After notching my first in round 6, I was hoping I had turned the corner, but in round 7, I was knocked back down by GM Jaan Ehlvest (2591 FIDE). Ehlvest and I have played a number of interesting games – this was the 5th game between us, and the 4th where he had the white pieces. The first two times around, I outplayed him in different lines of the French Defense, and since then, he has not tried 1.e4 against me.
The first few opening moves were the same as in my game against Altounian from round 4 (1.Nf3 d5 2.g3 c6 3.Bg2), and I decided to go with the …Bg4 setup instead of the …Bf5 setup I used in that previous game. I wasn’t expecting the King’s Indian Attack (Jaan can play just about anything), and I had not really reviewed the opening phase of that game in between, so I figured he had found what both Altounian and I suspected was there (a small advantage for white). Unfortunately, what I tried in this game also didn’t quite equalize.
r2q1rk1/pp1nbppp/2p2n2/8/4P1b1/2N2NP1/PPQ2PBP/R1B2RK1 b - - 0 11)
White has played the KIA with c4, cxd5, and then Qc2 and e4. With a little more space and central control, White has a definite plus. It’s nothing amazing, but Jaan likes to work with small positional advantages like this and it was definitely not what I was hoping to get from the opening. In the two games where I had the most success against him, I had quickly created dynamic imbalances and then outplayed him in complications later on. Here, I’d have to sit and defend for a while – not something I’m unfamiliar with, just not something that I’d like to do.
r1q1rbk1/1p3pp1/2p2n1p/p1n5/3NP3/1PN3P1/P1Q2PKP/2BRR3 b - - 0 19)
White has just played 19.b2-b3. I thought I had done well to get here from the first diagram above, as the bishop exchange has relieved some congestion and the flank pawn moves (…h6 and …a5) have covered my knights against some annoying threats. Still, White has the more space and a more harmonious position. Black’s main question is how to finish his development and I didn’t answer this properly.
I played 19…Qc7, and after 20.Nf5 Rad8 21.Bf4, I went back to c8 with the queen. Jaan didn’t increase the pressure accurately, and that is why I equalized later on.
Instead, 19…Rd8 was maybe a bit more accurate. I was loathe to give up all control over the e5-square, but for now, 20.e5 Nd5 isn’t a real concern. Meanwhile, on 20.Nf5, Black has 20…Qe6, which is a much more comfortable square than c7 or c8 is. Black doesn’t equalize immediately with 19…Rd8, but it’s probably a slightly more accurate way to go about things as White has to find something useful to do as well. A generally useful move like 20.f3 can be met with 20…Ne6 21.Nf5 Qc7, when Black starts to unravel (with the f4-square covered, he can safely keep the queen on c7 and exchange all the rooks on the d-file).
5bk1/1p3pp1/2p1qn1p/p1n5/2N1PB2/1P3PP1/P1Q3KP/3N4 b - - 0 26)
As I mentioned above, I played 19…Qc7 thinking that I could get away with 26…Qd7 in the above position (27.Nxa5 Nd3 was the plan, and both of us thought Black had adequate counterplay – he doesn’t). Instead, I missed my best chance for clear equality with 26…g5!! 27.Bc1 g4. Neither of us considered this, but it makes some sense.
While White is busy going after the a5-pawn, Black starts counterplay on the other side of the board and hits the base of the pawn-chain, a la Nimzowitsch. I guess my not reading My System caught up to me?! But more seriously, this is just a difficult thing to find over the board. White doesn’t have enough firepower to take advantage of the open kingside.
5bk1/5pp1/5n1p/1pn1N3/p3P3/3N1PP1/P2B1K1P/8 b - - 1 34)
Both of us were in some mild time pressure at this point, and this is where I started to lose the thread. Faced with big decisions on every move, I had to guess at what was the best option a few times.
After the game, this was the only place where Ehlvest felt I made a serious mistake. Right now, Black has a choice of whether to stop 35.Bb4 (with 34…Na6) or not. I decided to bring my Nf6 into the action with 34…Ne8, and Jaan thought 34…Na6 would have probably held.
After 34…Ne8 35.Bb4, Black has another decision – exchange knights on d3 or just bishops? Exchanging the knights looks like the wrong decision because with fewer pieces on the board, White’s king will have an easier time of getting into Black’s position.
Thus, 35…Ne6 36.Nc6 Bxb4 37.Ndxb4 followed. Here I made a mistake with 37…Nd6. I was debating between 37…f6 and 37…Nd6, but I decided to avoid any pawn moves that weren’t obviously necessary, as I thought that after 37…Nd6 38.e5 Nc4 39.f4 g5, I’d be able to break down White’s pawn chain. Unfortunately, I had forgotten about 40.Nd3 and the double-knight endgame becomes a bit problematic. With 40…gxf4 41.gxf4, I made time control and sat down to think.
As it turns out, this was my final chance – I didn’t see any good way to stop Kf3-e4-d5, and so I panicked and sacrificed a pawn, but in the resulting pawn-up endgame, I couldn’t get the pawns off quickly enough to survive. With 41…f6 42.Kf3 fxe5 43.fxe5 Kf7 44.Ke4 Nd2+! 45.Kd5 Nb1!, though, Black repositions the knight perfectly, guarding b5 and hitting a2. The endgame is drawn.
This was a tough game to lose, as after the game, neither of could pinpoint a clear error in my play (neither of us had seen the g5-g4 idea or the Nd2-b1-c3 maneuver). Going into the rest day with another loss was not what I been hoping for.
After the rest day, I got the white pieces against GM Gregory Kaidanov (2571 FIDE). Kaidanov was my last regular coach (I worked with him, one hour a week, from late 1997 through a good chunk of 2001). We had played once before in 2002, where I had a clear advantage but ended up forcing a draw by repetition on my way to my first GM norm.
This was an ugly game, only trumped in that department by my game against Gurevich in round 5. I was pretty much unable to think more than one or two moves ahead the entire game and ended up dropping an exchange right after the opening.
r3r1k1/p4ppp/1p6/1QnpPb1q/5B2/2P2P2/P3N1PP/2RR2K1 b - - 6 22)
With 22…Qg6, Black highlights a number of drawbacks in White’s position. White’s queen has no good squares (b4 and b2 both allow a fork on d3), and running away with the rook doesn’t help. For example, 23.Ra1 Bd3 24.Qb2 f6 leaves White in big trouble. I decided to play on down an exchange after 23.Rxd5 a6 24.Qb2 Nd3 25.Rxd3 Bxd3 26.Nd4, hoping to stir up some trouble on the kingside. Objectively, it’s just lost, but Gregory was also having a tough tournament and wasn’t up to the task that day.
2r1r1k1/5ppp/p7/4PP2/1p1N2bP/2P1R3/q4BP1/2Q3K1 w - - 0 35)
I’ve achieved my goals to some extent, as even though Black has won his pawn back and can win another on c3, I’ve introduced some confusion into the game with kingside threats. Of course, it could all disappear shortly if I don’t find some way to keep it going. Luckily I found 35.Re1!, clearing the queen’s path to g5 immediately. Delaying that transfer would be wrong – 35.f6 bxc3 36.Rg3 Qd2!, for example, keeps White’s queen from getting to the key square.
Black is still winning after 35.Re1, but with about 7 minutes left, Kaidanov didn’t find the computer’s path – 35…b3! 36.Qg5 Rxe5!! 37.Rxe5 b2 38.Qxg4 b1=Q+ 39.Re1 h5!! (not 39…Qb7 40.f6 or 39…Qbb2 40.f6!, winning) 40.Qxh5 Qbb2 and Black wins.
Instead, after 35…bxc3 36.Qg5 Qd2, Black forces White to give up the knight for the c-pawn, but White gets the bishop and can play on down an exchange with some kingside threats. I managed to parlay those threats into a perpetual around move 60 to escape with half a point.
My last round game was as black against GM Alex Lenderman (2598 FIDE). This wasn’t a particularly interesting game. I decided to surprise him with the Nimzo-Indian, but when faced with 1.d4 Nf6 2.c4 e6, he decided that I wasn’t properly prepared for the Catalan and played 3.g3. I played 3…d5 4.Nf3 dxc4, and this was where Alex was left to his own devices. While I hadn’t played the Nimzo before, I did have some experience against the Catalan and in those games, I had always gone for a Closed Catalan system (by not taking on c4). Not sure what I was up to, he played 5.Qa4+, but this leads to absolutely nothing after 5…c6 6.Qxc4 b5! 7.Qc2 Bb7 8.Bg2 Nbd7 9.Nc3 c5!. Note that 10.Nxb5 Qa5+ 11.Nc3 cxd4 wins a piece, so Black gets in the freeing …c5 break and actually enjoys the slightly more pleasant position. Recognizing that, Alex tried to find equality and managed to do so.
So, after having some high hopes going into the event, I crashed and burned my way to 3.5/9, a tie for 18th place with Lenderman, Khachiyan, and Altounian. I managed to put together only a few solid games in the event (against Shulman, Shankland, and Lenderman), while in all the other ones, I made some very strange decisions at various junctures.
In my past 10 tournaments, this was the only major disaster. The only event during that run at which I lost serious points was at Gibraltar in January, but even that was a relatively minor loss. Hopefully I can rebound during my upcoming trip to Spain.