Old Fort Bay Invitational 2019 part 2

My next game was against a young talent, Nathan Smith. He was just 13 when we played, though his birthday was later in the week. I'm always nervous about playing young people, because often their ratings don't reflect their actual playing strength. They can be inconsistent, but sometimes they show their power. Just watch how, despite one opening inaccuracy, Nathan basically outplays me through most of the game.

photo by Gurth Smith

Smith,Nathan (1603) - Cross,Ted (1998) [B33]
Old Fort Bay Invitational Nassau (5), 24.02.2019

1.e4 c5 2.Nf3 Nc6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 e5 6.Ndb5 d6 7.Nd5 
After the recent world championship match, I wonder if we should call this the Caruana variation? It has been around a long time, but it was the first time I recall seeing it in the title match.
7...Nxd5 8.Qxd5? 
Here Nathan shows that he didn't know the opening line, which of course is to take with the pawn. This move gives black some tempi, though I managed to not follow up correctly and squandered anything it gave me.
8...a6 9.Na3 Be6 10.Qd1 Be7? 
And just like that, with one normal-looking move, I lose my edge and we're back at equality. The move the computer likes isn't a natural one to find, so I'm not surprised I didn't see it. [10...Qh4 11.Bd3 d5 12.exd5 Bxd5 13.0-0 Be7 14.Be3 Rd8 15.f3 Nb4-/+]
11.Bc4 Qd7?! 
There are a lot of minor inaccuracies in this game, due to us being humans in a difficult position for humans to understand. [>=11...0-0]
12.Be3 0-0 13.0-0 b5 14.Bd5 Rac8?! 
This inaccuracy changes the evaluation from a minuscule black edge to a small white one. I don't really understand the computer's preference for Bf6. I assume it is to allow the knight to come back to e7, but the rook would still need to move from a8 first. [>=14...Bf6]
15.c4! 
This move was the first moment when I began to worry that I was letting the position slip away from equality.
15...b4?! 
My response is normally fairly typical in this opening, but the computer didn't like it this time. [15...Bxd5 16.cxd5 Nb8 17.Nc2 f5 18.exf5 Rxf5+/=]
16.Nc2 a5 17.Bb6?! 
Now the computer dislikes one of Nathan's moves. [>=17.Qd3+/=]
17...Bd8 
[17...Rb8! 18.Be3= (18.Bxc6?? Qxc6 19.Bxa5 Qxc4-/+) ]
18.Bxd8 Rcxd8?! 
As I said, a lot of tiny inaccuracies, but that is common in such complex positions. You'll generally see that the computer lines are not at all intuitive. [>=18...Bxd5 19.Qxd5 Rfxd8 20.Qd2 Nb8 21.b3 Qb7 22.Rae1 Na6+/=]
19.Ne3?! 
[>=19.b3+/=]
19...Ne7?! 
This one, though, is one I should have seen and played. The d4 square is crying out for a knight. [>=19...Nd4]
20.Bxe6?! 
[>=20.b3]
20...fxe6 21.Rc1 Qc6 22.Qg4 Rf6 23.Qg5 
A clever move. It threatens Knight to g4 and my rook can't go back to protect the knight on e7 because of knight to h6.
23...Qb7?! 
[23...h6 24.Qg4 Ng6=]
24.c5 Rg6 25.Qh4 d5? 
I saw the perpetual attack on his queen with the rook and briefly considered going for it. But I needed to win this game badly, so it was worth it to risk losing. I certainly wasn't happy allowing that passed c pawn though, but there was no other way to play on for a win. [25...Rh6 26.Qg5 Rg6=]
26.exd5?
Luckily neither of us saw the following tactic to win a pawn. [26.Ng4! Qc7 27.Nxe5+/-]
26...exd5 27.Nf5 Nc6 28.Nd6 Qe7?! 
I can honestly say I don't comprehend the computer's preference of Qa6. [>=28...Qa6 29.f3 Rf8=]
29.Qxe7 Nxe7 30.Rfd1 
[>=30.Rfe1 Re6 31.f4 Nc6 32.Rcd1 d4+/=]
30...d4?! 
Here the computer suggestion is a logical one, but I just didn't see it. [>=30...Re6 31.f3 g6=]
31.Kh1?! 
[>=31.Re1 Nc6 32.Re4 Re6 33.f4 Rf8 34.fxe5+/=]
31...Nd5?! 
Yep, I knew that Nc6 was the more rational move...if I was content to just draw the game. My move offers more tactical possibilities, which is what I needed to try to win. [31...Nc6=]
32.Nb7 Rf8 33.Rd2 a4 34.c6? 
I didn't fully grasp that this was a mistake at the time. [>=34.g3=]
34...Rc8 
And here is why. Actually, I saw the move Nc3 but gave it only a cursory look. It looked fascinating, but I assumed there was a flaw I was overlooking and I didn't want to risk losing when I had a solid continuation. Still, it would have been fun to have played Nc3. [34...Nc3!! 35.Na5 (35.bxc3? bxc3 36.Rdd1 Rxc6 37.Na5 Rc7-+) 35...Nxa2 36.Rc5 e4! 37.h3 (37.Rxd4? e3 38.f3 Re6 39.Rd1 e2 40.Re1 b3-+) 37...d3 38.Kh2 Re6=/+]
35.Rc5?? 
I expected Na5, which seems to peter out to a drawn endgame. Poor Nathan overlooked his back rank weakness. [35.Na5 Rd6 36.Rc5 Ne7 37.g3 Nxc6 38.b3 a3 39.Kg2 Kf7 40.Nxc6 Rcxc6 41.Rxe5 d3 42.Kf3 Rc2 43.Ke3 Rb2=]
35...Rgxc6 36.Rdc2 Rxc5 
My move is fine and completely winning, but the computer likes it's more clever move. [36...d3!]
37.Rxc5 Rxc5 38.Nxc5 a3 39.Nd3 e4 0-1


What a relief to survive that one! The first weekend finished well for me with three wins and a draw. But now there came three of the highest-rated opponents in a row.

My next opponent is one I dread playing. I call him a grinder, because he never goes down easily. It's always a protracted battle. He had a dreadful tournament result this time, but my game against him showed his mettle and I had to really struggle to pull it out.

Cross,Ted (1998) - Cox,Valentine (1872) [B23]
Old Fort Bay Invitational Nassau (6), 27.02.2019

1.e4 c5 2.Nc3 Nc6 3.f4 g6 4.Nf3 Bg7 5.Bb5 a6 
I don't think a6 can be very good. White wants to trade the bishop for the knight at some point anyway, and the pawn being on a6 doesn't typically add anything for black in the types of positions that follow, so I think it is a wasted tempo.
6.Bxc6 dxc6 7.0-0 e6 8.d3 Ne7 9.Qe1 0-0 10.Qh4 f5 11.Be3 b6 12.Rae1 
While my move isn't that bad, the rook is more at home on d1 instead. [>=12.Rad1]
12...Ra7! 
I give this move an exclamation point because this rook basically holds black's position together throughout most of my attack. I kept ruing that it was there.
13.Kh1 Bd7 14.Rg1 Nc8 15.Qf2 
[>=15.Qg3 c4 16.exf5 exf5 17.Rd1+/-]
15...Nd6 16.e5 Nb5 17.Ne2 
I was so intent on attacking on the kingside that I didn't really even look at queenside possibilities, to be honest. [17.Na4! Rb7 18.c4 Nc7 19.b4+-]
17...Nc7 18.b4 
Though I had to play this queenside move, because it simply wins a pawn.
18...Nd5 19.bxc5 Nxe3 20.Qxe3 b5 21.g4!? 
I was in attack mode, so forget about queenside moves! [>=21.a4+-]
21...fxg4? 
This mistake gives white a basically winning game. [21...Be8 22.a4 a5 23.c4 bxc4 24.dxc4 Rb7 25.gxf5 exf5 26.Rd1 Qe7 27.Rd6+-]
22.Rxg4 Qa5 23.Reg1 
I decided to simply abandon the queenside and go all-in on my kingside attack. I knew that if I failed I would probably lose, unless I could come up with a perpetual.
23...Qxa2 24.h4? 
My move changes the evaluation from 'winning' to merely a strong white edge. [24.Ng5! Rf5 25.R4g2+-]
24...Be8 25.Ned4 Qd5 26.Kh2?! 
I so wanted to get my king out of the pin that I missed a lovely tactical sequence. [26.h5! Qxc5 27.hxg6 hxg6 28.Ng5 Qd5+ 29.Kh2 Re7 30.Qh3+-]
26...Qxc5 27.h5 Rd7 28.c3?? 
I made a bad mistake here, going from nearly winning to equality. [28.Qe4 Qd5 29.Qxd5 exd5 30.hxg6 hxg6 31.Ne6 Rf5 32.Rh4 Re7 33.Nfd4 Rf8 (33...Rh5 34.Rxh5 gxh5 35.f5+-; 33...Rff7 34.Rxg6+-) 34.Nxf8+-]
28...Qd5? 
Luckily Cox didn't go in for taking the pawn. [28...Qxc3 29.hxg6 h6 30.Qe4 Re7 31.R4g2 c5 32.Nxe6 Rxe6 33.Qd5 Kh8 34.Qxe6 Qxd3 35.Qg4 Qe4= 36.Re1 Qxf4+ 37.Qxf4 Rxf4 38.Kg3 Rf5 39.e6 c4 40.Ra2 Bxg6 41.Nh4 Rg5+ 42.Kh3 Bh5 43.e7 Be8 44.Rxa6 Kg8 45.Rf1 Re5 46.Kg3 Rxe7 47.Nf5 Rb7 48.Ra8 Be5+ 49.Kg4 Rb8 50.Ra7 Kh8 51.Nxh6 Bg7 52.Nf5 Bf8 53.Kf4=]
29.hxg6?? 
Another big mistake for me, though it certainly looked fine over the board. The computer's idea of Qe2 to support a c4 push is very clever and hard to see. [29.Qe2! Rb7 (29...c5? 30.c4! Qa8 31.Nxe6+-) 30.hxg6 h6 31.c4 bxc4 32.dxc4 Qd7 33.Rd1 c5 34.Nb5 Qe7 35.Nd6 Rb4 36.Qe3+/=]
29...Bxg6? 
Cox plays the natural move, but he had the unusual h5 push, which leads to equality. [29...h5! 30.R4g2 c5 31.Ne2 Qxd3 32.Qf2 Qc4 (32...Qd5 33.Qg3=) 33.Qh4 Bh6 34.Ng5 Bxg5 35.Qxg5 Rg7 36.Qxh5=]
30.Ng5 Bf5 31.Rh4! c5 [31...h6 32.Ne4 Kh8 33.Qg3 c5 34.Nxf5 exf5 35.Nf6 Rxf6 36.exf6 Bf8 37.Qg6 Qf7 38.Qxf5+-] 32.Nxf5 Rxf5 33.Nxh7! 
Once I got in Nxh7 I finally felt like I was winning.
33...Kf7 34.Qg3 
I like my move a lot, and it's much easier to see than the computer's suggestion. [34.Ng5+ Ke8 35.Ne4 Bf8 36.Rg3 b4 37.cxb4+-]
34...Ke8 35.Nf6+ Rxf6 36.exf6 Bf8 37.Qg6+ Rf7 
It was over anyway, but this move allows me to finish quickly with a cute tactical sequence. [37...Kd8 38.Rh8 Qd6 39.Qe4 Kc7 40.Rgg8+-]
38.Rh7 Qd7 39.Re1 b4 40.Rxf7 1-0

It was a relief to get by Cox successfully and move to 4.5 out of 5. Now I had a Saturday ahead of me with two really tough opponents.