Crowthorne A v Guildford A, Surrey Border League Division 1.
Time control: 80' (+10").
1.e4 c6
I faced the Caro-Kann a lot last season, playing eight games and having a constant struggle for an advantage.
2.d4 d5 3.e5 Bf5
3...c5 has become more popular recently and personally I think is a better repertoire option.
4.h4
4.Nf3 e6 5.Be2 is the other critical variation nowadays.
4…h5
Black's main move. The advantage of including h4 and ...h5 for White is that the g5-square is weakened and h5 itself is a target for attack if Black castles kingside.
A: 4...e6?? 5.g4 Be4 6.f3 Bg6 7.h5+- is a famous Bishop trap.
B: 4...h6 is an alternative. White probably has a small advantage after 5.g4 Bd7.
5...Be4 6.f3 Bh7 7.e6! is a typical pawn sacrifice in this line, restricting the kingside development and weakening the kingside light squares.
6.h5 due to the extra kingside space.
5.Bd3
A: 5.Bg5 is less popular now due to 5...Qb6 6.Bd3 Qxd4! 7.Nf3 Qg4 which is supposed to be fine for Black.
B: 5.c4 is another major alternative.
5…Bxd3 6.Qxd3
At first sight it seems Black has successfully traded the bad Bishop and can enjoy the better minor piece, but White's extra space and slight lead in development often leads to a kingside attack and the "bad" c1-Bishop has an excellent square on g5. Hence Black's next:
6…Qa5+
which disrupts White's coordination.
6…e6 7.Nf3! This seems like the better and more practical line to me, although the alternative is more popular.
7.Bg5 Qb6 8.Nd2 c5 9.c4 Qxb2 10.Rb1 Qxd4∞ This is highly theoretical and about equal.
7…Nd7
7...Qa5+ transposes to 6...Qa5+ and is perhaps best.
8.Bg5 Now that d4 is defended, ...Qb6xb2 is less of a problem, so White has managed to reach the ideal setup after Nbd2 next.
7.Nd2
7.Bd2 Qa6 and White has no advantage in the endgame: 8.e6 Qxd3 9.exf7+ Kxf7 10.cxd3=
7…e6
7...Qa6 is also an important direction. Unlike after 7. Bd2 White now has 8.c4! to avoid the trade of queens. 8...e6 9.Ngf3 Nh6 10.O-O Nf5 11.b3 Be7 12.a4! Taking the h4-pawn is very dangerous due to Ba3 preventing castling and taking over the a3-f8 diagonal, so White perhaps maintains a tiny edge.
8.Nf3
8.Ne2 is yet another alternative.
8…Nh6 9.O-O Nf5
By keeping the Queen on a5 Black saves a move and avoids misplacing the Queen on a6.
10.Nb3
A: 10.c4?! I once essayed against Nigel Povah who surprised me with this line. Black is fine after 10...Be7 forcing 11.g3 as b3 + a4 + Ba3 is a move too slow now!
B: 10.a4!? is an interesting alternative to gain queenside space and prevent ...Qb5.
10...Nd7 11.Nb3 Qa6 12.Qd1 c5∞
10...Qa6
10...Qb5!? was introduced by Caruana a few years ago and is perhaps a better way to go: 11.Qd1 c5 Black gets the critical break in immediately.
11.Qd1
Queen trades are bad as we want to attack on the kingside. So back home we go.
11...b6!
For some reason this wasn't mentioned in the reading material I have on this line, despite being the most popular move, so I was on my own from now. It's very sensible: after ...c5, Black wants to play ...Nc6 rather than ...Nd7. From here there are active options on a5 or b4 for the Knight and ...N(c)e7 is possible to shore up the f5-Knight, typically Black's strongest piece, which attacks d4 and is critical to the defence of the kingside.
12.Bg5 c5 13.c3
13.dxc5 bxc5 looked interesting too. I was worried about b2 becoming weak after a future ...c4, but 13...Bxc5 is the main line as I've now discovered. The theoretical battle continues. 14.Nbd2 Nc6 15.c4!+= cuts across those plans and weakens Black's structure.
13...Nc6 14.Nc1
A typical manoeuvre in these lines. The b3-Knight is out of play, so Nc1-e2-g3/f4 is the plan to trade off the f5-Knight or attack h5.
14...Rc8 15.Ne2
I considered 15.a3 to prevent ...Nb4, but realised it wasn't an issue and that a3 might be a waste of time which weakens the queenside.
15...Qb5
Annoying, as b2 is a bit weak.
15...cxd4 16.cxd4 Nb4
16...Be7 17.Bxe7
17.Ng3 Bxg5 18.Nxf5 Be7 leads nowhere.
17...Ncxe7 and I was going to unrealistically claim a tiny advantage due to the extra space and weakness on h5. The game is quite equal.
17.Rc1 White's lead in development prevents Black from pursuing active operations on the queenside, so a3 will just force the Knight back to c6.
17.Nf4 might be even better.
16.b3
A: 16.Rb1 seemed passive but avoids weakening the queenside.
B: 16.Qd2 I didn't like because of 16...cxd4 17.cxd4
17.Nexd4 Nfxd4 18.Nxd4 Nxd4 19.cxd4 is actually quite strong for White due to the weakness on h5 and lead in development, but I didn't recognise this during the game.
17...Bb4 when the Queen has to move again.
16...cxd4 17.a4!
A critical inclusion.
A: 17.cxd4 Ba3! is awkward as White can no longer use the c-file. Due to the lead in development, this is often used as a second avenue of attack in addition to the kingside. This ...Ba3/Ba6 idea (or similar) is not uncommon as a method to control an open c-file.
B: 17.Nexd4 Ncxd4 18.Nxd4 Nxd4 19.cxd4
19.Qxd4?! Qc5
19...Ba3!=
17...Qa6
17...Qa5?? 18.b4 wins a piece: 18...Qa6 19.b5+-
18.cxd4
A fairly typical situation for the Caro-Kann in general: White is minutely better with some extra space, but Black is very solid and can hope to equalise or win if White overpresses or ends up in a worse endgame.
18...Bb4!
A: 18...Nb4 19.Rc1 is again nothing for Black.
B: 18...Be7 fills the e7-square, so ...Nce7 is no longer possible. I was looking at 19.Ng3
19.Bxe7 Ncxe7 20.Qd2+=
19...Bxg5
19...Nxg3 20.fxg3 is a favourable structural change despite the doubled pawns, as f7 is now a target.
20.Nxf5 Be7 and now 21.Nd6+!? looked interesting, trying to get to Black's weak dark squares. Of course, the d6-pawn will be quite weak.
21.Nxe7 is a safer option: 21...Nxe7=
19.Nf4
Aiming at h5, and allowing Nxd5 in some situations.
A: 19.Ng3 would be the normal idea with the Knight on d7 rather than c6, but here 19...Nxg3 20.fxg3 Ne7 brings the other Knight to f5 and White hasn't achieved anything.
B: 19.Rc1 O-O also allows Black to untangle.
19...Be7
Now Ng3 isn't possible, so this is sensible.
A: 19...g6? 20.Bf6 leaves Black with chronic weaknesses on the kingside dark squares.
B: 19...Bc3? Going after the d4-pawn fails tactically: 20.Rc1 Ncxd4
20...Bxd4 21.Rxc6 Bxf2+ 22.Rxf2 Rxc6 and now most direct is 23.Nxd5! exd5 24.Qxd5 Qc8 25.Nd4+-
21.Rxc3 Rxc3
21...Nxf3+ 22.Qxf3+-
22.Nxd4 Nxd4 23.Qxd4 Rxb3 Otherwise Black is down material for no compensation. 24.Nxd5! This one again. 24...exd5 25.Qxd5+- The twin threats of Qd8# and Qxb3 are too strong for Black to handle.
20.Bxe7
20.Qd2?! Bxg5 21.hxg5 g6=+ As I discovered in my game last year against Paul, with the g5-outpost filled by a pawn White has much less play. Black can make inroads on the queenside and against d4 in peace.
20...Nfxe7?!
The first step in the wrong direction. As mentioned previously, the f5-Knight is a very important piece for Black and when it disappears the kingside can collapse remarkably quickly.
20...Ncxe7 is probably just equal. 21.Qd2
21.a5? just helps Black bring the Queen into play: 21...Nc6 22.axb6 Qxb6∓
21...g6 22.Rfc1 O-O= and it's hard to make progress.
21.Rc1
A: 21.Ng5 forces ...g6, but Black will need to play this move anyway to finish development. Meanwhile d4 is now weaker, and going wild with 21...g6 22.g4?? hxg4 23.Qxg4 Nxd4-+ is just silly.
B: 21.Nh2 with a similar g4 idea (thematic, but typically only once Black castles) is also no good: 21...g6 22.g4 hxg4 23.Nxg4 Rxh4 24.Nf6+ Kd8-+ leads nowhere.
21...g6
21...Nf5 22.Qd2+= Intending to double Rooks and put pressure down the c-file.
22.Qd2
The kingside dark squares now look very weak.
22...Kf8?
Trying to play ...Kg7 to get the Rooks connected while avoiding Nxh5. The position is already difficult for Black - it's hard to suggest anything better.
A: 22...O-O? looks extremely dangerous with Nxh5 and g4 ideas in the air (both as a result of 4. h4 h5). Indeed, the computer spots a complicated win: 23.b4!
23.Nxh5?! gxh5 24.Qg5+ Ng6 25.Qxh5? Qd3! demonstrates why the Queen must be forced away: 26.Ng5 Nf4-+ holding h7.
23...Qxa4
23...Qb7 is better as the Queen is closer to the kingside, but similar bad things happen after 24.Nxh5!±
24.Nxh5! gxh5 25.Qg5+ Ng6 26.Qxh5+- Apparently there is no defence. This sort of slow kingside attack is typical for the variation and can be easy to underestimate. E.g.26...Kg7 27.Ng5 Rh8 28.Qf3 Rc7 29.Qf6+ Kg8 30.Nxe6+- winning.
B: 22...Qa5 23.Rc3!± leaves Black unable to coordinate, while White will just increase the pressure down the c-file.
23.Qd3 was the move I was mainly looking at. 23...O-O! Now that the White Queen is away from the dark squares this is ok for Black, although it still looks frightening.
23...Nb4 24.Rxc8+ Nxc8 25.Qc3 Ne7 26.Qc7+- is a typical example of White infiltrating down the c-file.
24.Nxh5 gxh5 25.Ng5 Nf5 26.Qe2 Ng7 and White should repeat with 27.Qd3=
C: 22...Nf5 might be best: 23.Rc3+= with continuing pressure.
23.Ne2!
The Knight has done its job on f4 and makes way for the Queen, while also preparing Ng3 to challenge a Knight appearing on f5.
23...Na5
Consistent but too ambitious - another piece moves away from the kingside.
23...Nf5 24.Ng3 Nce7
24...Nxg3 25.fxg3 Again this exposes f7.
25.Qg5 followed by Qf6 is nasty - Black is attacked on the kingside dark squares and down the c-file.
24.Qg5!
24.Rxc8+! is similarly deadly but less flashy (and therefore probably better!). 24...Qxc8
24...Nxc8 25.Qf4+- and Ng5.
25.Qg5+- I didn't like this quite as much as the Black Queen is one step closer to the kingside, but it doesn't make a difference.
24...Rxc1
A: 24...Qxe2 25.Qf6 is fatal: 25...Rh7 26.Ng5 Rg7 27.Nxe6++-
B: 24...Nxb3! is the best attempt: 25.Rxc8+ Qxc8 26.Qf6 Rh7 27.Ng5
27.Nf4 Ng8 and Black is still kicking.
27...Rg7 28.Nf4 I must admit I missed that 28...Nxd4 defends e6 here, but fortunately White still has a knockout: 29.Rd1 Nec6 30.Rxd4!
30.Nxg6+ Kg8 31.Nxf7 also wins.
30...Nxd4 31.Nxg6+! Kg8 32.Ne7++-
25.Rxc1 Ng8
25...Qb7 26.Qf6 Rg8 27.Ng5 Nec6 28.Nxe6+ Ke8 29.Ng5+- followed by e6, as 29...Qe7 loses to 30.Rxc6+-
26.Qd8+ Kg7 27.Rc8
Black must give the Queen to prevent Qf8+ and Ng5#. A nice game featuring White's typical plan of attack in these lines, and some of Black's defensive resources (the Knight on f5 is critical to maintain).
27.Nf4 Nh6 28.Nxe6+ fxe6 29.Qf6+ Kg8 30.Rc7 is even more accurate.
1-0