Game of the Month
>
> The Breakthrough
>
> Blocked positions are among the most fascinating, and frustrating,
> in chess. Considering a breakthrough sacrifice in these positions
> involves understanding the strength and mobility of the pieces hiding
> behind the pawn chains. The clearest type of breakthrough sacrifice is
> the destruction of a chain of pawns sheltering the king. In the 3rd
> round of the playoff match between Jan Timman and Predrag Nikolic for
> the Netherlands Championship, Nikolic first essayed a clearance pawn
> sacrifice to open a file. Timman reacted by munching the pawn and then
> hiding his king behind a new pawn chain. What he did not count on,
> however, was Nikolic's follow-up knight sacrifice which dissolved the
> White pawn covering, and set Black's pawns in motion.
>
> [Event "ch-NED m (playoff)"]
> [Site "Rotterdam"]
> [Date "1997.10.23"]
> [Round "03"]
> [White "Timman,J"]
> [Black "Nikolic,Pr"]
> [Result "0-1"]
> [ECO "A80"]
> [WhiteElo "2625"]
> [BlackElo "2630"]
>
> 1. Nf3 f5
>
> Predrag Nikolic is known as a very solid professional, and
> therefore the choice of the Dutch Defence in this important game is
> quite notable.
>
> 2. d4
>
> The move order 1 Nf3 f5 allows the Lisitsin Gambit, 2 e4!?,
> which I have had some success with. Then after 2 ... fxe4 3 Ng5 Nf6 4
> d3 e3! (4 ... exd3? 5 Bxd3 is too dangerous) 5 Bxe3 e5 6 d4! exd4 7
> Qxd4 White gets an initiative. In one game, after 3 Ng5, I was faced
> with 3 ... e5, with the idea 4 d3 e3 5 Bxe3 Nc6! and White cannot
> conveniently get d3-d4 in. Therefore, I reacted with 4 d4!!, getting
> decent compensation for the pawn after 4 ... e5xd4 5 Nxe4! Nc6 (not 5
> .. d5? 6 Qh5+) 6 Bc4.
>
> Nf6 3. g3 g6
>
> The Leningrad Dutch, where Black usually strives for ... d7-d6
> and ... e7-e5. But Timman plays essays an unusual response designed to
> force Black back into ... e7-e6 lines.
>
> 4. c3
>
> A strange idea of the Hungarian player Groszpeter, using an
> early queen sortie to be to interfere with Black's basic plan of
> development. But Groszpeter and others (including Nikolic! - see below)
> played it with the White bishop on g2 rather than the knight on f3, as
> then White still has the option of Ng1-h3.
>
> Bg7
>
> In Sorokin - Kramnik, USSR 1989 (with White having played
> Bf1-g2 rather than Ng1-f3), Black immediately clamped down on the light
> squares with 4 ... c6 5 Nd2 (Nikolic tried 5 Bg5 Bg7 6 Nd2 d5 7 Nh3
> against Ivanchuk at Manila, 1990) d5 6 Nh3 Be6?! but White was clearly
> better after 7 Nf3 Bg7 8 Nf4 Bg8 9 h4.
>
> 5. Qb3
>
> DIAGRAM
>
> e6
>
> A very calm approach. In Groszpeter - Knezevic, Maribor 1987,
> play developed (with White having played Bf1-g2 rather than Ng1-f3) 5
> .. d5 6 Nh3 (Lobac - Jagupov, Rossija 1992 featured 6 Nd2 Nc6!? 7 Ndf3
> O-O 8 Bf4 Ne4 9 h4 e6 with a reasonable game for Black) c6 7 Nd2 Qb6 8
> Nf3 Ng4?! 9 Nf4 e5 10 Nxe5 Nxe5 11 dxe5 Bxe5 12 Qc2! with a White edge
> due to the threat of h2-h4-h5.
>
> 6. Bg2 O-O 7. O-O Nc6 8. Bg5
>
> Timman decides to dump his dark-squared bishop. But the plan used
> in Sokolov - Malaniuk, Moscow 1994 may have been more effective: 8 Nbd2
> a5 9 a4 d5 10 Qc2 b6 11 b3 Ne4 12 Ba3 and the bishop finally takes up a
> useful post.
>
> h6 9. Bxf6 Qxf6 10. Nbd2 d6
>
> If 10 ... d5, White would be more comfortable with 11 e3 followed
> by c3-c4. Instead, Nikolic decides to build a dark-square pawn chain.
>
> 11. e4 f4 12. Rae1 Qf7 13. e5!
>
> DIAGRAM
>
> This push is the idea behind 8 Bg5 - Black's bishops are hemmed
> in. However, the Black kingside wedge with the pawn on f4 counts for
> something too.
>
> d5 14. Qc2 Bd7 15. Kh1 Ne7
>
> A maneuvering phase begins. Black would like to transfer his
> light-squared bishop to the g6-b1 diagonal.
>
> 16. Rg1 g5 17. g4
>
> The position is about even. The kingside is becoming blocked, and
> it is not clear who will have the better chances on the queenside.
>
> Qg6 18. Qd1
>
> Timman preserves the queens in the hope that his more flexible
> position will give him superior middlegame chances.
>
> Bb5 19. Bf1
>
> Nothing is to be achieved by 19 b3 Bd3.
>
> Bxf1 20. Rexf1 b6 21. h4 c5 22. Re1 Rac8
>
> DIAGRAM
>
> 23. Kg2
>
> Inviting Black intofrightful complications. 23 Nb3 would have
> preserved the status quo, as if 23 ... c4 24 Nbd2 b5 25 h5 the position
> is completely blocked.
>
> cxd4!? 24. Nxd4
>
> Black would have a clear advantage after 24 cxd4 Rc2. But now
> thestrong White knight on d4 will confer a definite advantage unless
> Black has a concrete followup.
>
> f3+!
>
> Not 24 ... gxh4 25 Ndf3! and White takes the kingside initiative
> as 25 ... Qxg4+? 26 Kh2 and Black drops the e6 pawn. But Nikolic's 24
> .. f3+ gives Black's rooks the very important square f4 for
> operations. Timman plays to close the kingside for good.
>
> 25. Kh3 Rf4 26. h5 Qf7 27. N2xf3
>
> Perhaps Timman should have tried 27 Rg3!, to take on f3 with the
> rook and exchange a pair of rooks to relieve the pressure. Then a
> strange line is 27 ... Nf5!? 28 Rxf3 (of course not 28 gxf5?? Rh4 mate
> or 28 ... Qxh5 mate!) Nxd4 (there is no strong follow-up after 28 ...
> Nh4 29 Rxf4 Qxf4 30 Qe2) 29 cxd4 Rc1!? (if 29 ... Rxf3+ 30 Nxf3 Rc1!?
> 31 Qe2!!) 30 Rxf4! (better than 30 Qxc1 Rxf3+ 31 Nxf3 Qxf3+ 32 Kh2
> Qxf2+ 33 Kh1 Qh4+) Rxd1 31 Rxf7 Rxe1 32 Rxa7 Re2 33 Nb3 and White is
> better due to the possibility of Ra7-e7.
>
> Rf8 28. Rg3 Re4
>
> Now Black has all the play on the f-file and White's pieces have
> become tangled up.
>
> 29. Qb1
>
> DIAGRAM
>
> Nf5!
>
> A destructive sacrifice breaking down White's protective pawn
> chain. For the knight Black gets a couple of pawns and a good chance of
> recovering material through pawn forks on White's jumbled pieces.
>
> 30. gxf5
>
> If 30 Rg2 Nxd4 31 Nxd4 Rxe5, the liberation of the Black bishop
> coupled with the awful pressure on the f-file combine to give Black a
> winning position.
>
> exf5 31. Kg2 f4 32. e6
>
> The immediate 32 Rg4 Qxh5 33 Nh2 f3+ is hopeless.
>
> Qxh5 33. e7
>
> But this further advance does not achieve anything, as the pawn
> was at least protected on e6. Best was 33 Rh3 Qg6 34 Kh2.
>
> Re8 34. Rh1?!
>
> Best was 34 Rh3. The plan to ensconce the rook on g4 is too
> unstable.
>
> Qg6 35. Rg4 R8xe7 36. Qd1 Bxd4 37. cxd4
>
> If the f3 knight leaves its post, ... h6-h5 snares the rook.
>
> Qf5 38. Rhh4 Rg7 39. Nxg5 hxg5
>
> If the h4 rook moves, 40 ... f3+ is the end.
>
> 0-1