On the Saturday of Thanksgiving Weekend 1998, I hopped into the fast schedule of the CCA's National Chess Congress in Philadelphia. In the penultimate round, I was paired with the talented former US Junior Champion Boris Kreiman, who seems to alternate between solid positional chess and wild attacking play. This time Kreiman teed off and, with a convenient pawn sacrifice, obtained a great position.

After somehow escaping by returning the pawn, I agreed to a draw in a still-crazy position. My idea was to bank everything on the last round instead, although, as it turned out, I then lost a grueling six-hour battle against IM Ziatdinov. Boris Kreiman - Michael Rohde, Philadelphia (National Chess Congress) 1997 B46 1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 Nc6

The Taimanov Variation. 5.Nc3 The only other real choice here is 5 Nb5 d6 6 c4 (or the sharp 6 Bf4), establishing a "Maroczy Bind" but losing time with the knight. The immediate 5 c4 Nf6 6 Nc3 Bb4 is thought to give Black fairly easy equality. 5...Qc7 I prefer this move order because on the straight Taimanov 5 ... a6, White gets a very comfortable attacking position with 6 Nxc6 bxc6 7 Bd3. 6.a3!? Kreiman is anxious to avoid lines involving ... Bb4. Now the task for me is to find a formation where White's a2-a3 is useless.

If instead White had played 6 g3, 6 Be2, 6 Be3 or 6 f4, my replly would have been 6 ... a6, a necessary precaution. But right now 6 Ndb5 Qb8 achieves nothing for White. 6...Nf6 7.Be3 Be7 8.Be2 On the spike 8 g4, a la the Keres Attack against the Scheveningen, 8 ... d5 is decent for Black, accepting an isolated central pawn, but leaving White with a kingside weakness. 8...d6 This was a good moment for 8 ... a6, as 9 g4 is still met by 9 .. d5. But this was also a decent time to finally transpose to the Scheveningen formation (with .. e6 and ... d6), as now 9 O-O Bd7 allows Black to play for ... Nc6xd4 and ... Bd7-c6 without ever bothering with ... a7-a6. 9.g4 During the game, I did not think this was that strong, for two reasons: (1) White has spent a tempo on a2-a3, and I thought I could find a line where that is irrelevant, and (2) White's bishop does not definitely belong on e2 in a Keres Attack. 9...a6 The alternative 9 ... h6 does not appeal, because, even though White's g4-g5 is held up a couple of moves, it comes with greater effect after 10 h4 a6 11 Rg1 with the idea g4-g5-g6. As White's bishop is on e2, Black cannot meet 11 Rg1 with the typical counter ... h6-h5. 10.g5 Nd7 11.h4 Na5 It seemed that only White would have good attacking prospects after 11 ... b5 12 Nxc6! Qxc6 13 Bd4 O-O 14 Qd2. Besides, I was still trying to prove that a2-a3 was a waste of time. 12.f4

12...Nb6?! Fixated on the c4 square. But Black should have gained space instead with 12 ... b5! and White's queen cannot find a good developing square - 13 Qd2 Nb6, or 13 Qd3 Nc5. If instead 13 O-O Bb7 14 f5 e5 15 Nf3 Nc4 Black has a good game. 13.Qd3 Nac4 Black would be too cramped on the waiting move 13 ... Bd7 14 b3! Rc8 15 Bd2 Nc6 16 Nf3. 14.Bc1 e5 This followup is mandatory. Else 15 b3 followed by Bc1-b2 with an overwhelming attack shaping up. 15.Nf5 Bxf5 16.exf5
exf4
 

17.Bxf4! An obviously strong pawn sacrifice which I underestimated when embarking on the mission to c4 with 12 ... Nb6. Kreiman simply lets the b-pawn go in return for a very strong presence in the center. Of course not 17 Nd5? Qc6 18 Bf3? Ne5 19 Nf6+ Kf8, and 17 O-O can be met by the wild 17 ... h6!? 17...Nxb2 Black's kingside would be destoryed on 17 ... O-O 18 f6!, and 17 Ne5 18 Qd4 is bleak. At least the text stops queenside castling by White, and softens up the knight on c3.

18.Qd4 Qc6 Probably best was the cold-blooded 18 ... O-O-O as if 19 Rb1 Qc6! 20 Rh3 (20 O-O d5! allows too much counterplay) N2a4. 19.Rh3

19...d5?? I had to play 19 ... O-O-O! with the possible continuation 20 Rb1 N2a4 21 Nxa4 Nxa4 22 Bf3 d5 23 f6 Bd6 24 Bxd5 Qxc2 with incalculable complications. 20.Qxg7?? Returning the favor. Black is simply flattened after 20 Re3 N2c4 (20 ... Kf8 is splattered by 21 f6! and a bishop check on h6 will end matters) 21 Bxc4 Nxc4 22 Re2 and the bishop on e7 is terminal. 20...0-0-0 21.Nb5?! I thought White still had the better of it on 21 Qxf7! Bd6 (to stop the trade of queens with Qf7-e6+) 22 Bxd6 Rxd6 23 Qg7! and the queen returns to the blockading square d4.

21...axb5 22.Qxb2 Nc4 23.Qb3 Rhe8
 

24.a4! Kreiman accompanied this move with a draw offer. Black seemed, for the first time, to be clearly better, but I only had 5 minutes left for the next 17 moves, while Boris had about 45 minutes. After 24 ... b4 Black would stand better because of his strong knight and White's airy king, but there would be nothing concrete. I spent 3 minutes analyzing 24 ... Qc5 25 O-O-O (the only response to the threatened ... Qc5-g1+; not 25 Kf1 Bd6) Qa3+ 26 Kb1 (White's bishop is hanging on 26 Qxa3 Bxa3+) bxa4 27 Qxa3 Nxa3+ 28 Ka2! and White is ok because of 28 ... Nc4 29 f6 Bd6 30 Bxc4.