Part 1 | Part 2

34.Rf2 Qxg2+ 35.Kxg2 Be4+ 36.Kg3 Nd3 37.Rxf8+ Kxf8 38.Rf1+ Kg7 39.Nf4 Be5 40.Nc4 Nxf4 41.Bxf4 Bc3 42.Rd1 Rc5 43.Be3 Rxd5 44.Rxd5 Bxd5 45.Bxa7 Be4 46.Bb6 d5 47.Na5 Kf6 48.Nc6 e5 49.Bc5 d4 50.Nxb4

Material equality is finally re-established but the d pawn is king.

50...d3, 0-1. A marvelous exhibition by Geller and probably sufficient to remove the idea of d4-d5 from the mind of the competitive player for generations. Can't White do better?

Very recently I had the chance to try something similar.

Martin - Mack [E60]
4NCL GB Team Ch, 2004
1.Nf3 Nf6 2.c4 g6 3.g3 Bg7 4.Bg2 0–0 5.0–0 d6 6.d4 Nbd7 7.d5!?

Another variation, another opportunity. Keith Arkell ventured this move against me in the 1997 British Championship and I have seen Rustemov play 7.d5 on occasion. As my opponent said in the post mortem: “From this moment on, I had no point of reference." 7.d5 seems to remain faithful to Petrosian's original thought. If Black wants to hit the pawn with ...c7-c6 or ...e7-e6 he will open the diagonal for the Bg2, and if not Black is cramped. There are difficult positional problems to solve for both players.

7...e5 8.dxe6 fxe6 9.Nd4?!

But this is not correct, despite looking good. White should keep it simple with 9.Nc3, when I prefer his position. The Black centre is backward and difficult to advance, meanwhile White can develop smoothly: 9.Nc3 Nc5 10.Be3 Nce4 11.Nxe4 Nxe4 12.Qc2 Nf6 13.c5! and White had the edge in Benko –Browne, Oberlin 1975.

9...Qe7 10.Nc3 Nb6 11.b3 d5!

Black hits out before White gets the chance to cement his advantage. He also justifies the position of the Nb6, which is always important. Suddenly and strangely, Black is better.

12.cxd5 Nbxd5 13.Bb2 Rd8?

13...Nxc3! 14.Bxc3 Nd5 puts Black in the driver's seat.

Probably White has to play 15 Bxd5, but it is not nice: 15.Bxd5 exd5 16.Nb5 c6 (16...Bxc3 17.Nxc3 c6 18.Qd4 isn't clear) 17.Bxg7 Qxg7 18.Nd4 Bh3 19.Re1 Qf6 20.f3 Rae8.

14.Qc1!

The move that saves the day, more than that as Black now has to face his structural difficulties.

14...Nxc3 15.Bxc3 e5 16.Nc2 Nd5 17.Bb2 Be6 18.Ne3 Nxe3 19.Qxe3 Bd5 20.Rad1

White has a small but obvious edge. Maybe this can't be converted against perfect defense, but who is up to the groveling  defensive task? Certainly not Mack, who seemed demoralized just looking at his game.

20...Bxg2 21.Kxg2 a5 22.Rxd8+ Rxd8 23.Rc1 Qd6 24.Qe4

Simply taking squares.

24...c6 25.Rc2

Waiting, maybe preparing Bc1 and Rd2. It's tough for Black to find a good move.

25...Qd5 26.Qxd5+ Rxd5

26...cxd5 27.Rc7 b6 28.Rc6 Rb8 29.e3! is strong for White.

His threat is Rd6.

27.Kf3 Kf7 28.e4 Rd7?!

I expected 28...Rd1 29.Ke2 Rh1 30.h4 Rb1 keeping Black's rook active, but after 31.Bc3 b6 32.Rd2 Ke6 33.f4 White is still quite a bit better.

29.Ke2 Ke6 30.Rc5! a4

There are no tricks after 30...Bf8: 31.Rxe5+ Kf7 32.Rxa5 Bb4 33.Ra4! Rd2+ 34.Ke3 Rxb2 35.Rxb4 and White wins.

31.bxa4 Rd8 32.a5 Ra8 33.Bc3 Kd6 34.Rc4 Kc7 35.a4 b6 36.axb6+ Kxb6 37.f4 Re8

Black's play must be seen against the background of serious time shortage.

38.g4 Bh8 39.f5 gxf5 40.exf5 Bf6 41.Bd2 Rg8 42.g5, 1-0. In my opinion, the idea of an early d4-d5 works very well here in the Fianchetto Variation. I shall try it again whenever I get the chance.

Now we proceed to an even earlier moment in the game where White unleashes his unexpected pawn advance.

Yucel - Ozdemir [E60]
Izmir, 2002
1.d4 Nf6 2.c4 g6 3.d5!?

Stops the Grunfeld, don't knock it! This will be a blow to many players. Surely 3.d5 cannot be good, but it is up to your opponent to show why. The pressure might well prove considerable.

3...c6

One way to chip at the pawn. For the moment, White is able to comfortably reinforce his center.

4.Nc3 cxd5 5.cxd5 d6 6.e4 Bg7 7.Nf3 0–0 8.h3!

I like this move very much, just taking g4 and preparing Nd4.

8...Nfd7 9.Be2 Nc5 10.0–0 Bxc3?!

The signs are that Black's judgment has been impaired by the arrogant d-pawn. He wants to punish White, but his intuition should surely have indicated that the dark-squared Bishop is worth far more than a measly pawn. I'm not sure what to suggest though; White is simply better.

11.bxc3 Nxe4 12.Qd4 Nf6 13.Bh6 Re8 14.Rab1

Open lines and all the dark squares on the board. More than adequate compensation.

14...Qc7 15.c4 Qc5 16.Qh4 Nbd7 17.Be3 Qc7 18.Bd4 Ne5 19.Ng5

Taking up a very threatening position with the idea  f2-f4,winning!

19...Bf5 20.Rb3! Ned7 21.Rc1 a6 22.Qh6! b5

22...Nf8 23.Bf3 Rab8 24.g4 Bc8 25.Ne4! N8d7 26.g5 should be the end of the game.

23.g4 e5 24.dxe6 Bxe6 25.Nxe6

25.Nxh7! was completely crushing.

It's a shame that he missed this finish.

25...Rxe6 26.cxb5 Qa5 27.Bf3

Of course this is also very strong.

27...d5 28.g5 Re1+

28...Qd2 29.Rd1 Re1+ 30.Rxe1 Qxe1+ 31.Kg2 Nh5 32.Re3 is clearly better for White.

29.Kg2 Ne8 30.Rxe1 Qxe1 31.Re3, 1-0. On the evidence of this game, 3...c6 doesn't look too hot. Doubtless there are many better ways for Black to handle it, but the d5-pawn is annoying.

To finish, one more from the master.

Petrosian (2640) - Hort (2600) [E64]
Wijk aan Zee, 1971
1.c4 g6 2.d4 Nf6 3.d5!? Bg7 4.Nc3 d6 5.g3

Here he shows a different way to the earlier Nf3, Be2. The fianchetto variation beckons, albeit in a somewhat different guise.

5...e5 6.Bg2 0–0 7.Nf3 c5

The Yugoslav Variation.

8.0–0 Nbd7?!

It's more coherent to play the Knight to a6 and then c7. Black then has two breaks to choose from (...b7-b5, ...f7-f5) instead of one. 8...Ne8!? was also possible, delaying a decision about the queenside pieces.

9.e4 h6

9...Ne8!? 10.Ne1 f5 11.Nd3 Nef6 12.f4! is another possibility.

10.Bd2 Ne8 11.Qe2 Kh7

11...f5 is a little premature. Black is opening the position whilst behind in development. It's no surprise that White ends up better after 12.Nh4! Qf6 13.exf5 gxf5 14.f4!

12.a4 b6

He could also have considered 12...f5, though after 13.exf5 gxf5 14.Nh4! Black's positional weaknesses are quite pronounced.

He can hardly play ...f5-f4, surrendering the light squares and if he doesn't, White will assume control with f2-f4 himself.

13.a5!

White already has a significant advantage.

13…Rb8

13...bxa5 14.Nd1.

14.axb6 Qxb6

14...axb6 15.Ra7! Rb7 16.Rfa1 is also very much in White's favor.

15.Rfb1 a6 16.Bh3!? Qb7

16...f5? 17.exf5 gxf5 18.Nh4 wins for White.

17.Nd1!

Slow and steady. White prepares the very effective regrouping Ne3 and then the pawn advance b2-b4! which drives Black into confusion.

17...Nb6

No better is 17...f5 18.exf5 gxf5 19.Nh4 Nb6 20.Ne3 f4 21.Qd3+ Kg8 22.Bxc8 Rxc8 23.Nef5.

18.Bxc8 Qxc8 19.Nh4! Nc7

Or 19...f5 20.exf5 gxf5 21.Ne3.

20.Ne3 Qh3

A token show of force.

21.b4!

Black's busted.

21…Nd7

Also winning for White is 21...cxb4 22.Bxb4 Rfd8 23.c5 (23.Ba5) 23...dxc5 24.Bxc5.

22.b5! axb5 23.Ra7 Rfc8

23...Na8 24.cxb5 Nab6 25.Rba1.

24.Ba5! Nb6

24...Na8 25.cxb5 Nab6 26.Bxb6! Rxb6 (26...Nxb6 27.Rxf7) 27.Nc4 wins.

25.cxb5

Beautifully played. Black is right out of good moves.

25...Qd7 26.Kg2! c4 27.Bxb6 Rxb6 28.Nxc4 Rxb5 29.Nb6 Rxb6 30.Rxb6 Bf6 31.Nf3 Be7 32.Rc6 Bd8 33.Nd2 Qe8 34.Nc4, 1-0. It's my belief that 3.d5 will have an unsettling effect on even the strongest opponents. Perhaps this game is enough to convince you but if not, I now include a few extra examples where White's idea succeeded.

Peev (2450) - Radev (2405) [E60]
Bulgaria, 1975
1.d4 Nf6 2.c4 g6 3.d5 b5 4.cxb5 a6 5.bxa6 c6 6.Nc3 cxd5 7.Nxd5 Qa5+ 8.Nc3 Bg7 9.Bd2 Nxa6 10.Nf3 Bb7 11.e3 0–0 12.Be2 Qf5 13.0–0 e5 14.Qb1 Qe6 15.b4 d5 16.Qb3 Rfd8 17.Rfd1 Qe7 18.Na4 Rab8 19.Rac1 d4 20.Bxa6 Bxa6 21.exd4 exd4 22.Nc5 Be2 23.Re1 Nd5 24.Bg5 f6 25.Nxd4 fxg5 26.Rxe2 Rxb4 27.Rxe7 Rxb3 28.Rxg7+, 1–0.

Kuzmin (2520) - Lputian (2550) [E60]
Kharkov, 1985
1.d4 Nf6 2.c4 g6 3.d5 b5 4.cxb5 a6 5.bxa6 Bg7 6.Nc3 0–0 7.e4 e6 8.Be2 Bxa6 9.dxe6 fxe6 10.Nf3 Bb7 11.Qc2 Na6 12.0–0 Nc5 13.Ng5 Qe7 14.Be3 h6 15.e5 hxg5 16.Bxg5 Qf7 17.exf6 Bxf6 18.Be3 Be7 19.b4 Na6 20.a3 d5 21.Na4 c6 22.f4 Bf6 23.Rad1 Nc7 24.Bc5 Nb5 25.Bxb5 cxb5 26.Bxf8 Rxa4 27.Bd6 Rxa3 28.Be5 Bd8 29.Bd4 Ra8 30.Qb2 Rc8 31.Rf3 Qf5 32.Rh3 Rc2 33.Qb3 Bb6 34.Rh8+ Kf7 35.Bxb6 Rxg2+ 36.Kxg2 d4+ 37.Kg3, 1–0.

Gheorghiu (2560) - Honfi (2390) [E60]
Baden-Baden, 1981
1.d4 Nf6 2.c4 g6 3.d5 c6 4.Nc3 cxd5 5.cxd5 d6 6.e4 Bg7 7.Nf3 0–0 8.Be2 b5 9.a3 a6 10.Be3 Nbd7 11.Nd4 Bb7 12.f3 Rc8 13.0–0 Ne5 14.Qd2 Nc4 15.Bxc4 Rxc4 16.b3 Rc8 17.a4 b4 18.Nce2 a5 19.Rac1 Qd7 20.Nb5 Ba6 21.Ned4 e6 22.Nc6 exd5 23.exd5 Rfe8 24.Bb6 Nh5 25.Bxa5 Bxb5 26.axb5 Bc3 27.Rxc3 bxc3 28.Bxc3 Rf8 29.Qd4 f6 30.b6 Ng7 31.Qb4 Nf5 32.Re1 Rce8 33.b7 Rb8 34.Nxb8 Rxb8 35.Bxf6 Rxb7 36.Qc4 Rc7 37.Qe4 Kf7 38.Bg5 Rb7 39.g4 Ng7 40.Qf4+, 1–0.

Sygulski (2425) - Zueger (2340) [E60]
Krefeld, 1983
1.d4 Nf6 2.c4 g6 3.d5 Bg7 4.Nc3 0–0 5.e4 d6 6.Nf3 Bg4 7.Be2 Nbd7 8.0–0 Nc5 9.Nd2 Bxe2 10.Qxe2 Nfd7 11.Nf3 Re8 12.Qc2 c6 13.Bd2 Rc8 14.Rac1 e6 15.dxc6 bxc6 16.b4 Na6 17.a3 Nb6 18.Qd3 Nc7 19.Bg5 Qd7 20.Rfd1 d5 21.c5 Nba8 22.e5 h6 23.Bh4 Rf8 24.Ne4 Ne8 25.Nf6+ Bxf6 26.exf6 g5 27.Qe3 gxh4 28.Qxh6 e5 29.Qg5+ Kh7 30.Nxe5, 1–0.

It's time to close our brief survey with the observation that an early d4-d5 hasn't really been seen very often on the modern master scene. Maybe it's the right moment for this space-gaining pawn push to be re-examined or even tried in your very own games. 3.d5 is worth playing!

Part 1 | Part 2