Anand pushes Carlsen to the wall

Garry Kasparov was among the privileged few who got to watch the third game of the world title match from a VIP seat. His old rival and sometimes “frenemy” Viswanathan Anand and his former protégé, Magnus Carlsen produced one of the finest fighting draws in chess history for his delectation.

Kasparov says he’s in Chennai purely as a chess tourist and, if one takes him at face value, he should have been well satisfied. The presence of a former world champion and living legend at a world title match should, in itself, scarcely be cause for speculation. However, theories about the “real reasons” for Garry Kasparov’s presence in Chennai, and the players’ reaction to his presence, dominated the post-game conference, which should really, really, have been about the game itself. 

Chess politics, which is labyrinthine and complex in nature, deserves to be discussed in much detail but that definitely wasn’t the appropriate time and place. One theory was however, laid to rest. Both players said that they had neither met nor spoken with Kasparov, which means that he did not help Carlsen with his preparation. GK also reiterated that he wasn’t helping Magnus.

One could have deduced that from the game itself. The opening was rather odd. White’s body language and clock use suggested, in fact, that something started going wrong fairly early.  Carlsen with the white pieces started yet again with a Reti, the non-commital 1.Nf3, and he transposed rapidly into an English, rather than going for a repeat of the first game.

Anand headed into a tabiya – a set of related positions – that he knows very well, when he captured the c-pawn with 3—dxc4. Carlsen took his first long think on move 4. The position resembled a Sicilian Dragon Defence – only the colours were reversed. 

The Dragon Variation is named for a fanciful resemblance between black’s kingside pawn structure (h7, g6, f7, e7, d6) and a dragon with a long neck. (Chess players have been known to use narcotic and psychotropic substances and it sometimes shows in opening nomenclature. There is also a Hippopotamus, a Hedgehog, an Orangutan and a Lammergeier.)

In the Dragon, the King’s Bishop is placed on the long open diagonal, and hopefully, breathes fire along it. In return, white gets space and attacking prospects.  Here, of course, white had the dragon structure and the bishop, while black had the attacking prospects. In fact, there was another slightly unusual detail in that black already had his King’s Bishop on the long diagonal and this may have favoured him. There are lines when it redeploys with loss of time to that diagonal.

As Parimarjan Negi pointed out with delight, the world champion played exactly as the young Indian GM has done in the past, until move 10. On move 10, Anand’s Knight jump 10.—Nd4 may have been a novelty. It led to a well-known type of position – still very much part of the tabiya.

Carlsen did not respond exactly. He had to swap the knight of course, allowing black a cramping pawn on d4. He was already taking a fair amount of time at this stage. After the knight trade, he moved his Queen’s Bishop to an unusual diagonal and then he pulled his queen back to an unusual square. At the end of this process, black was slightly better.

 If Carlsen had done it the other way around, by moving his queen first and then his bishop, he would gained a move via an intermediate threat. That would have been enough to retain the slight edge, which white usually has in the early opening.

Very soon, Carlsen was forced to make another concession and swap his bishop for a somewhat passive knight. That gave Anand something to build on in terms of a space advantage based on the strongpoint on d4, coupled to the bishop pair on an open board. 

The Knight and the bishop are very disparate pieces. They are generally reckoned to be roughly equal in strength. The Knight controls squares of either colour and it can leap over obstacles. But it is short-range – a helicopter rather than a strategic bomber. The bishop can only control squares of one colour and it cannot hop. But it is long-range. On an open board, a bishop usually dominates a knight and two bishops in combination can be devastating.

The next phase of the game was completely fascinating. Both players were playing across both sides of the board. Carlsen opened his Queen’s rook file creating prospects of counterplay for his rook. Anand started to expand on the queenside with a pawn–roller. Carlsen switched to the kingside trying to activate his queen and knight; Anand pushed the white queen into the corner, while conceding good play for the knight.

The crisis came after black’s 27—b5, a very committal move that is probably the best shot here. Anand later explained that he wanted to get a pawn roller flowing on the queenside while the white queen was stuck on the other side of the board.

One of black’s problems and white’s trumps is that white has undisputed control of the a-file and therefore, a more active rook.   If black can get his pawns rolling, however, he will either open lines to activate his own rooks, or create dangerous passed pawns, or both. White will find it difficult to counter this because his queen doesn’t have much influence in the key half of the board.

The downside is that, from this point onwards, Carlsen can eliminate the bishop-pair anytime by exchanging his knight for black’s light-squared bishop and probably damage black’s pawn structure in the process.  This will leave one set of minor pieces on the board – opposite coloured bishops. White has a light-squared bishop while black has the dark one.

Boris Spassky probably described the peculiar dynamics of opposite-coloured bishop actions best when he explained away one of his multiple divorces as “My wife and I had already become like bishops of opposite colours”. The two pieces cannot ever interact with each other.

This can render material balances irrelevant. In pure endgames where there are no other pieces in action, opposite bishops are drawish because the defender can set up impregnable blockades on “his” colour.  However, where heavy pieces are also in play, the man with the initiative has an extra attacking piece because his bishop cannot be opposed. 

Carlsen’s response to the roller was an act of practical desperation. He admitted that he hadn’t seen b5 at all and he knew he was in dire straits. He opened the centre with 28. e3, sacrificing a pawn to get counter-play. It may not have been an objectively good decision, but passive play would have seen him ground down without a fight.

Anand blinked, understandably, at this point. He could have taken the pawn; several hours of cool analysis with computer engines suggests that the pawn grab should work. In practice, white’s optical compensation is immediately obvious, and Anand said he reckoned it would be full compensation. White smashes the pawn structure and starts hitting weaknesses on the light-squares and along the e-file. Arcane tactical calculations suggest black can get on top anyway in the slugfest that will follow, but those lines cannot be seen or judged accurately by a human being with limited time on the clock.

The alternative Anand played was more than perfectly viable. It retained a strong attack for black with much-reduced prospects of white counter-play. By then, both players were getting short of time with extreme complications to negotiate.

On move 34, Anand missed a chance to chop down into a queen endgame with an extra pawn and winning chances. That was the last clear win, if it can be called that, since white retains serious drawing chances even in that endgame. 

Having missed that chance, on move 37, Anand decided to liquidate into a drawn endgame when he exchanged the last pair of rooks. It was a tacit admission that he couldn’t see realistic chances of transforming his residual edge. In fact, he offered a draw. Carlsen opted to play on, but the position had fizzled out.

While Anand deserves praise for creating pressure with black, he didn’t quite play precisely enough to take it away. Carlsen deserves his share of kudos for defending brilliantly in a very difficult position, though he needs to re-examine his opening strategy that got him into a mess.

Carlsen might contemplate abandoning the Reti / English systems he has tried twice and go for Plan B in Game 5 when he gets his next white.  Anand is obviously comfortable with this complex opening and the middle-games that arise from it.

The match has entered a second stage. The initial sparring and information exchange is over. One comment Kasparov made was that Anand has, so far, managed to keep Carlsen from getting the thing what he wants.

 Carlsen wants “calm positions with some advantage where he can play forever”. Instead,  “Anand’s strong reactions” has offered him the choice of “either a sharp fight, not to Carlsen’s preferred style, or calm positions where Carlsen is slightly worse”. 

In Game four, the onus will be on Anand to maintain momentum with his second white. Anand is equally comfortable pushing either central pawn. If he does play 1.e4 Anand will definitely be prepared for the Caro Kann. Carlsen played a very high-risk line the first time. He might deviate first, unless he is very confident about the depth of his personal “book”, against a man who knows the general contours of that opening better. The other possibility is, Anand will play 1.d4 just to find out what Carlsen intends to do against the Queen’s Pawn.

One point to be considered is that the colour sequence flips over on Game 7, at the halfway stage. This means Anand gets two back-to-back whites in Game 6 and 7 and Carlsen will not get time to repair any specific weakness in his defensive repertoire.  So, Anand might use his second white to go “wide”, probing the range of Carlsen’s preparation and then go “deep” in games #6 and #7 to exploit any specific weakness he picks up.

Game score

White: Carlsen,Magnus  Vs Black: Anand,Viswanathan
Game Three, World Championships, Chennai 2013.

1.Nf3 d5 2.g3 g6 3.c4 dxc4 4.Qa4+ Nc6 5.Bg2 Bg7 6.Nc3 e5 7.Qxc4 Nge7 8.0–0 0–0 9.d3 h6 10.Bd2 Nd4

A novelty. Negi has played 10. --Be6  11. Qa4 Nd4 here in an earlier game.

11.Nxd4 exd4 12.Ne4 c6 13.Bb4 Be6 14.Qc1 Bd5

More accurate was the same idea with inverted move order. By 13. Qc1 (hitting h6) Kh7 14. Bb4 and now 14.—Be6 can be met with 15. Nc5, white forces an “useless” 13. Kh7.

15.a4 b6 16.Bxe7 Qxe7 17.a5 Rab8 18.Re1 Rfc8 19.axb6 axb6 20.Qf4 Rd8 21.h4 Kh7 22.Nd2 Be5 23.Qg4 h5 24.Qh3 Be6 25.Qh1 c5 26.Ne4 Kg7 27.Ng5 b5!

Play on both sides of the board leads to a crisis. Black’s better but he needs to activate his rooks. White has a problem handling the queen-side pawn roller since his queen is misplaced.


28.e3 dxe3 29.Rxe3 Bd4!? 



The Bd4 move is good and keeps an edge while restricting counter-play. But the engines say 29...Bxb2 should win with lines like 30. Rae1 Rb6! 31. Bd5 Bd4 32. Rxe6 fxe6 33. Rxe6 Qf8! 34. Qg2 Rdd6. It’s highly complicated and white could go over into attack if there’s one mis-step.

30.Re2 c4 31.Nxe6+ fxe6 32.Be4 cxd3 33.Rd2 Qb4 34.Rad1 Bxb2 ?

The last clear chance was 34.- Rf8 35. Bxd3 Qd6 (with a threat of Qxg3) 36.Qg2 Rxf2! 37.Rxf2 Rf8 38.Rdd2 Rxf2 39.Rxf2 Bxf2+ 40.Qxf2 Qxd3. Black is a pawn ahead in a queen ending. White has drawing chances because the kings are open and he could deliver perpetual check. 

35.Qf3 Bf6 36.Rxd3 Rxd3 37.Rxd3 Rd8  Here 37...Bd4 keeps some pressure but 38. Qe2 Rf8 39. Rf3 Rxf3 is probably drawn. Open kings and opposite bishops devalue the material advantage. After the rook exchange, the position is dead.

 The game concluded 38.Rxd8 Bxd8 39.Bd3 Qd4 40.Bxb5 Qf6 41.Qb7+ Be7 42.Kg2 g5 43.hxg5 Qxg5 44.Bc4 h4 45.Qc7 hxg3 46.Qxg3 e5 47.Kf3 Qxg3+ 48.fxg3 Bc5 49.Ke4 Bd4 50.Kf5 Bf2 51.Kxe5 Bxg3+ (½–½) .

Devangshu Datta is an internationally rated chess-player and correspondence player. A senior journalist and columnist, he has also assisted Vishwanathan Anand in writing his autobiography.

Earlier Reports and Full Coverage