Winning with the Caro-Kann

The final game of the Under 2100 Section of the City of Chicago Class was an attractive matchup between two overperforming 1900 players. Emma Wang, one of our top high school players, was leading the section. Khalid Khan, formerly top board for Blue Cross in the Chicago Industrial Chess League and now my teammate on the group of slightly-over-the-hill CICL veterans known as Rogue Squadron, was half a point behind. 

Khalid had to win with Black against Emma's 1.e4, and he chose the Caro-Kann instead of the favorite weapon of desperate GMs and club players, the Sicilian.

Instead of settling for a draw that would clinch the tournament for her, Emma went for the perfect score. (And this is what ambitious young players should do!) But trying to break down the Caro-Kann is not an easy task. Players who overextend often wind up losing to their plodding opponents. 

It worked just fine for Khalid, but just barely. Emma almost clinched the tournament with a draw. She still won $175 for her share of second and third, while Khalid won $500 for clear first.

You can play through the game and commentary at this link:

Wang,Emma (1923) –

Khan,Mohammad Khalid (1900)

Caro-Kann Defense [B18]

City of Chicago Class

October 29, 2017

1.e4 c6

Compare the following variation of the French to understand the idea behind the Caro-Kann: 1...e6 2.d4 d5 3.Nc3 dxe4 4.Nxe4.In this variation of the French, Black has not solved the problem of the "bad bishop": the bishop on c8 is hemmed in by the pawn on e6. 4...Bd7 The "Fort Knox Variation" is Black's radical attempt to solve the problem of the light-square bishop. 5.Nf3 Bc6 6.Bd3 Black is willing to give up the two bishops, a minor concession, with a solid position. Recommended only for those who really enjoy hunkering down!

2.d4 d5 3.Nc3 dxe4 4.Nxe4 Bf5

This is the Classical Main Line of the Caro-Kann, one of the most solid defenses to 1.e4. Black has solved the problem of the "bad bishop" on c8.

There is a small cost: White has a central pawn, and Black doesn't, and White has a slightly larger lead in development than normal, as 1....c7–c6 is not a developing move. But the position is only "semi-open": neither factor is a particularly serious problem for Black.

5.Ng3 Bg6 6.f4

One of White's sharpest tries! White isn't really going to trap Black's bishop. But a kingside attack is definitely on White's menu, and White wants to claim total control over the e5 square.

On the other hand, playing both d4 and f4 so early in the opening weakens several critical squares in White's camp.


6...Nf6 is possible because Black has the trick 7.f5 Bxf5! 8.Nxf5 Qa5+ 9.Bd2 Qxf5 10.Bd3 with compensation for the pawn. But after the simple 7.Nf3 Black has to play 7...e6 anyway.

7.Nf3 Nd7 8.Bc4

8.Be2 was played by Judit Polgar against Alexander Khalifman in a very similar position in the very first World Cup (Las Vegas, 1999).; 8.Bd3 is the most common move, but attacking players are not fond of exchanging their light-square bishop.


Black wants to contest control of e5.

9.0–0 Ngf6

This might be a slight inaccuracy. Instead, 9...Qc7 10.Ne5 Ne7, after which Black can choose the side on which he wants to castle depending on his mood and on White's next moves.

White to play

White to play



10.Qe2! might give White a slight advantage because she now has the trick 10...0–0 11.Bxe6! Re8 (11...fxe6 12.Qxe6+ wins a pawn and continues White's attack.) 12.f5! Bh5 13.Nxh5 and now 13...Nxh5 does not win a piece because of the cool (13...fxe6 14.Nxf6+ Qxf6 15.fxe6 and White is just a pawn up) 14.Bxf7+! Kxf7 15.Ng5+

Instead, Black should prevent the sac on e6: 10...Qe7 11.Ne5 0–0 12.c3 White's extra space gives her a slight advantage.



Black should be at least equal now, as all of his pieces are harmonious. But White does have more space!


Probably too ambitous, and it's not consistent with White's strategic plan of locking down the e5 square.

Instead, White could make little moves that restore some harmony to her slightly drafty position. (The player with more space naturally has more weaknesses in her camp. Although this is a classical semi-closed postion, the logic is similar to hypermodern play.) 11.c3 Nb6 12.Bb3 c5 13.Ne5]

11...Qc7 12.Nxg6 hxg6 13.Qf3

White has to defend f4.


The immediate 13...c5! is stronger here. Often, the player with the two knights does not mind opening the position for enemy bishops when ahead in development! And when White has made weakening pawn moves like f2–f4, it's even more thematic.

14.Bd3 a6

Black has rearranged his pawns "Meran style": a6–b5–c6, with the idea of playing c6–c5. But Black does not have a Meran bishop on b7!


Now White could try to bail out by exchanging pieces and activating her bishops: 15.Ne4 Nxe4 16.Bxe4 Rfc8 17.g3 Rab8 18.c3 c5 19.Be3 and White has hopes to equalize.

15...c5! 16.dxc5 Nxc5

Now White wishes that f4–f2 were a legal pawn move. :-)

17.Rf1 Nxd3 18.Qxd3 Rfd8 19.Qc3 Qb7 20.Qb3

Black to play

Black to play


This natural move is good enough to win.. The sharp 20...Bc5+ 21.Be3 Rd2! would have won a pawn after 22.Rf3 Bxe3+ 23.Qxe3 Rxc2.

21.Kh1 Qc6

Khalid is a very modest and mild-mannered guy: going for the kill is not his style! Emma scrambles back into the game in the next few moves, so we should look for improvements to Black's play.

21...Rc4 hits the f4 target yet again. 22.Rad1 (22.Rf3 fails to a series of "computer moves" that increase Black's positional domination: 22...Bc7! 23.Rd3 Rd5! 24.Rd1 a5! and Black is just squeezing White off the board.) 22...Bxf4 23.Bxf4 Rxd1 24.Rxd1 Rxf4.

22.c3 Bc5

All of Black's pieces are better than all of White's pieces. Even the doubled g-pawns help freeze White's target on f4.

23.Qc2 Bb6 24.h3 b4 25.Rf3 Qd5 26.Be1 Rc6 27.Qb3 Qxb3 28.axb3 bxc3

The temporary pawn sac 28...Bc7! was stronger: no need to fear tripled pawns! 29.cxb4 Nd5 30.Rd1 Rb8 31.Ne4 Nxb4 Black is better, but not winning.


White wants to hold the position together by activating her bishop. Straightening the pawns with 29.bxc3 might be a better try, but it's a tough call.


Both player's pawn majorities suffer from doubled pawns. White's 2–1 queenside majority is absolutely crippled, while Black's 4–3 majority hopes to produce a passed e-pawn someday. Both a6 and f4 are targets.

30.Ra4 Nd5 31.Ne2 Rb8

And now White's b-pawns are targets, too.

32.Bd4 Bxd4 33.Nxd4 Rc1+ 34.Kh2 Nb4 35.Rc3!

I don't understand what just happened here, but White has crawled back into this game!


Black has shifted his attack to the queenside. Can Black win if all the queenside pawns disappear?

36.Rc4 Nd5 37.Rxa6 Rxb2

White to move

White to move


This slip costs White the game.

It's understandable that White wants to relieve the pressure on the second rank by doubling rooks to offer an exchange. But White should leave the c6 square open for a knight hop with discovered attack: Stockfish prefers 38.Rca4! Nxf4 39.Nc6 Rxg2+ 40.Kh1 wins a piece for a bunch of pawns. 40...Rxb3 41.Rxf4 Rg5 42.Ra8+ Kh7 43.Rh4+ Rh5 44.Rxh5+ gxh5 45.Kg2 and White has no trouble drawing: if necessary, the horsie can always give up its life for two pawns.

Black can try to press with the "computer move" 38...Rf8 (other rook moves away from the danger square b8 are also possible). For example, 39.Ra2 Rxa2 40.Rxa2 Nxf4 41.Nc6! (to control the queening square) 41...Nd5 42.b4 looks drawish, as the well-supported b-pawn is annoying.. But in a game between strong amateurs, any result is still possible here.

38...Nxf4 39.Rc2 Rxc2 40.Rxc2 e5 41.Nc6 Rb5

Again, White has an outside passed pawn as compensation for her pawn deficit. But in comparison with the variation after 38.Rca4! Rf8, however, Black's pieces are far more active here!

42.Rf2 g5 43.g3 Nd3 44.Re2 f6 45.b4 Rb6!

45...Nxb4?? 46.Rb2! would win the Exchange and save the draw for White.

46.Rc2 Nxb4 47.Ne7+ Kh7 48.Rc7 Na6 49.Ra7 Nc5 50.Kg2 Rb7 51.Ra5 Rxe7 52.Rxc5

"All rook endings are drawn," but not this one.

52...f5! 53.Kf3 e4+ 54.Ke3 g6 55.Rc2

Resignation is not premature, as White can't prevent Black from creating two connected passers: 55.Rc2 Kg7 56.Ra2 Kf6 57.Ra6+ Re6 58.Ra2 Rd6 59.Ra8 Rd3+ 60.Kf2 f4 61.Rf8+ Ke7]