| The following games were played at the Ken Smith Memorial in Dallas, Texas over the weekend of February 23-25, 2001. My performance in the tournament was neither good nor bad-I lost to everyone higher rated and beat everyone lower rated. Here are the games (the time control was 40/2, followed by SD/1): Kamberi, S. (2348) - Chalker, R. (1932) Ken Smith Memorial (1), February 23, 2001 1 Nf3 d5 2 c4 c6 3 e3 A tricky move order that I use to try and confuse Slav players. Although objectively 3 e3 is probably not as good a try for an advantage as 3 d4, its practitioners make up a veritable who's who of chess. Smyslov, Bronstein, Tal, Korchnoi, Larsen, Benko, Poulugaevsky, Gelfand, I. Sokolov, Speelman, Vaganian, M. Gurevich, Yermolinsky, Panno, Ivkov, Suba, Onischuk, Hodgson, Tukmakov, Malaniuk, Gulko, Ribli, Rohde, Chernin, Krogius, Damljanovic, and Gheorghiu have all included 3 e3 in their repertoires. You would think that any move that had such an impressive list of supporters would have been dealt with thoroughly by the theoretical works, but this is not the case at all. ECO A (1st ed., 1979), for instance, relegates 3 e3 to the following footnote: 3...Nf6 4 Nc3 Bg4 5 cd BXf3 6 QXf3 cd =. 3...Nf6 Most players play this as a matter of course, but there are some very interesting alternatives! For instance, the pinning move 3...Bg4 must be handled very carefully. The usual recipe against the early development of the c8-bishop, i.e. 4 cd cd 5 Qb3 leads to nothing after 5...Qc7! 6 Nc3 ?! (6 Bb5+ with an approximately equal position is better) BXf3 7 gf e6. In my opinion, this position is even slightly better for Black because of the poor White pawn structure and the ineffectiveness of White's bishops, particularly the one on c1. For instance, the game Rohde-Stalker, Harvard man-machine 1992 continued 8 d4 Nc6 9 Bd2 Nf6 10 Rc1 Rc8 11 f4 Bb4 12 Bg2 O-O 13 O-O Qe7 14 Rfd1 Na5 15 Qc2 Nc4 16 Be1 and now the simple 16...a6! followed by ...Rc7 and ...Rfc8 would have left Black with the better chances-he has a nice little queenside initiative while White's position doesn't offer anything comprable to crow about. Nor is 4 d4 anything worth playing for White-after 4...e6 5 Nc3 Nd7 Black was already fully equal in Bondarevsky-Botvinnik, Leningrad 1941 (note how this is just basically a queen's gambit declined where Black has already trivially solved what is usually his biggest problem-the development of the queen's bishop). At first I thought that White could get a tiny advantage with 4 h3, but I eventually discovered that Black has the very strong reply 4...Bh5!, a move which has never been tried in master practice. This powerful move deprives White of every last illusion, as 5 cd can be answered by 5...BXf3! (not 5...cd? 6 Qb3 Qc7 7 Bb5+ Nd7 8 BXd7+! QXd7 9 g4 Bg6 10 Ne5 with a very strong initiative for White) 6 QXf3 cd with easy equality for Black, as the White bishops lack scope. In view of the above analysis, I think White's best chance for advantage is Murey's move, the rarely played 4 Qb3!?. The idea is to answer either 4...Qb6 or 4...Qc7 with 5 Ne5, with perhaps a slight pull for White-this idea really needs some tests. In addition to 3...Bg4, 3...Bf5 is also very interesting. In practice, White's most common reply has been 4 cd cd 5 Qb3. When I asked the grandmaster staff of chessaid.com about this position, they said White was slightly better, but it seems to me that Black is equal after 5...Qc7. For example: (A)6 Nc3 e6 7 Bb5+ (7 d3 Nf6 8 e4 de 9 de Bg6 offers Black equal chances while 7 d4 Nf6 leads to a Slav position where it is well known that Black easily equalizes) Nc6 (naturally not 7...Nd7? 8 BXd7+! QXd7 9 Ne5 with a dangerous initiative for White) 8 Nd4 Bg6 9 Qa4 Nge7 10 b3 a6 11 Ba3 Rc8 12 Rc1 and now in Garcia Palermo-Meduna, 1990 Black had the pleasant choice between playing 12...Qd7 with safe equality or plunging into unclear complications with 12...ab. (B)6 Bb5+ (Alekhine's move) Bd7 (by now you should know that 6...Nd7? is bad because of 7 Nc3 e6 8 BXd7+! QXd7 9 Ne5) 7 Nc3 e6 8 BXd7+ NXd7 9 d4 Ngf6 10 Bd2 a6 11 O-O Bd6 12 Rfc1 Qb6. Pachman asseses this position, which was reached in Alekhine-Euwe (match 1937), as equal. 4 Nc3 e6 This move has been the most popular since the beginning of this line. For an example of 4...g6, which until recently was considered the main alternative to 4...e6, see my game against Umbarger from the Arkansas championship. In the past ten years, strong players have been playing the move 4...a6 more and more, so that today 4...a6 is played just about as often as 4...g6. It is interesting to note that the development of the light squared bishop is less effective now than without the moves ...Nf6 and Nc3 inserted. For instance, after 4...Bf5 (Matulovic's move) 5 cd NXd5 (5...cd 6 Qb3 Bc8 7 d4 was known to be difficult for Black even in the 19th century) 6 d4 White is slightly better. Also 4...Bg4 is not so easy to equalize with, e.g. 5 Qb3! (much stronger than the line given in ECO 5 cd BXf3 6 QXf3 cd with an easy game for the second player) Qb6 6 Ne5 Be6 7 d4 Nbd7 (Keres used 7...dc to make an easy draw in the game Hort-Keres, Petropolis Interzonal 1973. That game went 8 BXc4 BXc4 9 QXc4 e6 10 e4 Qa6! with easy play for Black, but with the better 10 O-O! White could have kept a slight advantage) 8 NXd7 BXd7 9 Na4 Qc7 10 Nc5 Bc8 11 cd NXd5 12 e4 Nb6 and White was better in the game M. Gurevich-Hauchard, Belfort 1998. After the game it was suggested that 12...Nf6 was an improvement for Black, but White should keep a slight edge with 13 e5 Nd5 14 Bc4 e6 15 Ne4. 5 d4 Nbd7 6 Bd3 dc The so-called Meran defense, named after the city where a famous tournament was played in 1924. In that tournament, Rubinstein introduced the idea of 6...dc followed by a rapid advance of his queenside pawns. Since Rubinstein's day many players of all levels have followed in his footsteps, as the Meran defense often leads to very complicated and interesting games. The bishop deployments, 6...Bd6, 6...Be7, and 6...Bb4 are all less popular but still common enough that they should be studied carefully: (A)6...Bd6 (this old move of Tchigorin's is by far the most common of the three bishop moves. Anand has used it to draw Piket, Akopian, and Bareev.) 7 e4 (the majority of GMs prefer this immediate central punch) de (theory considers 7...dc?! weak as after 8 BXc4 e5 9 de NXe5 10 NXe5 BXe5 11 QXd8+ KXd8 12 BXf7 BXc3+ 13 bc NXe4 14 O-O Black winds up in an unpleasant ending) 8 NXe4 NXe4 9 BXe4 e5! (9...O-O?! 10 O-O allows White a small edge according to Pachman, but personally I think he is underestimating Black's difficulties, as it is very hard to see to the defense of the Black king and even when Black solves this problem White's pieces almost always end up interacting much more successfully in the center than Black's. Look in any database and you will see Black scoring loss after loss in the 9...O-O?! line. One example from my own practice is Kamberi-Suarez, Dallas Action 2000: 10...Nf6 11 Bc2 c5 12 Qd3 h6 13 b3 cd 14 QXd4 Be7 15 Qh4 Nh7 16 Qg3 Bf6 17 Rb1 Kh8 18 Ne5 Qd4?? [a blunder in a bad position] 19 NXf7+! RXf7 20 Qg6 Kg8 21 QXh7+ Kf8 22 Ba3+, 1:0. After 9...O-O?! 10 O-O, the most popular line of play for Black is Pachman's recommendation 10...h6. The idea is to free Black's game with ...e5 without losing a pawn [10...e5 11 de NXe5 12 NXe5 BXe5 13 BXh7+! KXh7 14 Qh5+ Kg8 15 QXe5 and although the opposite colored bishops do give Black a decent shot at drawing nobody would consider this a decent result for the opening] After 10...h6 11 Bc2 e5, Pachman recommended 12 Re1, which was thought to lead to an edge for White until the famous game Kasparov-Deep Blue [Philadelphia 1996, 4th match game] which continued 12...ed 13 QXd4 Bc5 14 Qc3 a5 15 a3 Nf6 16 Be3 BXe3 17 RXe3 Bg4! [a valuable novelty-Pachman had only considered 17...Re8] 18 Ne5 Re8 19 Rae1 Be6 with equal chances for Black. Instead of 12 Re1, White does better with 12 Qd3! which I think does lead to a White edge. For example, 12...f5 13 c5 Bc7 14 de NXe5 15 Qb3+ Kh8 16 Re1 NXf3+ [the sacrifice 16...Be6? is unsound. After 17 QXe6 NXf3+ 18 gf Qh4 19 Qg6! QXh2+ 20 Kf1 Qh1+ 21 Ke2 Rae8+ 22 Be3 Qh2 23 Kf1 Qh1+ 24 Qg1 QXf3 25 Bd4 Qh3+ 26 Qg2 White was winning in Rahman-Suvrajit, Calcutta 1999.] 17 QXf3 Qf6 [17...f4 18 Qh5 Qg5 19 QXg5 hg 20 Bd2 also favored White in Gustafsson-Middelburg, Netherlands 1999 while 17...Qh4 18 h3 Qc4? runs into 19 Qh5! and Black is in trouble] 18 Bf4! BXf4 19 QXf4 QXb2 20 Bb3 and White had more than enough compensation for the pawn in Krasenkow-Chernikov, Katowice 1993.) Without the move 9...e5!, Tchigorin's variation would probably hardly ever be played anymore. After 9...e5!, play usually continues 10 O-O ed 11 QXd4 (Taimanov's idea 11 Bg5 leads to great complications. The best answer is probably 11...f6 [11...Qc7 12 QXd4 O-O 13 Rad1 was slightly better for White in Milov-Bisguier, New York Open 1997. Also 12...f6 would not have improved upon Bisguier's play because White can play 13 Rad1! Bc5 14 Bf4! BXd4 15 BXc7 BXb2 16 Bf5 when White has more than enough play for the pawn; 11...Nf6 12 Re1! O-O 13 QXd4 also gives White a slight pull; 11...Qb6 12 QXd4 QXd4 13 NXd4 leaves White a shade better due to his superior centralization.] 12 NXd4 fg! [This brave move seems like the best chance. 12...Ne5 13 Bf4 O-O was seen in Barczay-Paoli, Reggio Emilia when 14 Qc2! would have left White slightly better. 12...Nc5 13 Qh5+ Kf8 looks like a safer way to grab the piece, but surprisingly turns out to be weaker because the knight is sorely needed to block the e-file from e5. After 14 Rfe1! fg 15 Rad1, White has more than enough play for the piece.] 13 Qh5+ Ke7! [13...g6? 14 BXg6+ hg 15 QXh8+ Nf8 16 Rad1 gives White a winning attack] 14 QXg5+ [14 Rfe1?! is no longer so strong because of the reply 14...Ne5! and it is Black who is on top] Kf7 15 Bd5+ cd 16 QXd5+ Kg6 and I don't see anything better than perpetual for White.) Qf6 and now White usually plays 12 Rd1 with a slight edge, although there is also Pachman's recommendation 12 Qd1 which he says is better for White as well. (B)6...Be7 (Bogolyubov's system) 7 O-O O-O 8 b3 (According to Pachman, 8 e4 leads to equality only after 8...de 9 NXe4 b6 10 Qe2 Bb7 11 Rd1 Qc7 12 Bg5 c5) b6 9 Bb2 Bb7 10 Qe2 c5 11 Rad1 Qc7 12 Ne5 Rad8 13 f4 and White stands better according to Pachman. A trap Black can fall into is 13...Ne4? 14 Nb5! Qb8 15 NXd7 RXd7 16 dc NXc5 17 BXh7+ KXh7 18 Qh5+ Kg8 19 BXg7! with a winning attack for White, as in Junge-Kottnauer, Prague 1942. (C)6...Bb4 (the Romih variation) 7 a3 (according to theory this leads to an edge for White whereas Petrosian's plan 7 O-O O-O 8 Qc2 leads to equality only after 8...dc 9 BXc4 Bd6. Personally, I tend to think this may ultimately prove true, but it should be said that in this Petrosian line Black still has to play a few more careful moves before reaching full equality. After 9...Bd6, White's most common line of play is 10 e4 e5 11 de [11 d5 Nb6 12 Bb3 cd 13 ed Bg4 and Black had no problems in Oud-Flear, 1990] NXe5 12 NXe5 BXe5 13 h3. One example is Nielsen-Carlsson, Stockholm 1948 which continued 13...Qe7 14 Bd3 Re8 15 Ne2 Bc7 16 Bg5 h6 17 Bh4 Qe6 with approximately equal chances.) Ba5 (7...BXc3+ has been considered insufficient ever since the game Alekhine-Euwe, match 1937 which continued 8 bc O-O 9 O-O Qc7 10 Nd2 e5 11 Bb2 e4 12 Be2 b5 13 cb cb 14 a4 ba 15 c4 with advantage to White according to Pachman) 8 O-O O-O 9 Qc2 Bc7 10 b3 and White has the edge according to Horowitz. Note that the immediate 10...e5 would be well met by 11 cd cd 12 Nb5 Bb8 13 de NXe5 14 NXe5 BXe5 15 Bb2, when the isolated pawn position favors White. 7 BXc4 b5 8 Bd3 Bb7 Wade's move. One thing that has always struck me as curious is the way typical Meran books give the Old Meran, 8...a6, far more extensive coverage than 8...Bb7 despite the fact that the two moves are played equally often in practice. 8...b4, which Tartakower called the "Ultra-Meran", is played much more rarely even though it is not at all a bad move. After 9 Ne4 Black usually plays 9...Be7 and after 10 O-O he can at least play 10...Bb7, transposing back to normal Wade lines. It should be noted that after 8...b4 9 Ne4 c5?! is a little premature. For example, the game Elliott-Andersen, correspondence 1997 continued 10 Qa4 cd 11 NXf6+ gf 12 Be4 Rb8 13 NXd4 Qb6 14 O-O Bb7 15 BXb7 QXb7 16 Bd2 Rg8 17 f3 and White was better. 9 O-O This is the most common move, but 9 e4 remains a key alternative. After 9 e4 play normally continues 9...b4 10 Na4 c5 11 e5 Nd5 12 O-O with a slight edge to White. 9...a6 In the early nineties the line 9...b4 10 Ne4 Be7 was more popular, but today everybody plays 9...a6 and one has to wonder why 9...b4 is almost forgotten. In NIC Yearbook 51 (1999), Pelletier did an interesting survey on the 9...b4 10 Ne4 Be7 line, where he suggested that White's best try for advantage was 11 NXf6+ NXf6 12 e4 O-O 13 Qe2, arguing that White should be able to get an edge after 13...c5 14 dc Rc8 15 Rd1 Qc7 16 c6 BXc6 17 Nd4, Sadler-Ioseliani, World Team Championship, Lucerne 1997. Here in Dallas I once had to meet Ragozin's move, 9...Be7. The game Kamberi-Grancea, Dallas Action 2000 continued 10 e4 b4 and then I erred with 11 e5?! bc 12 ef, when Black could have obtained a very easy game with 12...BXf6 13 bc c5 14 Rb1 Bd5. Of course, 11 Na4 was much better, as after 11...c5 12 e5 Nd5 13 NXc5 NXc5 14 dc we reach a 9 e4 Wade variation where Black has lost a tempo. Here are some examples: (A)14...BXc5 15 Ng5 Be7 16 Qh5 BXg5 17 BXg5 and White's two bishops and development advantage gave him the edge in Taimanov-Ragozin, Leningrad 1957. (B)14...Rc8 15 Nd2 (the tempting 15...Nf4 also leaves White better after 16 Bb5+ Kf8 17 Nb3 NXg2 18 QXd8+ BXd8 19 c6 RXc6 20 BXc6 BXc6 21 Bd2. If Black weren't so far behind in development his play on the light squares might compensate for the loss of the exchange, but here, with Black's king rook still shut in White is definitely better.) BXc5 16 Ne4 O-O 17 Qh5 g6 18 Qh3 Be7 19 Bh6 Re8 20 Rad1 Qb6 21 Bg5 Bf8? (losing by force, but Black was in alot of trouble anyways) 22 Bb5! and White won in Sadler-Prie, Budapest 1993. 10 e4 c5 11 d5 ed ?! An old recipe of Sveshnikov's which was eventually discarded. The idea of ganging up on the d5-pawn and trying to win it is a reasonable one in principle, but in this particular position Black's poor development gives White too much counterplay for the pawn. In the past, there was considerable debate as to Black's best continuation. The two leading contenders were 11...Qc7, which Shirov enthusiastically supported and 11...c4 12 Bc2 Qc7 (in our post mortem Chalker recommended 12...Nc5?! as Black's best but then 13 Re1! leaves White better as it is hard for Black to develop given that 13...Be7? runs into 14 d6!. By the way trying to grab pawns with 12...b4 13 de fe 14 Na4 NXe4?! is pointless because after 15 Qe2 White will easily recover his pawn with the better game.), which Kramnik used to defeat Kasparov at Dos Hermanas 1996. But then there came the game Karpov-Anand, FIDE World Ch., Lausanne 1998 (1), which went 11...c4 12 Bc2 Qc7 13 de fe 14 Qe2 Bd6 15 Nd4 Nc5 16 f4 e5 17 NdXb5! (previously Black had been thought to be OK after 17 Nf5), and noone has been willing to try and defend this position since. That game continued 17...ab 18 NXb5 Qb6 19 NXd6+ QXd6 20 fe QXe5 21 Rf5 Qe7 22 QXc4 Rc8 (according to Kristensen who commented on this game at www.uschess.org 22...NcXe4 23 Bf4 Rc8 24 Ba4+ Kf8 25 Qd4 is very unpleasant for Black) 23 Qb5+ Ncd7 24 QXb7 RXc2 25 Bg5 Qd6 (an amateur would probably be sorely tempted by 25...QXe4? 26 QXe4 NXe4 27 Re1 Rc4, but Anand probably only needed about two seconds to see the refutation 28 b3 Rd4 29 Rf2! and White regains the piece with a winning position) 26 Qa8+ Kf7? (Kristensen gives 26...Nb8? as best, but this is not correct because after 27 e5! White goes over to a winning attack. Relatively the best chance was 26...Qb8 even though Kristensen asseses the position as unpleasant for Black after 27 QXb8+ NXb8 28 a4.) 27 QXh8 Qd4+ 28 Kh1 QXe4 29 Rf3 RXg2 30 KXg2 Ne5 31 QXg7+! (this must have been what Anand failed to spot when playing 26...Kf7?. If Karpov had played 31 Raf1? then the Indian grandmaster could have given perpetual after 31...NXf3 32 RXf3 Qe2+.) and White won. Of course, this victory by Karpov was very impressive, but it seems to me that this 11...c4 12 Bc2 Qc7 13 de fe line still poses some unanswered questions. For instance, after 14 Qe2 Bd6 15 Nd4 Black could try the wild piece sacrifice 15...BXh2+!? 16 Kh1 Nc5. Then 17 g3 Rd8 followed by 18...BXg3 gives Black enough play for a piece, so in the game J. Larsen- Van Orlschot, XIX World Correspondence Championship (1992) White tried instead 17 f4 and soon won after 17...O-O-O? 18 NdXb5! ab 19 NXb5. But after 17...e5 the position seems very unclear to me. In any event, ever since Karpov's famous win over Anand, the world elite have stopped discussing 11...c4 and have turned to 11...Qc7 12 de fe 13 Bc2 and now instead of 13...c4 which would transpose into Karpov-Anand, they are discussing 13...Bd6 14 Ng5 Nf8 15 f4 O-O-O (Illescas Cordoba-Shirov, Advanced Chess 2000). 12 ed NXd5 ? Chalker must have just miscalculated and thought that he was winning a pawn, as otherwise this move is just incomprehensible (it gives White a clear initiative for no material sacrifice). Although there was no longer a path to full equality, the size of White's advantage would have been less had Black adopted one of the following continuations: (A)12...Bd6 as in the game Kaidanov-Sveshnikov, 1987 which continued 13 Re1+ Kf8 14 b3. Now Black could have taken the d5-pawn with 14...Nb6 15 Bg5 NbXd5 16 NXd5 BXd5. Even though after 17 Bf5 White has more than enough for the pawn Black at least has reasonable chances for a draw in this line. (B)12...Be7 13 BXb5! (13 d6?! is weaker. After 13...BXd6, Piket, in his game with Van der Vliet, 1990 lethargically played 14 BXb5?! when after 14...O-O! Black would have even been slightly better as 15 QXd6 ab 16 NXb5 Ba6 17 Nc7 BXf1 18 NXa8 BXg2 favors him due to his safer king. Better than 14 BXb5?! would have been 14 Re1+! Be7 15 Qe2 with sufficient compensation for the pawn, but no more.) ab 14 d6 Ne5 15 NXe5 BXd6 (15...QXd6? 16 QXd6 BXd6 17 Re1 O-O 18 NXb5 and White went on to win in the game Farago-Sveshnikov, 1980) 16 Ng4 NXg4 17 QXg4 O-O 18 NXb5 Be5 with some compensation for the pawn, although White remains on top. While these lines are no bed of roses for Black, they are at least better than Chalker's 12...NXd5?, after which it is really hard for Black to save himself. 13 NXd5 BXd5 14 BXb5 ab 15 QXd5 Be7 16 Rd1 Ra6 ? Another mistake, as now it is very hard for Black's king to escape from the center. The last chance for survival was 16...Ra7. During the game I was looking at 17 Qf5? to stop castling, but this turns out to be a blank shot, because after 17...O-O! 18 Ne5 Black just plays 18...Qc8 and White cannot win a piece on d7 due to the weakness of his own back rank. The strongest reply to 16...Ra7 would have been 17 b4! when both 17...cb 18 Bb2 f6 (18...O-O?? 19 Qd4 and wins) 19 QXb5 O-O 20 Nd4 and 17...O-O 18 Be3 Qc7 19 bc NXc5 20 Ne5 leave Black in trouble. 17 Bf4 Qc8 18 Ng5 Rf6 18...BXg5 19 BXg5 Re6 20 Re1 O-O 21 RXe6 fe 22 QXe6+ is also winning for White. 19 Bg3 ? Overestimating my position, I made this lazy move and let Black back into the game. Much stronger was 19 Re1! after which all roads lead to victory for White: (A)19...RXf4 20 Qd6 O-O 21 QXf4 (B)19...Nf8 20 Bd6 Ne6 21 Ne4 Rg6 22 BXe7 KXe7 23 Qd6+ (C)19...Qc6 20 QXc6 RXc6 21 Rad1: (C1)21...h6 22 Bd6! RXd6 23 RXd6 hg 24 Ra6 (C2)21...Nf8 22 Bd6 Ne6 23 NXe6 RXd6 24 NXg7+ Kd7 25 RXe7+ KXe7 25 Nf5+ Ke6 26 RXd6+ KXf5 27 Rd5+ 19...Nb6 20 Qb3 b4 ? With this move, Black puts his head back onto the chopping block. 20...Qc6 21 Re1 h6 would have taken advantage of my weak 19th and allowed Black to retain some survival chances, but of course White's chances would still remain much better. But after 20...b4? the rest was trivial. White mopped up as follows: 21 Ne4 Re6 22 Nd6+ BXd6 23 BXd6 Qc6 24 Qg3 g5 25 QXg5 Black Resigned. After that win I was paired against a GM. I employed the so-called "Berlin Wall" as Black, and it held up impressively against the GM's best efforts. Unfortunately, I made a horrible mistake on move 26 and was ground down.. Kaminski, M. (2559) - Kamberi, S. (2348) Ken Smith Memorial (2), February 24, 2001 1 e4 e5 2 Nf3 Nc6 3 Bb5 Nf6 4 O-O NXe4 5 d4 Nd6 6 BXc6 Today, this is viewed as the only move promising good chances for an advantage. When the first edition of ECO C (1981) came out, the section editors (Keres and Olafsson) analyzed both 6 Bg5 and 6 de as leading to a small pull for White, but practice has not borne these assessements out. After 6 Bg5 Be7 7 BXe7 QXe7 8 BXc6 dc 9 de Nf5 10 Nc3 O-O 11 Qd2 Be6 12 Qf4 Rad8, Black has had no problems holding his own. And Black can answer 6 de with 6...NXb5 7 a4 d6 (but not Shamkovich's recommendation 7...Nd6?! because after 8 Bg5 f6 9 Re1 fg [9...Be7 is safer, but after 10 ed cd 11 Bf4 Ne5 12 Nd4 O-O 13 Nf5 White was better in Ivanovic-Z. Nikolic, 1981] 10 ed+ Kf7 [10...Ne7? 11 Ne5! killed Black in Wahltuch-Palmer, Manchester 1912] 11 Qd5+ Kg6 12 h4, White has more than enough play for the piece sacrificed. For instance, 12...BXd6? 13 g4! Bf4 14 Qf5+ Kh6 15 hg+ BXg5 16 Kg2 and White wins. Or 12...h6? 13 Qe4+ Kf7 14 hg BXd6 15 Qf5+ Kg8 16 g6 Qf6 17 Re8+ Bf8 18 Qd5+ and wins. Or 12...Qf6? 13 Qd3+ Qf5 14 h5+ Kf6 15 Qc3+ Kf7 16 Qb3+ Kf6 17 Nbd2 and Black cannot defend against the threats. Black's best is 12...gh! but even then I don't think Shamkovich would envy Black's position.) 8 e6 (after 8 ab NXe5 9 Re1 Be7 10 NXe5 de 11 QXd8+ KXd8 12 RXe5 Be6 the position was level in Rozentalis-Motwani, Hastings 1997) fe 9 ab Ne7 10 Nc3 Nf5 11 Nd4 NXd4 12 QXd4 Kf7 when Black had no problems in Ivanov-Yemelin, St. Petersburg Ch. 2001. 6...dc 7 de Nf5 8 QXd8+ 8 Qe2 Nd4 is well known to allow Black easy equality. 8...KXd8 9 Bf4 An interesting plan, which has hardly ever been used. The idea is to play h3, Bh2, Nd4, and f4, activating the kingside majority. 9...h6 Standard strategy, ruling out the use of the g5 square. 10 h3 Bd7 Kaminski criticized this move after the game, but I like it and would play it again. The idea is to evacuate the king to the queenside without allowing the light-squared bishop to be hit by Nf3-d4. 11 Nc3 Ne7 ?! To the unitiated, this undeveloping move looks downright nonsensical, but in fact it is a thematic and positionally well-motivated idea. It has three points: (1)To swivel the knight to a better square on c6 or g6.. (2)To give the light-squared bishop the option of moving to f5. (3)To get out of the way of a g2-g4 hit. Unfortunately for me, in this particular position 11...Ne7?! has one drawback that outweighs all of its positive points: it allows White to use his lead in development to work up an initiative on the d-file. The plan of ...Ne7, ...c5, ...Be6, and ...Nc6 was a good one, but Black should have first strengthened his position by 11...b6!. For example, 12 Rad1 Kc8 13 g4 (doubling on the d-file now poses no problems after 13 Rd3 Be6 14 Rfd1 Kb7 =) Ne7 14 Bg3 c5 15 Rfe1 Be6 16 Nh4 h5 17 f3? (17 Re4 =) Nc6 and Black was better in Lanka-Schlindwein, Cappelle la Grande 1995. 12 Rad1 Kc8 13 Bh2 ?! Logically carrying on with his plan, but now I get a chance to stabilize my position and equalize the chances. White could have got a small advantage with the more incisive 13 Rd2!. For example, 13 Rd2! Be6 14 Rfd1 Nd5 15 Nce2 with a slightly better position for White. 13...Ng6 I wanted to put my bishop on c6, but putting the knight there with 13...c5 followed by ...Nc6 was also good enough for equality. 14 Nd4 c5 With the idea of bringing the queenside forces into play by ...Bc6, ...b5, ...Kb7, and ...Rd8. I could have aimed to develop the kingside pieces instead with 14...Bc5?! 15 f4 Re8, but in that case White would have obtained somewhat better chances with 16 Ne4. 15 Nde2 15 Nb5 Be6 (but not 15...a6? 16 NXc7! and White is winning) gives Black no problems as the b5-knight will soon be put out of play by ...a6. 15...Bc6 16 Nd5 In notes sent to me after this game was played, Kaminski said this was a weak move which he only played because he failed to realize that 18 f5 would be met by 18...NXe5. But I can't find anything better. For instance, Kaminski's suggestion of 16 f4 would be met by 16...Nh4! 17 Nd5 Nf5 when the threat of 18...BXd5 is awkward to meet. After something like 18 Nec3 b5 I don't think Black would be worse. 16...b5 17 f4 Kb7 18 Nec3 18 f5 NXe5! = also gives White no advantage. And if White plays 18 b3, preparing to solidify the knight on d5 with c2-c4, then Black has 18...Rd8 19 c4 BXd5! 20 cd Ne7 21 d6 cd 22 RXd6 RXd6 23 ed Nd5 24 f5 Kc6 25 Rc1 BXd6 26 Nd4+ Kb7 27 NXb5 BXh2+ 28 KXh2 Kc6 with an equal ending. 18...Rd8 19 f5 Ne7 20 Ne3 Re8 ? This lame move hands the initiative to White. Instead, 20...Rd4! keeping the d-file was screaming to be played, after which the position would remain balanced. Kaminski disputed this arguing that after 21 f6 gf 22 ef Ng6 23 Nf5 he would have been better in any case, but I think he is overlooking the very strong reply 23...Nh4! which offers Black full compensation for the exchange after 24 NXd4 cd 25 Ne2 NXg2. 21 e6 ? This leads to a microscopic advantage at best. As Kaminski pointed out after the game, the slower 21 Rfe1 was much stronger, leaving White with a clear advantage due to his pawn majority, space advantage, and better coordinated forces. 21...Nc8 ! A good defensive move which minimizes White's edge by preventing the rook on f1 from working. 21...fe 22 fe b4 allows White to get some advantage with 23 Ne2! (23 Ncd5?! NXd5 24 NXd5 RXe6 25 Rf7 Bd6 26 RXg7 [26 NXc7 BXh2+ 27 KXh2 Re2 28 RXg7 Rh7!, drawn] BXh2+ 27 KXh2 Rd6 28 c4 bc 29 NXc3 RXd1 30 NXd1 Rd8 with an equal ending; 23 Nb1?! h5! also gives Black enough counterplay). 22 Ned5 White might try 22 Rd7!? as after 22...Re7 23 Rd8! (but not 23 Rfd1? fe 24 fe Nb6 25 RXe7 BXe7 26 Be5 Bg5 when Black's strong bishops leave him on top) White gets a very small advantage, but after 23...Nb6 I don't think White's winning chances would have been that great. 22...Bd6 23 ef Ref8 24 Ne4 RXf7 25 g4 Re8 26 Rfe1 BXd5 ? An absolutely horrible move, after which I doubt that Black's game can be saved. Much better, and good enough for a draw was 26...BXh2+ 27 KXh2 Kb8!. Then Black threatens 28...BXd5, followed by 29...Rfe7, winning. White has nothing better than 28 Kg3 Rd7 29 Nef6 RXd5 30 NXe8 RXd1 31 RXd1 BXe8 32 Rd8 Bf7, with a drawn position. 27 NXd6+ NXd6 28 RXe8 NXe8 29 RXd5 Kc6 30 Re5 Nd6 31 Kg2 Nc4 32 Re6+ Kd5 33 b3 Na3 ? This speeds up the loss a little by putting the knight out of play. Slightly more resistant would have been 33...Nd6. 34 c3 Nb1 35 Re5+ Kc6 36 Bg1 c4 37 Rc5+ Kd6 38 bc NXc3 39 cb NXa2 ? Losing a piece, but it makes no difference at this point. 40 Rc2 Black Resigned. In the third round, I notched a win against the solid NM L. Van Meter. That put me up against the promising Texas junior, A. Suarez, who had already knocked off two masters. He was no doubt looking to make me number three! Suarez, A. (2108) - Kamberi, S. (2348) Ken Smith Memorial (4), February 25, 2001 1 d4 Nf6 2 c4 g6 3 Nc3 Bg7 4 e4 d6 5 f3 O-O 6 Be3 This is still the most popular move, but 6 Bg5 is catching up. Formerly, 6 Bg5 would be met by 6...c5 7 d5 e6 with Benoni-like play, but in the 90s, the line 6...a6 became equally popular. The idea is that 7 Qd2 can be answered by 7...c5 8 d5 b5!? 9 cb Nbd7 with adequate compensation for the pawn and a complex game, Ivanchuk-Kasparov, Linares 1997. In my game with D. Hahn (Michigan Closed Championship, Ann Arbor 1998) White tried to stop the gambit with 7 a4 and after 7...Nc6 8 Nge2, I unleashed a novelty I had been saving up, 8...Nb4!?. That game continued 9 Qd2 c5 10 d5 e6 11 Nc1 (Dave could have gone pawn hunting with 11 de fe 12 O-O-O Qa5 13 QXd6 b5 14 QXc5, but after 14...Nd7 Black has sufficient compensation) ed?! 12 NXd5! NbXd5 13 cd when White's space advantage left him slightly better. But after the game D. Seawald pointed out that Black could have played stronger: 11...h6! 12 Be3 ed 13 cd (13 NXd5? NfXd5 14 cd f5 and Black is already better) Nd7 followed by ...f5 and Black is not worse. 6...Nc6 This move, the Panno variation, was the first thing Vasik Rajlich showed me when he introduced me to the King's Indian. Over the years, it has brought my opponents alot of grief. 7 Qd2 Normally White's seventh and eighth moves are Qd2 and Nge2-it hardly matters to me what order White plays them in-I just play 7...a6 and 8...Rb8 in either case. In Artz-Kamberi, Dallas Action 2001, my opponent threw me a curve by playing the line 7 Nge2 a6 8 a4. The most logical reply looks like 8...e5 9 d5 Nb4 with unclear chances, but not 9...Na5?! 10 Nc1 c5 11 g4! when the center is blocked and Black has no play anywhere. 7...a6 8 Nge2 Rb8 9 Nc1 A positional continuation which is currently the main line. Some of my opponents have given into greed and tried the pawn grab 9 Ng3 e5 10 d5 Nd4 11 BXd4?, which already allows Black the better chances. For example, my game against the strong midwest junior, R. Miller, Ohio team tourney 1997, continued 11...ed 12 Ne2 (after 12 QXd4? NXe4! White was already busted in E. King-Kamberi, Ann Arbor 1997) c5 13 dc bc 14 NXd4 Qb6 15 Rd1 Re8 16 Nb3 d5 17 c5 Qa7 18 Be2 de 19 fe Be6 20 Nd4 QXc5 21 Rc1 Qb4 22 QXb4 RXb4 23 NXc6 RXb2 and White's position had collapsed. The acid test of the Panno is supposed to be the aggressive 9 h4 h5 10 O-O-O b5 11 Bh6. In his excellent book, The Samisch King's Indian (1995), GM Joe Gallagher vehemently argues that this line is better for White, but in my opinion Black has sufficient counterchances. Here is my analysis: (A)11...bc!? (in my opinion this move, which has been unfairly condemned by theory, offers Black an adequate defense if handled properly) 12 g4?! (White usually resorts to this supposed "refutation", but it turns out to lead to complications favorable to Black. Instead White should play 12 BXg7 KXg7 13 Nf4 with about even chances) e5! (Giving Black the better game. At first I thought 12...hg? also favored Black, but some outstanding analysis by V. Rajlich has convinced me that this line favors White. The key line runs 13 h5 gf 14 BXg7? [given by Gallagher as good for White] KXg7 15 hg [15 Ng3 gives Black the standard counterattack 15...Ng4! 16 hg e5 with advantage] Rh8 16 RXh8 QXh8 17 Ng1 [17 Nf4? fg is much better for Black] Qh1, and while White does indeed have some swindling chances, it is hard to believe he has full compensation for being two pawns down. However, Rajlich came up with a big improvement on this line in the form of 14 hg!. Then both 14...fg 15 Nf4 and 14...fe 15 BXe2 fg 16 BXg7 KXg7 17 Qh6+ Kf7 18 BXc4+ e6 19 e5 favor White. Another line is 13...e5 as in the game J. Miles-Burlando, 1987 which continued 14 hg fg 15 BXg7 KXg7 16 Qh6+ Kf7 17 f4, but then instead of Burlando's 17...Bd7? which led to resignation after 18 fe de 19 Nf4! ef 20 BXc4+ Ke7 21 Qg7+, Black should play 17...Qe7!. I had a hard time finding anything better for White than taking a draw with 18 de de 19 Nd5 NXd5 20 Qh7+ Kf6 21 Qh4+, etc, but Rajlich pointed out a much stronger line here too with 19 f5! Rg8 [19...gf 20 Nf4 Ke8 21 Ncd5 also favors White] 20 Nf4 Qb4 21 fg+ Ke8 22 Rd2 ef 23 QXf4 with a vicious attack which more than compensates for the sacrificed material. Rajlich also found an improvement on White's play from the game Van Buskirk-J. Peters, Championship of Southern California 1997 which saw 13...BXh6!? 14 QXh6 g5! 15 QXg5+ Kh7 16 e5 gf 17 ef ef 18 Qf4 fe 19 BXe2 d5 20 Rdg1 Ne7 with an equal position which Black eventually even converted into a win. But White should have played 17 Ng1! with a clear advantage in all lines, e.g. 17...f2 18 BXc4 Ng4 19 Nh3). The game Rajlich-Schmaltz, New York Open 1998, continued 13 d5 Nb4 14 Ng3 c6? (After this silly move it is plain for all to see that Black's attack on the queenside will just be pocket fluff compared to White's strong kingside threats. The correct play was 14...hg! 15 h5 gf 16 BXg7 KXg7 17 BXc4 Ng4!. After this strong piece sacrifice which introduces the threat of ...Qf6-f4, Black is better. Rajlich tried to improve on this with 15 BXc4, but then Black obtains somewhat better chances with the same motif, 15...BXh6 16 QXh6 gf 17 a3 Ng4 18 Qd2 Qf6, etc.) and now if White had played the simple 15 gh! he would have been winning, e.g. 15...Qb6 16 BXg7 KXg7 17 Rh2! and Black will never make another threat, while White's kingside attack can proceed at leisure. (B)11...e5 (This central strike has been more played and analyzed than 11...bc. It should also offer Black adequate chances.) 12 Nd5?! (after the Palsson game below, Rajlich came to the conclusion that this move favors White, but in reality Black should get somewhat better chances) BXh6! (recommended by Gallagher. In Palsson-Rajlich, Toronto 1996 Black played instead 12...bc? 13 NXf6+ BXf6 14 d5 Nd4 [14...Nb4 15 Nc3 Re8 16 BXc4 a5 17 g4! also favored White in Lin Ta-Peng Xiaomin, China 1991] 15 NXd4 c3 when 16 QXc3 ed 17 Qd2 Re8 18 Bg5 would have left White on top due to White's attacking chances and the weakness of the pawn on d4.) 13 QXh6 bc 14 g4 BXg4! 15 NXf6+ QXf6 16 fg ed with more than enough counterattack for the piece. At first I thought White could advantageously prepare g2-g4 with 14 Ne3, but V. Rajlich pointed out that 14...ed 15 NXd4 NXd4 16 RXd4 c3! favors Black. 9...e5 10 Nb3 This is the most important alternative to the main line of 10 d5. Over the years, 10 Nb3 has found its way into the repertoires of Petrosian, Beliavsky, Van der Sterren, and Hjartarson. After 10 d5 Nd4 11 Nb3, I think Black should play 11...NXb3 12 ab c5. White's most popular reply to this has been 13 b4?! cb 14 Na4 b5 15 cb ab 16 QXb4 Ne8, but this has now been discredited. For example, 17 Nb6 (Watson gives 17 Nc3 Bh6 as being slightly better for Black) Bh6 18 BXh6 (18 Bf2 has been met in practice with the queen sacrifice 18...RXb6?! 19 Qa5 Ra6 20 QXd8 RXa1+ 21 Ke2 Rb1 when Black had sufficient compensation for the queen in Ibragimov-Delchev, Linares 1997. But 18...f5! with better chances for Black looks even stronger to me.) Qh4+ 19 g3 QXh6 20 NXc8 Qe3+ 21 Be2 RXc8 22 Qd2 Qb6 with slightly better chances for Black. In view of this, it seems White should prefer Karpov's move, 13 g4 when Gufeld recommends 13...h5 14 h3 Nh7 15 gh (otherwise 15...h4) Qh4+ 16 Qf2 QXf2+ 17 KXf2 gh with a position assessed as equal by ECO and as slightly better for White by Westerinen. In his book, Gallagher tries to argue that the pawn sacrifice 11...c5?! also offers adequate counterchances, but I think he is being a bit overoptimstic. After 12 dc bc 13 NXd4 ed 14 BXd4 the following all favor White: (A)14...Qa5?!. Traditionally, White has feared 15...c5 16 Be3 RXb2 17 QXb2 NXe4, and has therefore played 15 Rc1?! which does indeed allow Black some decent counterplay as Gallagher shows, but White has a much stronger idea in 15 Rd1! (now it will be exceedingly hard to get counterplay with ...d5). Dreev-Loginov, Novgorod 1999 continued 15...Be6 [15...c5 16 Be3 RXb2?? is now refuted by 17 QXb2 NXe4 18 fe BXc3+ 19 Bd2 with ruinous material losses for Black] 16 b3 c5 17 Be3 Ng4 18 Nd5 QXd2+ 19 BXd2 BXd5 20 cd Ne5 21 Kf2 when Black didn't have nearly enough for the pawn. (B)14...Re8 15 Be2 d5 16 cd cd (Gallagher mentions 16...Rb4?! as being worthy of investigation, presumably with the idea of 17 Bc5 NXe4 18 fe BXc3 19 QXc3 RbXe4 20 O-O RXe2, but after 21 dc White is of course better) 17 e5 Nd7 18 f4 f6 (18...Bh6 19 Qe3! also favored White in Wellman-Dirr, Dortmund 1993) 19 e6 Nf8 (Watson pointed out all the way back in 1982 that 19...RXe6 20 O-O leaves White with a positional advantage) 20 f5 gf 21 O-O and White is better. (C)14...d5 15 cd cd 16 e5 Nd7 17 f4 f6 18 e6 Re8 19 Be2 and we have transposed to (B) above where White is better. 10...ed 11 NXd4 NXd4 This move has the reputation of a safe equalizer, but in actuality it can lead to some real complications.. More popular, however, is 11...Ne5 12 Be2 and now I prefer 12...c6 since in the main line 12...c5 13 Nc2 Be6, I can't shake the feeling that White is a little better after 14 Na3!. After 12...c6, Black can answer 13 a4 with 13...d5 when he easily equalizes and 13 Rd1 with 13...b5 14 cb ab 15 b4 c5 when he gets enough sharp counterplay to compensate for material he may have to sacrifice (Beliavsky-Kasparov, Linares 1990). 12 BXd4 Be6 13 Be2 c6 Black's counterplay in this line is based on the thrusts ...b5 and ...d5. 14 a4 Stopping ...b5. Just as often White allows it with the line 14 O-O b5 15 cb ab but this also leads to approximately equal chances. For instance, in the latest game Nicoara Etchegaray-Shah, Womens' Olympiad, Yerevan 1996 White played 16 Be3?! and now Black could have seized the initiative with 16...d5!. The idea is 17 e5 Nd7 18 f4 f6 19 ef QXf6 20 Rad1 Nb6 when Black, with ideas of ...b4 or ...Nc4 is better. Or 17 ed NXd5 18 NXd5 QXd5 and Black is better because of his more active pieces. 14...Qa5 By far the most popular move in this position. In his 1982 monograph, Watson suggested that Black could equalize with the line 14...d5 15 cd (15 ed cd 16 c5 is mentioned by Gallagher as being worthy of investigation, but in that case Black could play 16...Qa5 17 O-O Rfd8 [threatening ...Ne4] 18 Rfd1 Ne8 19 BXg7 NXg7 20 Ne4 QXd2 21 Nf6+ Kh8 22 RXd2 a5 and Black should draw the ending without too much difficulty) cd 16 e5 Nd7 17 f4 f6. Gallagher and Gufeld concur with this, citing the game Beliavsky-Nunn, Reykjavik 1990 which continued 18 ef NXf6 19 Rd1 Qd7?! 20 O-O Rbd8 21 Bb6?! when Nunn unleashed the brilliant 21...d4!!. According to Gallagher 22 BXd8 gives Black quite sufficient compensation for an exchange, so Beliavsky played 22 QXd4 QXd4+ 23 BXd4 Bb3 24 Rd3 (24 Bc5 BXd1 25 Bc4+ Kh8 26 BXf8 RXf8 27 Rd1 Ng4 gives Black sufficient counterplay for a pawn) Bc4 25 Rd2, when he then analyzes 25...BXe2 26 NXe2 Ne4 27 Rdd1 BXd4+ 28 NXd4 Nc5 29 a5 Rd5 30 b4 Rfd8 31 bc RXd4 32 RXd4 RXd4 33 Rb1 Rc4 as drawn. Very impressive, but I don't find Nunn's play totally convincing. Much stronger would have been 21 Bf3 with an obvious positional advantage for White. Still, things are not all gloom and doom for this line-Nunn himself points out an improvement in the form of 19...b5! with equal chances for Black. GM Blehm gives the line 20 ab ab 21 BXb5 Bg4 22 Be2 BXe2 23 QXe2 Re8 24 Be5 Qb6 with good compensation. 15 O-O Qb4 ?! Stirring up a hornet's nest of complications. In Wisskirchen-Oechsleln, 1989 Black played more accurately and managed to solve his problems after 15...d5 16 cd [16 ed cd 17 c5 lets Black steer the game towards a draw as shown above] cd 17 e5 Nd7 18 f4 Nc5 when the position was balanced. 16 a5 ?! This allows a clever tactical response. Instead, White should have kept the tension with 16 Rfc1! Rfd8 (16...d5 17 e5 also gives White a slight advantage) 17 Kh1!. In that case, he would have kept a slight advantage as Black has no good breaks and his d-pawn remains a weakness. 16...c5 ?! This sets a devilish trap which Suarez falls into with both feet. Objectively stronger, however, was 16...b5! as was played in Huschenbedt-Laske, 1985. The idea is that after 17 ab c5 18 BXf6 BXf6 19 RXa6 Bc8 20 Ra2 Bd4+ 21 Kh1 RXb6 Black gets sufficient compensation for the pawn based on his dark-square control and the fact that White's extra pawn on b2 is a weakling that is very hard to mobilize. 17 Ra4 ? With an incredulous look on his face, Suarez played this blunder fairly quickly. S. Anderson, commenting in Texas Knights (May-June 2001), pointed out that 17 Be3 would have been better. After 17 Be3 Anderson makes two suggestions both of which should favor White somewhat: (A)17...Nd7 18 Qc1! (threatening 19 Ra4 trapping the queen) b5 19 cb ab 20 Rd1 and the passed a-pawn and weak pawn on d6 give White a clear advantage. (B)17...BXc4!? is suggested as being unclear by Anderson, but White should have a few ways to get the advantage here. For example, the simple 18 BXc4 QXc4 19 QXd6 must favor White a little since Black has a target on c5 and a weak square on d5. 18 Ra4 also looks favorable for White, e.g. 18...NXe4 19 fe BXc3 20 Qc2! Bb3 21 RXb4 BXc2 22 Rc4 BXe4 23 RXc3 and although Black has reasonable chances to draw this ending White must be better here. Since I can't make any suggestions for Black better than the moves proposed by Anderson, I must conclude that White would have been better after 17 Be3. 17...NXe4 ! This is the pretty trap that had lured me into playing 16...c5?!. White's position falls to pieces. 18 fe Everything loses. 18 NXe4 QXd2 19 NXd2 BXd4+ 20 Kh1 BXb2 leaves Black two pawns up with a winning position. And 18 Qc2 BXd4+ 19 Kh1 Nf2+ 20 RXf2 QXa4 21 NXa4 BXf2 22 Qd2 Bd7! (if White managed to get in Nb6 and Bd1-a4 he could put up some resistance) is a fairly easy technical exercise for Black since his rooks will invade decisively on the e-line. 18...BXd4+ 19 QXd4 QXa4 20 QXd6 QXa5 21 Qf4 Qb6 22 e5 Since White was totally lost, he figured he might as well try this desperate last attempt for an attack. 22...QXb2 23 Rf3 Rbd8 24 Ne4 QXe2 25 Nf6+ Kh8 26 h3 Rd2 27 Rg3 Rd1+ 28 Kh2 Qf1 29 Rf3 Qh1+ 30 Kg3 Rg1 White Resigned. |