Saturday, February 18, 2017

Playing an IM with a brand new opening

This next game happened in round 6 of the 2014 Budapest Spring Chess Festival. It is long and has some really crazy twists to the story. First, when I looked at the white repertoire of IM Szalanczy, I realized pretty quickly that I had no hope if I played my standard Sicilian as black. He was just too good with white, even doing very well against GMs. I couldn't find any weaknesses, plus he played so many different variations that I couldn't even try to prepare for any specific one.

That left me in a quandary, because I only play the Sicilian as black. Isn't it insane to try a brand-new opening for the first time against an International Master? How could I hope to survive such a thing?

Well, one thing I noticed when I went through his games in the database was that in one particular line of the French he had only played against it three times, and he hadn't won any of those games. I'm not fond of the French and it has so many complications, but I did like this one particular variation. So I studied it like crazy for hours and hoped he wouldn't vary. I got lucky. He played right into it. Let's see how it went.
IM Emil Szalanczy (hunonchess.com)
Szalanczy,IM Emil (2240) - Cross,Ted (2034) [C18]
Budapest Spring Chess Festival Budapest (6), 19.03.2014

1.e4 e6 2.d4 d5 3.Nc3 Bb4 4.e5 c5 5.a3 Bxc3+ 6.bxc3 Qa5 7.Bd2 Qa4
This was it, the line of the Winawer French that I was hoping for!
8.Qb1 Nc6 [8...c4]
But here already I managed to forget the correct way to play it. I needed to play c4 right away. Luckily he didn't punish me, which he could have done with 9.Bb5 Qa5 10.dxc5 Nge7².
9.Nf3 c4 10.g3 Nge7
I knew the correct move here was Bd7, but I suddenly became afraid that if he took the pawn on b7 I wouldn't remember how to punish it properly. [10...Bd7 11.Bg2 (11.Qxb7? Rb8 12.Qc7 Nge7 13.Be2 Qa6 14.0–0 Rb7 15.Qd6 Nf5 16.Qc5 Rb5 17.Bxc4 dxc4 18.Qxc4 Na5–+) 11...0–0–0÷]
11.Bh3 b6 12.0–0 Bd7 13.Nh4 0–0–0 14.Qb2 Rde8 15.Rae1 Rhg8 16.Bg4 Nf5
This may look odd, but while looking over all the GM games in this line, I saw that pretty much the only way black gets any play at all is by making this knight move at the proper time and trying to get some counterplay. Not saying I made it at the right time, but I made the offer and he refused it.
17.Ng2 
[¹17.Nxf5 exf5 18.Bh5 Be6²]
17...g6 18.h4 h5 19.Bf3
By refusing to take the knight on f5, white essentially allowed me to create a fortress that is very hard to break down. In fact, it looked so difficult that shortly white decided to sac an exchange!
[¹19.Bxf5 exf5 20.Nf4 Be6±]
19...Nfe7 20.Nf4 Rd8 21.Nh3 Be8 22.Bg2 Nf5 23.Ng5 Rd7 24.Nh7 Rh8 25.Nf6
I was very unhappy that I had allowed him to bring this knight to f6 where it could strangle my position. I took the time to get rid of it, but that meant allowing a pawn to throttle me on f6 instead. Still the fortress is hard to do anything against.
25...Rb7 26.Ra1 Nfe7 27.Rfb1 Bd7 28.Qc1 Ng8 29.Bf4 Nxf6 30.exf6 Re8
So now white decides he isn't going to get anywhere without desperate measures, so he sacs an exchange. As long as my pawn chains don't break down, the only route into my position is via e5, and there just isn't enough there for white to do anything.
31.Rb4 Nxb4 32.axb4 Qc6
32...Qb5! 33.Qa3 Bc6µ is what the computer says, but I struggle to understand it. My best guess is that the computer believes black can win by stuffing white's bishops with the pawn wall, while breaking through on the queenside.
33.Qe3
The computer doesn't like white's move and thinks he gets equality with 33.Ra6! e5 34.Bxe5 Bf5=. Now comes a long maneuvering battle where white tries to find some way to break down the fortress and fails.
33...b5 34.Qe5 a6 35.Ra5 Kd8 36.Kf1 Bc8 37.Ke2 Bd7 38.Kd2 Bc8 39.Bf3 Rh8 40.Ra1 Bd7 41.Qg5 Bc8 42.Re1 Bd7 43.Ra1 Bc8 44.Be2 Rb6 45.Rh1 Re8 46.Be5 Rh8 47.Qe3 Rb7 48.Bf4 Ra7 49.Ra1 Ra8 50.Bh6 Kc7 51.Bg7 Re8 52.Qh6?!
I knew full well that white was playing absurd moves like this only to try to goad me into doing something, anything other than simply maintaining the fortress.
52...Bd7
So I could have punished him here with 52...Qd6! 53.Qe3 e5 54.dxe5 Rxe5 55.Bf8 Qxf6 56.Qc5+ Kd7™–+. Frankly, though, look at the position. Over the board it looks a bit scary, so I did look at trying to break out using my extra exchange, but each time I felt it was dangerous, so I decided to play it safe and simply keep up the fortress. 
53.Qe3 Kb7 54.Bh6 Rh8
Here again I had a hard-to-fully-see line to push for a win. 54...Qb6 55.Qe5 a5 56.bxa5 Rxa5 57.Rxa5 Qxa5 58.Qd6 Bc6 59.f3 b4 60.Qxb4+ (60.cxb4 Qa1 61.Qc5 Ba4 62.Bd1 Rc8 63.Qa5 Qxd4+–+) 60...Qxb4 61.cxb4 e5 62.Be3 exd4 63.Bxd4 Bd7 64.c3 Kc6µ 
55.Bf4 Rae8 56.Bf3 Ra8 57.Qe5 Rhe8 58.Qg5 Rh8 59.Be5 Rae8 60.Qf4 Bc8 61.Bd6 Rd8 62.Be7 Rde8 63.Qg5 Qd7 64.Bc5 Qc6 65.Re1 Qd7 66.Re5 Kc6 67.Bg2 Bb7 68.Qf4 Qd8 69.Bf3 Qd7 70.Bg2 Qd8 71.f3 Qd7 72.Ke2 Qd8 73.Kf2 Qd7 74.Kg1 Qd8 75.Kh2 Qd7 76.Bh3 Bc8 77.Qg5 Bb7 78.Re1 Bc8 79.Bg2 Bb7 80.g4?
As with his earlier exchange sacrifice, white gets tired of trying to get anywhere and pushes a bit too hard.
80...hxg4
Perhaps a slightly better try at punishing him was ¹80...Qc7+ 81.Re5 (81.Kh3 Rh7 82.Qe5 hxg4+ 83.fxg4 Reh8–+) 81...Bc8 82.Kg3 Bd7 83.Qf4 hxg4 84.fxg4 Kb7 85.Bf3 Reg8 86.g5 Bc6µ
81.fxg4 Qc7+ 82.Re5 Rh7 83.Kg1 Qd7 84.Re3 Bc8 85.Kf2 Reh8 86.Rh3 Re8 87.Bf3 Qc7 88.Qe5 Bd7
I knew when he played the queen to e5 that my chance to win was by taking it, but I looked and looked and just couldn't see deeply enough, so again I decided to keep the draw in hand. [88...Qxe5 89.dxe5 a5 90.h5 gxh5 91.gxh5 a4 92.Be3 Bb7 93.Rh1 Ra8 94.Ra1 Kd7 95.h6 Bc6 96.Be2 Rah8 97.Rg1 Kd8 98.Ke1 Rxh6 99.Bxh6 Rxh6–+]
89.Be7 Reh8 90.Qg5 Kb7 91.Kg2 Bc6 92.Rh1 Kc8 [92...Ra8µ] 93.Bc5 Re8 94.h5 gxh5 95.Rxh5 Rhh8 96.Qe5 Kd7
The computer prefers 96...Qxe5 97.dxe5 Reg8 98.Kf2 Kc7 99.Bd4 Kd7 100.Bd1 Bb7 101.Bf3= and says black is better, but I think it just doesn't understand fortress positions. 
97.Rg5 Rh7 [97...Qxe5 98.dxe5 Rh7 99.Rg7 Rxg7 100.fxg7 Rg8 101.Bf8 Ke8 102.g5 Rxf8 103.gxf8Q+ Kxf8 104.Kg3=]
98.Kg3 Reh8 99.Rg7 Qxe5+ 100.dxe5 Rxg7?
I thought this was clearly the right way to play, because I didn't see that sacrificing a pawn here would actually be better for me. [100...d4! 101.Bxc6+ Kxc6 102.cxd4 Kd5 103.Kf3 c3! 104.Kf4 Rh2 105.Rxf7 Rf2+ 106.Kg5 Rxc2 107.Rc7 Rf2 108.Kg6 c2 109.Bb6 a5! 110.Bxa5 Rh1 111.Rc5+ Ke4 112.f7 c1Q 113.Rxc1 Rxc1 114.Bd8 Rc8 115.Bf6 Rfc2–+] Now that's a hard line to see!
101.fxg7 Rg8 102.Bf8 Ke8 103.Kf4 Rxf8
This is why I played this line, believing that I could give back the exchange to go into an endgame up a pawn. I had no idea what a weird resource was available to white. Can you imagine a position where taking a free rook with a pawn with check is wrong? Well here it is!

104.gxf8Q+?
Who could have thought this was a mistake? Check out the amazing double-exclam move white had here! And on move 104 no less.
104.Kg5!! Rg8 (104...Ke7 105.gxf8Q+ Kxf8 106.Kf6+–) 105.Kf6 Rxg7 106.Kxg7 Ke7 107.g5 Bd7 108.Bh5 Be8 109.Kg8+–
104...Kxf8 105.g5 Ke7 106.Bh5 Be8 107.Ke3 Kd7 108.Kd4 Kc6 109.Bf3 Kb6 110.Bg4 Kc6 [¹110...Bd7=]
 I overlooked his tactic here or I would have played Bd7. I could tell that the spectators and IM Szalanczy thought I had blown it and was lost. To be honest, I thought I had as well, but I took a long time after 111.g6 and examined the position deeply and realized it was still a draw.
111.g6 fxg6 112.Bxe6 g5 113.Bxd5+ Kc7 114.Be4 Bd7 115.Kc5 Be6 116.Kd4 Kb6 117.Bd5 Bf5
Even taking the bishop is still a draw, though during the game I felt that taking the pawn on c2 was a must.
[117...Bxd5 118.Kxd5 g4 119.e6 Kc7 120.Ke4 Kd6 121.Kf4 Kxe6 122.Kxg4=]
118.e6 Bxc2 119.Ke5
Nothing white does here makes any progress.
[119.e7 Bg6 120.Be6 Be8 121.Bf5 a5 122.bxa5+ Kxa5 123.Kc5 Ka4 124.Bc2+ Ka3 125.Bd1 Kb2 126.Kb4 Kc1 127.Be2 Kd2 128.Bg4 Kc2=; 119.Bf3 Bg6 120.Kd5 a5 121.bxa5+ Kxa5 122.e7 Ka4 123.Be4 Be8=]
119...Bg6 120.Kf6 Be8 121.Kxg5 a5 122.bxa5+ Kxa5 123.Kf6 b4 124.cxb4+ Kxb4 125.Ke7 Bh5
½–½

I was really proud of this result, amazed that I could play a totally new opening for me against an IM and survive.

No comments: