Anand—Carlsen

 Game 1: A hard-fought draw kicked off the match in Sochi—home to the 2014 Winter Olympics. ANAND was the one that came out with guns blazing in a Grunfeld (5.Bd2) ruining his pawn structure for activity and giving himself a bind on e6. CARLSEN slowly pushed it back and eventually it was Black that found himself with the initiative, attacking White's pawn structure on the kingside and the weak king on a1. Although CARLSEN had the possibility to put ANAND under more pressure with 42...Re3 the opportunity was not taken and the game was agreed drawn on move 48.
 Magnus CARLSEN:

I had a very good start. In the first game I had some minor problems in the opening. I realised during the game that the position was a bit more dangerous than we had thought before the game, so I decided to try and find something else – which I didn’t – but fortunately he didn’t find the most critical lines at some point and I could escape into a drawn position. So for the next moves I managed to gain more and more territory. He didn’t want to exchange queens, he didn’t want to exchange rooks, and as a result both my queen and rook became active and his king was a little bit weak as well. Unfortunately at one point I missed a key defensive resource – his move 44.Qh1. After that there was nothing, but nevertheless it was a good start.

1.d4 Nf6 2.c4 g6 3.Nc3 d5!?
 The Grunfeld must be one of the most popular ways of meeting 1.d4 at the top level of chess; however it is not part of the World Champion's common repertoire. It is clear that CARLSEN is bringing new things to the table and ANAND's team will have to adapt quickly.

4.cxd5 Nxd5 5.Bd2
 In the face of an opening that he did not expect, ANAND decides to play a variation that is relatively quiet and is far from being a theoretical duel.

5...Bg7 6.e4 Nxc3 7.Bxc3 0-0 8.Qd2
 This type of Grunfeld focuses on holding the central tension without committing White's pawn to c3, which leaves it vulnerable to the pressure from the g7 bishop. In this case White will rely on his pieces to support his d4- and e4-pawns.
 8...c5 has been seen in many grandmaster games, for example: 9.d5 e6 10.Bc4 was the very recent Tomashevsky-Kasimdzhanov from the Grand Prix last month.

8...Nc6
 CARLSEN played this relatively quickly, but it is a rare line.
 Besides the text move, 9.Ne2 is also possible but it blocks the development of the f1-bishop.

9.Nf3
 After this move CARLSEN went on a 15 minute think. Interestingly, CARLSEN this time around has already revealed that his seconds are "the Dane and the Hammer" referring to Nielsen and Hammer. However, many speculated that maybe Nepomniachtchi would help CARLSEN, and this line has been played by the Russian player recently.

9...Bg4 10.d5 Bxf3
 The game mentioned above (August 2014) Moiseenko-Nepomniachtchi went: 11.gxf3 Ne5 12.Be2 c6, which Black won.

11.Bxg7
 This is a novelty already.

11...Kxg7 12.gxf3 Ne5 13.0-0-0
 Black must note that 13...Nxf3? drops a knight to 14.Qc3†.

13...c6 14.Qc3 f6 15.Bh3 cxd5 16.exd5!
 ANAND ruins his structure! To compensate this he now has a strong hold over e6 and will be able to kick out Black's knight from e5. For Black, he will be able to reroute this knight somewhere else that is useful and his structure remains very solid.
 Variation
16.Rxd5 Qb6–/+ starts looking bad for White.
 His control of the d-file is worthless as long as the knight is stable on e5, since it guards d7.
Kicking out the knight with 17.f4? is refuted by 17...e6! when the strong threat of ...Rac8 would already be winning for Black.

 Variation
16.Rxd5 Qb6–/+ starts looking bad for White.
 His control of the d-file is worthless as long as the knight is stable on e5, since it guards d7.
Kicking out the knight with 17.f4? is refuted by 17...e6! when the strong threat of ...Rac8 would already be winning for Black.

16...Nf7 17.f4 Qd6
 Black's blockade on the dark-squres is keeping his position together. There is no easy way to penetrate and Black will create counterplay if left alone. The idea of pushing the a-pawn down the ranks comes to mind.

18.Qd4
 A fascinating strategical game. White has a clear clamp on the position, but on the long-term he is worse because of his ruined pawn structure. Practically speaking it seems as if White has the easier time finding moves, but by no means do I think Black is objectively worse. CARLSEN is known for for his resilience in finding passive but solid moves.

18...Rad8 19.Be6 Qb6!
 I like this plan. ANAND can now go into an endgame, but that seems counter-intuitive.

20.Qd2?!
 This might be the turning point. The concession seems natural; give up the good position of the queen on d4 to avoid the queen trade. However It seems that after this White has no good way of continuing.
 Variation
20.Qxb6 axb6 21.Kb1
 Even though this might still be far from an advantage, the possibility of quickly going to c7 with the rook and the passive position of the black rooks allow White to have some chances to create problems.
21...Nd6 22.Rc1 f5! 23.Rc7 Kf6 24.Bd7! Ne4 25.Rxb7 is only slightly unpleasant for Black.

 Variation
20.Qxb6 axb6 21.Kb1
 Even though this might still be far from an advantage, the possibility of quickly going to c7 with the rook and the passive position of the black rooks allow White to have some chances to create problems.
21...Nd6 22.Rc1 f5! 23.Rc7 Kf6 24.Bd7! Ne4 25.Rxb7 is only slightly unpleasant for Black.

20...Rd6 21.Rhe1
 Or 21.Kb1 Nd8 22.Bg4 (22...e6 23.Bf3 exd5 24.h4!? This kind of computer idea is hard to play over the board. It is, at least immediately, not clear to me why White has enough compensation for the pawn. There is some initiative on the queenside but Black's position does look solid.) 22...Nc6!? 23.h4 Nb4 24.Bf3 Rc8.

21...Nd8 22.f5 Nxe6 23.Rxe6
 Despite White's bind on the e6 square, it is Black that is now better. As soon as he solidifies his e7 pawn he can start working on grinding down White's structure.
 23...Rxe6 24.fxe6 Rc8† 25.Kb1 Qd6 is also rather appealing, but it does seem that Black might run out of ideas if the blockade must be preserved.

23...Qc7† 24.Kb1 Rc8?!
 Perhaps letting White establish a rook on e6 is not a good way to push for any advantage.

25.Rde1 Rxe6 26.Rxe6 Rd8 27.Qe3 Rd7 28.d6!
 A tactical way of getting rid of his weakness. ANAND is very close to equalizing.

28...exd6 29.Qd4 Rf7 30.fxg6 hxg6 31.Rxd6
 White's structure on f2 and h2 is still somewhat weak, but with his piece activity it does seem as if he should hold comfortably.

31...a6 32.a3 Qa5 33.f4
 Despite the computer's approval of this move, it is somewhat strange. Why push the f-pawn instead of the h-pawn?
 Variation
33.h4 Qf5† 34.Ka2 Qe5 and perhaps ANAND didn't want to deal with another rook endgame, considering how it haunted him last year: 35.Qxe5 fxe5 36.Re6=.

 Variation
33.h4 Qf5† 34.Ka2 Qe5 and perhaps ANAND didn't want to deal with another rook endgame, considering how it haunted him last year: 35.Qxe5 fxe5 36.Re6=.

33...Qh5 34.Qd2 Qc5 35.Rd5 Qc4 36.Rd7
 36...Rxd7 37.Qxd7† Qf7 would force the queen to permanently defend the second rank, leaving it as a simple draw.

36...Qc6
 37.Rxf7† Kxf7 38.Qe3 a5! This would be what we have been calling a "CARLSEN position." With the weak structure on the kingside Black has some chances of making progress by bringing in the king, as long as there is no perpetual, but of course with perfect play it should be drawn.

37.Rd6 Qe4† 38.Ka2 Re7 39.Qc1 a5
 White has no way to improve so Black tightens the noose around White's king ever so slightly. This doesn't have a particular threat in mind, it is simply improving Black's chances.

40.Qf1 a4
 With time control reached it is clear that the position is nothing but unpleasant for White. He is now passive, has to defend f4, and his king doesn't feel very comfortable.

41.Rd1 Qc2 42.Rd4
 Black has to set-up some kind of lethal threat to make progress. The main thing is that if he can create real problems to the White king, White might run out of resources to defend both the king and the pawns on the kingside.

42...Re2?!
 Variation
42...Re3
 The commentary team was very worried about this move appearing on the board: 43...Rxa3†! is a lethal threat.
a) 43.Rd7† Kf8 44.Rxb7 Rb3 45.Rxb3 axb3† 46.Ka1 Qxh2
 White's badly placed king and the need to keep the queen on the first rank (or at lest protect the checkmate threats on the first rank) make this an incredibly unpleasant endgame that might just be lost.
 47.Qd3 Qh1† 48.Qb1 Qf3 49.Qxg6 Qxf4

47.Qd1!
 47...Qxf4 48.Qd8† Kg7 49.Qd7† Kh6 50.Qh3† Kg5 51.Qxb3=.

47...Qh3!
 A hard move to find, especially from far away.

48.a4 Kg7 49.a5 Qe6! 50.Kb1! Qe4† 51.Kc1 Qxf4† 52.Qd2 Qf1† 53.Qd1 Qc4† 54.Kd2 Qd4† 55.Ke1 Qe5† 56.Kf1 Qxb2 57.Qd7† Kh6 58.Qh3† Kg5 59.Qe3† Kf5 60.Qf3† Ke6 61.Qg4† Kd6 62.Qf4† Kd7 63.Qg4† f5 64.Qa4† Ke6 65.Qc6† Ke5 66.Qc7† Ke4 67.Qc6† Kd3 68.Qb5† Kc2 69.Qe2† Kb1
 And Black has finally run away from the checks.

b) 43.Qd1 Qxd1 44.Rxd1 Rb3 is very unpleasant as Black's king will march in and once b5 is played the queenside is locked down.

 Variation
42...Re3
 The commentary team was very worried about this move appearing on the board: 43...Rxa3†! is a lethal threat.
a) 43.Rd7† Kf8 44.Rxb7 Rb3 45.Rxb3 axb3† 46.Ka1 Qxh2
 White's badly placed king and the need to keep the queen on the first rank (or at lest protect the checkmate threats on the first rank) make this an incredibly unpleasant endgame that might just be lost.
 47.Qd3 Qh1† 48.Qb1 Qf3 49.Qxg6 Qxf4

47.Qd1!
 47...Qxf4 48.Qd8† Kg7 49.Qd7† Kh6 50.Qh3† Kg5 51.Qxb3=.

47...Qh3!
 A hard move to find, especially from far away.

48.a4 Kg7 49.a5 Qe6! 50.Kb1! Qe4† 51.Kc1 Qxf4† 52.Qd2 Qf1† 53.Qd1 Qc4† 54.Kd2 Qd4† 55.Ke1 Qe5† 56.Kf1 Qxb2 57.Qd7† Kh6 58.Qh3† Kg5 59.Qe3† Kf5 60.Qf3† Ke6 61.Qg4† Kd6 62.Qf4† Kd7 63.Qg4† f5 64.Qa4† Ke6 65.Qc6† Ke5 66.Qc7† Ke4 67.Qc6† Kd3 68.Qb5† Kc2 69.Qe2† Kb1
 And Black has finally run away from the checks.

b) 43.Qd1 Qxd1 44.Rxd1 Rb3 is very unpleasant as Black's king will march in and once b5 is played the queenside is locked down.

43.Rb4
 Now White's defensive set-up also includes pressure against the b7 pawn. Black cannot be careless with his own king.

43...b5
 Paralyzing White's position, but ANAND still has an important resource.

44.Qh1!
 Incredibly important! Now Black's king is under threats of a perpetual.

44...Re7 45.Qd5 Re1 46.Qd7† Kh6 47.Qh3† Kg7 48.Qd7† ½-½
 Draw agreed.

 Game 2: The second game of the match saw ANAND pay homage to the anniversary of the fall of the Berlin Wall by employing the defence of the same name. CARLSEN achieved a small edge out of the opening and perhaps ANAND did not give Magnus's slow-burning kingside initiative the respect it deserved. CARLSEN had a clear advantage in a major piece ending and a blunder by ANAND in an already difficult position just before the time control saw him forced to resign.

In the second game it was kind of a quiet Anti-Berlin opening. He doesn’t have any particular problems from the opening, but the position is a little bit non-standard and in many lines I may gain a slight initiative. That’s what happened in the game as well. I managed to gain a slight advantage in the centre, swing my rook to the kingside and although objectively there were no particular dangers for Black it’s clear that I had the initiative and not very much risk.
 Once he tried to bail out into a position with no minor pieces it was clear that with my activity and superior king position I would have an advantage. When he missed the chance to prevent my rooks from doubling and eventually my queen joining and tripling on the e-file it was very difficult, and at the end he missed a tactical stroke that meant that he lost immediately.

1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.d3
 The Berlin! Betting money that it would be seen this match would have had sure dividends; pretty much everyone expected this opening to be seen at some point in the match. Carlsen chooses not to go into the Berlin endgame and instead chooses one of the "quieter" d3 systems.

4...Bc5
 This is the "point" of Black's play. Usually he has to commit to playing the bishop to e7 and only then does White go d3, a variation that is becoming increasingly popular in the Spanish. In this particular move order, the bishop has no reason to fear going to c5.

5.0-0 d6 6.Re1
 White has tried basically everything under the sun, but this peculiar move-order has yet to be employed. Normally, they start with the move c3 or Nbd2.

6...0-0 7.Bxc6 bxc6
 White cannot claim a real advantage. His pawn structure superiority is compensated by Black's solid position and pair of bishops. However, it is a completely playable position; if anything Carlsen is making sure that the game is simply "playable" for both sides instead of trying to get an advantage from the opening.

8.h3 Re8 9.Nbd2 Nd7 10.Nc4 Bb6 11.a4 a5 12.Nxb6 cxb6 13.d4 Qc7
 In many cases the presence of opposite colored bishops means that any endgame will be drawn. This is still the case here, but White has a few resources to put some pressure. He does hold more space at the moment.
 Here 14.Nh4 was a serious suggestion, but after 14...Nf8 Black seems to be too solid.

14.Ra3!?
 A creative rook lift. The queenside rook is trying to make its way to the kingside (or even to the center) to put some quick pressure on that flank.
 14...exd4 15.Nxd4 Nc5 16.Bf4+/=

14...Nf8 15.dxe5 dxe5 16.Nh4 Rd8?!
 Had Anand foreseen what happened to him in the game, he might have avoided this move altogether. There is no reason to force White's queen to the attack as the d-file holds no value.
 Variation
Setting up defenses as quickly as possible is necessary: 16...f6 17.Rg3 Ne6 18.Nf5 g6 19.Qh5 Ng7!
 Exchanging the powerful knight. White's attack is not nearly as strong without it.
 The impatient 20.Nh6† just leaves the knight stranded on h6 after 20...Kh8 21.Qd1 Ba6!.
20.Nxg7 Qxg7 21.Qh4 Ba6=

 Variation
Setting up defenses as quickly as possible is necessary: 16...f6 17.Rg3 Ne6 18.Nf5 g6 19.Qh5 Ng7!
 Exchanging the powerful knight. White's attack is not nearly as strong without it.
 The impatient 20.Nh6† just leaves the knight stranded on h6 after 20...Kh8 21.Qd1 Ba6!.
20.Nxg7 Qxg7 21.Qh4 Ba6=

17.Qh5 f6 18.Nf5 Be6?
 I believe this relatively careless move is the beginning of Black's problems. Vishy underestimates how quickly he has to repeal White's pieces.
 18...Qf7 is a another way of meeting White's threats, for example: 19.Qg4 Bxf5 20.exf5 Rd4 21.Qf3 Qd5=

19.Rg3 Ng6
 Variation
Still more resilient was 19...Rd7! though after 20.Bh6 g6 21.Qh4 Qd8! it is still slightly unpleasant, but I don't see any immediate threats for White.

 Variation
Still more resilient was 19...Rd7! though after 20.Bh6 g6 21.Qh4 Qd8! it is still slightly unpleasant, but I don't see any immediate threats for White.

20.Bh6!?
 This interesting move leads to a long, forced variation.
 For starters: 20...Rd7 21.h4 looks very threatening.

20...gxh6 21.Rxg6† hxg6 22.Qxg6† Kf8 23.Qxf6†
 a1) 23...Bf7 24.f4! and White's attack will crash through. And 24.Qh8† Bg8 25.Re3 is also good enough for a winning advantage.

 a2) 23...Qf7 24.Qxh6† Ke8 25.Qh8† Kd7 26.Rd1† Kc7 27.Qxe5† Kb7 28.Nd6† Rxd6 29.Rxd6 and White comes out with material advantage, though Black should be able to hold by creating his own threats with 29...Re8+/=.

20.h4
 Pressure is piling up on the kingside! It is not so obvious how Black can repeal White's attack.
 20...Rd7 21.Bh6! (21...gxh6 allows White to recuperate the piece with devastating consequences after 22.Qxh6 Qd8 23.h5 and here White clearly has the advantage.) 21...Bxf5 22.exf5 Nf8 23.Re4!? and White's pressure is nothing to scoff at.

20...Bxf5 21.exf5 Nf4 22.Bxf4 exf4 23.Rc3! c5 24.Re6!
 It is clear that White has tremendous pressure. The control over the e-file, the pressure on b6, the anchored rook on e6, and also importantly the complete lack of counterplay. Black is not lost yet but it is very unpleasant.

24...Rab8 25.Rc4 Qd7 26.Kh2
 Of course White has no interest in allowing Qd1† forcing a queen trade.
 Now 26...Qd1? loses to 27.Re8†+-

26...Rf8
 Passive, but what else to do? There is a lack of a clear plan for Black.

27.Rce4 Rb7 28.Qe2 b5!?
 A nice opportunity to get rid of the pawn on b6 and open the b-file, but Black's c- and a-pawns have now become targets.

29.b3
 Variation
Immediately invading on the seventh rank with 29.Re7! is also winning. (29...Qc6 30.Rxb7 Qxb7 31.axb5) 29...Qd6 30.f3 Rxe7 31.Rxe7 bxa4 32.Qe4 The queen invading b7 is a big threat. And following 32...Qb8 33.Qxa4 it is hard to believe Black can survive with absolutely no activity.

 Variation
Immediately invading on the seventh rank with 29.Re7! is also winning. (29...Qc6 30.Rxb7 Qxb7 31.axb5) 29...Qd6 30.f3 Rxe7 31.Rxe7 bxa4 32.Qe4 The queen invading b7 is a big threat. And following 32...Qb8 33.Qxa4 it is hard to believe Black can survive with absolutely no activity.

29...bxa4 30.bxa4 Rb4 31.Re7 Qd6 32.Qf3!
 The queen wants to start looking for ways of getting into the seventh rank.

32...Rxe4 33.Qxe4 f3† 34.g3
 The only good way of preventing the queen from coming to b7 is by attacking f2, but this gives up the f3-pawn, for example: 34...Qd2 35.Qxf3 Qxc2 36.Kg2 and Black's is close to lost, but not there yet.

34...h5??
 A horrible blunder in a very difficult position. As the World Championship contender Nigel Short pointed out on Twitter: "Blunders don't happen in a vacuum. They came after enormous sustained pressure."

35.Qb7 1-0
 Black resigns. Carlsen created something out of seemingly nothing and earned a great victory.

 Game 3: After the loss in game two, many were predicting an ANAND collapse as was seen in the 2013 match. However, ANAND's deep preparation in a Queen's Gambit Declined with 5. Bf4 clearly caught CARLSEN by surprise. CARLSEN later admitted his choice of opening had been a poor decision and that his opponent played extremely well. ANAND improved on the game Aronian - Adams from 2013, securing a powerful passed pawn on c7 which tied CARLSEN's pieces down to its defence. CARLSEN's legendary resilience was not sufficient to hold the position and he resigned on move 34.

The third game was, frankly speaking, a horrible experience. I changed my mind right before the game about what I was going to play. I was not well enough prepared for the line that happened in the game. He showed very good preparation and finished the game quite nicely. I didn’t have many chances.
 Clearly that was a big blow for me. After the first two games I did think that the match was going to be easier than I’d thought beforehand. I thought I was playing much better than him and that would be the decisive factor. The third game showed me that it wasn’t so easy after all.

1.d4 Nf6 2.c4 e6 3.Nf3 d5 4.Nc3 Be7 5.Bf4 0-0 6.e3 Nbd7 7.c5
 This style of the "Aronian Queen's Gambit" has become popular in recent years. In the Tashir super-tournament, we have seen his position several times.
 7...Nh5 has been the favorite reply of the Black players in Tashir.

7...c6 8.Bd3 b6 9.b4 a5 10.a3
 White's expansion on the queenside looks scary, but if Black can look it down, open the a-file successfully and trade off his light-squared bishop (which is many times simply dead), then he can hold comfortably. Of course, doing this takes a long time.

10...Ba6 11.Bxa6 Rxa6 12.b5!
 This creation of a passed pawn has been known for some time. All of this is still well-known theory.

12...cxb5 13.c6 Qc8 14.c7 b4 15.Nb5 a4 16.Rc1 Ne4 17.Ng5
 Taking twice on g5 is certainly impossible, but taking once might be necessary.
 Variation
17...Bxg5 18.Bxg5 (if 18...Nxg5?? then 19.Nd6 rips apart the blockade and wins the queen.) 18...Ra5 19.Be7!? (19...Rxb5 20.Bxf8 Kxf8 21.Qxa4 Ra5 22.Qxb4† is somewhat unclear. The passed pawn on c7 does compensate for Black's two knights against a rook. 22...Ke8!) 19...Re8 20.Bxb4 Rxb5 21.Qxa4+/= and the rook on b5 is trapped. This must favor White as Black's rook on e8 is very passive.

12...cxb5 13.c6 Qc8 14.c7 b4 15.Nb5 a4 16.Rc1 Ne4 17.Ng5
 Taking twice on g5 is certainly impossible, but taking once might be necessary.
 Variation
17...Bxg5 18.Bxg5 (if 18...Nxg5?? then 19.Nd6 rips apart the blockade and wins the queen.) 18...Ra5 19.Be7!? (19...Rxb5 20.Bxf8 Kxf8 21.Qxa4 Ra5 22.Qxb4† Ke8! is somewhat unclear. The passed pawn on c7 does compensate for Black's two knights against a rook.) 19...Re8 20.Bxb4 Rxb5 21.Qxa4+/= and the rook on b5 is trapped. This must favor White as Black's rook on e8 is very passive.

17...Ndf6 18.Nxe4 Nxe4 19.f3 Ra5
 20.Qe2 Qd7 21.fxe4 Rc8!=/+ was Aronian-Adams, 2013. Vishy must have taken a fresh look at this game.

20.fxe4
 Even though both players took a long time to get here (about an hour and a half to get to this position between the both of them) only this move is new.

20...Rxb5 21.Qxa4 Ra5 22.Qc6 bxa3 23.exd5 Rxd5 24.Qxb6
 A fascinating position. Material is equal, but White's position is clearly to be preferred. The reason is that the a-pawn is not as dangerous as the c-pawn, which needs to be blockaded immediately.

24...Qd7
 The computers were screaming for 25.Qa6 to be played in many occasions, but it was not always that clear.

25.0-0
 25...g5 26.Qb8! Rc8 27.Qxc8† Qxc8 28.Rb1±.

25...Rc8 26.Rc6
 Interestingly, this exact position was seen in the game Tomashevsky, Evgeny - Riazantsev, Alexander from the 2008 Russian Super Final. Except, in that game, White's pawn was on h3, and not on h2! Tomashevsky also won that game rather cleanly.

26...g5
 Black is running out of resources. He has to devote too much to stopping the c-pawn and this means that his a-pawn is not playing.

27.Bg3 Bb4 28.Ra1!
 An excellent move. There is no way to rip through the blockade immediately, so ANAND adds pressure on the a-pawn.

28...Ba5 29.Qa6!
 Keeping an eye on the a-pawn and especially the bishop on a5.

29...Bxc7
 Here 30.Rxa3 is also strong as the bishop is pinned regardless.

30.Qc4!
 The pressure on the bishop is huge. This will cost CARLSEN a piece. At this point he is also very low on the clock.

30...e5 31.Bxe5 Rxe5 32.dxe5
 As Svidler pointed out, Black has excellent chances to draw this game if he can break the pin and put pressure on White's weak pawns. But that, simply put, is not going to happen!

32...Qe7 33.e6!
 The easiest. Now Black's king is also a factor. There is no way to dismantle the pin, Black's position is simply hopeless.

33...Kf8 34.Rc1 1-0
 And it is over! ANAND does it! Excellent preparation by the Indian player and an absolutely precise and surgical game to beat CARLSEN very cleanly.

1.d4 Nf6 2.c4 e6 3.Nf3 d5 4.Nc3 Be7 5.Bf4 0-0 6.e3 Nbd7 7.c5 c6 8.Bd3
 It's quite remarkable that White could play the same position with an extra tempo after 8.h3 b6 9.b4 a5 10.a3 Ba6 11.Bxa6 Rxa6 12.b5 cxb5 13.c6 Qc8 14.c7 b4 15.Nb5 a4 16.Rc1 Ne4 and we reach a twin of the game's position with a key difference: Anand's idea of 17.Ng5? would be a mistake due to 17...Bxg5 18.Bxg5 b3 19.f3 Ra5!=/+ and here White can not play 20.Qe2 because of 20...Ng3. Therefore, 20.Qd3 b2 21.Rb1 Nxg5 22.h4 Ne4 23.fxe4 dxe4 24.Qc4 Rxb5 25.Qxb5 Qxc7–/+ followed by Nf6 and then Nd5 with a clear advantage.

8...b6 9.b4 a5 10.a3 Ba6 11.Bxa6 Rxa6 12.b5 cxb5 13.c6 Qc8
 Only move.

14.c7 b4 15.Nb5
 15...bxa3 16.0-0 followed by Qc2 with good compensation.

15...a4 16.Rc1 Ne4 17.Ng5!?
 Without a pawn on h3 this idea works better.
 Variation
17...Bxg5 18.Bxg5 (if immediately 18...Ra5, White is better after 19.Be7 Rxb5 20.Bxf8 Kxf8 21.Qxa4+/=) 18...b3± with an advantage since after 19.f3± here White would be considerably better as can be seen after 19...Ra5 20.Qe2!±. But not 20.Qd3 b2 21.Rb1 Nxg5 22.h4 Ne4 23.fxe4 dxe4 24.Qc4 Rxb5 25.Qxb5 Qxc7–/+.

15...a4 16.Rc1 Ne4 17.Ng5!?
 Without a pawn on h3 this idea works better.
 Variation
17...Bxg5 18.Bxg5 (if immediately 18...Ra5, White is better after 19.Be7 Rxb5 20.Bxf8 Kxf8 21.Qxa4+/=) 18...b3± with an advantage since after 19.f3± here White would be considerably better as can be seen after 19...Ra5 20.Qe2!±. But not 20.Qd3 b2 21.Rb1 Nxg5 22.h4 Ne4 23.fxe4 dxe4 24.Qc4 Rxb5 25.Qxb5 Qxc7–/+.

17...Ndf6 18.Nxe4 Nxe4
 18...dxe4 was a good alternative. Not 19.Bd6 Nd5!=/+. The main line would be 19.Nd6 Bxd6 20.Bxd6 (not 20...Re8 21.Bxb4+/=) 20...b3! 21.Bxf8 Kxf8 with compensation.

19.f3 Ra5?!
 Variation
A better continuation was 19...Nc3! 20.Nxc3 bxc3 21.Rxc3 Qd7 and here it looks difficult for White to make progress. For example, 22.Qd3 b5 23.0-0 Rc8 24.Rfc1 Rb6= is fine for Black since he will gain counterplay after pushing ...b4.

Anand improves: 20.fxe4! Rxb5 21.Qxa4 Ra5 22.Qc6 bxa3 23.exd5± Rxd5 24.Qxb6 Qd7
 25.Qa6 is better than the text but perhaps there is no big difference since Black is lost already.

25.0-0 Rc8 26.Rc6!
 White controls the position and can improve it. The main idea is to play Rb8 at some point but there are other threats also.
 If Black chooses to wait-and-see, White's plan could be 26...h6 27.Rfc1 Bf8 28.Qa6 g6 29.Rb6+-

26...g5 27.Bg3 Bb4 28.Ra1! Ba5?
 Loses immediately but passive defense would also fail and I don't believe there is a defense against precise play by White.

 If Black tried 28...h5 White could reply with 29.h3 Bf8 30.Rac1 Kh7 31.Qa6 h4 32.Bh2 Kg8 33.Rb1+-.

29.Qa6
 29...Bb4 30.Rb6

29...Bxc7 30.Qc4
 Now it's over and the rest is easy.

30...e5 31.Bxe5 Rxe5 32.dxe5 Qe7 33.e6 Kf8 34.Rc1 1-0

 Game 4: ANAND tried to stir things up by playing the Sicilian rather than going for a positionally sound line in the Spanish as was seen in game 2. CARLSEN avoided a sharp, theoretical confrontation by choosing a quiet line (1. e4 c5 2. Nf3 e6 3. g3!?) but obtained no advantage from the opening. With sound play from both sides the game was eventually drawn in 47 moves.

The fourth game was another bad game. I didn’t gain anything from the opening but the position was still one with quite a bit of life in it. I gradually started to outplay him a little bit again but then I missed the moment when I could have consolidated my advantage. I let him equalise and then I played some really poor moves and he could almost start to play for a win while I had little time. Fortunately he made a mistake – he blundered a pawn – but in the end it turned out there was no win.

1.e4 c5!?
 Already a sharper approach to the game. Carlsen showed excellent understanding in the positional and quiet waters of the Spanish in game two. Anand hits back in his second black game with the Sicilian.

2.Nf3 e6
 This move indicates that Black, in case of White going for the open Sicilian 3.d4 cxd4 4.Nxd4 will play one of the following:
 a) 4...Nc6 — the Paulsen;
 b) 4...a6 — the Kan;
 c) 4...Nf6 followed by 5...d6 — the Scheveningen.

Unfortunately we will not know which one Anand had planned until, maybe, the next game Carlsen has white!
3.g3
 This way of avoiding the theoretical battle has been very popular in the last few years. I go into an in-depth discussion of this variation in my recently released DVD on the Paulsen Sicilian that you can find at the ChessBase Shop.

3...Nc6 4.Bg2 d5 5.exd5 exd5 6.0-0 Nf6 7.d4 Be7
 There are many ways that Black can develop once he plays d5. Some of the more popular ones are as Anand played, but without Nc6, or playing Bd6 and Nge7. This specific variation is a little more rare, but the character of the position should not change.

 8.Nc3 0-0 9.dxc5 Bxc5 10.a3!? was played in Zvjaginsev-Rublevsky this year. Zvjaginsev is one of the experts on the white side of this variation.

8.Be3 cxd4
 The first move for which Anand took a small think. It was more or less forced, but he is trying to evaluate how to continue from here. Probably Anand expected 8.Nc3.

9.Nxd4
 9...0-0 10.Nc3 Bg4 11.Qd3 was Jones-Caruana from the Reykjavik Open from 2012.

9...Bg4
 Black develops with tempo as it is not trivial to defend the queen. The queen does not want to move, but hardly has a choice. The move-order difference allows Carlsen to put his knight on d2 rather than on c3, but it does not seem to make a big difference.

 10.f3 is too ugly to be played, as it locks in the bishop on g2 and weakens the bishop on e3.

10.Qd3 Qd7 11.Nd2 0-0 12.N2f3 Rfe8 13.Rfe1 Bd6
 White has play against the isolated pawn, but in return all of Black's pieces are well placed. He will soon bring his rook to c8 finishing his development comfortably and he has relatively little to worry about in the short-run. He has to keep up some kind of pressure so White doesn't simplify into an endgame where the pawn might be weak.

14.c3 h6 15.Qf1!?
 The queen did not feel comfortable on d3 as it was vulnerable to certain knight jumps. Meanwhile, the move also prepares Bh3 just in case.

15...Bh5 16.h3 Bg6 17.Rad1 Rad8 18.Nxc6 bxc6
 With the structure changing to hanging pawns d5 is not nearly as weak, and since the c-file is not open c6 is not a big target. White has to exert pressure on the center quickly or he risks being worse.

19.c4 Be4 20.Bd4
 A strangely annoying move for Black, and a move that allows White to keep equality.

20...Nh7
 The knight is ready to jump to g5!

21.cxd5
 21...cxd5 looks more natural to me but there is nothing wrong with capturing with the bishop either.

21...Bxd5 22.Rxe8+ Rxe8 23.Qd3 Nf8
 Black re-routes the knight. With Anand's passive play the pawn structure will give White an edge, but it is difficult to do much with it. That being said, if anyone can form an initiative out of nothing, it is Carlsen!

24.Nh4 Be5 25.Bxd5 Qxd5 26.Bxe5 Qxe5 27.b3
 Every trade makes the Black pawn structure more vulnerable.

27...Ne6 28.Nf3 Qf6 29.Kg2 Rd8 30.Qe2 Rd5!?
 A little sharp. With the passed pawn on d5 Black will be able to create some counterplay, but a lot of endgames are lost for him because of White's queenside majority which allows him to create an outside passed pawn. A simple pawn endgame, for example, would be lost for Black.

31.Rxd5 cxd5
 It made a lot of sense to try to trade queens as the knight endgame seems unpleasant for Black: 32.Qe5!? but 32...Qd8!? might be both more sedate and more reasonable. White retains a very slight edge, but maybe not enough to claim an advantage after 33.Nd4+/=.

32.Ne5 Qf5 33.Nd3 Nd4 34.g4!?
 A great practical decision! Black has three major options, and with the clock ticking this is not always easy.

 First option is 34...Nxe2 35.gxf5 Nc3!? and because of White's structure Black is probably fine in this position. However, it is still dangerous as White's pawns on the kingside look threatening. He may also immeidately bring the king to the center 35...Kf8!?.
 Second option is to trade queens on e4 34...Qe4+? 35.Qxe4 dxe4 36.Nc5± but it loses a pawn.

Third option is to retreat: 34...Qd7 35.Qe5
 White gains some space, but Black's pieces are well placed and White's king is now a bit vulnerable, allowing counterplay.

35...Ne6
 A nice position for the knight. If White ever goes f4-f5 he will be faced with a defenseless king, while if the knight is allowed to stay on e6 it controls many, many key squares.

36.Kg3 Qb5 37.Nf4 Nxf4
 Not a bad time to trade knights. With White's exposed king the queen endgame will almost inevitably be drawn.

38.Kxf4 Qb4+ 39.Kf3 d4!?
 Black sacrifices a pawn, but he hopes the activity of the queen and the passed pawn will give him enough to draw.

40.Qe8+ Kh7
 With time control reached, Carlsen took a breather to look at his alternatives.

41.Qxf7 Qd2!
 This move is actually forced. Grandmaster Rustam Kasimdzhanov, who is doing live commentary for the www.playchess.com website, mentioned that Anand would find this by process of elimination if nothing else. Rustam has a good feel for Anand's thought process as he was the Indian's second for a long time!
 Variation
Not good is 41...Qc3+ 42.Ke4 d3 43.Qf3! and White loses his pawn on d3. All he can do is trade it for the one on a2, but he would be simply down a pawn following 43...Qe1+ 44.Kxd3 Qb1+ 45.Ke3±.

And 41...Qe1 42.Qf5+ Kh8 43.Qd3 allows an ugly blockade.

42.Qf5+
 42...Kg8? is a clear mistake because it allows White's queen to go to e2 after 43.Qd5+ Kh7 44.Qe4+ Kg8 45.Qe2! Black has no good way of making progress with the d-pawn, will probably lose it and with it the game.

42...Kh8 43.h4
 From a practical point of view, Anand is not worried at all that White will win a long endgame, he will be worried however that he gets mated! Carlsen wants to play g5 and g6 which puts real pressure on the stranded king on h8.

43...Qxa2 44.Qe6!?
 Carlsen always likes to keep tension in the position. In this case the queen controls many squares, and White is still threatening g5 and g6.
 Dangerous would be 44...a5? 45.g5 (if 45...hxg5 46.hxg5 a4 47.Qe8+ Kh7 48.bxa4+-) 45...a4 46.Qe8+ is one of the ways that Black gets mated.

44...Qd2
 The queen swings to the defense. It is important to note that g5 must be covered.

45.Qe8+ Kh7 46.Qe4+ Kh8 47.Qe8† Kh7
 A well earned draw from both sides!

1. e4 c5!
 Hats off to Vishy! Somehow the endless 1. e4 e5 Berlins and Anti-Berlins feel like Carlsen's natural habitat. So Vishy takes the battle somewhere else: 1. d4 with White and 1...c5 with Black. Let's see how it goes.

2. Nf3 e6!
 This move offers the best chance of reaching an Open Sicilian.

 2...d6 has been Anand's main choice in the past when playing the Sicilian. I'm sure he would be happy to play the Najdorf but the problem is that Magnus has not been allowing him to get there, playing 3.Bb5+ instead which leads to a slow positional battle where the World Champion once again feels like a fish in water.

 2...Nc6 is the other main move here, but it does invite 3.Bb5 yet again.

3. g3!?
 "Sorry Vishy, you did your best, but there will be no open battle here either!"
 Anish Giri: It seems impossible for Anand to get Magnus into a real sharp Sicilian.

 This sideline is a Carlsen favorite — at least 4 games in recent times — and the most respectable deviation from 3. d4 nowadays. Yes, that includes 3. c3.
 White does not aim for anything concrete — he wants to go Bg2 and 0-0 before determining the situation in the centre. Of course Black can use that time to make that choice for him — delaying the central push for one move.

3...d5
 This is the main line, which tends to transpose to our game later.
4. exd5 exd5

a) 5. d4
 5...Nc6 6.Bg2 cxd4 7. Nxd4 Nf6 8. O-O Be7 9. Nc3 O-O 10. Be3 h6 11. Re1+/= 1-0 Carlsen,M (2837) - Kotsur,P (2548), Astana 2012.

5...Nf6 6. Bg2 cxd4 7. O-O Be7 8. Nxd4 O-O 9. h3 Nc6 10. Be3 Re8 11. Re1
 11...h6 ½–½ Short,N (2696) - Caruana,F (2675), Wijk aan Zee 2010.

11...Bf8 12. c3 Bd7 13. Nd2 Rc8 14. Nf1
 1-0 Carlsen,M (2881) - El Debs,F (2511), Caxias do Sul 2014.

b) 5. Bg2 Nc6 6. O-O Nf6 7. d4 Be7 8. Be3 O-O.

3...Nc6 4. Bg2 d5
 Grabbing the centre. This will almost inevitably lead to an isolated pawn on d5. Those who dislike such a scenario should try 4...Nf6, which was played by Caruana in his latest outing against this line.
 Variation
4...g6?! 5. d4 cxd4 6. Nxd4 Qb6 7. Nxc6 bxc6 8. Be3 Qc7 9. O-O+/= 0-1 Carlsen,M (2837)-Svidler,P (2749), Astana 2012.

4...Nf6!? 5. d3 d5 6. Nbd2 Be7 7. O-O O-O 8. Re1 b5 9. exd5 exd5= ½–½ Topalov,V (2772)-Caruana,F (2791), Flor & Fjaere 2014.

 Variation
4...g6?! 5. d4 cxd4 6. Nxd4 Qb6 7. Nxc6 bxc6 8. Be3 Qc7 9. O-O+/= 0-1 Carlsen,M (2837)-Svidler,P (2749), Astana 2012.

4...Nf6!? 5. d3 d5 6. Nbd2 Be7 7. O-O O-O 8. Re1 b5 9. exd5 exd5= ½–½ Topalov,V (2772)-Caruana,F (2791), Flor & Fjaere 2014.

5. exd5 exd5 6. O-O Nf6 7. d4 Be7 8. Be3 cxd4 9. Nxd4
 9...O-O transposes to the games given above.

9...Bg4
 A pretty obvious novelty, attacking the queen with tempo. Anyway, this position is not so much about single moves. The scene is set: Black has an isolated pawn and will have to compensate by playing actively. White has a small static advantage and will try to maneuver in order to increase it.
 10. Qd2 makes a lot of sense as well, intending to put the knight on c3.

10. Qd3 Qd7
 Stopping h3.

11. Nd2
 I am still unsure about this move. The knight heads towards f3, reinforcing the white blockade on d4.
 Variation
Philosophically, I prefer the alternative 11. Nc3 playing in the spirit of another great Scandinavian, Bent Larsen: "There is no point blockading the isolated pawn with all the pieces — let's try to win it!" At least I recall him saying something along those lines. 11...O-O 12. Rfe1 Rfe8 does leave Black with a pretty harmonious position, though. Winning the d5 pawn would be quite a feat....

 Variation
Philosophically, I prefer the alternative 11. Nc3 playing in the spirit of another great Scandinavian, Bent Larsen: "There is no point blockading the isolated pawn with all the pieces — let's try to win it!" At least I recall him saying something along those lines. 11...O-O 12. Rfe1 Rfe8 does leave Black with a pretty harmonious position, though. Winning the d5 pawn would be quite a feat....

11...O-O 12. N2f3 Rfe8?!
 Jonathan Rowson praised this move on Twitter: "Glad Vishy didn't worry about Nxc6 and Ne5; it's important not to duck every time your opponent raises his arms." I disagreed, preferring Bd6 instead.
 Of course, one can read too much into a single move. My theory was — and is — Vishy has a tendency to prefer simplifications into equal or even slightly worse positions to complicated good positions against Magnus. And I believe that tendency hurts his chances. He has fixed it in the opening: 1. d4 with White and 1...c5 with Black. We will see if it shows up in the later stages. Armchair General out.

13. Rfe1
 Declining the invitation to simplify.
 13. Nxc6 bxc6 14. Ne5 Qb7 15. Nxg4 Nxg4 16. Bd4 Bf6 17. c4 is what the discussion is about. While this is certainly not scary for Black, it is a simplified position where White does have a little something to work with.

13...Bd6! 14. c3 h6 15. Qf1!
 I ran my mouth on Twitter calling this a computer move and saying Magnus would not do this. I was silenced not only by Magnus in fact doing this, but also by the man who most consider the future challenger number one, Fabiano Caruana: "Number one priority for White is to play h3 and Kh2, so it seems like a reasonable option."

15...Bh5 16. h3?
 This one, however, is a clear inaccuracy. Allowing Bg6 gives all the black pieces a meaningful job. White just cannot be better after this.
 Variation
Critical was 16. Nh4!? to stop the Bg6-Be4 maneuver. Does the World Champion subscribe to, "a knight on the rim is dim" dictum? I doubt it.
16...Ng4 must have been the move that scared Carlsen away from playing his knight to h4.
17. Qb5! Nxe3 18. Rxe3 Rxe3 19. fxe3 and the d5-pawn will fall. While there is certain compensation — yes, as a Marshall Gambit player I have seen worse — this is still much better than what White gets in the game.

 Variation
Critical was 16. Nh4!? to stop the Bg6-Be4 maneuver. Does the World Champion subscribe to, "a knight on the rim is dim" dictum? I doubt it.
16...Ng4 must have been the move that scared Carlsen away from playing his knight to h4.
17. Qb5! Nxe3 18. Rxe3 Rxe3 19. fxe3 and the d5-pawn will fall. While there is certain compensation — yes, as a Marshall Gambit player I have seen worse — this is still much better than what White gets in the game.

16...Bg6! 17. Rad1 Rad8
 Black is fully mobilised now. Carlsen sees he can make no inroads by maneuvering from here and decides to change the structure by capturing on c6.

18. Nxc6 bxc6 19. c4
 Statically, White is still to be preferred. After cxd5, both the pawn on a7 and the pawn on d5 (or c6) will be isolated and weak. But Black has great pieces, while the white queen looks a bit funny on f1. How to use those pieces?

19...Be4?!
 Keeping the status quo.
 Variation
But Black should compensate for his worse structure by going for activity: 19...Bb4! (20. Bd2 Bxd2 21. Nxd2 Rxe1 22. Rxe1 Qb7 23. b3?! dxc4 24. Nxc4 Bd3) 20. Re2 Ne4 21. cxd5 cxd5 22. Bd4 Bf5 23. Kh2 Rc8 strikes me as one plausible line where piece power outweighs pawn weaknesses.

 Variation
But Black should compensate for his worse structure by going for activity: 19...Bb4! (20. Bd2 Bxd2 21. Nxd2 Rxe1 22. Rxe1 Qb7 23. b3?! dxc4 24. Nxc4 Bd3) 20. Re2 Ne4 21. cxd5 cxd5 22. Bd4 Bf5 23. Kh2 Rc8 strikes me as one plausible line where piece power outweighs pawn weaknesses.

20. Bd4 Nh7
 A decent move, but maybe a sign that his last move wasn't ideal. I am sure the Dark Knight would rather be on e4 (Gotham City) than on h7 (The Pit).

21. cxd5 Bxd5
 Concluding the trinity of mediocre decisions, similar to Universal allowing Michael Bay three sequels to Transformers.

22. Rxe8+ Rxe8 23. Qd3
 The tables have turned and Carlsen is now officially slightly better with no risk.

23...Nf8
 Here Carlsen starts to look like a predator about to go in for the kill!

24. Nh4
 Encouraging exchanges.
 24. Bc3 Ng6 25. Qd4 looked quite good as well and would have retained more pieces and winning chances.

24...Be5!
 This is a scenario we have also seen in the previous match. Vishy ends up in a slightly worse, unpleasant position, which could have been avoided. Once there, Vishy starts defending with great precision. He starts by exchanging the powerful d4-bishop.
 Anish Giri: "I have to admit I missed Be5! from Anand. Now he is no direct danger, but he still has to defend."

25. Bxd5 Qxd5 26. Bxe5 Qxe5 27. b3
 This is not fun for Black, although his drawing chances should still be much greater than White's winning chances — we do have equal material and a slightly weak white king, after all.
 27...Qe2 was a path to a holdable ending, but Anand prefers to keep the queens on the board.

27...Ne6 28. Nf3 Qf6 29. Kg2 Rd8 30. Ke2 Rd5!
 See note to move 24.

31. Rxd5 cxd5
 Vishy has handled this with great care and should not even be worse any more. He is dealing with Magnus Carlsen, though, so there is still no handshake in sight.

32. Ne5 Qf5 33. Nd3 Nd4!
 Should White start to be careful?
 34. Qe3 Qe4+ 35. Qxe4 dxe4 36. Nc5 f5 should end in a draw as well.

34. g4!?
 34...Qxe2 35. gxf5 Kf8 36. Nb4 Ke7! is fine for Black according to Houdini, but Anand sees no necessity for such adventures.

34...Qd7 35. Qe5 Ne6
 There is not much to talk about. White has better structure but Black is active enough.

36. Kg3 Qb5 37. Nf4 Nxf4 38. Qxf4 Qb4+?!
 38...g6 preparing to meet Qe8 with Kg7, strikes me as easier. Don't give that Norwegian Iron Man anything to work with!

39. Kf3 d4 40. Qe8+ Kh7 41. Qxf7
 A pawn is a pawn, but the queen ending is drawn.

41...Qd2! 42. Qf5+ Kh8 43. h4
 It ain't over till it's over!

43...Qxa2 44. Qe6 Qd2 45. Qe8+ Kh7 46. Qe4+ Kh8 47. Qe8+ Kh7
 Draw agreed.

 A typical Carlsen game, in a way. A quiet opening, nothing special. All of a sudden he creates chances out of nothing. Vishy ends up slightly worse, but defends very precisely and doesn't crack. That should give him confidence! Score is 2-2.

 The game had been tough for all concerned, so the press conference was hardly a barrel of laughs. A glum-sounding Carlsen repeatedly castigated his play: I missed several things but I think overall it was just not a very high quality game. There were no glaring blunders, but just in general not good enough.

 For Vishy, meanwhile, a draw with Black was obviously a satisfactory outcome, and he claimed to have been worried only for a brief moment on move 39: There wasn’t really a moment I was very concerned, except when I played 39...d4, but then I saw 41...Qd2 [instead of the intended 41...Qc3+ he realised might be losing] very quickly.

 It was left to Magnus to sum up how this match compares to Chennai: The score in Chennai was 2:2 after four games. Here’s it’s 2:2, so I don’t see any difference.

 Game 5: Round five saw something of a role-reversal. ANAND, who is known for his deep preparation may have been startled at the speed with which CARLSEN played his moves. ANAND obtained a slight positional advantage from the opening, a Queen's Indian. CARLSEN was under a little pressure as White's bishop and rooks dominated the position, though he was temporarily a pawn up. Maybe with a little more finesse ANAND could have made life harder for the World Champion, but as it was, the Norwegian had few problems to make a draw just before the time control at move 40.

The fifth game was the conclusion of a very difficult part of the match for me. In that game we prepared something quite different from what I usually play, but it turned out he was exceptionally well-prepared for this line as well and he chose the line that was most dangerous for Black, and one we had somewhat underestimated.
 So forcibly I had to go to either for an endgame with two pawns each on the queenside and three pawns each on the kingside and he had the advantage of a strong bishop vs. a weaker knight, or I would have a two pawn against one advantage on the queenside but with a crippled kingside. I chose to play with the crippled kingside and later on he could have created much more problems for me than he did. Frankly speaking, I was very surprised that he chose the way he played in the game, which caused me very few problems. I was lucky to have a free day before the sixth game. I got some help from some Russian friends and we prepared something much better against the Sicilian variation he had surprised me with in the fourth game.

1. d4
 We should make Vishy an honorary citizen of Hamburg. He avoids the Berlin as consistently as we do.

1...Nf6 2. c4 e6 3. Nf3
 Never change a winning team! This worked in Game 3, so Magnus is politely invited to another Queen's Gambit discussion.

3...b6
 And he declines, opting for the Queen's Indian instead. By playing b6, Black intends to develop his bishop to b7 and control the centre with pieces, not pawns. Nah, modern chess is too concrete for such generalisations. Pieces, pawns, whatever works. He is just trying to get the Indian out of his preparation but getting Vishy out of his preparation is easier said than done.

4. g3 Bb4+!?
 And here comes the rare move.
 This check is much more common a move earlier — or a move later, for example: 4...Ba6 5. b3 Bb4+.

5. Bd2 Be7!?
 Giving the check on b4 and provoking Bd2 before heading back to e7 is a common theme. The argument is that the white bishop is worse on d2 than on c1, for one of two reasons:

 (1) The bishop on d2 interrupts the queen's control of d5, making it more difficult for White to get in d4-d5.
 (2) The bishop can no longer go to its (sometimes) ideal square on b2 from d2 and might have to settle for c3 instead.
 Granted, it is not easy to come up with lines that illustrate these advantages in this very position, which might in turn explain why the idea is not all that popular.

 5...Bxd2+ featured in two games played by a much younger Magnus. One loss and one draw did not incline him to stick to this line.

6. Nc3 Bb7 7. Bg2
 Anand is not rattled and just quickly plays the most natural moves. We are still a long way from him being out of book.
 7...d5 8. cxd5 exd5 leads to a structure that is considered to be slightly better for White.

7...c6!?
 If a less experienced player made such a move, blocking his own bishop, own would probably start questioning his understanding of the very basics of chess.
 Does he know the board has 64 squares?
 Does he know there are 32 and not 24 pieces?
 Does he know you should not block your own bishop's diagonal?

 However, it's the World Champion we're talking about here — he knows all that stuff!
 Anyway, c6 makes sense. Black wants to go ...d5 and be able to play ...cxd5 after White takes it.
 8. O-O d5 9. cxd5?! cxd5 with equality, shows Black's idea.

8. e4
 Black's c6 did have the drawback of allowing this move.

8...d5
 9. e5 might look like the most natural move, but after 9...Ne4 Black should be doing fine.

9. exd5! cxd5
 The structure is fine for Black, who is one move away from going dxc4 and creating a pleasant IQP situation. But his problems are not solved yet — it's White to move!

10. Ne5!
 Exploiting the pin on the long diagonal. A wise man once wrote: "The relationship between bishops on g2 and b7 always favours the guy on g2." This is no exception.

10...O-O 11. O-O Nc6
 12. Bf4 was seen in Van Wely,L (2654)-Tkachiev,V (2625), Dubai 2014.
 Main alternatives are: 12. Qa4!? Qe8 is okay for Black, but not 12...Nxd4? due to 13. cxd5! hitting d4.

12. cxd5
 A very direct approach.
 12...Nxd5 13. Nxd5 exd5 14. Nxc6 Bxc6 15. Rc1 Rc8 16. Re1 Bf6 17. Bf4 leads to a symmetrical position where White keeps a nagging edge due to his more active pieces.

12...Nxe5
 13. dxe5 Nxd5 and Black is at least equal.
13. d6!
 The point. A cute little tactic, based on the fact that the bishop on g2 is covered, but his counterpart on b7 is not. It does not win material but it does keep Black under pressure.
 Bad for Black is 13...Bxd6? because of 14. Bxb7.
 13...Bxg2 14. dxe7 Qxe7 15. dxe5 and, all of a sudden, two pieces are hanging. And 15...Bxf1 16. exf6 Qxf6 17. Qxf1 favours White.

13...Nc6!
 The correct reply.

14. dxe7 Qxe7 15. Bg5!
 Anand has to keep playing concretely to compensate for his weak d4-pawn. By playing Bg5 he not only pins the knight but also more or less guarantees himself the advance d4-d5, getting rid of the weakness.
 If 15...Rad8 then 16. d5.

15...h6 16. d5
 Anyway!

16...Na5
 Carlsen kept blitzing out his moves. He came prepared!
 Variation
16...Rad8 looks natural, but I believe Carlsen rightly didn't allow Anand to sacrifice his queen: 17. Bxf6 Qxf6 18. dxc6! Rxd1 19. Rfxd1 when the passed c-pawn would not only create unpleasant memories of Game 3, but also cement the white edge, no matter what the engines tell us.

 Variation
16...Rad8 looks natural, but I believe Carlsen rightly didn't allow Anand to sacrifice his queen: 17. Bxf6 Qxf6 18. dxc6! Rxd1 19. Rfxd1 when the passed c-pawn would not only create unpleasant memories of Game 3, but also cement the white edge, no matter what the engines tell us.

 Here 17. d6?! might look tempting, but the pawn lacks support. For example, 17...Qd8 18. Bxf6 Qxf6 19. Qe2 Rad8 20. Rfd1 Rd7 and Black will reach a very compact position after ...Rad8.

17. Bxf6 Qxf6
 Now 18. Qe2!? just controlling the c4-square and preparing Rad1, was another way to keep an edge.

18. dxe6
 The position might look equal, but it's not. Black still has quite some problems to solve to neutralise the superior white pieces — the difference between the knight on c3 and the knight on a5 especially is quite telling.
 If I understood the press conference correctly, this and the next two moves were still part of Anand's preparation. You just cannot surprise him! Years of very serious opening work preparing for World Championship matches have made him bullet proof in this area. Carlsen was never regarded as an opening expert who had lots of computer lines ready everywhere, but he has always been an expert in choosing the right territory and making his opponent's opening knowledge seems useless. It feels different in this match so far. He has not found a place to hide after 1. d4 yet! Good thing for him Carlsen that he is White in games 6 and 7.

18...Qxe6 19. Re1
 19. Bxb7 Nxb7 20. Qf3 is another bonus option.

19...Qf6 20. Nd5!
 The point of White's play. The pawn on b2 can be left en prise due to some tactics. Black is not given a breather to get his a5-knight back into the game.
 20...Qxb2 21. Re2 Qa3 22. Re3 Qb2 23. Rb1 Qxa2 24. Ra1 Qc4 25. Rxa5! bxa5 26. Ne7+ Kh8 27. Bxb7 are the aforementioned tactics. Black might survive this, but there is no question who is better.

20...Bxd5 21. Bxd5 Rad8
 22. Re2 is met by 22...Nc4!.

22. Qf3!
 The only good move.
 22...Qxf3 23. Bxf3 was discarded by the world champ, who explained at the press conference why White is always better here: The bishop is stronger than the knight. The light squares on the queenside are weak. Just imagine a white king on a6 in the future.

22...Qxb2
 Facing a choice between two unpleasant positions, Carlsen goes for the one a pawn up.

23. Rad1
 The White forces are perfectly coordinated, while the knight on a5 is still out of play. Black is fighting for survival here.
 23...Rd7 24. Qf5 Rfd8 25. Bxf7+ is winning for White.

23...Qf6
 A good practical decision. Reducing material should increase the drawing chances.

24. Qxf6 gxf6 25. Re7 Kg7!
 Variation
25...Nc6 was an alternative. 26. Rc7 Ne5 27. Rxa7 Rd6 28. Bb3 Rxd1+ 29. Bxd1 Rd8 and here 30. Bh5!? was mentioned by Anand, solidifying White's edge.
 After 30. Bb3, the active 30...Rd2 looked like enough counterplay according to Anand, but even here 31. Rb7 ensures that Black will have plenty of work to do.

 Variation
25...Nc6 was an alternative. 26. Rc7 Ne5 27. Rxa7 Rd6 28. Bb3 Rxd1+ 29. Bxd1 Rd8 and here 30. Bh5!? was mentioned by Anand, solidifying White's edge.
 If instead 30. Bb3, the active 30...Rd2 looked like enough counterplay according to Anand, but even here 31. Rb7 ensures that Black will have plenty of work to do.

26. Rxa7?!
 It is always nice to restore the material balance, but this move loses a huge chunk of the advantage superior home preparation had provided.

26. Kg2 was one way to keep the pressure, but 26...Nc6 27. Rc7 Ne5 28. Rxa7 Rd6 29. Bb3 Rxd1 30. Bxd1 Rd8 31. Bb3 Rd2 might lead to enough counterplay to hold the game.

26. Rc7! was pointed out by Anand as well and might have been the strongest. The rook just helps to restrict the Black knight.
 26...Rd6 27. Bf3 Rxd1+ 28. Bxd1 a6 29. Bh5 b5 30. Ra7 Nc4 31. Rxa6 is another version of this ending where the champ would still have a lot of defending to do.

26...a6 27. Kg2!? and Black still has complicated problems to solve.
 Or 27. Rd3 Rc8 28. Rdc3 Rcd8 as discussed by the players.

26...Nc6
 Black is suddenly back in business!

27. Rb7?!
 He is not even trying to play for a win! Anand probably did not believe that there were serious chances any more and considered it a better decision not to lose any more energy on this game. In my Armchair General capacity I have to point out that I believe this would have looked very different had the colours been reversed. And that I believe that Anand will have to go after even such small chances in order to win the match.
 Variation
27. Ra4 stops Nb4 and further simplification, and was the only way to keep it going. Anand discarded this because of 27...Rd6 when he did not think he had enough, for example:
a) 28. Rc1! Rxd5 29. Rxc6 and this is by no means an easy draw, e.g. 29...b5 30. Rg4+ Rg5 31. Rf4 Rg6 32. Ra6 and while Black might and probably should hold, it can't hurt finding out if he will. Leading to an easy draw is 32. Rf5 Ra8.

b) 28. Bb3 Rxd1+ 29. Bxd1 Rd8 and it might not be enough, but it is certainly more than the nothing of the game after 30. Bh5.

 Variation
27. Ra4 stops Nb4 and further simplification, and was the only way to keep it going. Anand discarded this because of 27...Rd6 when he did not think he had enough, for example:
a) 28. Rc1! Rxd5 29. Rxc6 and this is by no means an easy draw, e.g. 29...b5 30. Rg4+ Rg5 31. Rf4 Rg6 32. Ra6 and while Black might and probably should hold, it can't hurt finding out if he will. Leading to an easy draw is 32. Rf5 Ra8.

b) 28. Bb3 Rxd1+ 29. Bxd1 Rd8 and it might not be enough, but it is certainly more than the nothing of the game after 30. Bh5.

27...Nb4
 This leads to the exchange of the remaining queenside pawns and is the reason why 27. Ra4 had to be played in order to continue the game. The rest requires no comment except for the fact that the engine Stockfish still gave +1 for White here, leading to some false hopes for Vishy fans. This is a dead draw.

28. Bb3 Rxd1+ 29. Bxd1 Nxa2 30. Rxb6 Nc3 31. Bf3 f5 32. Kg2 Rd8 33. Rc6 Ne4 34. Bxe4 fxe4 35. Rc4 f5 36. g4 Rd2 37. gxf5 e3 38. Re4 Rxf2+ 39. Kg3 Rxf5 ½–½
 A tale of two halves: Great opening knowledge by Anand once again. I am not sure we can call it preparation because it is certainly not something he could have expected for this game or maybe even this match. Improved opening preparation by Carlsen — he called it OK himself — but he once again ran into an opponent who knew more. Later on, Carlsen was his resilient self. The choices he was facing between 20...Bxd5 or 20...Qxb2 and 22...Qxb2 or 22...Qxf3 lead to similarly unpleasant situations, so he just took one. Both ...Qf6 and ...Kg7 strike me as good decisions.

 The bad news for Team Carlsen is there's no safe hiding spot against 1. d4 yet. The good news is that he has 2 Whites in a row now. Anand knew everything, saw everything, but did not push hard in the end(game). Choosing 26. Rc7, 26. Kg2 or 27. Ra4 was a freeroll. The outcome might have been the same, but there was some upside and no risk. The match is heating up, let's check out game 6!

 Game 6: Magnus CARLSEN won but could and should have lost. Magnus was playing White and got a good position from the opening (a Sicilian Kan) but later overlooked a simple tactical trick with which ANAND could have won. However, ANAND missed his chance. Unable to put up much resistance afterwards he lost without much of a fight. As CARLSEN said in the press conference: "I was very lucky."
 26.Kd2?? could have been disastrous, since 26...Nxe5! is close to winning for Black. Instead 26...a4?? followed.

Right from the opening I gained an initiative in the endgame. Soon the position was of a static nature where I had a clear advantage. It was not so easy to win but he had no counterplay at all. Then, as many of you know, I threw it all away in one move, but sometimes in chess you get lucky and he didn’t see it.
 After that the situation didn’t change much. I did have to allow, though, his pawn to advance from a5 to a3, not to fall for the same trick again. But still, it felt to me that his counterplay should not be sufficient and I should still be better, and that’s what happened. I managed to break through on the kingside and his counterplay was not enough.

1. e4
 We have the rare situation that both sides seem to be very happy with their first move and are sticking to it. Anand has used his superior 1. d4 opening preparation, Carlsen has used 1. e4 to steer the game towards the "slow" positions he likes. Both sides have been under pressure as Black.

1...c5!
 As in Game 3. I still consider this a better bet than 1...e5 against Carlsen. The e4 e5 territory where the price of a move is lower just fits like a glove with the champion's style.

2. Nf3 e6 3. d4
 An actual open Sicilian! This is quite a logical choice, actually. When Anand played 1...c5, 2...e6 the first time around, it came as a surprise, so Carlsen went for the less critical 3. g3. Second time around, no surprise. Three days have passed. Team Carlsen have done their work. Show me the money!

3...cxd4 4. Nxd4 a6
 The Kan is not something Vishy has done much in the past. One outing or two in rapid against weaker players. Surprise value — check.

5. c4
 Very much in Carlsen's style — he loves grabbing space and playing the Maroczy structure with White. But before the grind, the gods have put the theory.
 Variation
5. Nc3 or 5. Bd3 are all more en vogue than the text move.

5. Be2 Nf6 6. Nc3 Qc7 7. O-O Bc5 8. Be3 d6 9. Qd3 Nbd7 10. Nb3 Bxe3 11. Qxe3 b5 12. Qd4 Rb8 13. a3 e5 14. Qd2 Nb6 15. Bxb5+ axb5 16. Nxb5 Qc6 17. Nxd6+ Ke7 18. Rad1 Rd8 19. Na5 Qxd6 20. Qb4 Nbd5 21. Nc6+ Ke8 22. Qxd6 Rxd6 23. Nxb8 Rb6 24. exd5 Rxb8 25. b3 Bf5 26. Rd2 Kd7 27. f3 Kd6 28. Rc1 e4 29. b4 Ke5 30. c4 g5 31. d6 g4 32. fxe4 Be6 33. c5 Nxe4 34. Re1 Bd7 35. b5 Rxb5 36. c6 Bxc6 37. d7 Bxd7 38. Rxd7 f5 39. Rxh7 Ra5 40. Re7+ Kf4 41. Rf1+ Kg5 42. Re1 Rxa3 43. R7xe4 fxe4 44. Rxe4 ½–½ Fedorchuk, S (2673)-Anand,V (2785), Bastia 2014, CB43-2014.

 Variation
5. Nc3 or 5. Bd3 are all more en vogue than the text move.

5. Be2 Nf6 6. Nc3 Qc7 7. O-O Bc5 8. Be3 d6 9. Qd3 Nbd7 10. Nb3 Bxe3 11. Qxe3 b5 12. Qd4 Rb8 13. a3 e5 14. Qd2 Nb6 15. Bxb5+ axb5 16. Nxb5 Qc6 17. Nxd6+ Ke7 18. Rad1 Rd8 19. Na5 Qxd6 20. Qb4 Nbd5 21. Nc6+ Ke8 22. Qxd6 Rxd6 23. Nxb8 Rb6 24. exd5 Rxb8 25. b3 Bf5 26. Rd2 Kd7 27. f3 Kd6 28. Rc1 e4 29. b4 Ke5 30. c4 g5 31. d6 g4 32. fxe4 Be6 33. c5 Nxe4 34. Re1 Bd7 35. b5 Rxb5 36. c6 Bxc6 37. d7 Bxd7 38. Rxd7 f5 39. Rxh7 Ra5 40. Re7+ Kf4 41. Rf1+ Kg5 42. Re1 Rxa3 43. R7xe4 fxe4 44. Rxe4 ½–½ Fedorchuk, S (2673)-Anand,V (2785), Bastia 2014, CB43-2014.

5...Nf6 6. Nc3 Bb4
 7. e5 Ne4 8. Qg4 Nxc3 9. a3 Bf8 10. bxc3 d6 is fine for Black.

7. Qd3!
 This queen move is a Carlsen favourite. He used it in the last match in a very different position. In this one, it happens to be the best move as well! White protects e4 and c3 and threatens e5 and Nc2.

7...Nc6?!
 Of course this is home preparation, so I am not sure the punctuation is warranted. It is a logical follow-up to Bb4, so the ?! should be split between those two moves. By playing Nc6, Black skips the middlegame and heads straight for the ending. However, it's not a good ending. Deliberately going into slightly worse endings and trying to hold them is a strategy everybody and their dog was critical of after the last match. I think they were right.
 Variation
7...d6 8. Nc2! (8...Bxc3+ 9. bxc3+/=) 8...Ba5 9. b4 Bc7 has been seen more often, leading back to Hedgehog-type positions.

 Variation
7...d6 8. Nc2! (8...Bxc3+ 9. bxc3+/=) 8...Ba5 9. b4 Bc7 has been seen more often, leading back to Hedgehog-type positions.

8. Nxc6 dxc6 9. Qxd8+
 As we know, Carlsen is not the type to shy away from a pleasant ending.
 9. e5 is a serious alternative, allowing Black to avoid the queen exchange with 9...Nd7.

9...Kxd8 10. e5
 The only try. If Black were to get in e5 himself, he ain't worried about nothing.

10...Nd7
 Variation
10...Ne4 looks critical, but is met by the strong 11. a3! Bxc3+ 12. bxc3 Nxc3?! 13. a4! and the knight on c3 is short of squares. Bringing it home with 13...Ne4 14. Be3 f6 15. Bd3 Ng5 16. f4 Nf7 17. exf6 gxf6 18. O-O leads to tremendous compensation due to the bishops on an open board, the weak black king and the lead in development.

 Variation
10...Ne4 looks critical, but is met by the strong 11. a3! Bxc3+ 12. bxc3 Nxc3?! 13. a4! and the knight on c3 is short of squares. Bringing it home with 13...Ne4 14. Be3 f6 15. Bd3 Ng5 16. f4 Nf7 17. exf6 gxf6 18. O-O leads to tremendous compensation due to the bishops on an open board, the weak black king and the lead in development.

11. Bf4
 White has a simple plan: 0-0-0, Ne4, Nd6 and smother the opponent. Black has no choice but to eliminate the knight.

11...Bxc3+ 12. bxc3 Kc7 13. h4!
 Grabbing more space and giving Black a tough choice:

A) go ...h5 to stop that pawn
B) go ...h6 and stop it on h5
C) allow its march to h6

 My preferences are A and C, but B happened.

13...b6 14. h5 h6
 The reason this is the option I like least is that it creates a clear target on g7 for Carlsen to attack.
 14...Bb7 15. h6 g6 is no picnic either, and the thorn in the flesh on h6 is not much fun. Still, f7 strikes me as much harder to attack than g7.

15. O-O-O Bb7 16. Rd3!
 Immediately redeploying the rook to the kingside.

16...c5
 The nature of the struggle is well defined now. Black is solid but passive and will try to control the kingside situation. The chances to hold the position are not too bad.
 I thought Vladimir Kramnik said it best in the live commentary: "Maybe it is holdable, but the starting position is not that bad for Black that you have to go for such an ending."
 Coming from the man who put the Berlin on the map! Isn't that the same thing?
 Anish Giri: "The Berlin looks much more hopeful and double edged that this."

17. Rg3
 17...Rhg8? 18. Bd3 planning Bh7 and if 18...Nf8? then 19. Bxh6.

17...Rag8
 A necessity, even though the h8-rook might feel like Jodie Foster in Panic Room for a while.
 Anish Giri: "I like the plan Peter Svidler is suggesting, Rag8, c5, Nf8, possibly g5 somewhere. Black stands very tight."

18. Bd3 Nf8
 Anish, are you happy?

19. Be3!
 A touch of class. Black was getting ready for g6 or g5 with counterplay due to the attack on the f4-bishop by a knight appearing on g6. So Carlsen gets it out of the way ahead of time.

19...g6 20. hxg6 Nxg6 21. Rh5!
 And he's in time to keep everything under control. Placing the rook on h5 is a very nice move to make.

21...Bc6
 Here 22. Bxh6 does not work and Carlsen refrains from it over the next couple of moves, since pinning himself on the h-file is hardly the way forward. White has to be much better, but it is by no means easy to break through after 22...Nxe5.

22. Bc2
 Manoeuvring.

22...Kb7 23. Rg4
 Is this move any good? All this manoeuvring is way above my head.
 Anish Giri: "23.Rg4! is very cunning, gonna be a tough day for Anand."

23...a5 24. Bd1
 Still too subtle for me.

24...Rd8 25. Bc2
 Ok, he's just poking around.

25...Rdg8
 Since Bd1 didn't have an impact, so it's time to try something else:

26. Kd2??
 No! This move allows a simple tactic, leading to a winning position for Black. It's one of the biggest blunders we've ever seen on this stage, isn't it?
 Garry Kasparov: "Today's is surely a candidate for "worst WCh blunder ever" because of giant eval flip. From huge advantage to probably losing in one move."
 I could not agree more. 26. Kd2 is a much bigger blunder than Nf1 from the last match IMO.

 26. Kd1 was mentioned by Carlsen at the press conference. It achieves the same goal of bringing the king to his side with the minor difference that it doesn't lose the house.
 And 26. f3 looks like a sensible move as well.

26...a4??
 A rare case of mutual blindness. Anand explained that sometimes you don't accept a gift if you didn't expect it. While worrying about how to defend the position, you just don't look for tactical shots. Still, this tactic was child's play for both these players. I would be upset if I missed it in a blitz game. The reasons must lie in the realm of psychology. As usual, both players spotted it right after making their move. Pretty tough to get this one out of your head and continue the game normally!
 Oh, by going a4, Black is looking for potential counterplay with a3 and a later Ba4. The concept has pros and cons, creating a second weakness while generating some potential counterplay. What is quite certain is that it's inferior to Nxe5.
 Variation
26...Nxe5 would win on the spot! For example, 27. Rxg8 Nxc4+ 28. Kd3 Nb2+! 29. Ke2 Rxg8 and Black is two pawns up with excellent winning chances.

 Variation
26...Nxe5 would win on the spot! For example, 27. Rxg8 Nxc4+ 28. Kd3 Nb2+! 29. Ke2 Rxg8 and Black is two pawns up with excellent winning chances.

27. Ke2!
 Taking Nxe5 out of the position. The relief Carlsen must have felt here is hard to fathom.

27...a3 28. f3
 Back to normal. Carlsen is pressing, Anand is trying to hold.

28...Rd8
 Here 29. Bxg6 probably felt like cashing in too early to MC as after 29...fxg6 30. Rxg6 Be8 31. Rg7+ Rd7 32. Rxd7+ Bxd7 and Black has good drawing chances. The idea of Ba4-Bb3 is an important source of counterplay.

29. Ke1 Rd7 30. Bc1 Ra8 31. Ke2 Ba4?
 Sorry, I'm still shell-shocked by the Kd2-a4 sequence; can't really focus on the next moves. Neither could the players, apparently, since both sides were critical of their subsequent play. Ba4 was singled out by Anand as the crucial mistake during the press conference after the game.
 Variation
31...Ne7! kept the game going. 32. Rxh6 Ba4 33. Be4+ Bc6 still poses considerable technical problems, even though White keeps pressing. 34. Bd3 Ba4 35. Rg7 with an extra pawn and an edge.

 Variation
31...Ne7! kept the game going. 32. Rxh6 Ba4 33. Be4+ Bc6 still poses considerable technical problems, even though White keeps pressing. 34. Bd3 Ba4 35. Rg7 with an extra pawn and an edge.

32. Be4+ Bc6?
 The final lapse, if we can trust our engine friends. The way to stay in the game is admittedly not all that human, but since the alternative — the way the game went — is hopeless, it could have been tried nonetheless.
 32...Ka7 33. Bxa8 Kxa8 is surprisingly tough to play. White is a whole exchange up and can grab two pawns, but the small black army coordinates extremely well: 34. Rxh6 Rd1 35. Bxa3 Ra1! and good luck converting this!

33. Bxg6
 Now it's all simple: g6, e6 and h6 fall and Black isn't in time to whip up any play.

33...fxg6 34. Rxg6 Ba4 35. Rxe6 Rd1 36. Bxa3
 And a3 falls as well.

36...Ra1 37. Ke3!
 A small step for the king, a big step towards defending the title.
 37...Rxa2 38. Bc1 Rxg2 39. Rhxh6 wins.

37...Bc2 38. Re7+! 1-0
 Anand resigned. His king has to either step onto the a-file, where it blocks its own rook, or the back rank, where it's mated by Rxh6 and Rh8. Fabiano, how do you think they feel?

 Fabiano Caruana: "Shocking blunders. Vishy won't be able to sleep tonight. I imagine they're both horrified with how they played today, but Magnus can console himself with a point closer to the title."

 Game 7: Now that we've passed the halfway mark in this world championship match the pressure is firmly on ANAND who began the day a point down. ANAND chose to employ the Berlin against Magnus's Ruy Lopez but his position was being gradually eroded until he decided to take action and give up a piece for two pawns in a position many expert commentators said was lost. There then followed a marathon struggle as the younger Magnus tried in vain to break Vishy and his fortress, but the Indian never gave ground and the fortress remained intact. The game was agreed drawn on move 122, making game 7 the longest of the match so far.

The seventh game was an important one for me because I really wanted to put the match away at that point. I gained an advantage in the Berlin endgame. He saw nothing better than to liquidate to an endgame where I had a knight and rook and two pawns against a rook and four pawns.
 At the start I was absolutely convinced that my position should be winning and so I missed the moment very early on when I could have forced his pawn to advance to the a5-square, creating a weakness on the b5-square, where my knight could later hop in. That would have given me excellent winning chances. As it happened in the game it was much, much more difficult to win than I’d thought. Maybe it was not possible at all. I tried for a really long time, even with just knight and rook against rook. I felt there was absolutely no harm in playing but I didn’t manage to win.

1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.0-0
 Not avoiding the endgame this time around! We saw 4.d3 previously in the match.

4...Nxe4 5.d4 Nd6 6.Bxc6 dxc6 7.dxe5 Nf5 8.Qxd8† Kxd8 9.h3 Ke8 10.Nc3 h5 11.Bf4 Be7 12.Rad1 Be6 13.Ng5 Rh6
 14.Rfe1 h4

14.g3
 14...h4 15.g4

14...Bxg5 15.Bxg5 Rg6 16.h4 f6 17.exf6 gxf6 18.Bf4 Nxh4 19.f3 Rd8 20.Kf2 Rxd1 21.Nxd1 Nf5 22.Rh1 Bxa2
 23.b3 Bb1 24.Ne3 Nxe3

23.Rxh5 Be6 24.g4
 The game's novelty, and deep into the middlegame. The position had been reached three times in past GM games and they all ended in a draw. CARLSEN (or, to be more precise, CARLSEN's team) must have figured out that 24.g4 allowed him to retain some pressure.

24...Nd6
 Played after a five minute think, maybe ANAND was recalling his preparation, but from the amount of time he was taking on his moves it is possible he was already making them up over the board. CARLSEN, on the other hand, looked confident and played every move after a one or two minute think.

25.Rh7
 25...Bf7 is a little too passive. White has more than enough compensation after 26.Ne3! if 26...Kf8 27.Bxd6† cxd6 28.Nf5+/=. Or if 26...Rg8 27.Bxd6 cxd6 28.Nf5±

25...Nf7 26.Ne3
 White's pressure is annoying. He controls the seventh rank which forces Black to keep the h7 rook out of the queenside with his knight on f7, paralyzing it. The c7-pawn is weak and has to be defended by the bishop on f4. The rook on g6 is inactive and the bishop on e6 has no targets now that it has chowed down the a2 pawn. And yet, ANAND has good reasons to believe he will be OK.
 26...Ke7 27.Bxc7 Rh6 (28.Rxh6 Nxh6 looks slightly uncomfortable for Black, but may be defensible.) 28.Rg7 Kf8!–/+

26...Kd8 27.Nf5
 27...Bxf5 28.gxf5 leaves the knight on f7 hanging.

27...c5 28.Ng3
 CARLSEN played up to this point only taking two minutes per move at the most. After this move ANAND tanked for almost half an hour trying to figure out the position! He must have calculated the following:
28...b6 29.Nh5 keeps enormous pressure.
28...Kd7 29.Nh5 Kc6 30.Bd2!? preparing a fork on f4 also looks uncomfortable.
28...c4 29.Ne4 a6 30.Nc5 Bd5 31.Nxb7†!±

28...Ne5
 And now it was CARLSEN's turn to think. He thought for about 25 minutes before he decided to go for this check.

29.Rh8†
 Variation
He also had a very real alternative in taking on e5, for example: 29.Bxe5 fxe5
a) 30.Rh5
 30...Rf6 31.Ke3 Bd5 32.Ne4 Bxe4 33.Kxe4 Rf4† 34.Ke3 Rb4 35.Rxe5 and White's passers definitely look more dangerous than Black's.
30...Bxg4 31.fxg4 Rxg4 32.Rxe5 is again similar to the game.

b) 30.Ne4
 Any variation that does not involve taking on g4 looks too risky, for example: 30...Bd5?! 31.Nxc5 Bc6 32.g5!±. Or 30...b6 31.Rh5 Bxg4 32.fxg4 Rxg4 33.Rxe5 is similar to the game.
30...Bxg4 31.fxg4 Rxg4 32.Nxc5 Rc4 33.Ne6† Kc8 34.c3 is a much worse version of the sacrifice in the game.

 Variation
He also had a very real alternative in taking on e5, for example: 29.Bxe5 fxe5
a) 30.Rh5
 30...Rf6 31.Ke3 Bd5 32.Ne4 Bxe4 33.Kxe4 Rf4† 34.Ke3 Rb4 35.Rxe5 and White's passers definitely look more dangerous than Black's.
30...Bxg4 31.fxg4 Rxg4 32.Rxe5 is again similar to the game.

b) 30.Ne4
 Any variation that does not involve taking on g4 looks too risky, for example: 30...Bd5?! 31.Nxc5 Bc6 32.g5!±. Or 30...b6 31.Rh5 Bxg4 32.fxg4 Rxg4 33.Rxe5 is similar to the game.
30...Bxg4 31.fxg4 Rxg4 32.Nxc5 Rc4 33.Ne6† Kc8 34.c3 is a much worse version of the sacrifice in the game.

Sidestepping the check instead of blocking: 29...Kd7 leads to a White edge.
 30.Rh5!? stops Black from taking on g4: 30...Nxf3? 31.Rh7†+-
30.Ne4
 30...Kc6 31.Bxe5 fxe5 32.Rh5!± White's pieces dominate and his position will soon force Black to sacrifice on g4.
30...b6 31.Nxf6† Rxf6 32.Bxe5±

29...Rg8 30.Bxe5 fxe5 31.Rh5 Bxg4!
 I think this is the best practical decision that ANAND could have taken.
 Variation
The endgame in which he was not allowed to sacrifice his bishop already looked dangerous at best. For example: 31...Rf8 32.Ke3 Bd5 33.Ne4 Bxe4 34.Kxe4 Rf4† 35.Ke3 Rb4 36.Rxe5 Rxb2 37.Rxc5 b6 38.Rc3 and White should win the pawn race, making Black's position lost.

 Variation
The endgame in which he was not allowed to sacrifice his bishop already looked dangerous at best. For example: 31...Rf8 32.Ke3 Bd5 33.Ne4 Bxe4 34.Kxe4 Rf4† 35.Ke3 Rb4 36.Rxe5 Rxb2 37.Rxc5 b6 38.Rc3 and White should win the pawn race, making Black's position lost.

32.fxg4 Rxg4 33.Rxe5
 A fascinating endgame. I had the pleasure of doing live commentary for this game with super-GM Loek Van Wely, and we both had the feeling there should be a way to crack Black's defense, but we never found one! Somehow Black kept finding interesting resources that just barely drew; most of which involved counterplay and trying to trade off the pawns on the queenside, even at the cost of all of Black's remaining pawns. Nepomniachtchi mentioned in the live broadcast that he thought this should be a draw.

33...b6
 33...Rb4 34.b3 (34...c4 35.Re4) 34...b6 35.Ne2 a5 36.Nc3 c4 37.Na2 cxb3 38.Nxb4

34.Ne4 Rh4 35.Ke2 Rh6 36.b3 Kd7 37.Kd2 Kc6 38.Nc3 a6 39.Re4 Rh2† 40.Kc1 Rh1† 41.Kb2 Rh6
 You could say we have reached the starting point. The question here is how CARLSEN will attempt to make progress. It seems that Black's only weakness is his pawn on c7, and he must target that. A rook trade is fatal for Black, so he must avoid it unless White's king is so far away that the knight and pawn endgame can be calculated to a draw. This logically means that the knight must attack c7, so d5 is a prime candidate, but things are not so easy.

42.Nd1 Rg6 43.Ne3 Rh6 44.Re7
 ANAND always had to be careful of falling to a fork: 44...Rf6? 45.Rxc7† Kxc7 46.Nd5†

44...Rh2 45.Re6† Kb7
 46.Nd5 Rd2 doesn't help White.

46.Kc3 Rh4 47.Kb2 Rh2
 Variation
48.Ng4 Rg2 Putting the knight on e5 was an idea we toyed around for a while, but it did not lead anywhere: 49.Ne5 Rh2 50.Nd3 Rh5 51.c4 Rh2† 52.Kc3 Rh5 53.Re3 Rh2 54.Nf4 Rh4 55.Rf3 Rg4 56.Nd5 b5 and as much as we tried we could not break down Black's position.

 Variation
48.Ng4 Rg2 Putting the knight on e5 was an idea we toyed around for a while, but it did not lead anywhere: 49.Ne5 Rh2 50.Nd3 Rh5 51.c4 Rh2† 52.Kc3 Rh5 53.Re3 Rh2 54.Nf4 Rh4 55.Rf3 Rg4 56.Nd5 b5 and as much as we tried we could not break down Black's position.

48.Nd5 Rd2 49.Nf6 Rf2 50.Kc3 Rf4 51.Ne4 Rh4 52.Nf2 Rh2 53.Rf6 Rh7 54.Nd3 Rh3 55.Kd2 Rh2† 56.Rf2 Rh4 57.c4
 Finally White commits to c4. This locks down Black's queenside and creates a stronghold for the knight on d5. ANAND has two choices, he can continue without moving (by passing with his rook) or he can try to set up a c6 and b5 structure to hold.

57...Rh3 58.Kc2 Rh7 59.Nb2 Rh5!
 This move signifies the start of ANAND's commitment to play b5 and c6. This particular structure will leave c5 very weak, which is why the rook is now needed on the fifth rank.
 60.Na4 b5 61.Nc3 Rh4 62.cxb5 axb5 63.Nxb5 c4

60.Re2 Rg5 61.Nd1
 Variation
61.Na4 b5 62.Nc3 c6 63.Ne4 Rh5 64.Nf6 Rf5 65.Re7†
 This is strikingly similar to the line that was played in the game, but with the rook on f5 things change!
65...Kb6 66.Nd7†
 66...Kc7! 67.Rh7 Kd6 68.Nb6 Rf2†, Black holds.
66...Ka5 67.Ne5 Kb4 68.Nd3† the fact that the knight here controls f2 is of paramount importance: 68...Ka3 69.Ra7+-

 Variation
61.Na4 b5 62.Nc3 c6 63.Ne4 Rh5 64.Nf6 Rf5 65.Re7†
 This is strikingly similar to the line that was played in the game, but with the rook on f5 things change!
65...Kb6 66.Nd7†
 66...Kc7! 67.Rh7 Kd6 68.Nb6 Rf2†, Black holds.
66...Ka5 67.Ne5 Kb4 68.Nd3† the fact that the knight here controls f2 is of paramount importance: 68...Ka3 69.Ra7+-

61...b5 62.Nc3 c6 63.Ne4 Rh5 64.Nf6 Rg5
 64...Rf5 was also good.
 64...Rh4!? is another way of defending: 65.Re7† Kb6 66.Re4 Rh2† (67.Kd1 Rh1†) 67.Kd3 Rh3† 68.Re3 Rxe3+=

65.Re7† Kb6 66.Nd7†
 66...Kc7 67.Nxc5† Kd6 68.Ne4† does not work here!

66...Ka5 67.Re4
 Variation
67.Ne5 Kb4!
 68.Nxc6† Ka3= White's pieces are so awkward and the king is so exposed to checks on the ranks that it does not seem possible now for him to win! What a surprising turn of events.
68.Nd3† Ka3 69.Ra7 Rg2+=

 Variation
67.Ne5 Kb4!
 68.Nxc6† Ka3= White's pieces are so awkward and the king is so exposed to checks on the ranks that it does not seem possible now for him to win! What a surprising turn of events.
68.Nd3† Ka3 69.Ra7 Rg2+=

67...Rg2† 68.Kc1 Rg1† 69.Kd2 Rg2† 70.Ke1 bxc4!
 Very well calculated by ANAND. He realizes that with White's king on e1 this capture exactly works and he can draw comfortably.
 For example, 71.bxc4 Kb4 is pretty obviously drawn. Black just needs to push the a-pawn.

71.Rxc4 Rg3
 72.Rxc5† Kb4 73.Rxc6 Kxb3 leads to a knight and pawn endgame that is drawn, and it was basically what happened in the game after a much longer route.

72.Nxc5 Kb5 73.Rc2 a5
 White's pieces are completely paralyzed. With b3 being the only pawn remaining there is no way to untangle the queenside. The king cannot approach the b3-pawn to defend it and thus it is doomed to fall.
 74.Kd1 Rg1† 75.Kd2 Rg2+=

74.Kf2 Rh3 75.Rc1 Kb4 76.Ke2 Rc3 77.Nd3† Kxb3 78.Ra1 Kc4 79.Nf2
 CARLSEN goes for a position in which he can hope to win the pawns and then go into a drawn rook and knight endgame, but one in which he can torture ANAND. The rest of the game needs no commentary, ANAND held comfortably.

79...Kb5 80.Rb1† Kc4 81.Ne4 Ra3 82.Nd2† Kd5 83.Rh1 a4 84.Rh5† Kd4 85.Rh4† Kc5 86.Kd1 Kb5 87.Kc2 Rg3 88.Ne4 Rg2† 89.Kd3 a3 90.Nc3† Kb6 91.Ra4 a2 92.Nxa2 Rg3† 93.Kc2 Rg2† 94.Kb3 Rg3† 95.Nc3 Rh3 96.Rb4† Kc7 97.Rg4 Rh7 98.Kc4 Rf7 99.Rg5 Kb6 100.Na4† Kc7 101.Kc5 Kd7 102.Kb6 Rf1 103.Nc5† Ke7 104.Kxc6
 With the last pawn captured White has until move 154 to checkmate his opponent. The count begins.

104...Rd1 105.Rg6 Kf7 106.Rh6 Rg1 107.Kd5 Rg5† 108.Kd4 Rg6 109.Rh1 Rg2 110.Ne4 Ra2 111.Rf1† Ke7 112.Nc3 Rh2 113.Nd5† Kd6 114.Rf6† Kd7 115.Nf4 Rh1 116.Rg6 Rd1† 117.Nd3 Ke7 118.Ra6 Kd7 119.Ke4 Ke7 120.Rc6 Kd7
 Clearly White has made no progress and he calls it a day.

121.Rc1 Rxc1 122.Nxc1 ½-½
 Draw agreed.

 Game 8: if the previous day's marathon game had had an effect on the players' energy levels then it was not at all obvious to those watching around the world. Magnus repeated the Queens Gambit Declined from game three (which ANAND won), but this time was excellently prepared and confidently neutralised ANAND's initiative and space advantage. Having invested almost a full hour more in thinking time than CARLSEN, ANAND agreed a draw after a series of exchanges left little material and an equal endgame.

The eighth game was a difficult one. When I woke that day I felt kind of sick, nauseated. It was not a good situation to be in. So at the start of the game I felt pretty horrible. That’s why it appeared to some that I was sleeping at the board. Then at some point my medicines started to kick in a bit and I felt much better.
 The game itself there was not so much to talk about. My opening choice was kind of risky but the way he played – which was the most natural and also quite dangerous-looking – was perhaps not the most critical way and I managed to equalise and make a draw.

 After a 122 move game that we witnessed yesterday, the players might have been a tad tired. (Maybe that showed as Magnus almost dozed off to sleep in the middle of the game today!) But for Vishy this was maybe the most important game of the match. With the White pieces he has been successful in putting pressure on Carlsen in the last three games. After the two whites, this was Carlsen's black after almost three days. Team Magnus had done their homework. Carlsen had a stoic face as the game began. Maybe he was determined not to give any chances to the challenger today.

1.d4
 Anand sticks to what has worked so far for him.

1...Nf6 2.c4 e6
 The third time in the match that Magnus has played the same moves. The amount of flexibility is so huge from this position that you still may not know what variation is he angling for. Is it the QGD, the Bogo Indian, the Queen's Indian, the Benoni, the Blumenfeld?

3.Nf3 d5
 It's the QGD once again.

4.Nc3 Be7 5.Bf4 0-0 6.e3
 So far all of this had been played in the third game but now Magnus goes for the old Classical line.

6...c5
 Just to let you know how popular this line has been: more than 3000 games have reached this position. All the big guys of chess have played it. Kasparov has played it with White on six occasions and five times with Black. Karpov on seven occasions with white and 17 (!!) times as Black! and Kramnik 11 games as White and the same number as Black. So you can say that this is surely a line played by the World Champions!

7.dxc5 Bxc5 8.a3 Nc6 9.Qc2
 This is all pretty normal. Now the main move in this position is ...Qa5. But with Magnus you can always expect some rare sideline in a known variation. And he doesn't disappoint.

9...Re8!?
 In a position that occurred in nearly 1000 games in the MegaBase 2014, Magnus plays a move that has only been tried 32 times and never by a 2700 player! The idea of the move is very simple. Play ...e5 and gain space in the center.
 If White takes on d5 with 10.cxd5, after 10...exd5 the rook will be standing on a very useful half open file.

10.Bg5!
 Whether this is Vishy's preparation or not is unclear but this is definitely one of the best moves in the position not only removing the bishop from the direct threat of e6-e5 but also preparing to put more pressure on d5 with the help of Rd1 or 0-0-0.

10...Be7
 This is the first new move in the position. Carlsen must have looked at this position with fresh eyes, asking himself: "Why not ...Be7?" He found nothing wrong with the move and tried it. Previously d5-d4 had been played by the majority of the players. Magnus plays a move that looks pretty passive but there are a few ideas here. Black might want to go ...Qa5 later in reply to Rd1. As the bishop already retreated there would be no fork with b4.
 The computer evaluates these positions as +/=. Carlsen boldly goes into these lines and has the faith that he can neutralize White's advantage by playing accurately. On the other hand, White must be careful not to make even a small inaccuracy or the position would just peter out to equality.
 Variation
A game from the Indian National Championship of 2013 where the White player played extremely well went: 10...d4 11.0-0-0 e5 12.Nd5 Be7 13.Nxe7+ Qxe7 14.exd4 Nxd4 15.Nxd4 exd4 16.Rxd4 Qe5 17.Be3+/= 1-0 (17) Lalith,B (2569)-Thejkumar,M (2452) Jalgaon IND 2013 where Black has some compensation but maybe not enough.

 Variation
A game from the Indian National Championship of 2013 where the White player played extremely well went: 10...d4 11.0-0-0 e5 12.Nd5 Be7 13.Nxe7+ Qxe7 14.exd4 Nxd4 15.Nxd4 exd4 16.Rxd4 Qe5 17.Be3+/= 1-0 (17) Lalith,B (2569)-Thejkumar,M (2452) Jalgaon IND 2013 where Black has some compensation but maybe not enough.

11.Rd1
 The pressure on the d5 pawn begins to intensify.

 It is difficult to account for all of White's possibilities and definitely castling queenside looks dangerous for Black. But if your opponent surprised you, you usually do not want to go into the sharpest line:
11.0-0-0!? Qa5 12.Bxf6 Bxf6 13.cxd5
 If 13...Bxc3 14.Qxc3+/= or if 13...exd5 14.Nxd5+/=.
13...Ne5 14.Nxe5 Bxe5 15.f4 Bxc3 16.Qxc3 Qxc3+ 17.bxc3 exd5 18.Rxd5+/=
 White has pawn weaknesses but a pawn is a pawn.

11...Qa5
 A natural developing move that pins the c3 knight.
 Here 12.cxd5 Nxd5! 13.Bxe7 Ncxe7= would give White absolutely nothing.

12.Bd3
 12...dxc4 13.Bxc4 h6 is not just a transposition of moves because now White has the additional option of going to f4 instead of h4 and that is a more active square: 14.Bf4!? e5 15.Bg3+/=

12...h6
 13.Bf4? is impossible due to 13...e5! and if 14.Bg3 d4–+. Or if 14.Nxe5 Nxe5 15.Bxe5 dxc4–+.

13.Bh4 dxc4 14.Bxc4 a6
 In such symmetrical tructures all that counts is activity. Vishy has the more active pieces currently because all his pieces are developed. But that wouldn't last for long as Magnus is threatening to play b5 and Bb7 when the position would be close to equal. This is definitely the moment when Vishy must grab his chance.

15.0-0
 Variation
Immediately 15.Ba2!? was very interesting.
 The point is that White wants to setup the battery on the b1-h7 diagonal before Black can develop his pieces.

15...b5 16.Bb1
 Castling can wait. Mating the opponent's king is a much more pressing matter.
 16...Bb7? 17.Bxf6 Bxf6 18.Qh7+ Kf8 19.Rd7+- is game over. You could see how Black was not in time to challenge the white rook on the d-file.

16...g6
 This slight weakening of the kingside might be a small victory for White.
 17.Bxf6 Bxf6 18.h4 might be a little bit too much as Black can go 18...b4! 19.axb4 Nxb4 20.Qe4 Nd5–/+ and before White has even started his attack, his position is falling apart.

17.0-0 Bb7
 White has a small edge but with the rook coming to d8, I wonder how long will it last.
18.Bxf6 Bxf6 19.Rd7 Qb6! 20.Ne4 Be7 21.Nd6 Bxd6 22.Rxd6 Rad8=

 Variation
Immediately 15.Ba2!? was very interesting.
 The point is that White wants to setup the battery on the b1-h7 diagonal before Black can develop his pieces.

15...b5 16.Bb1
 Castling can wait. Mating the opponent's king is a much more pressing matter.
 16...Bb7? 17.Bxf6 Bxf6 18.Qh7+ Kf8 19.Rd7+- is game over. You could see how Black was not in time to challenge the white rook on the d-file.

16...g6
 This slight weakening of the kingside might be a small victory for White.
 17.Bxf6 Bxf6 18.h4 might be a little bit too much as Black can go 18...b4! 19.axb4 Nxb4 20.Qe4 Nd5–/+ and before White has even started his attack, his position is falling apart.

17.0-0 Bb7
 White has a small edge but with the rook coming to d8, I wonder how long will it last.
18.Bxf6 Bxf6 19.Rd7 Qb6! 20.Ne4 Be7 21.Nd6 Bxd6 22.Rxd6 Rad8=

15...b5 16.Ba2
 The Bishop is transferred to the long diagonal to create a battery with the queen. Combined with the bishop on h4, the entry into h7 looks quite possible but as Nigel Short rightly points out, the move ...Re8 has created a nice luft on f8 for the king.

16...Bb7 17.Bb1
 Capturing the knight on f6 followed by Qh7+ and Rd7! is a big threat right now but it is extremely easy to parry it.

17...Rad8
 Black is developed fully now. Positionally there seems to be absolutely no advantage for White now. The only way he can hope for an edge is by means of concrete threats. And that is exactly what Anand does now.

18.Bxf6 Bxf6
 The white queen has now direct access to h7 but Black's king is ready to run away to f8-e7.
 For example: 19.Qh7+ Kf8 (20.Qh8+ Ke7–/+) 20.Ne4 Rxd1 21.Rxd1 Rd8= White doesn't have much in this position.

19.Ne4
 By this point Vishy was down to 50 minutes having used up already more than an hour while Carlsen was still in his preparation with 1 hour and 44 minutes left.

19...Be7
 Carlsen keeps everything under control. The funny thing is that a discovered attack by the e4 knight will be devastating as now there is a mate on h7 followed by Qh8, but the the bishop on e7 controls almost all the squares of the e4 knight.

20.Nc5
 Variation
20.Ng3 was suggested by GM R.B. Ramesh. White has given up his important dark squared bishop in order to launch an attack, but if it misfires Black might get a long-term advantage.
 20...Kf8!? might be an interesting prophylactic option.
 But not 20...Bf6? 21.Nh5!± gives White a strong attack.

20...g6 21.h4! Rxd1 22.Rxd1
 Anand's threat of h4-h5 of course looks very dangerous but Black has this little tricky move now:

22...Rc8!
 23.Qe2 looks best but is not in the spirit of the position.

23.h5? Nd4! 24.Qd3
 This looks pretty scary for Black. But he can wriggle out of it.

24...Bxf3 25.gxf3 Nxf3+ 26.Kg2
 A) 26...Nh4+ 27.Kh3 Nf5? (27...Rd8!) 28.hxg6 fxg6 29.Nxf5 gxf5 30.Qd7+-.
 B) 26...Rd8! 27.Qc2 Rxd1 28.Qxd1 Nh4+ 29.Kh3 Qd8=/+ and Black is totally fine.

 Variation
20.Ng3 was suggested by GM R.B. Ramesh. White has given up his important dark squared bishop in order to launch an attack, but if it misfires Black might get a long-term advantage.
 20...Kf8!? might be an interesting prophylactic option.
 But not 20...Bf6? 21.Nh5!± gives White a strong attack.

20...g6 21.h4! Rxd1 22.Rxd1
 Anand's threat of h4-h5 of course looks very dangerous but Black has this little tricky move now:

22...Rc8!
 23.Qe2 looks best but is not in the spirit of the position.

23.h5? Nd4! 24.Qd3
 This looks pretty scary for Black. But he can wriggle out of it.

24...Bxf3 25.gxf3 Nxf3+ 26.Kg2
 A) 26...Nh4+ 27.Kh3 Nf5? (27...Rd8!) 28.hxg6 fxg6 29.Nxf5 gxf5 30.Qd7+-.
 B) 26...Rd8! 27.Qc2 Rxd1 28.Qxd1 Nh4+ 29.Kh3 Qd8=/+ and Black is totally fine.

20...Bxc5 21.Qxc5
 The structure of this position is akin to that of a Catalan. White would be very happy if his bishop were on g2 instead of b1. When Sopiko Guramashivili asked Peter Svidler in the commentary room whether this is the right time for Black to end his pre-game preparation, Svidler said maybe not yet. This shows that there are still some dangers lurking for Black if he doesn't play accurately but nothing much to worry for Carlsen fans, it seemed as if he was still in his preparation.
 Here 21...Rxd1 22.Rxd1 Rd8 23.Rxd8+ Qxd8 24.Bc2 Ne7 would have also equalized without much effort.

21...b4!?
 22.Qxa5 Nxa5 23.axb4 (23...Nc6?! 24.Be4!+/=) 23...Nc4 24.b3 Bxf3 (not 25.gxf3? Nd2–+) 25.Rxd8 Rxd8 26.bxc4 Be2 27.Rc1 Bxc4= was the critical line that Carlsen must have worked out at home.

22.Rc1 bxa3 23.bxa3 Qxc5 24.Rxc5 Ne7
 There is not much left in this symmetrical position. The rooks will be exchanged and a draw will be agreed.
 25.Rc7 (25...Bxf3 26.gxf3 Rd5=) 25...Bc6 26.Ne5 Rc8 27.Rxc8 Rxc8 28.Nxc6 Rxc6=

25.Rfc1 Rc8 26.Bd3 Rxc5 27.Rxc5 Rc8 28.Rxc8+ Nxc8
 Both players keep playing this position for another few moves. I do not think they had any doubts in their mind that the point would have to be split.

29.Nd2 Nb6 30.Nb3 Nd7
 Of course the knight had to be prevented from coming into c5.

31.Na5 Bc8 32.Kf1 Kf8 33.Ke1 Ke7 34.Kd2 Kd6 35.Kc3 Ne5 36.Be2 Kc5 37.f4 Nc6 38.Nxc6 Kxc6 39.Kd4 f6 40.e4 Kd6 41.e5+
 Draw was agreed. A very important game for the theory of the QGD. The idea of Re8 and Be7 is freshly baked in Team Magnus's oven and I am sure it will be subjected to many more tests in the future. But apart from that it was a pretty dull day for the viewers as there were absolutely no fireworks.
 In games 7 and 8 Carlsen was the one who proved to be the better prepared player. A big blow to Anand because it was one phase of the game where he seemed better than Magnus but not any more. The match seems to be slowly running away from the Challenger's grasp.

 Game 9: Undoubtedly the dullest game of the entire match. Magnus deviated from game 7 by playing 11.Ne2 (11.Bf4 was previously played) and for another few moves the game followed that of Dominguez - Ponomariov from 2012. At move 15 the World Champion set the stage for a draw by repetition, which he promptly seized and the game was concluded after just 20 moves. At this point in the match, with a maximum of three games remaining and CARLSEN holding a one point lead, it is up to ANAND to find a way to make a fight of it against an uncooperative opponent.

The ninth game – there’s not much to talk about there. My only real opening failure with White. After about 16 moves I didn’t see anything to play and I decided to force a draw by repetition.

1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.0-0 Nxe4 5.d4 Nd6 6.Bxc6 dxc6 7.dxe5 Nf5 8.Qxd8† Kxd8 9.h3 Ke8 10.Nc3 h5 11.Ne2 b6 12.Rd1 Ba6 13.Nf4 Bb7
 Variation
13...Rd8 14.Bd2 Nd4 15.Nxd4 Rxd4 16.a4 Bc8 17.a5 a6 18.Be3 Rxd1† 19.Rxd1 b5 20.Nd3 Be7 21.Bc5 Bd8 22.Nb4 Rh6 23.f4 f5 24.c3 Bh4 25.Rd3 Rg6 26.Kh2 Bb7 27.Nc2 Bc8 28.g3 Bd8 29.h4 Be6 30.Nb4 Bc8 31.Rd2 Bb7 32.Rd1 Bc8 33.Rh1 Bb7 34.Kg2 Be7 35.Nd3 Bd8 36.Kf2 Rh6 37.Re1 Bc8 38.Nb4 Kf7 39.Rd1 Ke8 40.Re1 Kf7 41.Re3 Rg6 42.Ke2 Rh6 43.Kd2 Rg6 44.b3 Rh6 45.c4 Rg6 46.Kc3 Rh6 47.Nc2 Re6 48.Nd4 Re8 49.Rd3 bxc4 50.bxc4 Bd7 51.Re3 Be7 52.Bxe7 Kxe7 53.e6 Bc8 54.Kb4 Kf6 55.Kc5 Bb7 56.Nxc6 g6 57.e7 Ba8 58.Re5 Bb7 59.Nd8 Bg2 60.Nc6 Kf7 61.Nb4 Rxe7 62.Rxe7† Kxe7 63.Nxa6 Kd8 64.Nb4 Ba8 65.Nc6† Kc8 66.a6 1-0 (66) Dominguez Perez,L (2726) -Ponomariov,R (2741) Leon 2012

14.e6 Bd6
 With hindsight one might be tempted to argue that here Black should have deviated with something like 14...fxe6 15.Ng6 Rg8 16.Bf4 but it looks quite dreadful and could only favor White. If there was a plan to duke it out, the decision came earlier.

15.exf7† Kxf7 16.Ng5† Kf6 17.Ne4† Kf7 18.Ng5† Kf6 19.Ne4† Kf7 20.Ng5† ½-½
 Draw by repetition.

 Variation
13...Rd8 14.Bd2 Nd4 15.Nxd4 Rxd4 16.a4 Bc8 17.a5 a6 18.Be3 Rxd1† 19.Rxd1 b5 20.Nd3 Be7 21.Bc5 Bd8 22.Nb4 Rh6 23.f4 f5 24.c3 Bh4 25.Rd3 Rg6 26.Kh2 Bb7 27.Nc2 Bc8 28.g3 Bd8 29.h4 Be6 30.Nb4 Bc8 31.Rd2 Bb7 32.Rd1 Bc8 33.Rh1 Bb7 34.Kg2 Be7 35.Nd3 Bd8 36.Kf2 Rh6 37.Re1 Bc8 38.Nb4 Kf7 39.Rd1 Ke8 40.Re1 Kf7 41.Re3 Rg6 42.Ke2 Rh6 43.Kd2 Rg6 44.b3 Rh6 45.c4 Rg6 46.Kc3 Rh6 47.Nc2 Re6 48.Nd4 Re8 49.Rd3 bxc4 50.bxc4 Bd7 51.Re3 Be7 52.Bxe7 Kxe7 53.e6 Bc8 54.Kb4 Kf6 55.Kc5 Bb7 56.Nxc6 g6 57.e7 Ba8 58.Re5 Bb7 59.Nd8 Bg2 60.Nc6 Kf7 61.Nb4 Rxe7 62.Rxe7† Kxe7 63.Nxa6 Kd8 64.Nb4 Ba8 65.Nc6† Kc8 66.a6 1-0 (66) Dominguez Perez,L (2726) -Ponomariov,R (2741) Leon 2012

14.e6 Bd6
 With hindsight one might be tempted to argue that here Black should have deviated with something like 14...fxe6 15.Ng6 Rg8 16.Bf4 but it looks quite dreadful and could only favor White. If there was a plan to duke it out, the decision came earlier.

15.exf7† Kxf7 16.Ng5† Kf6 17.Ne4† Kf7 18.Ng5† Kf6 19.Ne4† Kf7 20.Ng5† ½-½
 Draw by repetition.

 Game 10: Vishy ANAND brought the first opening surprise with his 5.Qb3 in the Grunfeld, a line he knows well, indeed until move 11 the game was following ANAND-Kasparov, 1999. However, it was Magnus' 7...Na6 which surprised the experts and commentators who were expecting the more usual 7...a6.
 Things were looking good for Vishy and a time advantage and frightening looking d6 pawn gave him a slight advantage. However, the ensuing complications and indecision took their toll on ANAND's clock and by now he was down to just 20 minutes. Whether failing to see a way to clearly improve, or feeling the pressure of the time and situation, some imprecisions led to mass exchanges and after 32.Rd2, with only the rooks left and a dead equal endgame they shook hands.
 CARLSEN thought for almost 34 minutes over how to respond to 19.Ng5 - with his opponent looking as though he meant business.

The tenth game was another difficult one. I decided to change the opening again with Black to play the Grünfeld Defence. He also played something different and we got a very complicated position. He had a passed d-pawn but I also had some good piece play. Potentially his d-pawn could become weak as well. After some more or less forcing moves we got to an endgame which I had previously thought to be pretty harmless for me. I thought there had to be a way to deal with the d-pawn and my pawn majority on the queenside should long-term be a very valuable trump as well, but I’d underestimated his somewhat peculiar looking construction on the kingside with a knight on g5 and a bishop on h6 – that it could actually cause my king quite a bit of trouble.
 Sometimes it’s hard to realise that when the queens are off the board there could be a mating attack against the king, and so I was forced to defend a position where he had the bishop pair and a passed pawn and he could certainly have done much better there. I was very surprised when he allowed me to take the e-file. That was a crucial moment. If he had taken the e-file himself I would not have had a chance to activate my pieces in the same way and I would have been condemned to a long and passive defence. As it happened, I managed to equalise and he decided to liquidate to a draw before things got any further.

1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.Nf3 Bg7 5.Qb3
 Here we go! The queen move is one of the sharpest ways of meeting the Grunfeld. White aims to quickly establish his grip on the center by forcing Black to trade on c4 immediately. Black on the other hand hopes to develop quickly and put pressure on the White center, many times at the cost of a pawn.

5...dxc4 6.Qxc4 0-0 7.e4
 7...a6 is an extremely sharp line that still is an important theoretical battlefield. However, the text move is also critical but is the older line.

7...Na6 8.Be2
 White has tried lots of moves, but the text remains by far the most popular. Now Black has to counter against the center with c5 for his setup to make sense.

8...c5 9.d5 e6 10.0-0 exd5 11.exd5
 After a few forced moves we have reached a position that has been a battleground of the Grunfeld almost since its inception. Black has two main choices that lead to somewhat different games. Normally Black is hoping to blockade the d5-pawn while retaining piece activity. His main danger is that the pawn will advance unchecked, but he also has to make sure the a6 knight returns to the game successfully at some point.
 And here 11...Bf5 is the main line, but the move in the text is nearly as popular.

11...Re8 12.Bg5!?
 If CARLSEN's team looked at ANAND's seconds to determine which variation of the Grunfeld might occur, there is no way they would have missed this move. Despite the fact that it is somewhat unusual (there are four moves that have been played more often) there is a key game between Wojtaszek and Ponomariov that the players are following.

12...h6 13.Be3 Bf5 14.Rad1
 14...Qb6 15.b3 Rad8 16.Rd2?! Ng4 17.Bf4 Qa5! 18.Rc1 g5?! and eventually Ponomariov beat Wojtaszek with the black pieces in 2012. Of course, one can only assume ANAND had a novelty prepared on move 15 or 16.
 An interesting try might be 18...b5! 19.Qxb5 Bxc3 20.Qxa6 Qxa6 21.Bxa6 Bxd2 22.Bxd2 Rxd5 gives White some compensation for the exchange, but that's it.

14...Ne4
 The improvement, one would assume. White is at an interesting juncture. He has to advance his d-pawn to make progress at some point, but he has many ways of doing so.
 ANAND here has three choices: he can advance immediately, which looks risky and is something that CARLSEN must have prepared. Or he can play 15.Nb5 and push the pawn later, or he can trade on e4 first.
 The computer evaluations that are shown in machines run by you and I will not easily determine what the best route is, deep analysis is needed for the position and sometimes it is even necessary to nudge the computer in the right path a few times.

15.Nxe4
 ANAND ends up playing a very logical continuation that is, after some thought, the number one choice of some engines.
 Variation
a) 15.d6!?
 15...Nxd6?? is suicidal and is met with 16.Qf4 Bf8 17.Bxa6 bxa6 18.Bxc5+-.
Somewhat better is 15...Nxc3 16.bxc3 Qf6 with Black's pawn on d6 about to come under fire and the c3 pawn hanging it is clear that he must act now, but it is not obvious how to continue. A sample continuation might be: 17.Qb5 Be4! 18.d7 Red8 19.Qa4 Bf5!?.

b) 15.Nb5 Nc7!? 16.Nh4! Bd7 17.Nxc7 Qxc7 18.Bd3 Nf6 19.Nxg6! b5 20.Qf4 Nxd5 21.Qxc7 Nxc7+/=

 Variation
a) 15.d6!?
 15...Nxd6?? is suicidal and is met with 16.Qf4 Bf8 17.Bxa6 bxa6 18.Bxc5+-.
Somewhat better is 15...Nxc3 16.bxc3 Qf6 with Black's pawn on d6 about to come under fire and the c3 pawn hanging it is clear that he must act now, but it is not obvious how to continue. A sample continuation might be: 17.Qb5 Be4! 18.d7 Red8 19.Qa4 Bf5!?.

b) 15.Nb5 Nc7!? 16.Nh4! Bd7 17.Nxc7 Qxc7 18.Bd3 Nf6 19.Nxg6! b5 20.Qf4 Nxd5 21.Qxc7 Nxc7+/=

18.Qxb2 Bxb2
 With the queens off the board the pawn can be more easily attacked by one of Black's rooks, but on the other hand White has gotten rid of his weakness on b2. Overall it seems that White has serious winning chances here.

19.Ng5
 Logical, but perhaps not the most accurate move order.
 Variation
19.d6 perhaps advancing immediately was more precise. If 19...Bf6 then 20.Bc4!? looks better for White. If instead 19...Nb4 20.d7 Red8 21.Bg5 f6 22.Bc4† Kg7 23.Be3 certainly remains uncomfortable for Black.

 Variation
19.d6 perhaps advancing immediately was more precise. If 19...Bf6 then 20.Bc4!? looks better for White. If instead 19...Nb4 20.d7 Red8 21.Bg5 f6 22.Bc4† Kg7 23.Be3 certainly remains uncomfortable for Black.

19...Bd4 20.Nxe4 Rxe4 21.Bf3 Re7 22.d6 Rd7 23.Bf4
 White's pair of bishops is strong, but Black has good control of the d4 square. There is no question that Black is clawing for equality, and that he is still far from it, but now the ball is in White's court on how he will break down CARLSEN's defenses.

23...Nb4!
 By bringing in his knight to c6 Black completes his defensive setup and he can start waiting to see what ANAND plans to do.

24.Rd2 Re8!
 Controlling the file is a nice touch. By no means is this equal, but it is not easy for White to keep improving his position. At some point he will have to consider giving up one of his bishops in order to go into a slightly better endgame.

25.Rc1 Re6 26.h4 Be5!?
 A cunning resource. White retains some advantage, but ANAND does not follow through.

27.Bxe5 Rxe5 28.Bxb7?!
 This simplifies the game too easily for CARLSEN.
 28.g3 simply keeping the pressure and slowly pushing the pawns on the kingside is by no means a win, but it does give Black some things to think about. For example: 28...Nc6 29.Bg4!? f5 30.f4 Re6 31.Bf3 Rdxd6 32.Rxd6 Rxd6 33.Rxc5 White is a little better, and the position is far from a dead draw.

28...Rxb7 29.d7 Nc6
 An important resource. Black catches the pawn right on time and White has no hope of winning the double rook endgame.

30.d8=Q† Nxd8 31.Rxd8† Kg7 32.Rd2 ½-½
 Draw agreed.

 Game 11: For the fourth time in the match ANAND rolled out the Berlin Defence and many were expecting a dull draw to ensue. However, ANAND had other ideas and unleashed the superb 23...b5! as a means of activating his light squared bishop, and at this point all three results were possible. However, just a few moves later Vishy seemed to mis-evaluate the position when he played 26...Rdb8 and 27...Rb4, and the bishop pair and passed pawn were not enough to compensate for his exchange sacrifice. To say that the rest of the game was straightforward for Magnus is too ignore the accuracy of his moves or to accord him with computer-like analytical abilities but, true to form, Magnus pressed home his advantage and at move 45 ANAND resigned the game and with it the match. Magnus retains the title and will remain world champion until at least 2016 where he will have to face a new challenger.

Then in the eleventh game he chose a different variation of the Berlin Defence – one that I’d studied a bit but we didn’t think was very likely to happen. I thought from the opening I had the more promising position and that my position was easier to play, but after a while it became more difficult to cover all my advanced pawns and pieces than I’d thought.

He managed to play this amazing b5-breakthrough which is always in the air, but which I’d missed at that particular moment. Then he had a lot of counterplay, lots of different possibilities, but fortunately for me he chose probably the worst one, sacrificing the exchange at the wrong moment.

That was a relief for me. I thought after he sacrifices the exchange I can in no way be worse and there should be a way for me to gradually extinguish his activity. As it happened things went very quickly – I sacrificed a pawn. I got very excited but I still managed to make a bunch of very good moves and that decided the game immediately.

1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.0-0 Nxe4 5.d4 Nd6 6.Bxc6 dxc6 7.dxe5 Nf5 8.Qxd8† Kxd8 9.h3 Bd7 10.Nc3 h6 11.b3 Kc8 12.Bb2
 An old Anand game went: 12...b6 13.Rad1 Ne7 14.Rfe1 c5 and was eventually drawn in Anand-Nakamura, 2010.

12...c5 13.Rad1 b6 14.Rfe1
 Played after a ten-minute think. It's possible that Carlsen was not entirely familiar with this position and already had to figure out over the board how to place his pieces.
 14.Nd5 a5 15.Nd2 Ne7 16.Ne3 was Kokarev-Leko in the Russian team championship earlier this year. The game was drawn without much happening.

14...Be6 15.Nd5 g5!?
 An interesting approach. This move severely weakens f6, but Anand claims that he can live with a knight there. On the other hand, on g5 Black protects f4, essentially cutting off reinforcements to the e5 pawn in the future. Also the White's kingside pawn majority has been successfully halted.

16.c4 Kb7 17.Kh2 a5 18.a4 Ne7 19.g4 Ng6
 So far Anand has played a very nice Berlin. His position is good, but of course White's central domination and extra space allow him to be at least equal.

20.Kg3 Be7 21.Nd2 Rhd8 22.Ne4 Bf8 23.Nef6 b5!!
 A bomb shell! This incredibly unexpected move comes with beautiful timing. The pawn cannot be taken or White already risks standing worse.

24.Bc3
 The safest idea for White is to ignore the pawn, but now Black's position is quite strong.
 Variation
a) If 24.axb5 a4 25.bxa4 Rxa4 allows Black to attack the weak c4 pawn and gives him control (for now) of the open a-file. Black is already better and it is not clear how White will hold his position together. For example: 26.Rc1 Nf4! 27.Nxf4?! gxf4† 28.Kxf4 Rd2! Black is better.

b) If instead 24.cxb5 c6! 25.bxc6† Kxc6 and now it is White's b3 pawn that is a big issue. The knight on d5 is destabilized which means it must retreat, exposing the weakness on b3, for example: 26.Ne3 Bxb3–+

 Variation
a) If 24.axb5 a4 25.bxa4 Rxa4 allows Black to attack the weak c4 pawn and gives him control (for now) of the open a-file. Black is already better and it is not clear how White will hold his position together. For example: 26.Rc1 Nf4! 27.Nxf4?! gxf4† 28.Kxf4 Rd2! Black is better.

b) If instead 24.cxb5 c6! 25.bxc6† Kxc6 and now it is White's b3 pawn that is a big issue. The knight on d5 is destabilized which means it must retreat, exposing the weakness on b3, for example: 26.Ne3 Bxb3–+

24...bxa4 25.bxa4 Kc6 26.Kf3 Rdb8?!
 Played after a 22-minute think. This interesting idea is not bad per se, but its follow-up is not nearly as good.

27.Ke4
 27...Rb3 28.Rb1 Rab8 29.Rxb3 Rxb3 30.Rc1 is playable for both sides, though White does have some annoying pressure against a5.

27...Rb4
 Black sacrifices the exchange, obtaining the pair of bishops, a passed pawn, and fixing his structure on the queenside. Overall he has good chances of obtaining compensation, but not for the full exchange unfortunately.

28.Bxb4 cxb4
 For Black to maximize compensation for the exchange, he absolutely had to back it up with activity.
 The only way of doing so was activating his rook on a8 with 28...axb4!.

29.Nh5
 The main issue for Black is that his position is still very solid, but it does not have an active way of making progress. If White trades all the minor pieces, eventually the extra exchange will just be an extra exchange.

29...Kb7 30.f4
 Here 30...Bd7! was an interesting resource, but White would have an ace up his sleeve following 31.f5 Bxa4 32.fxg6 fxg6 33.Nhf6 Bc2† 34.Kd4 Bxd1 35.Rxd1 c6 36.Nd7! And ignoring the pawn attack on d5 grants White a huge initiative and two monster passed pawns.

30...gxf4 31.Nhxf4 Nxf4?!
 This only helps White.
 Variation
31...c6 32.Nxg6 fxg6 33.Nf4 Bxc4 34.Nxg6 Bc5 was rather bad, but better than the game continuation.

30...gxf4 31.Nhxf4 Nxf4?!
 This only helps White.
 Variation
31...c6 32.Nxg6 fxg6 33.Nf4 Bxc4 34.Nxg6 Bc5 was rather bad, but better than the game continuation.

32.Nxf4 Bxc4 33.Rd7!
 The activity of the rook is starting to make itself felt. Black cannot dislodge this intruder.
 Black can try 33...Kc6 34.Rd4! Ba2 35.Rc1† Kb7 36.Rd7, but after the return of the rook to the seventh rank, things are much worse. For example: 36...Rc8 37.Nd5 b3 38.Nc3! b2 39.Nxa2 bxc1=Q 40.Nxc1 when Black's position is simply falling apart, starting with the f7 pawn.

33...Ra6 34.Nd5 Rc6 35.Rxf7 Bc5
 The material difference is too big. Anand is clawing for activity, but White's next stroke ends it.

36.Rxc7†! Rxc7 37.Nxc7
 If 37...b3 38.Nd5 b2 39.e6 this pawn will cost Black material while his on b2 is not going anywhere.
 And 37...Kxc7 loses to 38.Rc1 because after 38...b3 39.Rxc4 b2 40.Rxc5† followed by Rb5, White is on time to stop the pawn from queening.

37...Kc6
 One last try for a trick, but Carlsen is alert to it.
 If White carelessly plays 38.Rc1??, he loses to 38...b3 39.Rxc4 b2 and the black pawn queens

38.Nb5! Bxb5 39.axb5† Kxb5 40.e6 b3 41.Kd3 Be7 42.h4 a4 43.g5 hxg5 44.hxg5 a3
 White has to avoid 45.g6 a2 46.g7?? b2 will draw because the pawn will queen with check.

45.Kc3
 Black resigns as his pawns are no longer in play.
 For example: 45...b2 46.Kc2 Kc5 47.Kb1 Kd6 48.g6 Bf6 49.e7 and after Black gives up the bishop for this pawn the g-pawn will queen.

 The 11th game of the match turned out to be the last game. Vishy Anand was trailing by a point with the score being 5.5-4.5 in favour Magnus Carlsen. Vishy's solid choice of opening, Berlin, made it quite certain that he wanted to play it safe in game 11 and go for the win in game 12. The opening went wonderfully well for the Indian and after 20 odd moves he was verge of snatching the initiative. That's when things started to go wrong!

1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.O-O Nxe4 5.d4 Nd6 6.Bxc6 dxc6 7.dxe5 Nf5 8.Qxd8† Kxd8 9.h3
 This is the third time the exact position has been reached in this World Championship match. According to Carlsen the Berlin suits his style and he enjoys playing it with both the colours. And Vishy with his excellent opening preparation has made Berlin almost a bullet proof system.
 In the previous two games Vishy played 9...Ke8.

9...Bd7
 The idea of this move is to make way for the king to go to c8 where it will be safe. The bishop doesn't go to e6 immediately because it can be attacked by the knight from g5. Remember the key point: The light squared bishop is the key piece for Black in this position. Black as you will see takes utmost care that it remains on the board.
 Here 10.Ng5 Ke8 leads nowhere for White.

10.Nc3 h6
 The light squared bishop will be best placed on e6. It can be eliminated by Ng5 or Nd4. The move h6 prevents a future Ng5 and later Black goes c5 to prevent Nd4. All this is preserve his unchallenged light squared bishop.

11.b3 Kc8 12.Bb2 c5
 Controlling the d4 square. Now the bishop can go e6 without fear of getting molested.

13.Rad1 b6 14.Rfe1 Be6 15.Nd5
 This move was important as Black was threatening c5-c4 when his doubled pawns would be dissolved. It was necessary to stop it because that is an advantage that White gets in return for giving up his bishop. At the same time he threatens the move Nf4.

15...g5
 Didn't I tell you that the e6-bishop is the most important piece! Black will do just about anything to prevent it from getting exchanged.

16.c4 Kb7
 In general Black is behind in development but still he is fine because it is impossible for White to create any meaningful threats. The closed nature of the position gives Black a very nice position.
 Interesting is 17.g4 I show you this variation because I want to show you a very nice idea after 17...Ne7 18.Nf6 and now Black can play in the style of Benko Gambit with 18...b5!? sacrificing a pawn in order to open files on the queenside. And following 19.cxb5 a6!? 20.bxa6† Rxa6 Black has excellent compensation for the pawn.

17.Kh2 a5 18.a4
 Of course White must prevent the black pawn from opening lines on the queenside with a5-a4.

18...Ne7 19.g4
 This is the only real plan that White has in this position. He realises that the g5 pawn is a weakness and looks to take advantage of it with h4 in future. But before that he must fix this g5 pawn and hence g4 is a good move.

19...Ng6 20.Kg3 Be7
 At this point I felt that Black has achieved a very comfortable position. The h4 advance has been prevented and at the same time he has connected his rooks. Anand has absolutely no problems but White is not worse.

21.Nd2
 Carlsen starts to improve his f3-knight seeing that h4 is no longer possible.
 Variation
Instead 21.Nf6 could be met with the very strong 21...b5!, an idea which did occur in the game. White cannot take with the c-pawn as then b3 would hang and if he takes with the a-pawn then a5-a4 gives Black enough counterplay.
While 21.Nxe7 Nxe7 22.h4 gxh4† 23.Nxh4 Rag8 gives Black just too much counterplay.

 Variation
Instead 21.Nf6 could be met with the very strong 21...b5!, an idea which did occur in the game. White cannot take with the c-pawn as then b3 would hang and if he takes with the a-pawn then a5-a4 gives Black enough counterplay.
While 21.Nxe7 Nxe7 22.h4 gxh4† 23.Nxh4 Rag8 gives Black just too much counterplay.

21...Rhd8 22.Ne4 Bf8 23.Nef6
 Alternatives:
a) 23.Ndf6 Nf4 24.h4? Nd3! 25.Re2 Nxb2 26.Rxd8 Rxd8 27.Rxb2 Rd3†.
b) 23.Rd2 c6 24.Ndf6 Rxd2 25.Nxd2 Rd8.

23...b5!!
 A brave move by Vishy. He shows he is unafraid to take extreme amount of risk even in such a crucial encounter. Of course he already had a very comfortable position but this move just turns on the heat on Carlsen.
 The fact that there were also other good moves makes this decision by Vishy commendable. For example: 23...c6 24.Ne3 b5 was quite satisfactory as well.

24.Bc3!
 Such move is a mark of a World Champion. While many players would be quite disturbed by a beautiful move like ...b5!, Carlsen is not one among them. He maintains his cool and plays a move that puts the ball back in Vishy's court. That is the reason why beating Carlsen is extremely difficult. He never really gives you the game on a platter!
 Variation
If instead 24.cxb5 then 24...c6! was the brilliant idea. Following 25.bxc6† Kxc6 the main problem is the loose White center and also the weak b3 pawn which will fall pretty soon: 26.Nc3 Rxd1 27.Rxd1 Bxb3.
 Variation
Or if 24.axb5 a4! 25.bxa4 Rxa4 26.Rc1 Nf4! 27.Nxf4 gxf4† 28.Kxf4 Rd2 with a lot of activity for Black.

 Variation
If instead 24.cxb5 then 24...c6! was the brilliant idea. Following 25.bxc6† Kxc6 the main problem is the loose White center and also the weak b3 pawn which will fall pretty soon: 26.Nc3 Rxd1 27.Rxd1 Bxb3.
 Variation
Or if 24.axb5 a4! 25.bxa4 Rxa4 26.Rc1 Nf4! 27.Nxf4 gxf4† 28.Kxf4 Rd2 with a lot of activity for Black.

 Variation
If instead 24.cxb5 then 24...c6! was the brilliant idea. Following 25.bxc6† Kxc6 the main problem is the loose White center and also the weak b3 pawn which will fall pretty soon: 26.Nc3 Rxd1 27.Rxd1 Bxb3.
 Variation
Or if 24.axb5 a4! 25.bxa4 Rxa4 26.Rc1 Nf4! 27.Nxf4 gxf4† 28.Kxf4 Rd2 with a lot of activity for Black.

24...bxa4
 It was inadvisable for him to close the queenside with b5-b4 so this capture is a logical move.

25.bxa4 Kc6!
 At this point, almost everyone including the engines are very optimistic about Black's chances. He has pressure on the d5 point and makes the idea of infiltrating on the b-file with Rdb8 possible. Carlsen had to be very careful. He was visibly tensed because if he lost this game then the scores would be level with he having the disadvantage of Black pieces in the last game. While most players crack under pressure, guys like Carlsen come up with fantastic concepts!

26.Kf3!!
 The double exclamations are not for the objective merits of the move but for the explanation that Carlsen gave at the end of the game. This is what he said: "I felt that my center was disintegrating. The points e5 and d5 were under quite a lot of pressure. I saw that my king was doing nothing on g3, so why not bring it to e4 where it would protect everything?" So easy to understand when he explains it yet simply impossible for mere mortals to find it over the board. This gives us a gateway into the way Magnus thinks.

26...Rdb8
 Played after a 22-minute think.
 Here Anand considered 26...Bg7 which is also a reasonable move.

But in my opinion 26...Be7! was definitely the best move and would have given Black an excellent position. The idea is very simple. Black puts pressure on the f6 knight and intends to capture it. If the d5-knight takes back then the c4 pawn falls and if the pawn takes on f6 then d5 becomes weak. White is under pressure here.
 If now 27.Ke4 Bxf6 28.exf6 Bxd5† 29.cxd5† Kd6 Black is okay.
 While 27.Nxe7† Nxe7 when c4 is just too weak.
 Also possible is 27.Ne3 Rxd1 28.Rxd1 Bxf6 29.exf6 Rb8 30.Bxa5 Ne5† 31.Ke4 Nxc4.

27.Ke4 Rb4?!
 An extremely emotional decision by Anand. As he described after the game that he lost the control on his nerves. At the start of the game he was quite objective but here he sensed that it was his chance to force the matters and hence went for it. It turns out that this was not the best move. And practically speaking this move doesn't set White enough problems which normally such exchange sacrifices should pose.
 Better was 27...Rb3 but maybe sufficient only for equality after 28.Rb1 Ra3 29.Ra1 (29...Rxc3!? 30.Nxc3 Bxc4 31.Rac1 Bb3) 29...Rb3 30.Rab1 Ra3.

28.Bxb4
 Carlsen snatched the rook without much thought. As he said in the press conference, he usually loves to snatch material whenever he gets the chance.

28...cxb4
 Anand's decision is quite understandable. He makes way for the bishop to come to c5. And at the same time he thought he should have good compensation. But as Anish Giri rightly mentions: "Anand just felt it. He did not see it." The perfection with which Carlsen finishes off the game now is worth studying.
 Variation
I had a feeling that 28...axb4 was the better option because the a4 pawn is quite weak, for example: 29.Nh5 Rxa4 30.Ra1 Rxa1 31.Rxa1 when White definitely is better but Black at least has a pawn for his worries.

 Variation
I had a feeling that 28...axb4 was the better option because the a4 pawn is quite weak, for example: 29.Nh5 Rxa4 30.Ra1 Rxa1 31.Rxa1 when White definitely is better but Black at least has a pawn for his worries.

29.Nh5!
 Carlsen said he was quite proud of this idea with Nh5 with f4 that he found and maybe that was his best move in the entire game. Anand might have missed this.

29...Kb7 30.f4 gxf4?
 30...Bd7! 31.f5 Bxa4 32.fxg6 fxg6 33.Nhf6 Bc2† 34.Kd4 Bxd1 35.Rxd1 c6 was the last chance for Black to stay in the game.

31.Nhxf4 Nxf4 32.Nxf4!
 Very accurate.
 Capturing with the king: 32.Kxf4 c6 33.Ne3 Kb6 keeps White's advantage to a minimum.

32...Bxc4 33.Rd7 Ra6 34.Nd5 Rc6 35.Rxf7 Bc5 36.Rxc7†!
 The final little tactic which clinches the crown.

36...Rxc7 37.Nxc7 Kc6 38.Nb5 Bxb5 39.axb5† Kxb5 40.e6 b3 41.Kd3 Be7 42.h4
 Creating a second passed pawn.

42...a4 43.g5 hxg5 44.hxg5 a3 45.Kc3
 Here Anand stretched his hand forward which meant that Carlsen had successfully defended his title and won the match with a score of 6.5-4.5. Had Anand kept his nerves and drawn this game, chess lovers would have been treated to a very interesting 12th game. At the age of 23, Carlsen successfully defends his title. This is an amazing feat. He was surely the better player as Anand himself conceded in the press conference. The Carlsen reign will not end so easily! Who will be his next challenger? Majority of the people are talking about the Italian-American Fabiano Caruana. But the next World Championship is two years away, in 2016. Till then it's Carlsen all the way! Long live the chess king!