paulmorphy

The following games were played at the Paul Morphy Open in New Orleans, Louisiana. Although I didn't have a good showing in the tournament, I did get to play some interesting games. All games were played with the standard CCA time control of 40 moves in two hours followed by one hour sudden death. In the first round, I found myself paired against an expert from the Florida panhandle:

Kamberi, S. (2339) - Frye, R. (2078)
Paul Morphy Open (1), January 5, 2001

1 Nf3 Nf6 2 c4 g6 3 Nc3 Bg7 4 e4 d6 5 d4 O-O 6 Be2 e5 7 O-O Nc6 8 d5 Ne7

Back in Michigan, we used to call this the "death variation," because the first player to deviate from existing theory usually dies. Although I have a good deal of experience playing the Black side of this position, this was my first serious tournament game with White.

9 b4

Less than ten years ago, this move was classified as a harmless sideline. Today, however (thanks to the introduction of some new ideas by Ivan Sokolov), 9 b4 strikes terror in the hearts of King's Indian players all over the world. Have you noticed how Kasparov hardly ever plays the King's Indian anymore and that Shirov recently said he doesn't think the King's Indian is fully correct? Alot of people are speculating that 9 b4 is the reason why.

9...a5

An important move, which I've played several times myself. In addition to 9...a5, there is also the main line 9...Nh5. Several years ago I played 9...Nh5 against senior master A. Chow, and the game got referenced in a theoretical article in Timman's magazine, New In Chess.

10 Ba3

This move is recommended for White by Burgess in Beating the Indian Defences (1997). White's main alternative, 10 ba, is generally reckoned to be less crucial. Personally, I've twice met 10 ba with Shirov's line 10...c5!? (10...RXa5 is more popular). After 11 dc (Gufeld recommends 11 Bd2) bc, expert Armin Burghart played 12 Qc2 against me in the 1999 Texas Championship and eventually lost. One year later, the Dallas master A. Dokoutchaev tried 12 Bg5 (J.C. Thompson Memorial, Dallas 2000) and that game was eventually drawn.

10...ab 11 BXb4 b6

11...Nd7 as played by Kasparov and Ivanchuk is the main line, but Frye's 11...b6 is thought to be an important alternative. The Latvian super-GM, Alexei Shirov, has played 11...b6 at least twice.

12 a4 Nd7 ?

This weak and illogical move allows White's initiative to develop freely. The theoretical move 12...Ne8 is considered stronger for several reasons:
(1)13 a5? is stopped because of 13...c5! 14 dc ba with an easy game for Black, a trap that Michigan's strongest player, Ben Finegold, fell into against Har Zvi, Groningen 1992.
(2)The c7 and d6-squares are better protected.
(3)The action of the f5-bishop is not blocked off.
Just think. Supposed you had played 11...Nd7 and White responded 12 a4 (as he usually does). White is sitting there trying to figure out how he's going to open the queenside. Would you go ahead and play 12...b6? just giving him an easy way to prise open the queenside with a4-a5? No way!

13 a5 ba

Alternatives also leave the initiative in White's hands. For example, the game Mikanovic-Desjardins, Montreal 2000 continued 13...Nc5 14 ab RXa1 15 QXa1 cb 16 Qa7 Bh6 17 BXc5 bc 18 Rb1 Qd7 19 QXd7 BXd7 20 Rb7 with more space and activity for White. I haven't been able to find any games with the agressive 13...f5, but that too should lead to an edge for White. For instance, 13...f5 14 Ng5! Nc5 15 ab RXa1 16 QXa1 cb 17 Qa7 and White is better due to his more active heavy pieces.

14 RXa5 RXa5

14...Rb8 was quickly demolished in Blanco Fernandez-Arribas, Havana 1993. That game finished: 15 Rb5 RXb5 16 NXb5 f5 17 Ng5 Nf6 18 Ba5 Ne8 19 c5 Bf6 (19...dc 20 d6 is clearly better for White) 20 cd BXg5 21 de BXe7 22 Qb3 Kh8 23 Rc1 Ba6 24 d6 NXd6 25 BXc7 Qd7 26 NXd6 BXd6 27 Qd5 Rc8 28 BXa6 RXc7 29 Rd1, 1:0.

15 BXa5

Black finds himself faced with a singularly unpleasant choice. The threat is 16 Nb5 and if 16...Nb6 17 NXc7! winning a pawn. And if Black lashes out with 15...f5 then 16 Ng5 leaves White much better.

15...Ba6

At least this lets Black hack off the knight if it comes to b5. I can't really recommend anything better.

16 Nb5 ?!

This is not bad-White gets a position with the two bishops and a backward c-pawn to shoot at. Still, there was an even stronger move, namely 16 Qa1!, threatening 17 BXc7. Then Black has no reasonable continuation, for example:
(A)16...Nc5 17 Bb4! and Black's position is seriously compromised.
(B)16...f5 17 Ng5! Nc5 18 Bb4 Bc8 19 BXc5 dc 20 Qa7:
(B1)20...NXd5 21 NXd5 QXg5 22 QXc7 Kh8 23 ef BXf5 24 QXc5 with a good extra pawn for White.
(B2)20...Nc6 21 dc QXg5 22 QXc5 and the extra pawn and more active pieces leave White much better off.
(C)16...Bh6 (preparing ...f5 by ruling out Nf3-g5) 17 BXc7 QXc7 18 QXa6 Nc5 19 Qa2 and Black hasn't got nearly enough compensation for the pawn.

16...BXb5 17 cb Nc5 ?!

After this Black's position becomes very passive. Black's most active defensive setup comes about when he achieves the ...f5 break with his bishop on h6 stopping the White rook from coming to c1. He could have achieved it with 17...Bh6! 18 Qc2 Nb6 19 Ra1 f5, although White still remains a little better off after 20 BXb6 cb 21 Ra7 fe 22 QXe4-the annoying rook on the seventh and the exposed Black king are more important than the weak pawn on d5.

18 Qc2 Bh6 ?

As usual, this is positionally well motivated, but here it meets with a tactical refuation. Also bad was 18...f5? 19 Ng5! fe 20 Ne6! NXe6 21 de since the c7-pawn drops off and the b-pawn will soon be decisive. So Black should have settled for 18...h6 19 Bb4 Nb7 20 Rc1 Qa8! (so as to meet 21 QXc7?? with 21...Rc8) 21 Nd2 f5 22 Bd3 with a worse position, but at least White would not have had any decisive blow and Black could struggle on.

19 NXe5 !

That's it. Black's position now falls to pieces. The game finished:

19...f5 20 ef RXf5 21 Nc6 ! NXc6 22 dc Bg7 23 Rb1 Ne6 24 Bg4 Nd4 25 Qa2+ Rf7 26 b6 d5 27 bc RXc7 28 BXc7 QXc7 29 QXd5+ Kh8 30 Qd7

Black Resigned.

After a quick draw in the second round with New Orleans master A. Carlin, I was paired against Louisiana's top player, J. Bick:

Kamberi, S. (2339) - Bick, J. (2259)
Paul Morphy Open (3), January 6, 2001

1 Nf3 Nf6 2 c4 e6 3 g3 Be7

A slightly offbeat move order, but there isn't any way to take advantage of it. Most games with 3...Be7 soon reach a standard Reti/Catalan position after the further moves 4 Bg2 O-O 5 O-O d5.

4 Bg2 O-O 5 O-O d5 6 d4

This move transposes to a heavily analyzed position from the Catalan opening. Instead, the equally popular 6 b3 would have given the most theoretically important position for the 2...e6 Reti. There are two popular replies to 6 b3:
(A)the ultratheoretical 6...c5
(B)6...b6, intending 7...Bb7 and 8...Nbd7 before playing ...c5. This line will always be well remembered by chess fans because it was Karpov's choice in the 24th and last game of his 1987 Seville match with Kasparov. Karpov only needed to draw that game to win the title, but he lost and the rest is history!

6...dc

Black had at least seven reasonable moves to choose from, but in my experience Black chooses this 6...dc line more often than all the others put together.

7 Qc2

This is considered the strongest way to recover the pawn. Now 7...b5? is weak because after 8 a4 c6? 9 ab cb? 10 Ng5!, White wins.

7...a6

Preparing ...b5. Other moves are now regarded as totally obsolete.

8 QXc4

8 a4 is given as the main line by ECO E (2nd ed., 1991). The most common continuation after 8 a4 is 8...Bd7 9 QXc4 Bc6 10 Bg5, when I believe White's chances are slightly better. At this time, theory has not yet come to a definite conclusion as to whether 8 a4 or 8 QXc4 is better. Slightly more players favor the 8 a4 line, but the 8 QXc4 line got a big endorsement when Kasparov used it in the fifteenth game of his London match with Kramnik.

8...b5 9 Qc2 Bb7 10 Bd2

The most common move-Kasparov himself played this way against Kramnik. Nevertheless, other moves also have their followers. IM John Donaldson recommends 10 Bf4, and Mikhail Tal used to play 10 Bg5.

10...Be4

The ECO E recommends this as best. According to ECO, the less commonly played alternatives, 10...Nc6 and 10...Nbd7 both lead to an edge for White.

11 Qc1 Nbd7

Again ECO recommends this as the only move to fully equalize, but in practice the top players have placed more faith in 11...Bb7 (for instance, Kramnik played 11...Bb7 against Kasparov). Whether they are of the opinion that the Encyclopedia understimated 11...Bb7 or that it overestimated 11...Nbd7 I do not know.

12 Ba5

Freeing the d2-square for a knight and making Black's thematic freeing move ...c5 harder to achieve.

12...Ra7 ?!

Here Bick finally deviates from the ECO, which analyzes 12...Rc8 as leading to an equal game for Black. No doubt Bick had made a very careful database study of this 12...Ra7?! move and concluded that Black gets an easy game. In the 12...Ra7?! database games that I found, Black got good positions by making very simple moves. 12...Ra7?!, however, has one major drawback-if Black can be stopped from advancing his c-pawn the rook on a7 will end up being very poorly placed. After analyzing 12...Ra7?!, I believe that White can accomplish the goal of restraining the c-pawn, and that 12...Ra7?! leads to play that is slightly better for White.

13 Nbd2

Completing development.

13...Bd5 14 Qc2

Connecting the rooks and threatening e4.

14...Qa8

It is important for Black to keep control of the e4-square.

15 b4 !

A strong novelty, which perhaps was not fully considered by Bick. In two previous games, White had played more weakly, allowing Black easy play:
(A)15 Nb3 Be4 16 Qc1 Rc8 (threatening ...c5 freeing Black's game) 17 Nc5 NXc5 18 dc (so far Cuevas Rodriguez-Korneev, Linares 1994) Nd7! and Black is better because it is hard for White to defend the c5-pawn. 19 b4 is met by 19...Bf6 winning the exchange and 19 Bb4 by 19...a5 driving off the defender of the c5-pawn.
(B)15 Rfc1 c5 16 dc BXc5 17 Nb3 Bb6 with equality, Tratar-Van Beers, Vejen 1993.
15 b4! is much stronger. Even though it leaves the a5-bishop out of play, this is more than compensated for by the fact that the rook on a7 will have a very hard time getting back into the game.

15...Nb6

White's last move did leave the c4-square a little weak, so naturally Bick rushes to occupy it.

16 Rac1

Control of this file is a key part of White's strategy.

16...Ne4

Not only threatening 17...NXd2 18 QXd2 Nc4, but also contemplating ...Nd6 followed by ...Nbc4.

17 BXb6 ?

I just don't know what got into me here. This move is positionally horrible, but I thought it was tactically justified beccause of my next move (which was nuts because my next move wasn't so strong, even if Black didn't have 18...NXg3!).
Much stronger was 17 Bh3, breaking the pin. Then Black would have had some problems to solve, for example, 17...Bf6 (putting some pressure on the d4-pawn) 18 Qb2! (guarding the pawn) Nd6 19 Ne5 BXe5 20 de Ndc4 21 Qd4 NXa5 (on 21...Rd8 I think the queen sacrifice 22 Rfd1! is better for White, e.g. 22...Bf3 23 QXd8+! QXd8 24 NXc4 Nd5 25 ef bc 26 f4 with better chances for White-Black's pieces are not very effective) 22 ba Nc4 23 NXc4 BXc4 24 Bg2 Qb8 25 Rc2 and White's better activity is more important than Black's structural advantage. Admittedly, this is just one variation, but I've looked at tons of lines and they all seem to share the common theme that White ends up better off due to the ineffectiveness of Black's queen and queen rook.

17...cb 18 Ne5 ??

This was the so-called "justification" of my last move. I thought that I was about to get the advantage after 18...NXd2 19 BXd5 QXd5 20 QXd2 because of the "weakness" of the c6 square. Unfortunately for me, I was wrong on two fronts. First of all, even in this 18...NXd2 19 BXd5 QXd5 20 QXd2 simply 20...Bd6 would leave Black better. The c6-square cannot be put to any good use, while Black can play ...a5 with the pleasant choice (after ba) between capturing on a5 with the rook to pressure the a2-pawn and capturing on a5 with the pawn to make a good queenside majority. And secondly, and far worse for me, I overlooked a devastating tactical resource at Bick's disposal.
Instead of going 18 Ne5??, I should have tried to make the best of a bad job with 18 Ne1, but after 18...NXd2 19 QXd2 a5, Black's pieces on the a-file come alive, with advantage to Black.

18...NXg3 !

This crushing move came as a total shock. I could have resigned here, but I staggered on a few more moves:

19 fg BXg2 20 Rf2 f6 21 RXg2 fe 22 Nf3 Qd5 23 Qc3 ed 24 QXd4 ?

24 NXd4 would hardly have changed the result, but at least it would have allowed me to play on a little longer.

24...QXd4 25 NXd4 Bg5 ! 26 Rc3 Rc7 !

White Resigned.

After that horrible loss, I was justly banished to the bottom boards. There I encountered my next opponent, the very strong Texas expert, M. Dejmek. Interestingly Dejmek and I had played before. Back in 1994, when I was an up and coming expert we played in the first round of the U.S. Amateur Team Midwest. In that game, Dejmek played the Black side of a c3 Sicilian and eventually ground me down. Now I had a chance to even the score:

Dejmek, M. (2172) - Kamberi, S. (2339)
Paul Morphy Open (4), January 7, 2001

1 e4 e5 2 Nf3 Nc6 3 Bb5 Nf6 4 O-O NXe4 5 Re1 Nd6 6 NXe5 Be7 7 Bd3 !

This has been know to be the most dangerous move since the nineteenth century. To see an example of the tame 7 Ba4 ?!, see my game against Galant from the Bill Lange Memorial. The other bishop move, 7 Bf1, can be met by 7...NXe5 8 RXe5 O-O 9 d4 Bf6 10 Re1 Nf5. Now White should probably try the relatively untested 11 d5!? d6 12 Nd2, since the standard 11 c3 is known to be ultra-drawish. Here are a couple of examples after 11 c3 d5 12 Bf4 c6 13 Nd2 Nh4:
(A)14 Nf3 Bg4 15 Be2 BXf3 16 BXf3 NXf3+ 17 QXf3, 1/2:1/2, Ivkov-Spassky, 1987.
(B)14 Bd3 Bf5 15 BXf5 NXf5 16 Nf3 Qd7 17 Qd2 Rae8 18 RXe8, 1/2:1/2, Kupreichik-Smyslov, 1981.

7...NXe5 8 RXe5 O-O 9 Re3

This move is a little illogical because moving the rook a second time with the entire queenside still in the box goes against basic principles. For this reason, almost all White players choose the simple developing move 9 Nc3. After 9 Nc3, theory considers model play for Black to be 9...Bf6 10 Re3 g6 11 b3 Bd4 12 Re2 b6 13 Ba3 c5 with equality, Sherwin-Bisguier, USA (Ch.) 1962/3.
As far as I know, Dejmek's 9 Re3 is a novelty. Even though 9 Re3 has a funny look and feel to it, Black can't reasonably expect to punish White because of it. The 5 Re1 variation of the Berlin defense is not the sort of line where Black has alot of counterchances!

9...f5

Of course I was thinking about stopping BXh7+ in some variations, but I was also hoping for a chance to use the f-pawn as an instrument of aggression later on. Naturally, 9...g6 and 9...Bg5 were also good.

10 Nc3 b6

Completing development and thus securing full equality. Dejmek's next move move is the only real attempt to mix things up, but it shouldn't secure any advantage against accurate play.

11 Nd5 Bg5 12 f4 Bh6 ?!

A very interesting move, but I don't think putting the bishop out of play like this is correct. Instead, just 12...Bb7! would have equalized. After 12...Bb7!, the desperado 13 NXc7? would have just led White into trouble after 13...BXf4 14 NXa8 BXe3+ 15 de QXa8 16 Bf1 Rf6 with a big initiative for Black. Therefore, White would probably have met 12...Bb7! with 13 fg BXd5 14 Qh5 with about level chances.

13 Ne7+

13 Qh5 ?! is weaker because White gets nowhere after the simple reply 13...Rf7.

13...Kh8 14 NXc8

This leads to a small edge, but I think Dejmek missed an even better opportunity with 14 Qh5!. I was planning to reply with the tricky 14...Bb7 (there isn't anything better for Black). Then White shouldn't take the f5-pawn because 15 NXf5? (15 BXf5?? Rf7! and Black wins a piece) NXf5 16 BXf5 Qf6 17 g4 Qd4 gives Black more than enough counterplay for the pawn. Instead, he should simply add more fuel to the fire with 15 b3!. A likely continuation would then be 15...Rf7 16 Bb2 RXe7 17 QXh6 Qf8 18 Rh3! Qg8 19 Qh5 when White's bishop pair and kingside pressure give him an edge.

14...RXc8 15 Qf3

White has to take a move to protect f4.

15...Qf6

Stopping b2-b3 and preparing to take over the e-file.

16 Rb1

It is very important for White to fianchetto his c1-bishop-otherwise he can't complete his development properly.

16...Rce8 17 b3 RXe3

On 17...Ne4 White must be careful to avoid 18 Bb2? because after 18...Qd6 he has double trouble at f4 and d2. Instead White should play the reinforcing move 18 g3! when the pair of bishops combined with the out of play bishop on h6 are slightly more important than Black's strong knight on e4 and good grip on the e-file.

18 de Re8 19 Bb2 Qe6 20 Re1 Ne4 ?

This impetuous leap should have put me in a world of hurt. It would have been much better to simply improve the position with the handy little move 20...c6. The ...c6 move would give me some extra options to seal off the a2-g8 diagnol with a timely ...d5 that I didn't have in the game. Of course, White's slight advantage would remain even after 20...c6 because Black has no active plan and must simply wait for White to try something. Still, Black's strong outpost on e4, the weakness on e3, and the fact that White would probably have to try g2-g4 sooner or later to win would all conspire to make a White victory hard to achieve.

21 Qh5 !

This strong move creates tons of overloading themes against my queen. For starters, there is the threat 22 BXe4 fe 23 f5, winning. I also have to worry about the idea of 22 g4 followed by 23 g5.

21...Qg6 ??

Another error which should have led to an easy victory for White. Although there were no really good moves, I should have just admitted my mistake and withdrawn my knight with 21...Nd6. After 22 Be5 Black would have a rather unpleasant position, but at least I could have fought on and tried to make White work to ovecome my defenses. By the way, I should just mention that 21...d5? would not have really solved my problems because of the sequence 22 c4! c6 23 g4! and there is no defense against the further advance of the g-pawn.

22 QXg6

Game Drawn by Mutual Agreement ??

After trading queens Dejmek offered a draw which I immediately accepted! This was an amazing stroke of good luck for me as all Dejmek had to do was to follow up the trade of queens with 23 h4!, and my h6-bishop turns into a big pawn. In that case, White's victory would have been guaranteed. Even so, I'm not sure that things turned out so badly for Dejmek-in the next round fortune smiled on him and rewarded him with a pairing against a C-player! I, on the other hand, was pitted in a rematch against the Oklahoma expert, S. Galant:

Kamberi, S. (2339) - Galant, S. (2084)
Paul Morphy Open (5), January 7, 2001

1 Nf3 g6 2 e4

Most 1 Nf3 players aim for a queen pawn opening when facing the Modern defense, but I like to try and throw my opponents off every once in a while.

2...Bg7 3 d4 d6

Galant chooses the most standard continuation, but there were two key alternatives. First of all, he could have made the game into a Sicilian with 3...c5. Arkansas's strongest player, R. Holyfield, likes to play this way. The idea is that after the natural 4 Nc3, Black can play 4...cd 5 NXd4 Nc6 with a better than usual Sicilian. And if 4 c4, trying to set up a sort of Maroczy bind, Black responds 4...Qa5+ and it is very hard for White to get an advantage as John Nunn points out in Beating the Sicilian 2 (1990). In grandmaster practice, the most common reply to 3...c5 has been 4 d5. Up until recently, this line has been thought to be quite good for White after 4...Nf6 5 Nc3 d6 6 Bb5+ (or 6 Be2). In the last few years, however, dynamic Black players have injected new life into this variation with the move 5...O-O!?. Not only does this rob White of the Bb5+ opportunity mentioned above but it also sets up the possibility of meeting 6 Be2 with 6...b5!?. The game Ziatdinov-Ehlvest, Las Vegas 1999 continued 7 e5 Ng4 8 Bf4 b4 9 Ne4 d6 10 ed ed 11 O-O Nf6! and Black had full equality. Yermolinsky recommends meeting 5...O-O!? with 6 e5 Ne8 7 h4 d6 8 e6 fe 9 h5, but this is also not so convincing because of the cold-blooded reply 9...gh!. After 10 Ng5 ed, White has nothing better than to give perpetual check with 11 NXh7 (Yermolinsky gives 11 NXd5 but his analysis is none too convincing) KXh7 (trying to win with 11...Rf5 is not worth it-Black would be more likely to lose than to win!) 12 QXh5+ Kg8 13 Qh7+ Kf7 14 Qh5+, etc. In the game DeVreugt-Malakhov, Lausanne 2000 White tried for more by playing 14 Bd3? and was soon in a lost position after 14...Rg8 15 NXd5 e6 16 Qh5+ Kf8.
If this 5...O-O!? line continues to hold up, White may start to turn his attention to Nunn's recommendation, 4 dc. After 4...Qa5+ 5 c3 QXc5 6 Be3 Qc7 7 Bd4 Nf6 8 e5 Ng4 9 Na3!, White preserves a modest advantage according to Nunn. Another interesting try for White is 4 c3, with the idea 4...cd 5 cd d5 6 ed!. After 6...Nf6 (6...QXd5 7 Nc3 Qa5 8 Qb3! also favors White) 7 Bb5+ Nbd7 8 d6 ed 9 Qe2+ Qe7 10 Bf4 QXe2+ 11 BXe2 White's better development gives him a slight edge.
Black's second third move alternative is Andrew Soltis's recommendation, 3...c6. 3...c6 prepares ...d5 with a very solid Caro-Kann like central pawn structure for Black. White has many possible replies (4 Bd3, 4 Be2, 4 c3, 4 c4, 4 h3, and 4 Nbd2 are all frequently seen in master play), the most popular continuation being 4 Nc3 d5 5 h3 Nf6 when both 6 e5 and 6 Bd3 lead to a very small edge for White.

4 Nc3

This so-called Two Knights variation is considered the most respectable. A favorite of Karpov for many years, it strikes a middle ground between the two main alternatives, the sharp 4 Bc4 and the quiet 4 c3.

4...Nf6

Most players consider this move automatic since it makes alot of sense to castle and see a little more of White's piece deployment before settling on a definite plan. The main alternative to 4...Nf6 is 4...Bg4, as IM Brooks played against me in the last round of the 2000 North American Open. Black's plan is to launch an immediate assault on the d4-square. After 4...Bg4 the most frequent continuation is 5 Be2 Nc6 6 Be3 e5 7 de de 8 QXd8+ RXd8 9 Nd5. In practice, White often manages to turn his lead in development into something more tangible, but my own opinion is that after the best move 9...Rc8! Black should soon be able to reach equality. For example, the game Gundersen-Gullaksen, Norway 1991 continued 10 O-O-O a6 11 h3 Bd7 12 Rd2 Nf6 13 NXf6+ BXf6 14 Bc4 Rd8 15 Rhd1 Bc8 16 RXd8+ and the players agreed to a draw. In my game with Brooks, I avoided this ending and went after the bishop pair instead, but also failed to gain an advantage after 8 O-O Nf6 9 h3 BXf3 10 BXf3 O-O. Black's control of d4 gives him sufficient counterplay to equalize the position-here are some sample lines:
(A)11 Nd5 NXd5 12 ed Nd4 13 c3? (13 BXd4 with equal chances is better) NXf3+ 14 QXf3 f5 15 c4 e4 16 Qe2 f4 17 Bc5 Re8 and Black's threatening pawn mass on the kingside gives him the better chances.
(B)11 Nb1!? (in our post mortem, IM Brooks suggested that this funny-looking knight move would have given me the advantage, but I don't see it) QXd1 (11...Nd4 immediately is also good enough to equalize) 12 RXd1 Nd4 13 BXd4 ed 14 c3 (and not 14 RXd4? Nd5! when Black wins material) dc 15 NXc3 and the endgame is even.

5 Be2

Naturally White had a very wide choice here. A couple of years ago, I was preparing one of my students to play this position as Black so we analyzed all the most important continuations-5 Be2 (considered the main line), 5 Bd3, 5 Bc4, 5 Be3, and 5 h3. In the process we found some really neat ideas, but I obviously can't share those here.

5...O-O 6 O-O

By transposition, we have reached one of the most important positions in the Pirc. Open up any book on the Pirc and you will see that the two main lines for Black are 6...c6 and 6...Bg4. Galant, however, has another idea.

6...c5

This is less usual than the two moves mentioned above, but it has been played by several good players, including Anand and Kramnik. Fortunately for me, I was fairly familiar with the idea since the Michigan master A. Lin frequently played it against me in speed chess. Those Lin games taught me that Black gets an easy game after 7 dc dc 8 QXd8 RXd8 9 Be3 b6 10 Rad1 Nc6 11 RXd8+ NXd8 12 Rd1 Bb7.

7 d5 !

Experience has shown this to be the best answer to 6...c5. 7 d5! transposes to the Schmid Benoni, an opening known to favor White at least slightly.

7...a6

It is unusual for Black to play ...a6 so early in this line, but noone has ever been able to prove that 7...a6 is weak. Much more common continuations are 7...Na6 and 7...e6.

8 a4 b6

Galant decides not to allow me to fix his queenside with a4-a5. Whether or not to play ...b6 after the mutual advance of a-pawns is a decision that Benoni players have to make on a regular basis. Of course, the decision will depend on the particular position, but my own observation has been that most Benoni players don't worry about this a4-a5 move, reasoning that after a later ...b5 they will end up a getting a useful half-open b-file. Still, there exists a sizable minority of Benoni players who have a strong aversion towards allowing the move a4-a5, and Galant seems to belong to this group!

9 Bf4

My plan for the next few moves can be summed up in one word-centralization! I will follow up by simply moving my queen to d2, and my rooks to d1 and e1. With my forces so well centralized, Black will constantly have to worry about the possibility of a central breakthrough with e4-e5, and this should limit Black's freedom of action to some extent. In the meantime, I will be able to use my space advantage to maneuver and try to create an opportunity for an advantageous pawn break.

9...Bb7 ?!

This is illogical in my view, since it makes ...b5 really hard to achieve. And without ...b5 where is Black's counterplay? Better, in my view, would have been 9...Nbd7 10 Qd2 Rb8 11 Rfe1 Ne8, followed by 12...Nc7 with only a slight advantage for White.

10 Qd2 Nbd7 11 Rad1 Qc7 12 Rfe1 Ng4

Galant sees the e4-e5 break coming, so he decides to try and trade his king knight for one of my pieces that support e5 (in this case the knight on f3).

13 h3 Nge5 14 NXe5 NXe5 15 Bg3

Preparing f2-f4 to evict the knight from its strong square on e5.

15...Rac8 16 f4 Nd7 ?

This makes a bad position even worse. Much better defensive chances would have been offered by the active 16...Bh6! as White would have needed to invest a couple of tempi to get out of the pin. Of course, White would have still been clearly better after 16...Bh6! since Black lacks counterplay, but at least Black could have posed the maximum amount of problems this way. After 16...Nd7?, on the other hand, White develops a truly alarming initiative in just a few more moves.

17 Bg4 Rcd8 18 Qe2 Nf6

Black can't really avoid the problems coming his way by stopping e4-e5 because after 18...f6 19 Be6+ Kh8 20 f5! g5 21 Qh5 White has a winning kingside attack.

19 Bf3 e6

Else there comes e4-e5 and Black is in trouble. But now how does White keep his initiative going?

20 de !

If I had followed this up correctly it would have been very strong.

20...fe 21 Qc4 ?

A tactical oversight which throws away all my advantage. I was looking at variations like 21...Qf7? 22 Qb3! when White wins the pawn on b6 or 21...Rfe8? 22 f5 with advantage to White. But I missed the strong reply 21...d5! 22 ed ed 23 NXd5? BXd5 24 BXd5+ NXd5 25 RXd5 Qf7 26 Red1 Bd4+ 27 R1Xd4 cd 28 RXd4 QXc4 with a better ending for Black.
Instead of 21 Qc4?, I should have kept a clear advantage by playing 21 f5!. Then the following variations all leave White with a nice edge:
(A)21...gf 22 ef BXf3 (22...e5 23 BXb7 QXb7 24 Bh4 is much better for White since after the exchange on f6 the White knight will be far superior to the Black bishop) 23 QXe6+ Kh8 24 BXd6 RXd6 25 QXd6 QXd6 26 RXd6 with a much better ending for White.
(B)21...e5 22 fg hg 23 Bh4 and White's better pawn structure gives a clear advantage.
(C)21...ef 22 ef BXf3 23 QXf3 gf 24 QXf5 with an advantage for White based on the weak square d5 and a safer king.

21...Kh8 ?

Missing his chance to get back on even terms (with 21...d5!). 21...Kh8? is an interesting move which has the advantage of leading to complications which are impossible to fully analyze over the board, but it turns out to be unsound.

22 Bh4 ?

To show the inadequacy of 21...Kh8?, I had to call his bluff. Despite a long think, I just couldn't summon up the courage to play the very strong 22 QXe6! because 22...c4 would have slammed the door shut on my queen. But 22...c4 could have been answered by 23 f5! with a winning position for White:
(A)23...Rfe8 24 BXd6 Qc6 25 e5 and White wins, as Black's queen has no square.
(B)23...Bc8 24 BXd6 and Black has nothing better than to transition to a lost ending by trading queens.
(C)23...b5 24 BXd6 Qb6+ 25 Kh1 Rfe8 26 Qf7 Rd7 27 Bc5! again with a winning position for White.
After the weaker 22 Bh4?, Black gets adequate counterplay.

22...Rde8 ?!

A natural move which unpins the f6-knight and guards the e6-pawn at the same time, but it would have allowed White to retain some edge with accurate play. Instead, the sharp 22...d5! was called for, as was pointed out by the GM staff of www.chessaid.com. After 23 ed ed 24 NXd5 RXd5! 25 BXf6 RXd1 26 BXg7+ KXg7! (but not 26...QXg7? 27 BXd1 QXb2 28 a5 ba 29 Re7! Qa1 30 QXc5! and the complications turn out in White's favor) Black holds the position.

23 Bg3 ?!

With Black now threatening ...d5 I really didn't want to leave the f4-pawn unprotected, and I also figured that with the d8-rook no longer guarding the pawn on d6, a bishop on g3 might eventually get to capture it. Despite these factors, however, it must be admitted that moving the bishop back and forth like this is not very logical, and is a sure sign that White is starting to drift.
The best move was the surprising 23 Rf1!. The idea is that 23...d5 is prevented because of 24 ed ed 25 NXd5 NXd5 26 BXd5 BXb2 27 c3! when Black's dark-squared bishop is shut out of play (note how 23 Rf1! made this line possible by providing extra protection to the f4-pawn). Having prevented ...d5, White can turn his attention to maneuvering to increase pressure on the enemy d-pawn, with a slight edge.

23...d5 ?!

Finally, but now this move is not so strong as it would have been on the previous two moves. The chessaid folks suggested instead 23...Nd7 24 Rf1 Qc6 with the idea of ...Bd4+ and ...e5, in which case Black would have got adequate counterplay.

24 ed ed 25 NXd5 RXe1+ 26 RXe1 NXd5 27 BXd5 BXd5 ?

Amazingly, this harmless looking exchange gives White fantastic piece coordination and puts Black in a lost position! The right play was pointed out by the chessaid staff: 27...b5 28 ab ab 29 Qe4 BXd5 30 QXd5 BXb2 31 Kh2. With the idea of 32 f5 White has some advantage, but it is not yet necessarily decisive. Another better move was 27...BXb2, although after 28 Kh2 the activity of White's pieces and the weakness of Black's queenside give White the advantage.

28 QXd5 BXb2 29 Re6 !

Black's king suddenly finds itself in great danger! Personally, I don't think Black can successfully defend any longer, and to make matters even worse Galant was down to just about 10 minutes to make it to move 40-not even close to enough time to navigate all the land mines awaiting him in this position!

29...Rd8 30 Qe4 Qd7 31 Kh2

Of course, 31 RXb6 is quite good as well, but given his time pressure I wanted to go for his king.

31...Bd4 32 Re7 QXa4 ?

This is the sort of time-trouble blunder I was waiting for! 32...Qc8 or 32...Qf5 would have lasted longer, but Black was lost in any case.

33 c3 Bg1+ 34 KXg1 QXe4 35 RXe4 b5 36 Bf2 Rd1+ 37 Kh2 Rc1 ? 38 BXc5

Black Resigned.