In this year's game from the 1994 U.S. Championship between former Champions Alexander Shabalov and Walter Browne, Browne demonstrated ingenious repertoire planning. By picking an off-beat line of the wild 6 Bg5 Najdorf, he invited the always obliging Shabalov into desperate complications. Moreover, by improving on the Black side on a game which Shabalov had previously lost as Black, he could be fairly confident that he would reach the desired position. The Najdorf again proves to be full of great "branching" opportunities for ambitious players.

Shabalov, Alexander - Browne, Walter, US Ch. 1994 (rd 11) 1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5 The reason this variation has lost some popularity for White is not that it isn't strong, but that Black has an incredible number of choices, and White must be heavily prepared for all of them. Now, aside from 6 ... e6, Black can steer the game into the Richter-Rauzer with 6 ... Nc6 7 Qd2 (White can vary with 7 Bxf6) e6, or play 6 ... Nbd7, but then 7 Bc4 has a good repuration. 6...e6 7.f4 DIAGRAM Now the big choices for Black are 7... Be7 (the Main Line), 7 ... b5 (the Polugaevsky), 7 ... Qb6 (the Poisoned Pawn), 7 ... Qc7 (a Kasparov favorite which prepares ... b7-b5) and 7 ... Nbd7 (generally transposes to the Main Line). Lesser known are 7 ... Nc6 (as in this game), 7 ... h6 (which can transpose to the game or to the Goteborg Variation after 8 Bh4 Be7 9 Qf3 g5!?) and the highly unusual 7 ... Bd7. 7...Nc6 8.e5 Along with 8 Nxc6, the critical move. Generally White can insert Nd4xc6 at any point in this series of moves; Black is not about to help White develop by playing ... Nc6 xd4 himself. Other 8th move alternatives for White are 8 Qd2, transposing to the Richter-Rauzer, and the excessively quiet 8 Be2 h6 9 Bh4 Qb6 10 Nb3 Qe3 11 Bxf6 gxf6 12 g3 (better is 12 Rf1) Be7 13 Qd3 with equality in Ljubojevic - Andersson, Stockholm 1980. 8...h6 9.Bh4 g5 In Ljubojevic - Ivanchuk, Tilburg 1989, Ivanchuk played the different idea 9 ... dxe5 10 Nxc6 Qxd1 11 Rxd1 bxc6 12 fxe5 Nd5 (Weak is 12 ... Nd7 13 Ne4) 13 Ne4 and White stood better after 13 ... Be7 14 Bf2!? (steadier are 14 Bxe7 or 14 Nd6+) Rb8 15 Bd4. In Brodskij - Rechel, Groningen Open 1993, Rechel tried Ivanchuk's recommendation 13 ... Rb8; after 14 b3! (not 14 c4? Rxb2! with advantage) DIAGRAM he followed it up with the wild 14 ... g5 - this enables Black to play Bf8-b4+ without allowing c2-c3 - 15 Bg3 Bb4+ 16 Ke2 f5?! 17 ef e5 18 Kf3! g4+ (18 ... Bg4+ 19 Kxg4 Ne3+ 20 Kf3 Nxd1 21 Bxe5 and White wins) 19 Kf2 Bf5 20 Bxe5 and White later consolidated and won. 10.fxg5 Nd5 DIAGRAM A question here is why not 11 Nxd5, as Black has to recapture with the e-pawn. If this is good for White, then it is a good argument for holding off on an eralier Nd4xc6, and it may theoretically force Black into the Ljubo-Ivanchuk line above. It turns out, however, that the opening of the e-file is not that impressive for White, and Black is quite active after 11 ... exd5 12 exd6 Bxd6 (or 12 ... Bg7?!), or 12 Bg3 Qxg5. 11.Nxc6 bxc6 12.Ne4 An interesting alternative here is 12 Nxd5 and if 12 ... cxd5 (12 ... exd5 is also possible, as in the last note) 13 Qg4 and 13 ... dxe5 14 g6 f5 15 Bd8! fxg4 Bf6 was good for White in Panajotov - Sandler, Albena 1989, but Black could have played 13 ... Be7. 12...Qb6 13.Bd3 A strange idea here is 13 c3, which has the virtue of blunting some of Black's counterplay and keeping the d-file open, but it does not seem White has time for such luxuries. 13...hxg5 A wrongly-timed simplification was 13 ... Qe3+ in Vitolins - Sandler, USSR 1989, which continued 14 Qe2 Qxe2+ 15 Bxe2 dxe5 16 g6! with a serious advantage. Also weak is 13 ... Qxb2 which lets White castle without first blocking the f-file with Bh4-f2: 14 O-O Qxe5 15 Qf3 is good.
14.Bf2

The theoretical move, from Vitolins - Sandler, USSR 1990, as expounded in Informant 50, Game 285. On 14 Bxg5 Qxb2 is OK now, as 15 O-O is met by 15 ... Qxe5 followed by f7-f5. 14...Qxb2 15.0-0 Qxe5 16.Bg3 Qd4+ 17.Kh1 Now the game Vitolins - Shabalov(!), USSR 1990, which appears as a note to the Vitolins - Sandler game in Inf. 50, continued 17 ... Ne3 18 Qf3 Nxf1 19 Rxf1 f5 20 c3! g4 21 Qe2 Qg7 22 Rxf5! with advantage. Did Browne prepare this line especially for Shabalov, with the following deviation in mind?

If so, this was truly masterful preparation. Or did Browne simply run into a minefield by accident, and find an ingenious escape? 17...f5!! 18.c3 Nxc3 19.Nxc3 Qxc3 20.Rc1 Qa5

The position is unclear. Black is up 3 pawns, and has a wall of center pawns covering his king, although his remaining pieces are completely undeveloped! Shabalov does not want to take the c6 pawn as this would open the long light-square diagonal. For example, 21 Rxc6 Bb7 22 Rc2 f4, and Black is gaining momentum. However, on Black's 20th, much better was 20 ... Qg7!, which likewise guards c7, and assists in defensive and offensive operations on the kingside. 21.Qe2 Be7 Not a good idea is 21 ... f4 22 Be1 because after 22 ... Qe5 23 Be4!,

23 ... d5 is impossible due to 24 Bg6+, and meanwhile 24 Bc3 is threatened. 22.Bxf5! Black was threatening to castle out of trouble. 22...exf5 23.Bxd6 Qd8 Black cannot afford the disorganization inherent in 23 ... Ra7 24 Rxc6 or 23 ... Rh7 24 Bxe7 Rxe7 25 Qh5+. 24.Rcd1 Vexing problems are posed by 24 Rxc6 on 24 ... Ra7 the creative 25 Qb2 does not achieve much after 25 ... O-O! as on 26 Bc7? Rxc7 27 Rg6+ Kf7 the king slips back out. But very annoying would be 25 Re1! which threatens to take on e7 because of the loose bishop on c8. If 25 ... Rh7, then 26 Qc4! hurts. Also 24 ... Bb7 fails to 25 Bxe7 Qxe7 26 Re6. But the amazing 24 ... Rh4!! may hold. One possibility then is 25 Re1 Re4 26 Qh5+ Kd7! and everything starts getting exchanged off. 24...Rh6!

Another nice defensive shot. If instead 24 ... Rh7 (to defend against the threatened 25 Be5) then 25 Qc4 would be most unpleasant. With the text Black envisions 25 Bf4 gxf4!! (otherwise the bishop returns to g5 with a vengeance) 26 Rxd8+ Kxd8 27 Rd1+ Rd6! and Black's huddled rook + 2 bishops against the queen give him the advantage. 25.Be5 Qa5 26.Bg7 Looking to create an entree for the queen on h5. 26...Rh7 27.Rfe1! A very nice idea - 27 ... Rxg7 28 Qh5+ Rf7 (28 ... Kf8 29 Qh8+ Rg8 30 Qh6+ Kf7 31 Qh7+ Rg7 32 Rxe7+ and White is winning;

in this line 30 ... Ke8 also fails to 31 Qh7) 29 Qh8+ Rf8 30 Rxe7+! Kxe7 31 Qg7+ with a terrible attack. 27...Ra7! 28.Bf6 g4??

Denying access to h5 and preserving the g-pawn, but there was no time (on the board) for this. Unfortunately, at this point, both players also had very little time on the clock. Much better was 28 ... Kf8 29 Bxg5 Bxg5 30 Qe8+ although White still has some attacking chances. 29.Qc4 Kf8 30.Bxe7+?? Probably winning for White was 30 Qxc6 as the immobility of the bishop on c8 finally tells; White would also get his queen into contact with f6 so as to threaten 31 Bxe7+ followed by 32 Qf6+ in most lines. After the text,

White never gets the opportunity to munch on c6. 30...Raxe7 31.Rf1 Qc7 Suddenly threatening mate! 32.Rf4 Rhf7 33.Rfd4 Rd7 White no longer has real compensation for the piece. 34.Qc1 Qe5 35.Qh6+ Kg8 36.Rxd7 Bxd7 37.Qg5+ Rg7
38.Qd8+ Kh7

The back rank mate holds the guy on d7. Now if 39 Qh4+ Kg6 and White's pieces have run out of space to roam. 39.h3 Qe7 40.Qc7 gxh3 41.Qh2 Qh4