Wednesday, January 2, 2019

My Best Games: vs. IM Kevin Wang

Happy new year! In this sequel to my last post, I'll be analyzing one more of my best games, this time against International Master Kevin Wang, who I played in the 4th round of the 2018 Washington Chess Congress, held in Falls Church, Virginia. In contrast to my game against Dragun, where we had an uneventful opening that skipped right into the endgame, this game had a little more action!

Wang-Paciorkowski; 2018 Washington Chess Congress (round 4)

1.e4 g6!?

This is a highly provocative opening choice, especially against a strong player. To provide some context, at this point in the tournament I had 2.5/3 after having beaten GM David Berczes with black in round two and then drawing GM Alexander Shabalov in round three; I very much wanted to maintain my momentum and was playing for a win with black. However, instead of going into any sharp hyper-aggressive lines right out of the opening, my opponent opted for a calmer approach.

2.Nf3 Bg7 3.g3 d6 4.Bg2 Nf6 5.d3 (D)

White sets up a reversed King's Indian formation which soon turns into almost a kind of Closed Sicilian. The game continued 5. ... c5 6.0-0 Nc6 7.c3 0-0 8.Nbd2 (D)

Normally in the Closed Sicilian, white's knight is actually on c3 and later (after Nf3-e1, f2-f4 and sometimes g3-g4) re-routes either to e2-g3 or to d1-f2 combined with a pawn storm on the kingside. Instead of that, my opponent wants to play more directly in the center and makes preparations to play d3-d4. Meanwhile, my position is solid and I decide to harmoniously complete development.

8. ... Qc7 9.a3 This is a useful prophylactic move in these kinds of positions to give extra cover to the b4 square. For example, on the immediate 9.d4?! white always has to consider the consequences of 9. ... cxd4 10.cxd4 Nb4!? (10. ... Bg4! is also a strong move that should be considered), aiming at c2 and d3. On the minus side though, 9.a3 wastes a move with a pawn and opens up a hole on b3, which may or may not become important later...

9. ... Rd8 10.Re1 White still was not ready to play 10.d4?! since after 10. ... cxd4 11.cxd4 the thematic response 11. ... Bg4! puts uncomfortable pressure on the fragile center pawn duo. If one of them is forced to prematurely advance, they could become weak targets rather than strengths.

10. ... Ne5!? (D)

Offering white to exchange on e5 so I can open the d-file for my rook.

11.Qe2 As I had anticipated, White declines to make the trade. After 11.Nxe5 dxe5 white has to figure out what to do about his backward pawn on d3 which could turn into a weakness later. Now, white plans d3-d4 with tempo to finally establish a strong duo of pawns in the center. In response I quickly activate the rest of my pieces.

11. ... Bg4 I was less satisfied with the more passive 11. ... b6!? 12.d4 Nxf3+ 13.Nxf3 cxd4 14.cxd4 Bb7, when white could gain space with either 15.d5 or 15.e5!?, perhaps with some advantage.

12.h3 Bxf3?! 12. ... Bd7 13.d4 Nxf3+ 14.Nxf3 e5!? was strictly speaking a better option, not giving up the bishop pair. However, I had a specific plan in mind which will be revealed shortly. 13.Nxf3 Rac8 14.Be3 (D)

This is an important moment. I have willingly given up the bishop pair, but at least white had to waste a tempo with h2-h3 to get it, and in the meantime I have fully completed development. However, if I dawdle around, white will continue with something like Rac1 and d3-d4 occupying the center, after which my position will quickly become cramped and unpleasant. So to avoid running into a worse position like that, I have to come up with some kind of plan for counterplay either right this move, or on my very next turn. Fortunately, I did in fact have such a plan that I had been thinking about since move 11. It may not have been objectively the best way to play, but it certainly led to an interesting position after a couple more moves.

14. ... Nfd7 A preparatory move... 15.Nd2 (D)

White also makes a preparatory move. Instead, on the immediate 15.d4!? I can play 15. ... Nxf3+ 16.Bxf3 cxd4 17.cxd4 and continue with either Qc4 or Qc2, getting a little counterplay via the c-file. Now that white's knight is on d2, my Ne5 will having nothing to take on f3 and so will be forced to retreat after the imminent d3-d4. Now is the time for me to reveal my plan...

15. ... c4!? This is my ambitious idea. If white bypasses with 16.d4, then I plant a strong knight on d3 and can wreak all sorts of unpredictable havoc in white's camp from there. If white exchanges with 16.dxc4, then after 16. ... Nxc4 17.Nxc4 Qxc4 18.Qxc4 Rxc4 we get a complex ending  - white has some advantage thanks to the bishop pair, but black is not without play and can create something on the queenside with a timely b5-a5-b4 minority attack. My knight on d7 is well prepared to meet the Bxa7 pawn grab with b7-b6, trapping the intruder.

My opponent chose not to enter the endgame and we continued with 16.d4 Nd3 17.Reb1 b5 (D)

The game has now become double-edged. It is difficult to say at the moment if the knight on d3 is more of a strength or a weakness. Right now its powerful influence radiates in all directions deep in enemy territory, but if white has time to undermine my supporting pawn chain with b2-b3 and a3-a4, then my intrepid explorer could get in trouble since he has no retreat squares! In any case, this is actually the most critical moment of the whole game and fortunately for me, my opponent did not make the best decision.

The game continued 18.Bf1?! (D)

My opponent's last move has the plan of Qd1, Bxd3, Qd1-f1 and finally Qf1-xd3 to win a pawn, but this is way too slow. It actually would have been quite strong to play more energetically with 18.a4! when after 18. ... a6?! 19.axb5 axb5 20.b3! white opens the queenside for his otherwise awkwardly placed rooks and starts undermining the support system for the Nd3. It seems that white would have a significant advantage in that position.

From the diagram, I wasted no time and vigorously opened the center to clear lines for my pieces: 18. ... e5! 19.Qd1 White was unwilling to play 19.d5!? himself, liberating the c5 square for either one of my knights. 19. ... d5! (D)

It felt great to be able to get this position over the board. Everything just works beautifully for black and the practical difficulties proved too much for my opponent to handle. Neither pawn capture is particularly desirable for white.

If 20.dxe5 then 20. ... dxe4 21.Nxe4 N7xe5 connects my knights and opens the d-file leaving white's queen uncomfortably staring into the gaze of my Rd8.

20.exd5 is not quite as bad, but even so I can calmly recollect the pawn with 20. ... Nb6 and if white completes his lengthy pawn grab with 21.Bxd3 cxd3 22.Qf1 Rxd5 23.Qxd3 black can continue playing actively with 23. ... f5! creating the unpleasant threat of f5-f4, trapping the bishop on e3. Considering white's dormant rooks on a1 and b1 and the immense pressure on d4, black has more than sufficient compensation for one measly pawn.

Instead, white opted to keep the tension and immediately go for 20.Bxd3 cxd3 21.Qf1? (D)

Instead it was necessary to insert 21.exd5 Nb6 and only then play 22.Qf1. Now, I have a neat tactical resource to preserve my pawn on d3.

21. ... dxe4 22.Nxe4 f5 Intending to follow up with e5-e4 after the knight moves, securing a powerful advanced protected passed pawn. White plays the most testing way with 23.Ng5 (D)

Threatening a fork on e6, but I had seen in advance that I can simply cover all my bases with 23. ... Qc4 Defending both e6 and d3. I now intend to continue with e5-e4, when the knight on g5 could even get trapped with h7-h6. If white stops me by taking on e5, then all of my pieces spring to life after ...Nxe5.

24.Nf3 e4 25.Nd2 Qc6 (D)

This is a highly unusual position. We both have protected passed d-pawns but the black one on d3 has a much stronger cramping effect than the white one on d4. If white does nothing, then I will apply further pressure with Nb6 and eventually a5-b4 to undermine the d4 pawn. I suspect my opponent grimly realized at this point how far south things had gone and rather than sit passively to watch, he frantically tried to mix things up.

26.a4 bxa4 27.d5 Pure desperation. 27. ... Qa6 There is no need to allow 27. ... Qxd5!? 28.Rxa4 when a7 falls next and white gets some counterplay (although black still has a large advantage). 28.b3 Desperately trying to activate the rooks. 28. ... Bxc3 29.Rxa4 Qb5 (D)

White has succeeding in making things messy, but in the process I have won a pawn and all my pieces are actively placed. Black should be winning now.

30.Nc4 Threatening a cheap fork on d6. 30. ... Qxd5 31.Rc1 Bd4 32.Ra5 Rc5 33.Rxa7 Ne5 (D)

Simple centralization usually gets the job done. If 34.Nxe5?, then 34. ... Rxc1 35.Qxc1 Bxa7 picks up an exchange. Otherwise, I have a triple attack on c4 which is only defended twice.

34.Ra4 Now I simplify the position by trading pieces, which generally should be the primary goal when one is ahead material: 34. ... Nxc4 35.Bxd4 Qxd4 36.Raxc4 Rxc4 37.bxc4 (D)

From here, I could win slowly by simply going after the c-pawn with 37. ... Rc8 and then queening the d-pawn by brute force, but there is a much more expedient method: simply ripping apart the white king's pawn shelter and playing for mate.

37. ... e3 38.fxe3 Otherwise e3-e2. 38. ... Qxe3+ 39.Kg2 Rb8 Activating all my forces in the attack. A rook and a queen may only be two pieces, but when working together they are more than capable of checkmating an exposed king. 40.h4 Rb2+ 41.Kh3 (D)

41. ... Qe4 The final quiet move to finish things off. White cannot defend against Qg4# without allowing Qg2#. 0-1

Next time I think I'll go through some more endgames!

Wednesday, December 5, 2018

My Best Games: vs. GM Kamil Dragun

Hello again and welcome back! In my next few posts I want to share some of my best games against strong players and see what we can learn from them. I play fairly regularly in open tournaments around the east coast and get plenty of opportunities to play GMs and IMs. That experience is invaluable for me but also makes for some great instructive games to analyze. Although I would primarily say this post is designed for players rated 1000-2000, even experts and some masters would probably find it an interesting read.

The first one we'll look at is my game with white against the Polish grandmaster Kamil Dragun (FIDE 2578). This was played in the 2018 Atlantic Open in Virginia. The opening was relatively benign (there's no need to go through it in any detail), and after an early queen trade we reached the following endgame:

Paciorkowski-Dragun; 2018 Atlantic Open

1.Nf3 Nf6 2.c4 e6 3.g3 b6 4.Bg2 Bb7 5.0-0 Be7 6.d4 0-0 7.Nc3 Ne4 8.Nxe4 Bxe4 9.Ne1 Bxg2 10.Nxg2 d5 11.Qa4 Qe8 12.Qxe8 Rxe8 13.cxd5 exd5 (D)

This is where I want to start the discussion. Theoretically, this endgame is supposed to be equal, right? Equal material, both sides have a pawn in the center, both sides don't have any glaring issues with their king, and so on. However, "equal" is not synonymous with "drawn". There are some important subtleties in this position that make the endgame rich with ideas; I explain them below.

1) We have an asymmetric pawn structure. White has an e-pawn vs. black's c-pawn. This simple enough fact by itself already gives the position some life and creates an imbalance.
2) Black's pawn is on b6, not on b7. Well, why on earth should this matter? It seems like such a tiny little detail that nobody would even bother to think of it. Yet, if you put this position in front of a grandmaster, I guarantee you that they will immediately highlight this factor as being one of the most significant talking points about the position.

Imagine that the pawn was back on b7 for a moment. Now, compare the position above to this position:

What's the difference? Well, in both cases, white's only open line for his rooks is the c-file. However in the second position, black's pawn on c6 is solidly defended by the pawn on b7; in fact, the entire b7-c6-d5 chain is quite resilient and it will take a large, organized effort for white in the form of a minority attack with b2-b4-b5 to put any pressure at all on black's structure.

Now look again at the first diagram. Because black's pawn is on b6, he can never set up this b7-c6-d5 chain. In fact, as soon as white puts a rook on c1 he will immediately be putting pressure on the backwards pawn on c7, which black can only ever solve by playing c7-c5. At that point, white would take d4xc5, and black either recaptures with a piece, in which case he is left with an isolated d-pawn, or he recaptures with the pawn, in which case he is left with this hanging pawn structure, which I show below.

Now, if black can get a position like this, he will have essentially fully equalized the game. Granted there is plenty of play left, but black's slightly inferior pawn structure (3 islands vs. 2 for white) is fully compensated for by his control of the center, potential targets on b2 and e2, and the potential for good piece activity.

So with all that in mind, let's go back again to the starting position for my endgame against Dragun:

We know that if black can achieve c7-c5 without any major concessions, he should be completely fine. But if white can stop him from doing that, then black will essentially be forever saddled with this backwards pawn on the c-file. So, from this position, we should know that white's objective is to not allow black to play c7-c5, and if white can succeed in this, then in theory, the position won't be equal, but white will be slightly better.

Now, let's finally start going through the game. It's white to move in the diagram, and I continued with 14.Rd1 (D)

To the untrained eye, unaware of white's strategy, this move looks more or less random. But hopefully now you can see that the point is to prevent 14. ... c5 because after 15.dxc5 bxc5 16.Rxd5 white simply snaps up the undefended d5 pawn. Dragun replied with 14. ... Bd6, which looks attractive because it attacks e2, but believe it or not I would argue that this move is already slightly inaccurate because it does nothing to help black play his c5 break.

In 1987, none other than Anatoly Karpov himself actually had this exact same position with black against the strong Swedish GM Ulf Andersson (who famously made a career out of playing these kinds of endgames with white for a win). Instead of 14. ... Bd6, Karpov continued with 14. ... Na6! 15.Be3 Rad8! defending his d-pawn. He soon followed up with c7-c5 and equalized the game, which ultimately ended in a quick draw.

Now, back to my game. After 14. ... Bd6, my e-pawn is hanging, so I shield the e-file and develop my last minor piece with 15.Be3 Notice how even this bishop has its eye on the c5 square. The game continued further with 15. ... f6 16.Rac1 g5 (D)

Now I am fully developed, and most of my pieces are directly fighting against the c5 break. With black's last move (g7-g5), he takes away the f4 square from my knight. At the same time, it opens up a hole on f5 and weakens his king a little, and moreover black still has to develop his queenside pieces.

So how to proceed with white? We've developed everything and for now black cannot play c5, but what to do next? Well, look at white's position and find the worst-placed piece: that's the Ng2, which is just sitting there doing nothing. Now ask yourself where that knight should go. The most obvious square to me was f5, which has just been weakened with black's latest g-pawn thrust. But you'll notice it's kind of tricky to maneuver it there; e3 is already occupied by my bishop, and if my bishop moves away I hang e2. So I came up with an interesting plan of f3, Kf2, Bd2, and then finally Ng2-e3-f5. In the meantime black decided to embark on quite an ambitious expedition with his a-pawn on the queenside.

From the diagram above, the game continued 17.f3 a5 18.Kf2 a4 19.Bd2 a3 20.b3 (D)

Notice how I just completely ignored black's demonstration on the queenside. I could have stopped him with a2-a4 or a2-a3 at some point, but both of those moves would have created holes in my position on either b4 or b3 which his knight could eventually maneuver to. Now the queenside remains semi-closed and black still cannot play 20. ... c5? since the simple reply 21.Ne3! just wins the d-pawn; after 21. ... cxd4 22.Nxd5 white has all of his pieces mobilized and threatens both Nxf6+ and Nxb6 with a winning position.

From the diagrammed position, it's black to move, and my opponent played 20. ... Kf7 (D)

The idea behind his last move is to meet 21.Ne3 with Ke6, defending both the d5 and f5 squares. So what should we do next? Black's kind of stopped our knight maneuver for now, so maybe we should look for another plan in the meantime. Later, we can always play Ne3 to hit d5 and threaten to jump into f5.

How do you find a plan in positions like this? If you've already more or less optimized your pieces, then you should move on to consider pawn breaks. This makes logical sense. If you are better mobilized and developed, then by simply opening up the position with a pawn break, you can often take advantage of your superior piece activity to win some material.

So what are white's pawn breaks here? I see three of them: e2-e4, f3-f4, and h2-h4. All of those moves are pawn breaks because they create pawn tension - if you play any three of them, you will be threatening pawn takes pawn to open the position. Go ahead and take a moment to see which one is the best for white, and then I'll talk about each one in detail.

The e4 break might seem attractive at first glance because it's in the center, but the strategic problem with it is that it would allow black to liquidate one of his big weaknesses - the pawn on d5. Suppose white got two moves and played Rd1-e1 and e2-e4. Then black would absolutely play d5xe4 and it doesn't matter how white recaptures; black's next two moves are Nb8-d7 and c7-c5, after which white's advantage is probably only minimal, if existent at all. Remember, black cannot play c7-c5 right now because then his d5 pawn would become exposed and vulnerable to capture. If we ever play e4, black could exchange pawns and his weak d5 would disappear. Taking all of that into consideration, e2-e4 is not the strongest plan, at least not right now.

The f4 break is interesting, and if black took, you could recapture with the knight and put some problematic pressure on d5. However, f3-f4 is positionally risky, because you create a hole on e4. Although black's knight can't get there now, it could later in the future and then you'll wish you still have a pawn on f3 to kick it out. In fact, after 21.f4, black could even immediately play Re4, attacking my d-pawn, and then follow up with h7-h6, reinforcing the pawn on g5. So f4 is not correct here either.

That leaves h2-h4! This is much stronger than the other two breaks simply because it doesn't create any weaknesses in my position. My plan is straightforward: I want to open a second front on the h-file and start invading there with my rooks.

So from the last diagram, it's white to move and I played 21.h4! to which Dragun replied 21. ... h6. After 22.Ne3 Ke6 23.hxg5 hxg5 24.Rh1 (D) we reached the following position:

What should black play here? Take a moment and think about it; we'll revisit this position at the end of the post.

The actual game continued 24. ... Nd7 25.Rh7 Rh8 Black contests the open file 26.Rch1 Rxh7 27.Rxh7 Nf8 28.Rg7 (D)

So thanks to my opening of the h-file I've managed to invade on the 7th rank, and you should notice now that black's king actually has no squares! Also, it's move 28 and black's rook on a8 still hasn't moved. In this position, black finally achieved his pawn break with 28. ... c5 but at this point white has enough of an advantage that this move doesn't equalize anymore. The focus of the game has shifted. Now, instead of c7-c5 being the most important factor, my active rook on the 7th cutting off his king on e6 has become the most important factor. I simply ignored my opponent and brought another piece closer to the attack with 29.g4! Rc8 30.Nf5 (D)

Now I have the dangerous threat of Rg7-b7, attacking b6 and threatening Ng7 mate! Objectively, black should play 30. ... Rc7 to trade the rooks off, but even then I can simply continue with 31.Rxc7 Bxc7 32.Bc1! and pick up the a3 pawn, giving white excellent winning chances.

Instead, Dragun played 30. ... cxd4 (D)

Now white can actually win material by force, and there are two different ways to do so (see if you can find both of them). I only saw the second-best way, but it is still good enough: 31.Nxd6 Kxd6 32.Bb4+ This is the point: if black plays 32. ... Ke6 he is mated in one with 33.Re7, but if he goes anywhere else then I play Rg7-g8 and there is no way for him to defend the pinned Nf8. Black is lost, and the game finished with 32. ... Ke5 33.Rg8 d3 34.Ke3 One last accurate move to stop black's counterplay. 34.exd3?? allows Rc2+ which unpins the Nf8 with advantage to black. 34. ... dxe2 35.Kxe2 Kd4 Or 35. ... Rc2+ 36.Kd3 with a double attack on rook and knight. 36.Kd2 (D)

Black resigned as there are no more tricks and the Nf8 is lost. Strangely enough black is actually up a pawn in the final position!

Now I want to revisit a couple key moments.

Position 1: After the move 24.Rh1

I would call this a critical moment. In the game black played 24. ... Nd7 so that after 25.Rh7 he could play 25. ... Rh8, challenging the h-file, which is logical and what I expected. However, in actual fact black could have ignored the h-file and gone for counterplay with 24. ... c5! My intention (and what black may have been worried about) was 25.Rh6 cxd4 26.Ng4, when I thought white would be doing well since I am threatening both Rxf6+ and Bxg5. However, in reality black can simply play 26. ... Nd7! defending f6 and after 27.Bxg5 Rh8! material is equal and the position is unclear.

Position 2: After 30. ... cxd4

I mentioned that white actually has two ways of winning material. The way that I played is good enough, but well done if you spotted the brilliant 31.f4!! when the threat is 32.Nxd4# and 31. ... gxf4 runs into 32.Bxf4! when black cannot take the bishop because of mate on e7! After 32. ... Bc5 the final blow is 33.b4! overloading the bishop, which cannot move without allowing either Nxd4# or Re7#. It should be no surprise that the strongest continuation involves getting white's last piece into the attack.

I hope you enjoyed that game, and I look forward to sharing a couple more like this!