2001 Texas State Championship

As usual, this was the strongest tournament of the year-I got to play 4 strong masters rated over 2300 and three tough experts as well. I didn't have a particularly good result, but the games were nevertheless very interesting.

Sarkisian, D. (2192) - Kamberi, S. (2347)
Texas Championship 2001, May 25, 2001

1 d4 Nf6 2 Nf3 g6 3 c4 Bg7 4 Nc3 O-O

Against this move order I always play this instead of 4...d6 because it lets me play lines I like such as 5 Bg5 c5 6 e3 cd 7 ed d5 and 5 Bf4 d5 6 cd NXd5 7 NXd5 QXd5 8 BXc7 Nc6. With most opponents, however, it all winds up as a transposition anyways.

5 e4 d6 6 Be2 e5 7 O-O Nc6 8 d5 Ne7 9 Bg5

Position after 9 Bg5

I have not scored very well against this tricky line, which is a favorite of the Dallas master, Jeff Ashton. One advantage of 9 Bg5 is that standard King's Indian hack attacks don't work so well in this line.

9...Nh5

More common is 9...h6 10 BXf6 (10 Bh4 g5 11 Bg3 NXe4!? 12 NXe4 f5 13 NeXg5 [13 Nc3 f4 14 Nd2 fg 15 hg Nf5 followed by ...c6 would leave the chances approximately equal] hg 14 NXg5 f4 15 Bh4 Nf5 16 Bd3 [16 g3?! Bf6 17 Ne6 BXe6 18 BXf6 QXf6 19 de QXe6 and Black's attacking chances on the kingside give him the advantage] Qe8 17 BXf5 BXf5 and Black, with moves like ...Qg6 and ...Bf6 to take advantage of the awkward placing of White's pieces, has adequate compensation for the pawn) BXf6 11 b4 and now:
(A)11...Bg7?! (the most natural but it seems inadequate) 12 c5 f5 13 Nd2 Bd7 (certainly not a hack attack with 13...f4? 14 Rc1 g5? 15 cd cd 16 Nb5 with a big advantage for White. In this last line, 14...Bd7 would be stronger, but the advantage would still lie with White after 15 a4 with ideas of Nd2-c4 and Nc3-b5) 14 Nc4 Rf6 15 a4, Duhayon-Slahaan, Belgium 1997. White is better because Black's kingside attack lacks its usual potency, there being no real way for it to get off the ground.
(B)11...Kg7! (the beginning of an excellent reorganization whereby the knight will go to g8 and the bishop to e7) 12 c5 Ng8 13 a4 Be7 14 c6 bc 15 dc f5 and Burgess evaluates the situation as unclear.

10 Ne1 Nf4 11 Nd3 NXe2+ 12 QXe2 h6 13 Bd2 f5 14 f3

A novelty. The main line 14 f4 ef?! 15 NXf4 g5 16 Nh5 Bd4+ 17 Kh1 f4 18 g3 fg (18...Bh3 19 gf! is given as good for White by Burgess) 19 RXf8+ QXf8 got an important reinforcement recently from the game P.Lukacs-Rajlich, Hungary 2001 where White found 20 Nb5!. Then 20...BXb2? 21 Rf1 Qd8 22 Qf3 Bh3 23 Qf7+ Kh8 24 Rf6! would be a slaughter, so Rajlich played 20...Qf2 but after 21 QXf2 BXf2 22 NXc7 Bh3 23 hg Rc8 he would have been in trouble had Lukacs played the very strong 24 Nb5! instead of the weaker 24 Kh2?!. The idea is that after 24 Nb5! RXc4 White plays 25 Bc3! Nc8 26 Kh2 and Black is forced to submit to the awful 26...g4 leaving his light-squared bishop in a sorry-looking state. White can then play something like 27 Rd1 when Black will have to worry about the e4-e5 breakthrough as well as the state of his bishop on h3.
Black could try to improve on Vas's play with 23...Rf8 with the threat of ...Bg4. For example, 24 Bc3? Bb6! when 25 Nf6+ RXf6! 26 BXf6 BXc7 27 BXe7 Kf7 is a nasty piece trap which would leave Black at least equal. But 24 Bb4! would leave White with the advantage. Since 14...ef?! is currently in trouble, I would recommend playing the less popular 14...fe 15 NXe4 Nf5 16 fe Nd4!. Here, I would have to concur with Burgess' evaluation of =.

14...Bd7

Black has two principal plans at his disposal here:
(1)Playing ...f4 followed by ...g5, ...h5, and ...g4 in the spirit of a standard King's Indian hack attack. White, in turn, can respond to this with either:
(a)Just advancing his pawns on the queenside, for instance 14...f4 15 c5 g5 16 a4 Rf6 17 b4, intending b4-b5 and c5-c6. In this case, there would ensue a very sharp game with about equal chances.
(b)Meeting ...f5-f4 with g2-g4!? which results in a pawn structure mainly seen in the Benko/Pinter system of the Classical King's Indian. Surprisingly, after dozens of hours of analysis I have ended up concluding that this plan brings about very drawish positions. I say this is surprising because Benko/Pinter structures normally don't have a reputation for drawishness. The difference here is that Black's king knight has been exchanged for White's king bishop. This exchange makes it considerably easier for White to defend his kingside. White can set up with pawns on h3, g4, and f3, a rook on h2, and a knight on f2. Then when Black plays ...h5 and ...h5Xg4, White recaptures with f3Xg4 and Black does not have enough pieces remaining for a good sacrifice on h3 or g4 which he would need to breakthrough. Since White can easily defend against Black's kingside attack, I therefore at first assumed that a plan with ...f4 was bad, but this is not the case either. What ends up happening is that White is so preoccupied defending on the kingside that he can never get any good queenside play going like he normally can in a Benko/Pinter. Therefore, with correct play neither side should ever get too much of an initiative and the position should remain drawish.
(2)Black can also play in the center and queenside with 14...c6!?. This also leads to about equal chances, but has the added virtue of avoiding the drawish position type mentioned in (1)(b).
With my 14...Bd7 I was intending to adopt the first plan. As IM Rajlich points out, 14...Bd7 is a little inaccurate if I want to go with plan (1) since 14...f4 is definitely necessary while 14...Bd7 may or may not be needed. Still, I'm not going to give 14...Bd7 any harsh punctuation since if White adopts the plan in scenario (1)(b) both 14...f4 and 14...Bd7 really lead to the same position type.

15 c5 f4 16 g4 !? g5 17 Rac1 h5

This h-file attack is a standard Black plan in Benko/Pinter type structures.

18 h3

Setting up a solid defense on the kingside. A strange alternative of about equal merit is 18 gh!? which intends to distract Black on the kingside long enough for White to generate some queenside play. Black should respond 18...Kf7! intending ...Rh8 in which case the chances should be approximately equal. Other replies would be weaker. For instance, 18...Rf6? fails to 19 Qg2 and if Black plays 18...Qe8 19 Nd1 Bh3 20 Re1 QXh5 then 21 N1f2 should lead to better chances for White after either 21...Ng6 22 NXh3 QXh3 23 Nf2 Qd7 24 Ng4 (bad bishop on g7 and no kingside attack) or 21...Bc8 22 cd cd 23 Rc7 Ng6 24 Ng4 (play on the c-file and because of White's light-squared control on the kingside Black's attack is not very dangerous). By the way, after 18...Kf7! a firm warning should be issued against 19 Qg2? Rh8! 20 QXg5?? Bf6 and the White queen will fall.

18...Rf6 19 Rf2

This rook is to be used for defense on the kingside.

19...Rh6 20 Rh2 Kf7 21 a4 a6 !

This is an excellent move for two reasons, only one of which I properly appreciated at the time I made it. First of all, there is the obvious point that the White knight is kept out of the b5 post, from where it might serve as the launching station for a White queenside attack. Secondly, and much more subtly, since Black has no way to break through on the kingside, he must be able to match White on the queenside pound for pound when the right time comes. Thus, an ideal pawn formation for him is pawns on a6 and b6 so that Black can meet b5 with ...a5 and a5 with ...b5. Then the only avenue of attack for White on the queenside would be the open c-file and it would be hard to make this alone decisive because White always has to bear in mind the potential of the h-file to open at an unpleasant moment. I think this whole idea is demonstrative of the sort of equilibrium between White's queenside play and Black's kingside attack that exists in this position.

22 b4 Qh8 ?!

An overly aggressive move which fails to respect the proper balance between defense and attack. Instead, I should have played 22...b6! with the idea mentioned in the note above. Then the position would have remained perhaps slightly more pleasant for White, but Black would have been in no real danger of losing. After 22...Qh8?! on the other hand White could have played 23 b5! strongly. If Black responded 23...b6 there would follow 24 c6 Bc8 25 a5! (IM Rajlich pointed out this move which is much stronger than 25 ba BXa6 26 Nb5 BXb5 27 ab Ra3 when Black has ideas like ...Qa8 controlling both the a and h files which fully compensates for his poor pawn structure.) ba 26 Ra1 and White recovers the sacrificed material with a clear advantage. And if Black tried to use the a-file for his own purposes with 23...ab 24 ab Ng6 then after 25 c6 White would again have the advantage. White just maneuvers a knight to f2 for defense and then can, depending on Black's play either attack down the c-file or move his queen to f1 and play Ra1 seizing the a-file.

23 Be1 ?!

A precaution against my h-file attack: Drew clears the second rank so that his heavy pieces will be able to defend each other. 23 Be1?! does not change the fundamental nature of the position, which still remains drawish, but it does let slip the good opportunity of 23 b5! and allow Black to call the tune for a while.

23...Qh7 24 Nb2 Rh8 25 Rc2 Qg8 26 Qd1 Qh7

Just trying to gain some time on the clock by repeating moves before committing to a definite plan. Amusingly, I thought the time control was 40/2, but in this early round it was really game in 2 hours! This lack of attention on my part led to my playing the last 35 moves of the game in severe time pressure!

27 Qe2 Qg8 28 Qd1 R8h7 29 Rcg2 Nc8

Since progress has come to a halt on the kingside, I figured it was time to turn my attention to the queenside. 29...Nc8 intends to maneuver the bishop to f8 to threaten the pawn on c5. Naturally, 29...Ng6 with the same idea was possible as well, but somewhere in the back of my mind I was thinking that it might be useful to have the option of recapturing on d6 with the knight if White were to exchange his c-pawn for my d-pawn.

30 c6 ?!

Eager to get some play of his own, White gives up a pawn to open some lines on the queenside. The ironic thing, however, is that line opening on the queenside is just what Black needed since he had run out of ideas on the kingside!

30...bc 31 dc BXc6 ?

There was no need for greed. This pawn cannot be held in any case. It was much better to work on nursing the positional advantage with 31...Be6! followed by ...Ne7 and the queen can come over to b8. In that case, Black would be clearly better since he would be more active on both the queenside and the kingside.

Position after 31...BXc6 ?

32 Nd5 ?

Time is of the essence and this approach is much too slow. Instead, White could have played either 32 Qd3! or 32 Qe2! with the twin threats of 33 Qc4+ and 33 QXa6. Either of these moves would have regained the pawn forcibly and left White with the better chances.

32...Qh8 ?

A mistake, allowing White all sorts of nasty counterplay. Instead, 32...hg! would have won easily after 33 hg Qh8 since if White moves his queen to defend h2 then Black trades everything on h2, trades on d5, and then plays either ...Ne7 or ...e4 winning outright. Things are a little trickier if White plays 34 RXh6 RXh6 35 NXc7, but Black still wins with 35...Rh1+ 36 Kf2 BXe4! 37 fe Qh4+ 38 Ke2 RXe1+ 39 QXe1 f3+ 40 Kd2 fg.

33 gh !

After the game, NM Andrew Whatley mentioned he was surprised I didn't see this and play 32...hg! to stop it, remarking that he thought it was a standard motif all King's Indian players would be familiar with, but let me assure you this is not exactly your garden variety idea for White in the King's Indian! When I was showing this game to IM Rajlich, he was at first disbelieving that this idea could ever be any good. Likewise, during the game I saw this move but simply assumed that such a crazy looking move couldn't possibly be any good. And, indeed, most of the time in Benko/Pinter type structures the move would be bad, but in this particular case White has a good attack and 33 gh! buys him some time to carry it out. By the way, the move 33 Rc2! would also have been very strong, giving White the better chances.

33...RXh5 34 h4 BXd5

Sarkisian criticized this, but did not suggest an alternative. In any case, I don't think alternative moves would have saved the game.

35 QXd5+ Ke8 36 Rc2 Bf6 37 Rhg2 Bd8

After this the passed a-pawn is certainly decisive, but the alternative 37...Kd8 seems to give White a winning attack in any case after 38 Nc4 with the idea of 39 Na5.

38 Qc6+ Kf7 39 QXa6 Ne7 40 Qc4+ Kf8 41 b5 gh 42 Rg4 h3 43 Kh2 Rg7 44 Qe6 Rh6

Sarkisian points out 44...RXg4 45 QXg4 Kf7 as a better try, but White still wins after 46 a5 Nf5!? (after other moves White would win more prosaically) 47 ef Rg5 48 RXc7+ Ke8 49 QXg5 BXg5 50 a6 Bd8 51 a7 Qg8 52 Rc2 Qa2 53 Bf2 (threatening 54 Ra8!) Kd7 54 Nd3, etc.

45 Qd7 Kf7 46 b6 ?

Remarkably, this turns out to be a serious blunder, which allows Black a teriffic drawing resource. Instead, 46 RXg7+ KXg7 47 b6 cb 48 Rc8 Kf7 49 RXd8 Qg7 50 Qg4 was a straightforward win.

46...RXg4 47 fg cb 48 Rc8 Qe8 ?

The final blunder in this exciting game. I never saw the draw that was there for the taking with 48...Qh7! 49 RXd8 QXe4 when White has nothing better than to force a perpetual with 50 Qe8+ Ke6 51 Qd7+.

49 RXd8 QXd7 50 RXd7

The rest obviously requires no comment.

50...Ke8 51 Rb7 d5 52 Bb4 Nc6 53 53 RXb6 de 54 g5 Rg6 55 KXh3 Kd7 56 Kg4 e3 57 Nd3 Kc7 58 a5 Rg8 59 Rb5 e2 60 Bc3 f3 61 BXe5+ NXe5+ 62 RXe5 Rd8 63 KXf3 RXd3+ 64 KXe2 Rg3 65 Kf2 Rg4 66 Kf3 Rg1 67 Kf4 Kd6 68 Rb5 Kc6 69 Rb6+ Kc7 70 Rf6 Kd7 71 Kf5 Kc7 72 Kg6 Ra1 73 a6 Ra5

White won.

In the second round, I got another tough pairing-the Waco expert Rudy Tia.

Kamberi, S. (2347) - Tia, R. (2119)
Texas Championship 2001, May 25, 2001

1 Nf3 d6

This contains some interesting transpositional possibilities. The first time somebody played this against me on the internet I replied 2 e4, only to realize that after the reply 2...c5 I had been tricked into entering one of the sharpest lines of the Siclian. Another move that has never really appealed to my tastes is 2 c4. After 2...e5 we are in a line of the English opening that is of course perfectly sound for White, but I don't like the fact that White's king often ends up enduring some nervous moments.

2 d4 Bg4 3 c4 Nd7 4 e3

An extremely tame continuation, but I wanted to play it safe since I wasn't too familiar with this line. When I reviewed this game with the Israeli GM Alex Finkel, he mentioned 4 g3 as probably White's best try for an advantage. After 4 g3 play usually continues 4...BXf3 5 ef g6 6 Bg2 Bg7 7 O-O Nh6! when the knight is headed for f5 to put maximum pressure on d4. One interesting example, Ivanchuk-Speelman, Debrecen 1992 continued 8 f4 Nf5 9 BXb7 Rb8 10 Bg2 NXd4 11 Nc3 O-O 12 Rb1. White followed up with b2-b3, Bc1-b2, and Nc3-e2 clearing all the pieces off the a1-h8 diagnol. Then he advanced his h-pawn to create weaknesses on the dark squares around Black's king, eventually winning in 36 moves.

4...e5 5 Be2 c6 6 Nc3 Nf6 7 O-O Be7 8 b4

Immediately advancing on the queenside has proved White's most popular plan in this position, but Finkel's suggestion of 8 Qc2 to prevent ...e5-e4 has also been frequently played.

8...O-O 9 a4 ed 10 ed ?!

A natural move which keeps the advantage in central space, but 10 NXd4! was better. After 10...BXe2 11 QXe2 White would have kept a small advantage based on the weakness of the d6-pawn as well as his space advantage on the queenside and more active pieces. An important point to note is that Black has no good pawn breaks to improve his position with. ...d6-d5 will leave him with an isolated pawn on d5 and ...c6-c5 will give up the d5-square to White.

10...c5 ?

This is a mistake since after White captures on c5 with his b-pawn and pushes d5 we are left with a position with a few advantages for White-more space and a little bit of pressure on the queenside. Black, on the other hand, doesn't have any real trumps that he can boast about. The situation would have been quite different had Black played 10...d5! 11 c5 Ne4. White remains with his extra space on the queenside but now Black has lots of trumps as well such as:
(1)the e4 square for his knight.
(2)the open e-file for counterplay.
(3)possible pressure against d4 with ...Bf6.
(4)the possibility of playing ...a5 at a good moment so that if White responds b4-b5 Black can capture on b5 and create a passed a-pawn.
(5)finally there is a good possibility that all the White minor pieces will be exchanged except for the dark squared bishop so that Black would get a good bishop or knight versus bishop endgame.
In my opinion, all these advantages would be more important than White's queenside space and White would have to play very carefully to avoid getting the worst of it.

11 bc dc 12 d5 BXf3 ?

Why give up the bishop pair, especially without any prompting? I would prefer 12...Bd6 with just a small advantage for White as Black will probably eventually feel compelled to play ...b6 to secure his queenside and White will have the lever a4-a5. Nimzovich would probably recommend blockading the passed d-pawn with a knight instead of a bishop which could be done with 12...Ne8 and 13...Nd6, followed perhaps by ...Bf6 and ...Re8. I like this plan less than blockading with the bishop, though, since after 12...Ne8 13 Qb3 Nd6 14 Bf4! it is hard for Black to improve the position of the knight on d7 and the bishop on e7.

13 BXf3 Bd6 14 Rb1 Nb6 15 Be2

Protecting the pawn on c4 and getting ready to reroute the bishop to the more aggressive b1-h7 diagnol.

15...Qe7 16 Nb5 Nc8 ?

Coughing up the second bishop leads to an overwhelming positional superiority for White. Instead, Black should have played 16...Be5 although White remains clearly better. By the way I should mention that after 16...Be5 there are enough tactical tricks for Black to survive if White plays 17 d6?! BXd6 (If the queen moves 18 f4 leads to a gigantic advantage for White.) 18 NXd6 Rfd8 19 Bf4 Ne4 20 Nf5 Qf6 21 Qc1 QXf5 22 a5 Na4 23 RXb7 Nac3 24 Bf3 Rd4! (the saving move-otherwise White's bishops would dominate). Therefore, White does better to play the simple 17 Bd3, maintaining his positional edge based on the bishop pair and extra space.

17 NXd6 NXd6 18 Bf4

Two bishops against two knights in a fairly open position give White a huge edge.

18...Nfe4 19 Bd3 !?

Position after 19 Bd3!?

Objectively, this move is good but it does lead to immense complications, and as GM Finkel pointed out, when you have a large positional advantage tactical complications are the last thing you want. In that vein, he suggested the simple 19 Rb3 after which White could gradually collect the full point. In my defense, however, I should say that I was not seeking complications. I thought 19...Nc3 would be easily refuted by 20 BXh7+! Kh8! (20...KXh7 21 BXd6! QXd6 22 Qd3+ and wins) 21 Qh5?? but did not see the rejoinder 21...Ne2+ 22 Kh1 NXf4 when it is Black who wins!

19...Qf6

This just allows White to win a pawn cleanly. More testing would have been 19...Nc3 20 BXh7+! Kh8! when in order to win I would have needed to find 21 Re1! which would have won as follows:
(A)21...NXd1 22 RXe7 Nc8 23 ReXb7 Nc3 24 Bd3 NXb1 25 RXb1 when White, with two pawns and a pair of monster bishops for the exchange would be easily winning.
(B)21...Qf6 22 Qg4! NXb1 (22...KXh7 23 Be5 wins back the piece with a winning position.) 23 BXb1 Kg8 24 Qh3 g6 25 Be5 Rfe8 26 f4! and White wins.
(C)21...Qh4 22 Qc1 Na2 (22...NXb1 23 BXd6 Rfd8 24 Be7 and White finishes a pawn up.) 23 Qd2 NXc4 24 QXa2 QXf4 25 Re4 Qc7 26 RXc4 KXh7 27 Qc2+ Kg8 28 RXc5 and White wins again.

20 Qf3 g5

20...Rfe8 21 BXe4 NXe4 22 RXb7 also wins a pawn for White.

21 Be3

Now White threatens to simply take on e4 as well as to take on f6 and then either g5 or c5. There is no defense.

21...Qg6 22 BXc5 f5 23 BXd6 NXd6 24 c5 Ne4 25 d6 Nd2 26 Qd5+ Qf7 27 QXf7+

27 Qd4 was another easy win since both rooks are off limits due to the move Bd3-c4.

27...RXf7 28 Rbd1 ?

This allows Black to continue thrashing about, whereas 28 Rb2! NXf1 29 Bc4 would have finished matters at once.

28...NXf1 29 Bc4 NXh2 30 KXh2 Kf8 ?

Now finding a win for White is child's play. 30...Rd8 would have lost as well, but at least would have been a tougher exercise.

Sarkisian, D. (2192) - Kamberi, S. (2347) Texas Championship 2001, May 25, 2001