Garry Kasparov's Forgotten Weapon Against The Queen's Gambit (part 12)

KASPAROV won with white in game nine of his Candidates final against Vasily Smyslov, meaning he led 3-0, with six draws, when game 10 began.

Smyslov (2600) - Kasparov (2710)

Candidates Final, Game 10 (Vilnius, USSR) 1984

1.d4 d5 2.Nf3 c5 3.c4 e6 4.cxd5 exd5 5.g3 Nf6 6.Bg2 Be7 7.0-0 0-0 8.Nc3 Nc6 9.Bg5 cxd4 10.Nxd4 h6 11.Be3 Re8 12.a3 Be6 13.Kh1 Bg4 14.f3 Bh5 15.Nxc6!?

Smyslov varies from 15.Bg1 of game eight.

15..bxc6 16.Na4

As usual in the Tarrasch, White's capture on c6 radically changes the play. Instead of targeting the d5 pawn, White now targets the c6 pawn and the c5 square, the latter becoming an outpost for white pieces.

16...Qc8!?

Taking control of the c8-h3 diagonal before it can be occupied by White's light-square bishop. That in turn suggests White could have played 16.Bh3?!, but then Black has a pleasant choice between 16...Bxa3 and 16...c5.

17.Bd4 Qe6

Supporting c6 from a flank, pressurising e2 and freeing c8 for possible occupation by a black rook.

18.Rc1 Nd7 19.Rc3 Bf6 20.e3 Bg6

Thanks to a rook being on c3, White can meet 20...Bxd4 with 21.Qxd4 or 21.exd4.

21.Kg1!?

It is surprising, at least at first glance, that Smyslov can find nothing better than bringing the king closer to the centre. An obvious plan is to double on the c file by 21.Qc1, or 21.Qd2 with Rfc1 to come. But Smyslov presumably decided there is little to be gained thereby as long as a black bishop covers c2.

21...Be7 22.Qd2 Rab8 23.Re1

As an anonymous ChessBase annotator points out, 23.Bxa7 does not win a pawn, viz 23...Ra8 24.Nc5 Qd6 25.Nxd7 Rxa7 26.Nc5 Bf6 27.Rcc1 Qe5, when Black is getting his pawn back and has excellent play.

23...a5 24.Bf1 h5 25.Rec1

25.Bd3?! Bxd3 26.Qxd3 h4 leaves the white king's position looking draughty.

25...Ne5!? 26.Bxe5 Qxe5 27.Rxc6 Bf6 28.R6c5

Not 28.R6c3? d4 29.exd4? Qxd4+ 30.Qxd4 Bxd4+.

But Stockfish12 and Komodo11.01 reckon the position is an example where greed is good and that White should hang on to his extra pawn with 28.Kf2!?, although they believe Black has sufficient compensation.

28...Qxe3+ 29.Qxe3 Rxe3 30.Rxd5 Rxf3 31.Be2 Re3 32.Bxh5

White re-establishes his pawn advantage and gets rid of the black bishop-pair, but after ...

32...Bxh5 33.Rxh5 g5!

... White's pieces are uncoordinated.

34.Nc3!?

34.Rc2 Kg7 35.g4? runs into 35...Re4.

Counterplay with 34.Rc5!? Re2 35.Rxa5 divides the engines. Stockfish12 reckons 35...Rd8 36.Kf1 Rc2 37.Ke1 Kg7 wins for Black, but Komodo11.01 is less sure.

34...Rd8!?

After 34...Rxb2 35.Nd5 Bd4! 36.Rxg5+ Kh7 37.Kh1 Rxa3 Black seems a little better thanks to having bishop-v-knight on an open board with rival passed pawns.

35.Rc2 Kg7 36.Kg2?!

Presumably played with the idea of g4 and Rh3, but this is easily foiled, so 36.Kf2 is probably better.

36...Kg6 37.g4 Rd4 38.h3 ½–½

The final position - should Kasparov have played on, or was Smyslov better?

*****

*****

*****
*****

*****

Maybe the match situation influenced Kasparov's decision, as Black seems better. One likely continuation given by the engines goes 38...Rdd3 39.Ne2 a4 40.Nc1 Rd1 41.Ne2 Rb3 42.Nc3 Bxc3 43.bxc3 Rd3 44.h4 gxh4 45.Rxh4 Rxa3, when Black has got his pawn back and is better coordinated. However, this is a long line and there are many alternatives along the way.