Wednesday, September 9, 2020

My Games from the 2020 New York State Championship

The 2020 New York State Chess Championship was held online this year from September 5-7 on the Internet Chess Club. With a combination of good opening preparation, tactical awareness and a little luck I managed a clean 6-0 sweep! I've selected a few of my games to analyze.

Let's start with round 2, which was definitely my sloppiest game of the event:

Paciorkowski, Lev (2443) - Ehsani, Yassamin (2095); Round 2

White to move

The opening has gone rather adventurously and my king did not end up finding a secure home. So naturally, I was very relieved in this position to be able to play 19.Qg5 trading the queens. After 19. ... Qxg5 it is difficult to say which recapture is best, but I went with the more ambitious one 20.fxg5!? gaining a tempo to win the b5 pawn: 20. ... Nf5 21.Bxb5 Be6 (D)

White to move

White is up a pawn but the pieces are so badly coordinated that black has pretty good compensation. My advanced kingside pawns have some cramping effect, but they also could turn out to be weaknesses, so this is double-edged. The game continued 22.Ke2 Rac8 23.Bd3 Bb6 24.Rc1 Rxc1 25.Bxc1 Rc8 26.Nf2 and now 26. ... d4! 27.e4 Ne3 (D) upped the stakes even more.

White to move

Black has gotten the so-called "octopus" knight on the third rank, and things do not look so rosy as there is also a threat of ...Rc5-xe5 regaining the pawn. I can play 28.Bxe3, but after 28. ... dxe3 neither 29.Ng4 Bxg4+= nor 29.Nd1? Rc1! -+ are what I want. However, it turns out that white has a surprising idea in this position. 28.Bb2! Preventing Rc5 by attacking d4... 28. ... Bd7 black naturally prepares Re8-xe5 instead. 29.Bb1!? Re8 30.a4! Rxe5 31.Kd3! Now the point of 30.a4 becomes clear: it was necessary to prevent ...Bb5+ in this position. I at first thought white was just winning, because the threat is Bxd4 when the Ne3 is suddenly trapped. I then realized black has an only move, which my opponent found: 31. ... Re6! planning a pin if I take on d4. Things are once again far from clear and after 32.Ba2 Rc6 33.Rc1 Rxc1 34.Bxc1 (D) I actually offered a draw, having just 7 minutes to my opponent's 39 minutes (time control is G/90 + 10).

Black to move

It looks like black has all the cards with a superb knight and lots of weak pawns to attack on the kingside. My engine assures me this is still equal, but during the game I thought black was simply much better. At this point I had an unbelievable stroke of luck: my opponent declined the draw, and then in four moves managed to give himself a lost position! 34. ... Ng2 35.Bd2 Bc7 36.h3 Nf4+? 37.Kxd4! Nxh3+? 38.Nxh3 Bxh3 39.b4! (D)

Black to move

On move 36 something like ...Be6! would have been much stronger, maintaining the tension and freezing my queenside. Now white's active king, 2-1 queenside majority and black's inability to make a passed pawn on the kingside give me a decisive advantage. With 5 minutes left I was able to convert: 39. ... Kg7 40.e5 stopping ...f6 Kf8 41.a5 Bf1 42.Bf4 Ke7 43.Bc4 Bg2 (43. ... Bxc4 44.Kxc4 is winning for white: 44. ... Ke6 45.Kc5 Kf5 46.e6! +-) 44.Bd5 Bf1 45.Kc5 Ba6 46.Kc6 Bb8 47.Bxf7 Bd3 (47. ... Kxf7 48.e6+ Kxe6 49.Bxb8 +-) 48.Bd5 Ba6 49.Bf3 Bc8 50.Bg2 Ba6 51.Bh3 Kd8 52.b5 Bc8 53.Bxc8 Kxc8 54.e6 1-0

Not my best game, but I'll take wins any way I can get them.

After winning my next two games against experts, it turned out that Jason Liang (2403) was the only other player with a perfect score. He had just won with black against GM Vladimir Belous (2635), and GM Timur Gareyev (2692) had drawn with FM Justin Chen (2390). All this meant my round 5 pairing would be black against Liang:

Liang, Jason (2403) - Paciorkowski, Lev (2443); Round 5

1.Nf3 Earlier in the tournament Jason had played 1.e4, but I saw he also has been playing d4 openings recently. 1. ... Nf6 2.c4 g6 3.Nc3 This is a very specific move order that white often chooses to avoid the main lines of the Grunfeld. If black now plays 3. ... d5 4.cxd5 Nxd5, white has lots of options including 5.e4, 5.Qb3, 5.Qa4+ and more recently even 5.h4! I was not very well prepared for any of this, so decided to enter a KID instead. 3. ... Bg7 4.d4 To my eyes this is a curious move order, because now Jason is allowing one of the mainlines of the Grunfeld if I wanted it -- instead 4.e4 would all but force a KID. Anyways, we got to the same position regardless. 4. ... 0-0 5.e4 d6 6.Be2 Na6 7.0-0 e5 8.Be3 Qe7!? (D)

White to move

I rarely play the King's Indian nowadays, and only used it here because of this very specific and unusual idea I wanted to try out. Normally if black chooses to move the queen it goes to e8, but on e7 it has the chief advantages of not blocking the Rf8 and controlling c5. There are some other details, but suffice it to say this is actually the top choice of the modern neural network engines Leela and Stockfish NNUE so it can't be all that bad! Black's top 3 alternatives are Ng4, c6 and Qe8; Qe7!? is the sixth-most common in my database with just 91 games played.

9.Qc2! Jason played this rather quickly, which came as an unpleasant surprise because I think it is the most challenging response, and the main line of NNUE. The idea is to clear d1 for the rook, so now white threatens 10.dxe5 dxe5 11.Nd5 Qd8 12.Rad1 +/-. I cannot play 9. ... c6? on account of the thematic 10.c5! dxc5 11.dxe5 +/-. This means the following transformation is virtually forced: 9. ... exd4 10.Nxd4 Re8 11.f3 c6 12.Bf2 (D) stopping the ...d5 break. At this point I was out of my prep.

Black to move

As a result of Jason's 9th move, we have gotten into an open KID structure. These positions are known today to be simply better for white, and black basically shuffles his pieces around playing for tricks. White's main plan is to start a pawn roller on the kingside: h2-h3, then f3-f4 and g2-g4 to gain more space. I wasn't very comfortable in this position, but in any case the game continued...

12. ... Nc5 13.Rad1 preparing Bg3 to hit d6, so I maneuver a N to e5: 13. ... Nfd7 14.Rfe1 a5 15.Bf1 Ne5 16.Kh1 a4 17.Qd2 Qc7 18.h3 Qa5 (D)

White to move

The first critical moment. Here I was most concerned about 19.f4! Ned7 20.g4!, which gives white a serious plus - it exposes the king but black lacks the piece coordination to properly take advantage of it. Instead white played 19.Nc2? allowing the thematic sacrifice 19. ... Bxh3! 20.Qxd6 (20.gxh3? Nxf3 21.Qxd6 Nxe1 -+ is bad for white; so is 20.f4?! Bg4! 21.fxe5 Bxd1 22.Rxd1 Rxe5! -/+ 23.Qxd6?? Rh5+ 24.Kg1 Be5 -+) 20. ... Bf8 21.Qf6 Bg7 22.Qd6 Bf8 (I calculated 22. ... Ncd3? for a while before finally realizing it just loses after 23.Bxd3 Rad8 24.Qb4! +-) 23.Qf6 Bc8!? declining the repetition; black has exchanged his weakness on d6 for white's valuable h3 pawn, so there is no reason not to continue playing.

We shuffled around for another several moves: 24.Bd4 Bg7 25.Qh4 Ne6 26.Be3 h5 27.Qf2 Qc7 28.Rd2?! (D) Instead 28.a3! would have kept the game balanced.

Black to move

Here is another thematic trick: 28. ... a3! 29.Nxa3 Rxa3 30.bxa3 (I was also ready for 30.Bb6 Qe7 31.bxa3 Nxf3 32.Qxf3? Qh4+ picking up the loose Re1) 30. ... Nxf3 now there are not even any intermezzos because of the mate threat on h2. 31.gxf3 Bxc3 32.Red1 Bxd2 33.Rxd2 Black is clearly better now and I stayed in control, pressing my advantage to a win: 33. ... Qe5 34.Bh3 If white was able to exchange his light squared bishop for my knight he might have counterplay on the dark squares, but I don't allow this. 34. ... Nf4 35.Bxc8 Rxc8 36.Bd4 Qg5 37.Bc3 Ne6 38.Rd1 Rd8 39.Rxd8+ Qxd8 40.Qd2 The endgame is hopeless for white, but keeping the queens on wasn't really an option either since the king on h1 is so exposed. 40. ... Qxd2 41.Bxd2 White's many pawn weaknesses and my passed h-pawn are decisive factors. 41. ... Kf8 42.Kg2 Ke7 43.Be3 Kd6 44.f4 f5 45.exf5 gxf5 46.Kg3 b5 Clearing an avenue for the king 47.cxb5 cxb5 48.Kh4 Kd5 49.Kxh5 Ke4 50.Bb6 Nxf4+ 51.Kg5 Nd5 52.Bg1 f4 53.Kg6 f3 54.Kf7 Ke5 55.Bh2+ Kd4 56.Bg1+ Kc4 57.Ke6 Nc3 58.Kd6 Ne4+ 59.Kc6 f2 60.Bxf2 Nxf2 61.Kb6 Ne4 62.Ka5 Nc3 63.Kb6 Nxa2 64.Kc6 Nc3 65.Kb6 Nb1 66.a4 bxa4 0-1

Heading into the final round, I was leading the field with 5/5 and only Evan Park (2305) had 4.5/5 after having won against Timur Gareyev with black in the previous round. At this point just a draw would secure me clear first.

Paciorkowski, Lev (2443) - Park, Evan (2305); Round 6

Black to move

After declining two draw offers in the opening, we reached this interesting position. Early on I had sent my h-pawn all the way up to h6, but the main question here is whether or not that pawn is a strength or a weakness. I had just castled last move, so there is no longer a rook defending it.

I thought an ambitious way of playing for black would be 16. ... g5!?, when I honestly wasn't entirely sure what to do. The idea is simply ...g4 and ...Bg5 to start rounding up my pawn. During the game I was planning to respond 17.d4 g4 18.Ne5, but I wasn't sure how clear that was since black can still play 18. ... Nxe5 19.fxe5 Bg5. I have to sac a pawn and from afar hadn't seen if I can make strong enough counter threats by doubling on the d-file and putting a rook on d6.

In the game, my opponent did not sense the danger he was facing and played nonchalantly 16. ... Qe7?! 17.Rfd1 Rbc8? (17. ... g5 was essential at this point to get some counterplay) 18.d4 exd4 19.exd4! (D)

Black to move

By far the strongest recapture. I actually think white is already just winning in this position, since black has no way to stop the d4-d5 break, when the center blows open and the king on g8 comes under attack. The Nc6, Bd7 and Qe7 are on perhaps the worst possible squares, as they will only be targets after I put my rooks on e1/d1 and push d5.

The only tactical variation to calculate was 19. ... b5, but it just doesn't work after 20.Bxb5 when I saw 21. ... Nb4? 22.Bxb4 +- and 21. ... Nxd4 22.Nxd4 Bxd4 23.Rxd4 Bxb5 24.Qd2 +- when black's king is fatally weak on the dark squares. Evan had been playing fairly quickly up until this point but started to consume a lot of time as he no doubt started to realize the trouble he was in.

19. ... Kh8 there are hardly better alternatives; 19. ... Na5 20.Bb4 wins the exchange. 20.Qb2 Rcd8 21.Re1 Qd6 22.Rcd1 Ne7 (D)

White to move

With all pistons firing, white is ready to go: 23.d5 exd5 (23. ... Nxd5 24.Bxf6+ Rxf6 25.Qxf6+ Nxf6 26.Rxd6 +-) 24.Bxf6+ Rxf6 (24. ... Qxf6 25.Qxf6+ Rxf6 26.Rxe7 dxc4 27.Rdxd7 +-) 25.Bxd5! From move 23 it had taken me a few minutes to appreciate the strength of this simple recapture, but once you see it, it becomes clear that black is lost. The threat is Be6 or Re6 which cannot be stopped except for 25. ... Nxd5, allowing 26.Rxd5 Qf8 (26. ... Qc6 27.Rxd7) 27.Rxd7! Rxd7 28.Qxf6+! winning. On 25. ... Rf8 I had also seen the cute interference 26.Bf7! +-.

25. ... Bc6 26.Be6 1-0