Thank you for the letter.I do not speak German, but here is the google translation:

If so, Anand pre-empts all that by the pawn sacrifice 15...d5!! To begin with, White's N on c3 is overworked. If takes the d5 pawn it cannot then defend the e4 pawn. On 16 exd5 Bb4 17 d6 Qd7 ( Black can't allow 18 d7 and 19 d7-d8/Q) the d6 pawn is going to fall.

It may take no more than a few tiny adjustments to Black's play to transform the evaluation of the position from a one sided looking win for White into little or no advantage for White.

After the move 13 Nh4 it is true that it seems much easier for White to play the pawn advance f2-f4 than it seems easy for Black to play the pawn advance ...f7-f5.

This is because it is much easier for White to free his f2 pawn by playing the move Nh4 than it is easy for Black to free the f7 pawn to move by playing the move ...Nh5.

However can Black do nothing else that is useful with his king's knight?

One answer is that the N on f6 supports the counter-advance 13...d5! If Black can play safely this advance this may defuse any prospects for a successful White king side attack arising out of the pawn advance f2-f4.

It is understandable that with Black deprived of counterplay and backward in development Kramnik should have wanted to attack quickly by c5.

However Black has not only been deprived of counterplay on the King side, White has in fact taken the initiaive there from Black, so that White has the initiative on both wings.

Now on either wing, the greater the advantage which White has over Black there, the better the chances of success are for his attack

White can muster in fact a greater superiority of material force over Black on the King side than he can arrange on the Queen side.

Therefore it is on the King side that White must proceed energetically and on the Queen side where he must play more conservatively.

Therefore 18 c5? followed the wrong policy. On the Queen side it was better to play more cautiously and to attack the King side instead.

So I thought before.

However the analysis given by <Hesam7> suggests another possibility.

What if the Queen side attack c5 assists White's King side attack instead of diverting resources away from it?

If the analysis given by <Hesam7> is right, this suggests that Kramnik's mistake was not 18 c5 but 20 Qh5 instead of 20 Qg4 and for this reason: Because it is after 20 Qg4 that White's Queen side attack is able to assist properly White's King side attack.

It is after the move 20 Qg4 that White's Queen side attack which includes moves such c5 and Rb1 followed by the sacrifice Rxb7! is able to act properly in concert with White's King side attack which includes moves such as fxg6.

In other words, it is the choice of the move 20 Qg4 instead of 20 Qh5 which enables White's Queen side attack to cooperate properly with White's King side attack.

1. Black's Rook on e8 is potentially overworked. If it has go to to g8 in reply to Rg4 it cannot then tie White's Queen's Rook to the defence of the e4 pawn so that this Rook becomes free to go to d1 to attack Black's d6 pawn.

On the chessdom website GM Naiditsch says of the position after Carlsen has set up a pawn centre that Aronian has no counterplay at all.

This gives us an indication of the explanation for Aronian's defeat.

After White has set up his pawn centre, Aronian has, by playing the flank pawn advances ...h5 and ...a5, made it more difficult for Black to play the pawn advances ...f7-f5 or ...c7-c5 so as to attack White's centre.

Another thing is that after 15 e4 the move 15...dxe4 concedes the centre and an advantage in space to White. It is a suboptimal move according to the chessdom analysis. An alternative to 15...dxe4 is to try to hold on to the centre by eg 15...Nb6.

On the move 19 gxh5 to quote from his remarks < 19 gxh5? Correct is 19 Ba2 and Black has to play very precisely to keep out of trouble...>

Reshevsky's comment on the move 22...Ne5 is < By giving up a pawn, Black has the initiative. The two bishops and the exposed position of the White King are more than sufficient compensation for the pawn sacrificed>

This suggests the following question: Why does White's acceptance of the pawn transfer the initiative to Black?

If we look at the position after the move 22...Ne5 we can see that Black has removed White's king's bishop so that it no longer controls the square g8.

Black's King enjoys therefore the use of the square g8 as a flight square, so that the check Qh4+ is no longer much of a threat.

With White's KB controlling the square g8, the move Qh4+ would indeed be a threat.

One answer is that if we look at the position reached in the game after 17 exd5, the d5 pawn is not only isolated but also not very strong. Black in fact ends up winning it and gains a draw.

If in the position after 17 exd5 we transfer the c3 pawn to e3 we can see that the d5 pawn can be supported by the advance e4.

That is not all. It is true that after 17 exd5 as in the game white has a c pawn which may advance to c4. However it is opposed by a Black c pawn and in fact Black plays ...c4 and prevent c4. An e3 pawn would lack a black opponent to obstruct the advance e3-e4.

Before the move 19 Nd1: < A very familiar type of position has been reached. White has <hanging> pawns at d4 and c4 and is more or less compelled to play for a King'a side attack, which has, however, under the circumstances,every chance of succeeding. White usually works for d4-d5 or f4-f5 in such positions. Botvinnik tackles the problem in an altogether different way: he brings his QKt over to the King's wing, where it decisively strengthens the attack. Each of his next few moves deserves an exclamation mark.>

On the move 22 Ng5: <The piano of the opening passed into a crescendo in the middle game and now becomes a fortissimo of attack. Throughout the next ten moves sacrificial combinations are always in the air; the black king's stronghold is smashed open with titanic power>

At the end of the game Euwe remarks < Though Botvinnik is primarily a position player, and though his construction of the game differs vastly from that of Alekhine, his play reveals, in his discernment of attacking chances, the greatest possible resemblance to the brilliant style of the world champion.>

An alternative to 19...Ra7 is the pawn sacrifice 19...b5! 20 cxb5 axb5 21 Bxb5 Ba3. There are three justifications for this:

1. White's a pawn and d4 pawn are both isolated and this gives Black at least partial compensation for the pawn.

2. White's 19th move Nd1 withdraws the knight to the back rank and disconnects the rooks, and this suggests Black's opening lines

In the position after 14...Rad8 White has the bishop pair.

Lasker manages however to bring about a position in which White's bishop pair counts for nothing.

Lasker does this in the following way: After the black squared bishops are exchanged following the advance ...c5 White's king's bishop enjoys no scope.

This suggests the double question of how the player facing the bishop pair may bring about such a state of affairs, and how the player possessing the bishop pair may prevent it.

It is conceivable that just one or two corrections may make all the difference to the side which can manage them.

For example suppose that taking the sequence of moves 13-16 as played in the game, White can manage the move Bf4 in place of the move Be3 and perhaps the move Kh1 in place of the move f3, while Black plays the same moves.

Then one variation is 13 Rfd1 Qa5 14 Bf4 Rad8 15 Kh1 c5 and now 16 Nb5 threatens 17 Bc7 or 17 Nd6 while 16 dxc5 Bxc5 17 Nb5 threatens 18 Bc7 and 18 Nd6.

In the final position both of the files remain obstructed.

It is along the diagonals h5-e8, a2-g8 and h4-d8 that Black's king has become subjected to attack.

This suggests that in order to gain a winning attack against an opposing king in the centre it may be not necessary to open a central file.

Fine says that every exchange of pieces takes the game closer to an endgame which favours Black's queen side pawn majority.

Recent games have indicated however a problem with this.

In the ending Black has indeed a queen side pawn majority. but White has more space.

White's advantage in space may result in White's king gaining a considerable lead in development over Black's king. In fact White may end up playing with an extra king.

<The move 11...h6 is open to question, as it moves a pawn in the opening. White's king's knight is going to head for e3 to control d5. Can Black afford to do nothing to hinder the manoeuvre? Suppose that Black tries instead 11..Qc7. Then on 12 0-0 Black has the skewer 12..Bc4>

I would like to point out that 11...h6 contributes to the fight for control of the d5 square. Black would like to play ...Nf6, but white can play Bg5 followed by Bxf6 and Nd5, and black will be unable to contest the d5 square with his bad dark squared bishop, which is a common theme in this type of structure. In the game, after 11...h6 Anand is able to conserve his knight and exchange the dark squared bishops.

If Black accepts it this displaces his e pawn and so opens the e file. This suggests that White must make it unhealthy for Black to open the e file, so suppose we place White's queen on e4 and the queen's rook on e1. Then on Qe8+ black's king can go to g7, so we want to take the g file away from black. Suppose we play g3. Then with White's queen on e4 as well, h4 is attacked thrice and on ...hxg3, h4 attacks the bishop on g5 which obstructs the g file. if the bishop moves, we have the g file.

This suggests the plan of 27 Qe4, followed by Re1 and then g3 and then h4 and then Nd4!! followed by Qe8 mate.

The first move of the pair is the move 31...Bh3 and the second move of the pair, namely, 32...Bg2 mate.