16...Nb6 Structurally this is even better, but after possible 17.Nxb6 Qxb6 18.Bxe6 fxe6 19.Qd3 Black will have to think about the a5-knight's employment.

17.Nxb6 Qxb6 18.Bxe6 fxe6 19.Re2 Anand protects on f2, and is able now to move his knight, however, the knight's destination is yet to be found.

19...Rc6 20.Qd3 Black has a sensible plan - 20...Qc7 21.Rd1 b5, and the knight goes to b7, from where it protects the d6-pawn and is ready to jump to c5.

20...Qc7 21.Rd1

21...Nc4?! This allows White to improve his pawn structure: 22.b3 Nb6 (it is crucial that after 22...Na3 White has 23.c4 b5 24.c5!) 23.c4.

22.b3 Nb6 23.c4 One can assess this position as follows: White stands slightly better. Black is rather cramped, and after White advanced his b- and c-pawns, one has to keep an eye on c4-c5 threat (which will be executed on, for example, 23...а5). Moreover, it is not clear what should Black do with his knight. White may cover the c5-square by b3-b4. Still, Leko's position is quite solid.

To Raffi: well, this game goes on, so we'll stick to it. Enjoy its positional subtleties. :)

To Thomas: Topalov-Mamedyarov is indeed an exciting game. Topalov's exchange sac is interesting, and the knight was very important piece, but an exchnage is an exchange. The chances (after 21 moves) are 50-50 in my opinion. 

23...Nc8 24.Red2 h6 25.Qe2

It is pointless to give any concrete variations here. Black's position is still solid, but quite passive. His main problem is bad c8-knight. White may vary plans and gradually improve his pieces' and pawns' locations - bringing the knight via g4 to e3, maybe advancing b3-b4-b5 or jumping to g4 with the queen, from where it creates some tension on the kingside and on e6...

Topalov-Mamedyarov: Shakhriyar's move 21...Kf8 is very suspicious. He did not need to catch on the h4-pawn. Stronger was 21...Bd6, fighting for dark squares, but even 21...0-0 would have been stronger than the text.

To Alejandro: yes, I don't like 21...Nc4. I think 21...b5 is better idea.

To JaiDeepBlue: here is my view on a quiet course of some Anand-Leko games including this one. Great players like Kramnik, Leko and Anand for some reason feel too much respect and awe to each other, which often leads to unmotivated rejection of challenging continuations. I think Anand could score more points if he played sharper.

25...Kh7 26.h4 White's plan is to play g3, Kg2, h5 and then Nh4, securing g4-square for the queen and g6 for the knight.

26...Qb6 27.h5! Black's position is already difficult. Leko handles this game unsatisfactory. Compare two knight, one on c8, and another one - on g6.

A few words about other games. Karjakin's attack looks deadly. Mamedyarov spoilt his game by a couple of second-best moves, and stands worse, but the situation is still very complicated. Gelfand and van Wely agreed to a draw in an even position - I don't think White has any problems with achieving a draw. Kamsky's position is bad - he has weaknesses on d6 and b4, plus a bishop is better than a knight in such positions. In Tiviakov-Sokolov White presses without risk. Aronian and Adams have mutual chances, I don't think Black is much worse. White has a bishop pair, but the c1-bishop is out of play, and the time required to activate it must be cleverly used by Adams to create some counterplay.

27...Qc5 28.Ne1 It seems Vishy wants to carry out c4-c5 with a tempo: Ne1-d3 and c4-c5! Black's position is bad.

28...Rc7? 29.Nd3 Qb6. Now 30.c5! makes White's advantage almost decisive. Black's position is good only on a- and h-files, while he is in trouble on the files from b to g.

30.c5 The games Karjakin-Bacrot and Carlsen-Beliavsky were very one-sided. Both teenagers celebrated nice opening victories, as their opponents were for some reason completely unprepared. And good opening preparation plays more and more crucial role in modern chess... Ivanchuk's position looks technically won, I don't think Kamsky can survive, as Vasily is nearly unstoppable with this kind of advantage.

30...Ne7. White can play 31.b4 threatening a4 and b5. Capturing on d6 is premature.

To Alex: actually, we commented the same 3rd round game with Shipov. We do not make any arrangements, but the field in Wijk aan Zee is more diverse, which makes the choice richer.

31.Qg4 Rf6 32.b4

32...d5 Black was almost in zugzwang, as White had multiple threats like a4 and b5, or cxd6 and Nc5. Now he drops a pawn.

33.Nxe5 Qa4 Anand's in completely winning. He has many ways to play a better position with an extra pawn, and the most precise is probably 34.Qg3 with the idea Ng4, and if 34...Rc8, then 35.a3, keeping all advantages of his position.

34.Qg3 Rc8 35.Ng4 Rf7. 35...Rf8 allowed to extend the struggle, although this would hardy alter the result. Now White wins by 36.Qd6! followed by Ne5, and if 36...Rc6, then 37.Qd7.

36.Qd6! Rcf8 

37.Ne5 Nc8 38.Qxe6 +-.

37.Qxe6 Qxb4 38.exd5 Qxc5 39.d6 Nc6 Leko is in a serious time pressure. Anand's 37th move is less perfect than his presequent play, but is sufficient to win.

40.d7 Nd8 41.Qe4+ The control has been passed, and it is time for Black to give up.

Seb asks about critical moment of this game. In my opinion, Leko made two main mistakes. I didn't like his idea with 16...Nb6, when he succumbed to certain passivity. And 21...Nc4? was just awful. After this Black's position is so passive that could be held only by the computer. I think no human player could defend it with such a terrible knight on c8.

41...Qf5

Gustavo, I looked at the final position in Navara-Naiditsch, and the result looks justified for me.

42.Re2 Ivanchuk's position is technically won. Adams sacrificed a pawn, but did not get compensation for it. Ivan Sokolov is a pawn down, but can make a draw with Sergey Tiviakov. The most interesting of the remaining games is Topalov-Mamedyarov. I am not convinced with the plan chosen by Topalov, he might obtain technical difficulties with winning the game. Advancing the f-pawn at some point was interesting.

42...Qxe4 43.Rxe4 

I don't think this game will offer us anything more than just converting a decisive advantage. Only a miracle can save Leko. The winning plan is Re8, Ne5, Ng6 - and then to play depending on Black's actions.

A curious situation arose in Topalov-Mamedyarov. If Black takes on f3, White gets winning chances (although his win is not certain, the endgame is tricky), but it is possible that White simply can't force the opponent taking on f3! Shak can just keep his rook on the 4th rank, and it seems to force a draw.

Therefore, I am leaving. Tomorrow we'll see more of Anand: Kamsky-Anand. Bye!

43...b5 44.f3 a5 45.Ne5 Rf6 46.Ng6 Rg8 47.Re8 Rf7 48.Rd5 b4 49.Ne7 Black resigned. 1-0