which prevents or at least delays the knight maneuver. I happily sat next to the following Black success: 7.0–0 Be7 8.Be3 Nc6 9.h3 Nf6 10.g4 0–0 11.Nh2 d5! (The timing is crucial, for White would be happy if he had time for g4-g5 followed by f2-f4) 12.g5 d4 13.gxf6 Bxf6 14.Ng4 dxe3 15.Nxf6+ Qxf6 16.fxe3 Qe5 17.Qg4 Ne7 18.a3 Ng6 19.Rf2 f5 20.Qg5 Rae8 21.Raf1 Rf6 22.exf5 Bxg2 23.Qxg2 Nh4 24.Qg5? Rg6 25.Qg4, 0–1, Galego-Gulko, Bled Olympiad 2002.  [Meanwhile on board two, I won on time in a boring rook endgame.  It concluded my worst Olympiad performance, in which no opponents resigned to me…I just won two games on time.]

My opponent in this game had even thought to play …Be7 even earlier – on move six, to be exact. This invites White to try to disrupt Black’s development with 7.e5!?, which proved innocuous in Sutovsky-Smirin, Dresden Zonal 1998 after 7...d6 8.exd6 Qxd6 9.a4 Qb6 10.0–0 Nd7 11.axb5 axb5 12.Rxa8+ Bxa8 13.d4 Ngf6 14.dxc5 Nxc5 15.Be3 b4 16.Na4 Qc7 (drawn in 27 moves).  Later Najer improved with 11.Ng5, threatening 12.axb5.

He received first a positional advantage and later a mating attack:  11...b4 12.a5 Qa7 13.Nce4 Ngf6 14.Nd2 h6 15.Nge4 0–0 16.Qe2 Nd5 17.Nc4 Bc6 18.f4 Qc7 19.f5 exf5 20.Rxf5 Rae8 21.Bxh6 gxh6 22.Qg4+ Kh7 23.Raf1 N7f6 24.Rxf6 Nxf6 25.Rxf6 Bxf6 26.Nxf6+ Kh8 27.Qf5 Re1+ 28.Kf2 Re2+ 29.Kxe2 Re8+ 30.Kf2 Kg7 31.Qh7+ Kxf6 32.Qxh6+ Kf5 33.Bh3 mate, Najer- Handke Cappelle La Grande, 2000

In the position after 6...Be7 7.0-0 d6, (or 6...d6 7.0-0 Be7), White could continue with the simple 8.Ne1.

After a sensible move like 8...Nc6 or Nd7, White would continue 9.f4 and return the knight to f3. White would have fairly good attacking chances, if not an advantage. But 8...Nf6? loses a bit of material: 9.e5 Bxg2 10.exf6 Bxf1 11.fxe7 Qxe7 12.Kxf1. It is quite useful to remember this trap and use it should the occasion arise!

Now back to our main game:

7...h6

Black can kick back the knight with 7...Be7, but the knight might not retreat – 8.f4 is perfectly playable at once.

8.Nh3 Ne7!? 9.0–0 g6

We have an unusual turn of events, as Black is now developing like a conventional Closed Sicilian. Now the intended plan of f2-f4 did not appeal to me. Black has half his army directed at the f4-f5 advance; after due preparation, he could stop me in my tracks with f7-f5. Reckless attacking could certainly backfire on me, e.g. 10.f4 Bg7 11.g4 Nbc6 12 f5 gxf5 13.gxf5 exf5 14.exf5 Bxc3 15.bxc3 Rg8 and who’s zooming whom?

10.d4

Of course, this advance looks quite nice if Black captures. But this is the rare case in the Sicilian that White advances d2-d4 and Black leaves it there without consequences.

10...Nd7 11.Be3

Fritz 8 suggests 11.d5, but 11...exd5 12.Nxd5 Bxd5 13.exd5 0-0 looks fine, as does simply 11...e5.

11...Bg7 12.Qd2

12.f4 prevents the maneuver executed in the game, but it doesn’t pressure Black much.  He could profitably trade queens with 12...cxd4 13.Bxd4 Bxd4+ 14.Qxd4 Qb6 15.Rfd1 Qxd4+ 16.Rxd4 Nb6.

I hoped to tie Black down to the h6-pawn, but there’s no chance of that.

12...cxd4 13.Bxd4 Ne5 14.f4 Nc4 15.Qf2 Bxd4 16.Qxd4 Qb6 17.Qxb6 Nxb6 18.Nd1