November and December Madness

They’re already two rounds into chess’ version of March Madness – in November and December … in Siberia (Khanty Mansiysk, Russia) – but I figured I’d post my predictions for the remainder of the brackets. The only American player left is GM Gata Kamsky, and while I expect him to get past GM Wesley So without too much trouble, he might well face a recent nemesis of his (GM Pavel Eljanov) after that. Kamsky won this thing the last time around.

The best brackets I’ve seen are at: http://en.wikipedia.org/wiki/Chess_World_Cup_2009

Based on that, here are my predicted winners of each section:

Section 1: Maxime Vachier Lagrave (isn’t there always a dark horse who makes it this far?)

Section 2: Alexander Grischuk

Section 3: Shakhriyar Mamedyarov

Section 4: Sergey Karjakin

Section 5: Vugar Gashimov

Section 6: Etienne Bacrot

Section 7: Peter Svidler

Section 8: Gata Kamsky

After that, I say Grischuk, Karjakin, Gashimov, and Svidler all advance (with wins over Vachier Lagrave, Mamedyarov, Bacrot, and Kamsky, respectively). Then, in the final four: Grischuk over Karjakin and Gashimov over Svidler, with Gashimov winning it all in the finals. Amusingly, if Gashimov wins the World Cup, he’ll qualify for the Candidates final which he probably will already be in based on his FIDE Grand Prix results and the fact his home country (Azerbaijan) gets to nominate somebody.

The Splat! Heard ‘Round the US(CL)

Last week the SF Mechanics squared off against the Miami Sharks in the Division finals, with the winner going on to meet the winner of the New York – New Jersey match (New York ended up advancing). While the final score was 2.5-1.5 in favor of the Sharks, we weren’t really in serious danger of being in the match. With draw odds and white on boards 1 and 3, we went into the match with high hopes, but as it was, history repeated itself and for the 3rd time, we lost in the playoffs with draw odds to Miami. I say we kick them out of the Western Division!

Here are the positions from our match after 15 moves (of course, we all reached this mark at different times):

Board 1: Kraai – Becerra

On board 1, Jesse was worse against Becerra’s surprise Grunfeld. Black equalized pretty quickly in the opening, but it wasn’t too bad for White. Unfortunately, Jesse found it difficult to back up the weakening f2-f4 thrust and soon shed the e3-pawn for no compensation. He was only saved when Becerra took a draw in a winning Rook and Pawn endgame to clinch the match for Miami.

Board 2: Lugo – Bhat

On my board, I was doing alright. Lugo surprised me with the Two Knights, but after some slightly non-standard maneuvers, I had played …c6. With …d5 soon to follow, I thought I had equalized. White doesn’t really have any kingside initiative, and after the pawn exchanges on d5, White has fewer pawn islands, but he can’t get at the d5-pawn so easily and his queenside pawns are a little weak.

Board 3: Pruess – Moreno Roman

On board 3, David was pretty much lost against Moreno Roman. David likes to play the King’s Gambit (even though Fischer refuted it ages ago!), and every so often, he produces a brilliancy like he did in France a month ago against GM Bogdan Lalic (take a look at the game here). Unfortunately, this time, he was on the receiving end of a miniature, as he allowed …Qh4+ in the opening and then had to defend against a furious onslaught. I’m not sure if the position was defensible to begin with, but from the above diagram, it ended in another move after 16.Qe1 Ng3.

Board 4: Rodriguez – Liou

On board 4, Yian was keeping it together against Rodriguez. With a big time advantage and a position that was about equal, I figured we’d have to rely on our two relatively equal positions to turn into wins. Sadly, after 16.Rxh8 Bxh8 17.Nd3, Yian allowed White to favorably change the structure by playing 17…Bg6. Now 18.Nxe5 dxe5 19.g4 (19.e4 looks even better) Qh7? 20.gxf5 Bxf5 21.e4 netted White a pawn, and Rodriguez cleaned up in a nice fashion. Instead of 17…Bg6, 17…Nc6 looks quite reasonable for Black.

Thanks to a little tactic, I turned my equal position into a clearly better one:

I played 20…h5! here, and Blas took a wrong turn with 21.Nh2. After 21…Bf6, Black is hitting the b2-pawn, but more importantly, he is threatening to play 22…Be5, trapping the knight on h2. White would have no choice but to give up his g-pawn then with 23.g3, but that’s a rather important pawn. Thus, Blas played 22.Qf2, but after 22…Be5 23.Nf3 Bg3, he lost the exchange. Instead of 21.Nh2, he could have played 21.Nf2, which is a little better. Still, I think Black’s position is better after 21…Bf6 22.c3 d4!. I ended up winning this game, but the finish of this game took place after the match was already put away by Miami.

So, as has been the case most years except for 2006 (when we won it all), we wait for next year. I think the team could have done better this season, but we did alright given the fact we were piecing together a lineup for pretty much every match.

From a personal standpoint, I was satisfied with my play. I lost my first two games of the year (the game against Stripunsky was a bad one, but the game against Barcenilla was pretty good and I should have won that one), but then scored 3.5/4 the rest of the way. Thanks to the 2 losses in 6 games, this was the first time in 5 years that my performance rating in the league was below 2550 FIDE. I clocked in at 2495 FIDE, marginally above my 2492 FIDE rating average for the season. Team captain John Donaldson has recapped the team’s performances in more detail at the team blog.

The finals match is scheduled for December 7th, and features Miami and New York. My guess is that New York will carry the day. With a double-GM lineup and then the underrated (for the league) Yaacov Norowitz on board 4, they seem to have the more dangerous lineup. Still, Miami has scored some big upsets themselves to get to the finals, so it won’t be an easy match.

Me Win Pretty One Day

Last night the SF Mechanics faced off against the Arizona Scorpions in the Western Division quarterfinals. As the 2nd place finishers during the regular season, we had draw odds (meaning we advance on a 2-2 tie) while they had choice of color on boards 1 and 3.

We had a topsy-turvy match against them earlier in the regular season (which they won, 2.5-1.5), but this time, we controlled the tempo from the start.

I was white on board 2 against IM-elect Daniel Rensch. We’ve played 3 times over the past 5 years, but I had black in all those games (with 2 draws and 1 loss). I think this was his first appearance on board 2 in the USCL, but the Scorpions probably wanted to get a master on board 4 to deal with Yian Liou, our underrated anchor.

Playing a Queen’s Indian, Danny went astray pretty quickly and chose a line that I don’t think is particularly good for Black (the full game can be replayed here):

Instead of the normal 8…c6, which commits White to a real pawn sacrifice, Black played 8…Ne4. After 9.cxd5 Bxh4 10.Bxe4 Bf6, Black is stuck with a worse pawn structure and a bad bishop on b7. This is similar to a line after 7.Re1 (instead of the 7.d5 that I played) that goes 7…c5 8.d5 exd5 9.Nh4 Ne4 10.cxd5 Bxh4 11.Bxe4 Bf6 – in this structure, though, Black’s c-pawn is on c5 already, so what he has is a slightly odd Benoni structure with his bishop on b7. It’s maybe not the best line against 7.Re1, but it is certainly quite playable.

Anyways, in the game, we reached the following position after 16.Bg2:

White has two main threats: one is 17.Ne4, exploiting the weakened kingside dark squares, while the other is b2-b4 at some point, sidelining the knight on a6. I think White has a big positional plus, for example, 17…Qe5 (not 17…Qe7 18.d6!, winning a lot of material) 18.Rad1, and Black is going to struggle to activate his minor pieces or deal with White’s central pawn roller with e4 and f4.

However, Danny may have missed the strength of Ne4 as he played 16…c5, trying to fix his queenside pieces and structure. After 17.Ne4 Qe5 18.f4 Qd4+ 19.e3! Qxe3+ 20.Kh1, Black has no good way of dealing with the knight hops to d6 and f6, and so he’s forced to give up the exchange with 20…Rxe4.

With relatively equal positions on the other boards at this point, this put a lot of pressure on the other Arizona players to try and make something of nothing, as they had to score 2.5 to advance. Danya’s game on board 3 was the only one I thought had decent chances of being decisive, and in the end, that was the first game to finish – Danya outplayed Adamson in a complicated position in mutual time pressure.

Meanwhile, I was doing my best to screw things up. Instead of playing for checkmate with 33.Qxh7, I played 33.Rxd5 Bxd5 34.Qxd5. I was now up a piece for a couple pawns, which was completely winning, but like a complete idiot, I botched the endgame in epic fashion.

In the above position, I played 46.g4??, after which it’s Black who is winning! After 46…d3 47.b3, instead of 47…cxb3??, Black can win with 47…Kd4!. I only realized this after I played 47.b3. After 47…Kd4 48.bxc4 b3, Black has 3 passed pawns, and White’s king and bishop can’t hold them all off.

Instead of 46.g4??, though, White is winning with 46.gxf4+ Kxf4 47.Bd5. The pawns are picked up after 47…c3+ 48.bxc3 dxc3+ 49.Kd3 Kg3 50.Bb3 Kxh4 51.Kc4. White picks up all the queenside pawns and gives his bishop up for Black’s h-pawn. That was my original plan, but then I saw the b3 idea, and I figured that was even simpler. Oops. Luckily for me, Rensch didn’t spot the …Kd4 idea, although it probably wouldn’t have made a difference for the overall result.

Ramirez and Wolff were exchanging draw offers on board 1 – Wolff was playing on a computer without the sound on, and so he didn’t hear the draw offers, and ICC doesn’t show the move number for the draw offer (not sure why they don’t implement this simple change), and so he kept noticing the draw offers too late. Meanwhile, Yian ended up winning the drawn endgame on board 4, so we ended up with a big 3.5-0.5 victory.

The other Western quarterfinal was a massacre, as Miami beat Seattle 3.5-0.5. GM Julio Becerra slaughtered GM Hikaru Nakamura on board 1 in 12 moves (12 moves!!!) – despite only starting 15 minutes before us, that game was essentially over before we had even played 10 moves. Here’s the final position in which Nakamura resigned:

That’s nasty.

Thus, next week we face off against the Sharks. While we’ve generally done well against them in the regular season, we are 0-2 against them in the playoffs (they eliminated us in 2005 and 2007 by a 2.5-1.5 score in each match). Hopefully the third time is the charm!

Backing into the Playoffs

Yesterday was the last round of the 2009 USCL regular season. Going into the match, we were tied with the Arizona Scorpions for 2nd place in the Western Division, a full match point behind Seattle and a point ahead of Miami.

Our primary goal was to secure draw odds in at least the first round of the playoffs, and to do that, we needed one of the following scenarios to play out:

(1)   A win in our match, coupled with an Arizona loss and a Seattle loss would give us the 1st seed in the West, because while we’d be tied on game points, our opponent’s average rating was higher than Seattle’s.

(2)   A win or draw in our match, coupled with an Arizona loss and a Seattle draw or win, would give us the 2nd seed.

(3)   A draw in our match would give us 2nd place at best, and only if Arizona lost their match (again, our opponent’s average rating was higher than Arizona). If Arizona drew their match, then they’d have more game points than us, and so it wouldn’t get to the opponent’s average rating.

Dallas has traditionally been a pretty tough match for us, so we weren’t expecting a cakewalk. An hour or so into the match, it wasn’t really clear to me how we were doing.

My game wasn’t particularly interesting – Bercys surprised me with the Queen’s Gambit Declined. I had been expecting a King’s Indian, so this was a big departure from the norm. As it was, we repeated a game of mine against a British GM earlier this year for a little less than 20 moves. That game ended in a draw, and this one was headed for the same result. I played a bit too loosely on the kingside, but with a big time advantage, I wasn’t in too much trouble.

By the time it was clear that I wouldn’t have any winning chances on my board, our board 4 phenom, Yian Liou, beat WFM Zorigt on board 4 in a strange Dragon endgame. At that point, we were a little better on board 3 (Naroditsky was up the exchange, although I thought White had decent compensation) and clearly better on board 1 (where Wolff had turned around a dubious-looking opening into a big endgame advantage). I quickly offered a draw, and Bercys was kind enough to accept.

In this position after 23…Rb8 from Ludwig-Wolff, White is already in some trouble:

White has to cover the b2-square, and so Ludwig played 24.Kc2. Wolff played 24…Ba6, threatening …Rb4 in some lines, so Ludwig covered that with 25.a3. Patrick then rerouted his knight nicely with 25…Na8!, heading for b6. This further inconvenienced White, who had to play 26.Na4 to cover the b6-square. With the e4-pawn no longer attacked, Black had a free hand to come in via the f-file with 26…Rf5!. With a series of jabs, White has been backed into a corner.

Wolff could have capped his effort off with a nice little tactic in the following position after 29.Rg4:

Instead of 29…Bc8 (which maintains a clear plus, because 30.Rxe4 loses to 30…Bf5 31.Re7? Rc2#!), Black had 29…Nxd5+!, taking advantage of the fact the Bf1 has one less defender. After 30.cxd5 Bxf1 31.Rxe4 Bg2, White is toast – Black’s rooks and bishop are too active, and White’s king is too exposed.

On the plus side, at this point, Arizona had already gone down to Miami, so we were playing with house money in a sense as the 2nd seed was ours. However, the Seattle match was up for grabs, and if they lost, we would have liked to get at least 2.5 from our match to take the top seed!

Unfortunately for us, after some good defense from Ludwig and mistakes in mild time-pressure, Wolff found himself having to defend an exchange-down endgame.

This was the final chance in my view for Black to try and save the game. Patrick played 48…Bh3, going after the c4-pawn. While it does win a pawn, it frees the d5-pawn for White and that is the more important factor. Black’s connected passers never became a factor and Ludwig pushed his d-pawn all the way. It seems to me that Black can try for a draw with either 48…Bc8 (forcing the rook to take the a7-pawn with Ra8xa7, while Black pushes on the kingside) or 48…Kf6 (centralizing the king, and again planning to push the kingside pawns).

In any case, the result of this game was a moot point as Seattle held on for a draw. Thus, we took 2nd place in the Western Division, behind Seattle and ahead of Arizona and Miami. After winning the division in 2005 and 2006, we’ve taken 2nd place the past 3 years. Our best regular season record of 8.5/10 however has gone untouched, as both Seattle and New Jersey fell just short of the mark this year.

Next week, we’ll face the Scorpions with draw odds (they’ll choose the color they want on board 1 tonight). We have a 1-1 record against them historically.

The Joy of Standardized Testing (Not!)

With a bit of a lull in my chess schedule (I haven’t played a tournament since the SPICE Cup ended on September 27), I had decided to take the GRE to get it out of the way. I’m not applying to graduate school this fall, but as I’m likely to at some point, it’d be nice to take it when I’m not particularly busy with a tournament or job.

It’s been a while since I’ve taken a test of any kind and even longer since I’ve taken a standardized test. This was also the first time I took a computer-adaptive test. While the GRE is supposed to be an exam that pretty much all prospective graduate students take, I was surprised by the subject content.

On the one hand, the verbal section was quite difficult. I was happy with my score, but it could easily have gone the other way had I not been exposed to some of the more difficult words. Is a vocabulary test the best measure of how well a prospective graduate student is going to do? I highly doubt it.

One of the more difficult words I had to identify (and did) on a practice test was nostrum. While some words were part of a practice vocabulary list culled from a prep book, nostrum wasn’t such a word. Luckily, though, I have seen pretty much every Seinfeld episode and had done a little research.

In one of the episodes where he pretends to be a doctor, Kramer plays the role of Dr. Peter Van Nostrum. Take a look at: http://www.youtube.com/watch?v=pyossoHFDJg

He pretends to be a similar doctor when he “diagnoses” George Costanza’s boss with a potentially dangerous mole. A little trivia behind the episode reveals that nostrum means “a usually questionable remedy or scheme.” Luckily that pointed to panacea as the antonym and I was in business. But, as the test is adaptive, answering such a question incorrectly could lead to a much lower score, and so once you miss a question, it lowers your score and gives you a simpler question. Especially near the start of the test when the program’s “prediction” of your score is moving around much more, this takes on increased importance.

Meanwhile, the math section is quite sad in my view. Despite being a test for college graduates, the math on the section is not really any more difficult than on the SAT and is at a basic high-school level. It tests basic arithmetic, algebra, and geometry – if you’re doing well, you might get a simple statistics or probability question. Why even bother making a math test like this? Getting a perfect score means you’re only in about the 95^th percentile. These schools can already get our SAT score without too much trouble – why make people jump through the same hoop so many times?

The only section that seems to test something really relevant is the analytical writing section. You have to write two essays, one presenting an argument in favor of or against a claim and then one analyzing an argument presented to you. This actually does test your ability to think logically while also allowing you to demonstrate a command of the language. For pretty much any graduate student, this seems like an important skill. The math and verbal sections, though, are pretty much useless in my view. Instead, a specific subject test of your desired field seems to be more appropriate.

Another strange part of the exam experience is the strict rules at the testing center. No food or drink is permitted in the testing hall (which I guess makes some sense, since they don’t want people spilling stuff on the computers), but they even have policies on the number of tissues you can take in! Another test-taker had a runny nose and took 3 tissues from the Kleenex box. The proctor immediately told him that there was a limit of 2 tissues in the room at any given time, and that if he wanted to blow his nose, he would have to exit the room (which involves showing ID and signing in and out)!

By the way, I gave an example of a word from a practice test because as part of a confidentiality agreement that every test-taker is bound by, I can’t talk about any question that I actually received on the test. This might partly be a result an ETS mistake, whereby their question pool is relatively small, and so questions can get repeated if you sit for the exam more than once. This was more of an issue in the mid-1990s, but I would hope it’s been resolved by now. I don’t plan on taking the GRE again, but if the question pool is still pretty small, then that could introduce a lot of bias into the scores.

Booking Our Ticket for the Playoffs – Part 2 of a Weeks 8 and 9 Recap in the USCL

On Monday, the SF Mechanics faced off against the Tennessee Tempo. The Tempo have 2 GMs on their roster (Jaan Ehlvest and Alex Shabalov), but I guess neither one was available for this match, and as a result, the Tempo lineup had a much lower average rating than normal. Still, the games aren’t played on paper, and even though we outrated them by almost 150 USCF points at the moment, it was not an easy match.

My own game was the first to finish. I had the black pieces against FM Todd Andrews, who I last faced in the 2008 USCL season (see the post and my annotations here). In that one, Andrews played 1.d4 and the game went into a Semi-Slav. This time around, he played 1.e4, and the game plodded along the normal Closed Lopez lines. The full game can be seen here.

Instead of my usual Graf Variation of the Chigorin (with 9…Na5 10.Bc2 c5 11.d4 Nd7), I played 9…Nd7 10.d4 Bf6, to get to the position above. I’m not quite sure what this is called, but I’ve seen it labeled as the Karpov System in some places, so I’ll go with that. I have played it before, against GMs Friedel and Becerra, but it’s not my main line of defense in the Lopez. Andrews may have been a bit surprised by it and responded with 11.Qd3!? in the above position. White’s dilemma is that the d4-pawn is under serious pressure, so the usual maneuver of Nb1-d2-f1 isn’t available just yet. If White doesn’t want to commit to d4-d5 just yet, then he can either go with 11.a4 (the main line), 11.a3 (Becerra’s choice against me), and 11.Be3 (guarding the pawn, but allowing …Na5 and …Nc4 with tempo). Andrews’ Queen move tries to solve that d4-pawn problem in a novel fashion. I decided to take the game into more traditional Chigorin waters with 11…Na5 12.Bc2 c5 13.d5. After some maneuvers, we reached the following position:

White has just played 17.Qd3-d1, getting the queen out of the way of …c5-c4 advances while waiting to see what Black does. I decided to go with 17…a5, avoiding the (more) natural 17…c4 because after 18.Be3, Black doesn’t have the c4-square for his knight. After 18…Nc5 19.Qd2 then, White is ready to play Bh6 and try and make some inroads on the kingside. However, after 17…a5, 18.Be3 (or 18.Bd2) can be met quite easily with 18…Nc4, when White’s best would be admit he has nothing and retreat with 19.Bc1. I kept with my plan of avoiding …c4 to get to the following position:

I just played 19…Bc8-d7, developing the bishop and guarding the potentially weak b5-pawn. At this point, White should really be trying to claim some squares. White played 20.Ng4?, which in a way is already the decisive mistake. Strange to say that, but after 20…Qh4!, Black is threatening 21…f5 (now the h6-square is under Black’s control). The game continued 21.Ne3 c4!, with a  clear advantage to Black. It’s a bit odd to have a c4-b5-a4 pawn structure, but Black has now staked out serious territory on the queenside. The knights and Bc2 are stuck preventing …f5, but this leaves the rest of White’s queenside without any prospects. White tried 22.Qf3 Nc5, but felt that he had to play 23.a3, irrevocably weakening the b3-square. I was planning to play …a3 myself if given the chance, since after 24.bxa3, White’s queenside both weak and immobilized. The bishop on c1 can’t move anywhere good without losing the a3-pawn (and the rest of his queenside would be very weak). Thus, Black has a free hand to play …Nb6-c8-e7, supporting the …f5 advance.

Instead of 20.Ng4, White had to play 20.b3, taking the c4-square and some queenside space before it’s too late. After 20…axb3 21.axb3, Black has the a-file, but it’s not especially amazing at the moment. Meanwhile …c4 can be met with b4, keeping Black’s knights from any nice central squares.

The game after 20.Ng4 Qh4! 21.Ne3 c4 didn’t go particularly well for White, as he was powerless to stop …f5. I slowly built up the advance and diverted all my pieces to the kingside, and once that side of the board was opened, the curtain fell pretty quickly.

As for the other games, on board 3, Danya was pretty much always a bit better, but not quite enough to win the game. On board 4, though, we thought we had a good matchup. However, David Justice took out the higher-rated Yian Liou in a pretty solid effort to bring things to a tie at 1.5 points apiece.

That meant it was all down to GM Patrick Wolff on board 1 against IM Ron Burnett. Burnett isn’t so high rated now, but he’s been a solid player for a long time. He was at least equal for most of the game, but as time pressure loomed, he fell victim to some nice knight hops:

White just played 33.Nc6-b4, aiming for the hole on d5. After 33…Re5 34.Nd5, Black blundered with 34…Qc5?. Wolff quickly responded with 35.Rxg6+! fxg6 36.Nf6+, forking king and rook. After 37.Nxd7, White was forking queen and rook to go up an exchange! Wolff finished the game off to take us to a 2.5 – 1.5 match victory.

Special thanks to Payam for bringing drinks and pastries to the match for us! After watching the NJ Knockouts sweep a powerful Boston team with the help of some donuts, I figure we could do worse than to copy them.

Back on Track – Part 1 of a Weeks 8 and 9 Recap in the USCL

After our loss to Arizona in Week 7, we fell into a 3-way tie for second place in the division with Arizona and Miami. In Week 8, we faced the Queens (NY) Pioneers. On paper, the Pioneers look like they should have a solid team, but for whatever reason, they don’t seem to have performed up to those expectations. Here’s a somewhat late summary of that match.

We won the match 3-1, with the lone loss being my doing on board 1.  My game was rather forgettable, as I turned in my worst game in my USCL career. I was white against Stripunsky, and after a pretty normal Meran Variation that I normally play as black, we got the following position:

Black has just played 19…Qxe7xd6. White’s advantage is pretty minimal, but I think there is some advantage due to the difference in the power of the light-squared bishops. Black’s pawn on c6 cuts down his own bishop, and while White’s e4-pawn does the same, White’s e4-pawn can more easily move forward. If Black plays …c5, then the b5-pawn will drop. Thus, White should watch out for …c5 tricks from Black.

With this in mind, both 20.f4 and 20.Rad1 are pretty normal looking moves. Pushing the f-pawn prepares 21.e4-e5, while 20.Rad1 brings a piece to the center and threatens a discovery against Black’s queen. The rook move would have maintained a small plus.

I didn’t spend much time here though and quickly played 20.f4?, failing to notice Black’s strong response 20…c5!. Stripunsky has played these sorts of positions a lot, and he wasn’t so quick to miss that resource. After 21.e5 Qb6, Black is threatening 22…c4+ and 23…cxd3, so White can’t simply take on f6. After 22.Rf2 c4 23.Bf5 Nd5, White can’t safely take on h7 because after 24.Bxh7+ Kh8 25.Be4 Ne3 26.Qb1 Red8, White has huge problems with his back rank and pieces. Black’s already clearly better. However, in mutual time pressure, Stripunsky made a couple mistakes to give me a chance to get back into the game.

Black has just played 29…Ng2-e3, hitting the queen. I quickly played 30.Bc6??, failing to notice that after 30…Rxd6 31.Qxd6, Black can just play 31…Qxc6, since 32.Qxe7 walks into 32…Qh1 mate! Oops.

Instead, 30.Bxb5 would have equalized quite simply and I saw this move. For some reason, I thought that 30.Bc6 was even stronger. After 30.Bxb5, a couple sample lines might be:

(1)   30…Rf8 31.Qd4! Qxb5 32.Qxe3 (32.Nc3 is also fine) Qxb2 33.Nd4 with no problems for White – he even has an initiative on the kingside that forces Black to go for a draw with 33…Qa1+ 34.Rf1 Qa2 35.Rf2 Qa1+

(2)   30…Rxd6 31.Qxd6, and now 31…Qxb5 doesn’t threaten mate, so Black has to play 31…Qxd6 32.exd6 Re6 33.d7, when again White is doing just fine.

Luckily, my mistakes in this game didn’t come back to hurt the team, as they all won their games. IM John Donaldson, FM Daniel Naroditsky, and NM Greg Young all won on boards 2 through 4.

There was one interesting endgame moment in the game between IMs Lev Milman and John Donaldson on board 2. Both players were in time pressure, and had been for some time, when White played 40.g4? here.

The endgame was about equal when it started, but Lev seemed to be pushing too hard and found himself headed towards a worse endgame. However, John mistakenly traded down into a drawn rook and pawn endgame to bring the correct result back into drawish territory.

The endgame should be a draw – one idea for White would to play 40.Kd3 here. If Black plays 40…Rb3+, then 41.Ke2 guards the f3-pawn and leaves the c2-pawn undefended. Black’s best chance would be to play 40…c1/Q 41.Rxc1 Rg2, but then 43.Ke3 Rxg3 44.Kf2 Rh3 45.Rc4 Rh2+ (otherwise Kg2 traps the rook) 46.Kg3 is a simple draw. Black’s extra pawn can’t be realized here.

Another way to try and draw this would be to play g4, trade the g4-pawn for the h5-pawn, and then go after the c2-pawn, sacrificing the f3- and h4-pawns in the process. The resulting rook endgame with extra f- and h-pawns is a theoretical draw, but given the time constraints, could easily swing Black’s way.

Lev seemed to start out on this path with 40.g4, but Black has a trick here to make sure that never comes to pass. He can play 40…g5!!, forcing the creation of a passed pawn, no matter what pawn White takes. After 41.gxh5 gxh4, the point is that White’s king can never cross onto the 2nd rank because of …c1Q with check. Meanwhile, the rook is tied to the c-file. White can play 42.h6, but then 43…Kg8 stops the pawn dead in its tracks. After 44.Kf4 h3 45.Kg3 h2, one of the two pawns will queen.

Unfortunately, John missed this nice finish and played 40…Ke7. Trading on h5 seems safest here, but Lev played 41.Kd3. This doesn’t throw the draw away, as after 41…Rb3+, White should play 42.Kxc2 Rxf3 43.gxh5 with a theoretical draw. Unfortunately for Queens, though, Lev played 42.Ke4, but after 42…Rb4+ and 43…hxg4, Black had two connected pawns to none on the kingside. John duly won that endgame and we were headed to our first match win since week 4.

Bizarro Day in the USCL

This last week’s match in the USCL was one that the SF team would rather forget. Although we ended up losing 2.5 – 1.5, the match could pretty easily have turned in our favor in a number of ways.

We got off on the wrong foot when on board 4, Greg Young blundered a piece to David Adelberg (the full game is here). In the following position, Black executed a simple tactic with 9…e5:

After 10.Nb3 d4, Black is just up a piece. At this point, I sort of chalked the game up as a loss for us, although, as we’ll see, nothing was what it seemed …

Meanwhile, in our other white on board 2, IM John Donaldson didn’t seem to get too much against IM Dionisio Aldama (the full game can be seen here):

Still, the position is generally without risk for White, and after a dozen moves or so, they had traded down into a minor piece ending that while looking marginally better for white, didn’t look so amazing. John decided to agree to a 3-time repetition, but as it turns out, according to IM Mark Ginsburg (as can be read here), White was winning in the final position:

Now 32.Nxh7 doesn’t work because of 32…Ke8, and the knight isn’t going to get out from h7 very easily.

However, 32.Kc4 looks quite strong – White threatens to exchange on d7 and then invade with his king on b5. If now 32…b5+ 33.Kd4 Ke8, then 34.Nxd7 Kxd7 35.Bf1 b4 36.Kc4, and White wins by invading with his king on the queenside. If 33…Be8 (instead of 33…Ke8), then 34.Nxh7 wins, as Black’s king can’t cut off the knight’s re-entry into the game via f8. Meanwhile, if 33…Bc8, then 34.Bf1 Ke8 35.Ne6 b4 36.Bb5+ is also a winning endgame for White.

A missed chance, but this would be the theme for the match. At this point, it was 0.5 – 0.5, with 3 games going, but I figured Young would lose on board 4. That left it up to Naroditsky (black against Rensch) and me (black against Barcenilla).

On board 3, Naroditsky didn’t seem to play the opening correctly. If nothing else, his time management seemed a bit suspect, but as I’m not usually one to talk about such things, I won’t harp on it too much. He did get a playable position, and slowly outplayed Rensch, reaching a completely winning endgame (the full game can be seen here):

Around this point, Greg got up and announced that he had won his game. I was completely floored – wasn’t he down a piece? Well, this was a match which showed both sides exhibiting “great” technique …

Back to Danya’s game – Black had won a pawn on the queenside, and then slowly walked his king up to b5 while giving up a relatively unimportant pawn on g6. Now it’s a pretty straightforward win for Black – he can play 61…Kc4, and White can’t stop 62…Kd3. White’s only chance is 62.Rc8+ Kd3 63.Ne1+, but then 63…Kxe3, and White’s pawn chain falls like the proverbial ripe apples.

Two possible lines then are: 64.g6 Re2 65.g7 (65.Nc2+ Kd3 66.Nxb4+ Bxb4 67.g7 Bc3 is curtains) Rxe1+ 66.Kc2 Re2+ 67.Kb3 Rb2+ 68.Ka4 Rg2 69.g8=Q Rxg8 70.Rxg8, and now either 70…Kxd4 or 70…Kxf4 gives Black too many passed pawns for White to deal with.

Instead, Danya played 61…b3? 62.Rb8+ Ka6? (62…Ka4 might still have held the balance) 63.Na1 (63.Nb4+ was also winning), and Black’s prized pawn has to be put back in the box. The game soon followed.

It was now tied at 1.5 apiece, with my game hanging in the balance. I will humbly submit that our game was reasonably well played until we each started playing on the increment (just after move 30). Barcenilla surprised me with a Scotch, and I in turn, surprised him with 4…Nf6.

I have exclusively played 4…Bc5 against the Scotch, and I figured that having played it twice against Barcenilla’s teammate Danny Rensch, he would be expecting that line and have it well prepared. Meanwhile, he had never played the Scotch from what I remembered, so it seemed like at least now both of us would be on unfamiliar territory rather than just one of us. The opening gamble paid off, as I don’t think he played a very critical line. The full game can be seen here.

So far it’s all normal, but here, 8….Ba6 is much more popular than the move I played, 8…Nb6. Still, the knight retreat is probably not so bad and seems to be theoretically quite playable. The game continued 9.Nc3 Qe6 when Barcenilla played a move which seemed a bit odd to me – 10.b3. Instead, 10.Qe4 is by far the most common move.

After 10.b3 a5, White either has to allow Black to make inroads on the queenside with …a5-a4, or stop that with a2-a4 and give Black the b4-square. He chose the latter, and after developing and castling, we reached the following position after 19.Kc1-b1:

I wanted to play 19…c4 here, immediately breaking open the queenside, but I couldn’t quite make it work after 20.bxc4 Bg4 (20…dxc4 21.Bxc4 doesn’t leave me with too many open lines, and in fact, the f7-pawn could become rather weak) 21.cxd5 (21.Be2 Bxe2 22.Qxe2 dxc4 23.Qxc4 Rec8 24.Qe4 Rab8 gives Black a strong initiative) Bxd1 22.Qxd1. The two bishops plus strong central pawns seemed to give White reasonable compensation for the exchange in my view. After a 20-minute think or so, I decided on 19…Bb7, simply developing a piece and waiting to play …c4 later. I finally got my chance after to play …c4 in the following position:

After 27…c4 28.bxc4, I played 28…Ba6. Not 28…Bc8 because of 29.Bxc8 Rdxc8 30.f5!, when Bh6 is a rather annoying threat. Black has to play 30…Qxe5, when 31.fxg6 hxg6 32.Bh6+ Kg8 (Black can’t go to the e-file due to the threat of a pin) 33.Rxg6+ with a forced draw by repetition.

After 28…Ba6 29.Bd3 Rb8, Barcenilla alertly played 30.f5 – not 30.Ka2 first, as then I wasn’t going to play 30…Bc3 31.f5 Qxe5 transposing back to the game, but 30…f5!, clearing the f7-square for the queen to hit the c4-pawn. Black is then winning.

After 30.f5 Bc3+ 31.Ka2 Qxe5 32.fxg6 fxg6 (I didn’t want to open the f-file, but 32…hxg6 again leads to a draw after 33.Bf4!, so I didn’t feel like I had a choice) 33.Re4, we got the following position:

As Adamson writes on the Scorpions’ blog, I had a chance for USCL glory (and maybe GotW, but I’m sure at least one judge hates me already by now – Jeff, your prayer that I lose this game was answered!) with 33…Qxe4!. The point is that after 34.Bxe4 Bxc4+ 35.Ka3, Black has 35…Be6!!, and White is helpless to stop Bb4+, winning the queen back with at least one extra pawn for the endgame. Of course I saw 33…Qxe4, but with about a minute on the clock, I missed 35…Be6, and so I decided to go for the game continuation.

The game continued: 33…Qd6 34.Ba3 Bb4 35.Bxb4 Rxb4 36.Qf2+ Rf7 (36…Kg7 was better) 37.Qxd4 Rxa4+, and in my calculations on move 33, I figured this position was winning. All of Black’s major pieces are involved in the attack and there are a lot of open files for them to pursue White’s king. Unfortunately, after 38.Kb1 Qb4+ (38…Rb7+ is met, paradoxically, by 39.Kc1, allowing a check on a3, after which White is ok!) 39.Qb2, it’s not so easy for Black to continue giving checks – e1 and f1 are covered!

I played 39…Bb7, thinking to myself that he had to keep both squares covered, and so on something like 40.Re4-e3, I would play 40…Bxg2 winning a pawn. However, he again found the correct move with 40.Re2, as after 40…Rf1+ 41.Kc2, there’s no good follow up for Black! After some more mistakes from me, we reached an endgame that was marginally better for White (passed pawn supported by king, rook, and bishop), but it should have been drawn.

In the following position, I played 56…g3:

I could have given up my bishop for the c-pawn earlier, and then hoped to liquidate the kingside pawns and reach a R vs R + B endgame. However, I didn’t want to defend that while on the increment – there are some players (namely GM Josh Friedel), who seem to know that endgame backwards and forwards, but I figured that given fatigue and time pressure, I would better avoid it.

The decision (to push my h- and g-pawns) was objectively correct, but at this point, I blundered horribly. After 57.hxg3, I played 57…h3??. After playing this move, I noticed that 58.Bd7 Rxc7 59.Kxc7 h2 60.Bc6 was simply winning for White! I also missed that Barcenilla’s chosen move, 58.Bc6, was also winning for White, albeit in a more difficult fashion.

Had I seen that I wouldn’t be queening my pawn, I would have played 57…hxg3, after which the draw is pretty simple. After 58.Bd7 Rxc7 59.Kxc7 g2, Black’s king gets to f2 or h2 fast enough to escort the pawn through. Thus, White has to give up his bishop for the pawn with Bh3xg2, but the resulting R vs B endgame is a trivial draw (it’s much simpler than R vs R + B).

In any case, after Barcenilla’s mistake, we got to a Q vs R endgame where I had some chances to hold out for a draw, especially had I not blundered after about 39 moves with 99…Kg7? (99…Kf8 would have kept some drawing chances). But so it goes, especially for this match.

Thus, we lost 2.5 – 1.5 and dropped into a 3-way tie for 2^nd place in the West with Miami and Arizona. With Dallas 1.5 games back, we’ve got to win a couple matches in the final 3 weeks to make sure we make the playoffs.

Black Magic

After 4 rounds, I had used up 3 of my whites (with 3 draws) and was sitting on 2.5/4, half a point behind the leaders.

In round 5, I was black against IM Gergely Antal (2487 FIDE). Antal had been on a roll coming into the event, winning the national Collegiate Championship and then the strong Southwest Open prior to this event. He was also tied for the lead at that point, with 3 wins and 1 loss from 4 games.

In our game, though, he didn’t play particularly well. In the following position, it was my turn after he played 11.Bc1-g5?:

The opening was a Scotch with 6.Qf3 (that is, 1.e4 e5 2.Nf3 Nc6 3.d4 exd4 4.Nxd4 Bc5 5.Nxc6 Qf6 6.Qf3), and while I didn’t play the theoretically most-approved solution with …bxc6, I did get a reasonable position. However, 11.Bg5 saw Antal take one liberty too many with his development and he quickly paid for it.

After 11…Nf6, Black is threatening to take on e4, so he played 12.f3. If 12.e5 instead, then 12…Ne4 hits the queen and bishop. After 12.f3, though, I have 12…Nh5 when 13.Qh4 is forced (not 13.Qe5 Rde8, when White is powerless to stop a discovery on the e-file or the fork with …f6). After 13…f6 14.Bd2 (14.g4 fxg5 15.Qxh5 Qf3 16.0-0-0 Qxf3 is no fun for White either) Qxg2, White is down a pawn without any compensation. To add insult to injury, he can’t even guard the f3-pawn with 15.Rf1 because of 15…Qh3!. White’s queen is trapped, so he has to exchange queens, but then after 16.Qxh3 Bxh3 17.Rh1 Bg2, the rook is trapped as well! Note that Black’s bishop on b6 covers the g1- and f2-squares here. Thus, he played 15.0-0-0, but then 15…Qxf3 left me with two extra pawns, after which I had no troubles winning.

So now I had 3.5/5 (or +2), with both wins coming as black. In round 6, I was white against the Brazilian GM (and current champion) Andre Diamant (2526 FIDE). Diamant was having a rough go of it in his first US tournament, but I couldn’t take him lightly.

He surprised me by switching away from his King’s Indian Defense, instead opting for a Queen’s Gambit Accepted. I thought I was better after the opening, and then prepared to play d4-d5 in the following position with Qd1-c1 and Rf1-d1.

I decided to push the d-pawn here with 14.d5. Now if he had played 14…exd5, I would have continued with 15.Nxd5 Nxd5 16.Rxd5 Qe8 17.Rxd8+. After either recapture on d8, White will play 18.Qc3 with some advantage. White is better developed and Black’s king is more likely to be attacked.

However, he surprised me with 14…Nxd5. I can transpose into the above variation with 15.Nxd5 exd5 16.Rxd5, but I thought I could take advantage of his move order with 15.Ba4!?, pinning the Nc6 and leaving the Nd5 pinned. White is hoping to play Nd4xc6 at some point, followed by Nxd5 and Bxc6. In any case, if White can play Nd4xc6, Black’s king’s shelter will be ripped open, which is easily worth a pawn.

He hadn’t see 15.Ba4 at all, but luckily for him, it just barely works for Black! After 15…Be7, White has some trouble actually playing 16.Nd4 because of 16…Nxc3! 17.Nxc6+ Qxc6! (this is why 15…Be7 is important – the rook on d8 is now protected!) 18.Bxc6 Ne2+!, forking White’s king and queen! I played 16.Nf4 instead, but after 16…h3 17.Ncxd5 exd5 18.Rxd5 Qe8, Black was barely hanging on. The game later ended in a draw where both of us had to accept a repetition or end up much worse.

In round 7, I was black against FM Daniel Rensch (2386 FIDE). I have played Danny a few times, although only once recently. That encounter was back in December 2008, where he beat me as white in the first round of the 2008 Berkeley Masters (see http://vbhat.wordpress.com/2008/12/17/the-berkeley-fight-club/). I repeated my opening from that encounter (another Scotch!), and in the following position, Danny played his prepared novelty 14.Kf2:

Back in December, Danny had played 14.Be2, but after 14…f5 15.e5 Nd5 16.Bd2 Be7 17.b3, I could have equalized with 17…f6!. This motivated his search for something better, but in general, this endgame is just equal because while Black’s pawns look funny, they do control some key squares. Meanwhile, White’s knight on a4 is a pain to bring back to the game and White is behind in development.

After 14.Kf2, the game continued: 14…c5 (threatening 15…Bd7) 15.c4 (allowing the knight to return, but giving Black the d4-square) Nc6 16.Nc3 f5 17.e5 Bf8 (avoiding e7 because of White’s next move) 18.Nd5 Nd4!. Black shouldn’t play 18…Bxd5 19.cxd5 Rxd5 20.Bc4 Rd7 21.e6, which was Danny’s preparation. However, after 18…Nd4, he was on his own and quickly realized that he didn’t really have any advantage. The game was agreed drawn in about 5 more moves.

With 4.5/7, I was white against GM Eugene Perelshteyn in round 8. Both of us were on 4.5, trailing the leader, Ben Finegold, by a point. I would have liked to fight, but Eugene was well-prepared, and by essaying a sideline of the Ragozin Defense that IM Kuljasevic played against me in round 1, he caught me off guard. I didn’t react particularly well and decided to agree to a draw after only 16 moves. In the same round, Finegold had a quick draw with GM Diamant to get to the magic mark of 6 points, enough to secure his last GM norm. Meanwhile, IM Robson won to join Eugene and myself in joint 2^nd place with 5 points.

Thus, going into the last round, there was a chance for a 4-way tie for first. GM Perelshteyn had white against IM Finegold (which Perelshteyn ended up winning); I had black against IM Ippolito; and IM Robson was black against FM Rensch (a game which ended in a draw).

After playing 14.a3 to reach the following position, Dean offered me a draw:

With first place potentially on the line, and an interesting position to play (where I thought was at least ok, and maybe even a bit better), I decided to continue with 14…0-0 15.b4 Rac8.

After 16.Qb3, I played 16…Ne5!. The tempting 16…Ne4 doesn’t really do anything, as after 17.Bb2 Rxf3 18.Bxf3 Nd2 19.Qd1, Black has only succeeded in exchanging some pieces. Instead, after 16…Ne5!, Black is threatening to meet 17.Bb2 with 17…Nc4, when the other knight will join its colleague in the center with 18…Ne4 with advantage. Thus, Dean played 17.Nxe5 Bxe5 18.Bb2, but then I played 18…Qd6!. It’s important to try and claim the a1-h8 diagonal, and if the bishops get exchanged on e5, Black’s queen is beautifully centralized on e5.

Dean reacted poorly with 19.f4? (both 19.Bxe5 and 19.g3 were better, although I think 19.g3 was really the best move by far). After 19…Bxb2 20.Qxb2 Qb6 21.Qe5 Ne4 22.Qd4 Qxd4 23.exd4, Dean surprisingly thought he could hold this endgame. However, after 23…g6! (putting an end to ideas of Bg4 and f5), Black is in total control. White has a number of weak pawns and squares, and it is only a matter of time before Black wins.

After about 10 moves, we reached the above position. Now if white plays 34.Bxb7, he is completely lost after 34…Nxd4. He’s only down a pawn, but more importantly, he has no way to meet 35…Ne2+. The knight check will either win the unprotected rook on c4, or set up a checkmate on f1 after the king slides over to h1.

Thus, Dean played 34.Ra4, but after 34…b6 35.Bd1 Nc3 36.g3 Nxd1 (not the only way, but visually more pleasing) 37.gxf4 Nb2, White’s rook is trapped on a4! Ippolito resigned, rather than play on down a whole piece with absolutely no hope.

So with my 3^rd win with the black pieces (in 4 games, as compared to 5 draws in 5 games as white!), I finished with 6.0/9. As mentioned earlier, Finegold lost to Eugene and Robson drew with Rensch, so there was a 3-way tie for first with Finegold, Perelshteyn, and me. Perelshteyn won the trophy on mathematical tiebreaks. Finegold made his final GM norm while Rensch made his final IM norm – congrats to them both! Thanks also to Susan Polgar for organizing a fun, professionally done event.

Ups and Downs at Check ‘Em Tech

I recently competed in the SPICE Cup (Group B) in Lubbock, Texas. Hosted by Texas Tech University, it was a Category 11 round-robin (average rating 2503 FIDE). The top group was a Category 16 (average 2631 FIDE) double round-robin.

As the 3rd lowest rated player in the event, I didn’t expect it to be an easy event, but unlike my previous event in Montreal, the players here were much more tightly bunched in terms of strength. The highest rated player in the group, IM Gabor Papp (2562 FIDE) of Hungary, was not so much higher rated than me, while the lowest rated player, FM Daniel Rensch (2386 FIDE), had beaten me two times in three previous encounters.

I started the event off with white against a nemesis of mine – IM Davorin Kuljasevic (2547 FIDE). He was at UTD for a few years before starting at Texas Tech this fall as a graduate student. We had only played once before in person, but in USCL play, he was one of only two players with a higher performance rating in league history and had beaten me down a couple times. After 14.dxc5, we reached the following position:

Strangely enough, I had the position after 14 moves before, in 2006 against GM R.B. Ramesh. Somehow, though, Davorin missed this in his preparations and so he had already spent a good deal of time to get to the diagrammed position!

Anyways, Ramesh played 14…f6, which avoids the problem of 14…Nxc5 – namely that the knight capture walks into 15.Rfd1!. The d5-pawn is pinned (since capturing on c4 would lead to mate on d8), but after 15…Be6, White has 16.Qb2!. With the pin broken, the d5-pawn is again under attack, but now the g7-pawn is as well.

After a short think, Kuljasevic played 14…dxc4, which is a better move. After 15.Rfd1 0-0 16.Be7 Re8 17.Bd6 Nf6, I felt I was better, but I struggled to find a way to keep my advantage. On 18.Rac1, I decided that Black would likely play 18…b5 (if 18…Be6 19.Nd4 Bd5 20.Qb2 looks annoying, threatening both Nf5 and Nb5). White would then like to play 19.a4, but without a rook on a1, Black can simply play 19…bxa4.

With this in mind, I played 18.Rdc1?!. Now 18…b5 walks right into 19.a4! bxa4 20.Qxc4 Qxc4 21.Rxc4 Be6 22.Rcxa4 with a winning position. However, with the rook’s departure from the d-file, the Bd6 is slightly loose, and Kuljasevic was alert to this fact. After 18…Be6 19.Nd4 Ne4! 20.Nxe6 Rxe6 21.Qxc4 Qxc4 22.Rxc4, Black gets his pawn back with 22…Nxd6. Unfortunately, White’s other rook is sitting on a1 instead of d1 …

In round 2, I was white against IM Ben Finegold (2513 FIDE). Ben has long been one of America’s strongest IMs, but he somehow has never managed to get the GM title. He surprised me in the opening, but my biggest mistake was thinking that I had the advantage (and so should press). In reality, I was worse, and I should have been playing for equality. By the time I realized what was up, I was already clearly worse, and soon found myself in the following position after 24…Bb4-e7!.

Black is now threatening to take twice on g3, followed by …Bh4, pinning and winning the rook. White can’t easily defend the g3-pawn and pushing it to g4 isn’t an option because f4 becomes irremediably weak. Meanwhile, 25.f4 doesn’t help as after 25…Nf6, White can’t really guard the e4-pawn properly. The Queen is needed to guard the b5-knight; the b5-knight needs to stay put because the b3-knight is undefended behind it.

In time pressure, I decided to play 25.Nd2. The idea is to guard the g3-pawn from f1. Finegold continued with his plan: 25…hxg3+ 26.hxg3 Nxg3 27.Rxg3 Bh4 28.Nf1 Nf4 29.Qc4! (not 29.Bxf4 Qxf4, and White can’t deal with the threat of …Rd2 later) Nh5 30.e5! (hitting the Bh4, so not giving Black time to calmly pick up the e5-pawn). After 30…Bxg3+ 31.Nxg3 Nxg3 32.Kxg3 Qxe5 33.Qf4, White has escaped the worse and I managed to draw the endgame after the queen exchange.

However, instead of 25…hxg3+, Black had the very strong 25…Rxd2!. The point is that on 26.Qxd2, Black gets two minor pieces for the rook with 26…Bxb5; meanwhile, after 26.Bxd2, Black has 26…Qb6+!, picking up the bishop on a6. It was this move that Ben missed. I saw this before playing the knight move, but as I didn’t see what else to do, I figured I might as well try. Luckily, it paid off.

With two uninspiring draws from my first two games, I faced IM Ray Robson (2527 FIDE) as black in round 3. Ray is one of America’s most promising talents in a while, and he beat me in a very complicated game in July at the World Open (http://vbhat.wordpress.com/2009/07/12/taking-a-ride-on-the-reading-the-first-half-in-philly/).

This time, I decided to go with a Ruy Lopez instead of the French Winawer. However, Ray knew the opening line I chose better than me and quickly got an advantage. After 22.e5, we reached the following position:

White controls more space across almost the entire board and has the bishop pair to boot. Among other things, he’s currently threatening 23.Bxg6 fxg6, when Black will never be able to really challenge White’s central dominance and will be stuck with rather weak kingside pawns. I played 22…Nf8, which isn’t particularly inspiring, but Black’s position is a bit tougher to crack than it looks like at first glance.

For all his pluses, White has two problems – Black’s knight on g4 controls some nice squares around White’s king, and White’s central pawns, if they can’t make any threats, can quickly become weak if Black puts pressure on the d-file.

Ray decided to solve this by going after the knight on g4 (and the pawn on h5), but landed himself in some trouble. After 23.Bf5 Qb6 24.Kg2 Rbd8 25.Nh2 Nxh2 26.Kxh2 c4 27.Be3 Qb7, I was threatening to take on e5 and then d5. Thus, Ray played 28.e6, and after 28…Bf6 29.axb5 Qxb5 30.Qxh5 Re7 31.Re2, we reached the following position.

Black is now down a pawn and it looks like White is in charge, but unlike the previous diagram where Black was retreating, Black now has an opportunity to move forward. After defending for a while, I didn’t need a second invitation.

With 31…fxe6 32.dxe6 d5, I got my pawns moving. Surprisingly, White’s position is not so easy. One threat, for example, is 33…d4 34.Bf2 Rd5 35.g5 g6, trapping White’s queen! Ray played 33.Bd2 d4 34.Ra5 Qb6, but there is still no real attack on the kingside, while Black is definitely threatening something with his pawn mass. After 35.g4 d3 36.Rg2 Bxb2 37.Be4 Qd4 38.Re5 c3, we reached the following position:

Black’s pawns are rolling, so Ray found the only way to make things interesting with 39.Be3!. By this point, by the way, we were both in horrible time pressure. With no second time control, all we could rely on was the 30-second increment. Thus, intuition has to take over, as the position is still rather complicated and you don’t have time to calculate everything.

I immediately played 39…Qxe3 40.Bh7+ Nxh7 41.Rxe3 d2, but after 42.Qc5, I had to think a little bit. I have two minor pieces for the queen, but with my d- and c-pawns, I felt my position had to be winning when I gave up my queen.

Still, it’s not trivial. For example, 42…Ree8 43.e7 Rd7 would lose to 44.Qc4+! Kh8 45.Qf7!, and White’s threats carry the day. Similarly, 42…Kf8 also loses, this time to 43.Rd3!! Rxd3 44.Qc8+ Re8 45.e7+ Kf7 (not 45…Kxe7 46.Re2+) 46.Qc4+ and after 47.Qxd3, Black’s pawns are frozen.

With that in mind, I played 42…Rxe6!, and after 43.Rxe6 d1/Q 44.g5 (trying to restrict Black’s knight on h7) Nf8 45.Re7, I found 45…Qh5. White’s h4-pawn is surprisingly weak, and moving the king up to the 3^rd rank entails allowing Black to play …Ng6, hitting the pawn again. Thus, Ray tried 46.Qc4+ Kh8 47.f5, but at this point, I took advantage of some nice chess geometry to play 47…c2!. The pawn is going to queen, but if White plays 48.Rxc2, then 48…Rd4 wins the h4-pawn and White is not going to survive that attack. Ray tried 48.f6, but after 48…c1/Q (double promotion!) 49.fxg7+, Black has 49…Bxg7!. The end was 50.Qxc1 Qxh4+ 51.Kg1 Bd4+ 52.Kf1 Qh1+, and White is going to lose his queen, so Ray resigned.

Phew!

In round 4, I was white against Gabor Papp. This time I got in an effective opening surprise and he reacted poorly. After 24.Qd2-h6 Rf8-e8, we reached the following position:

White’s position looks dominating, but how does he make progress. My original plan was to play 25.Nf5 here, but while 25…gxf5 loses after 26.Qg5+ Kf8 27.Rxe8+ Qxe8 28.Rd8, 25…f6 was a considerably tougher nut to crack over the board. I looked at all sorts of variations for close to half an hour before deciding that nothing was particularly clear.

I then spent another chunk of time to find something else to do, and decided on 25.c4. It’s a rather tricky move, but sadly, not the right one. After something like 25…Nc3 26.R5d6 Qxe2, White has the amazing 27.Qd2!! which wins a piece! Other moves similarly lose after 26.R5d6. White also gets the advantage after two other natural replies by Black: 25…Ra8 26.Rxa8 Rxa8 27.Nf3! is good for White, while 25…Nd4 26.e3! is also pleasant. Papp admitted after the game that he didn’t see what was wrong with 25…Nc3, but that after my 40 minute think, he figured I had found something. Buying the act, he decided to play the correct 25…Nc7!.

After 26.R5d6, he again reached correctly with 26…Qxe2 27.R8d7 Qxc4!. Black needs to break the pin on the c7-knight, so he needs to free the b5-square for the knight. He also threatens Re1+ and Qf1+ himself, so I played 28.Nf3. After 28…Nb5, though, I had no choice but to go for a repetition with 29.Ng5 Re1+ 30.Kg2 Qf1+ 31.Kf3 Qe2+ 32.Kg2 and the game was agreed drawn shortly.

Back in that diagram, though, White did have a win with 25.Nf5 – after 25.Nf5 f6 26.Rxe8+ Qxe8 27.Qd2! Ra8 28.Nh6+ Kh8 29.a4!, White is winning. 29.Rd7 would have failed to 29…Nd4!!, but by kicking the knight away first, White’s rook and queen invade with decisive impact.

So with a few ups and downs, I found myself with 2.5/4. At the time, GM Perelshteyn was leading with 3.0/4, along with IMs Finegold and Antal.