But it seems to me like people focus too much on that mutual blunder on move 26. I don't find that blunder interesting at all. What interests me, is whether Carlsen could break Anand defense in case Anand would have played simple waiting moves ?

Lets say that instead of the blunder, Carlsen played 26 Kd1 and continued with his plan to move his K to king side. If Anand just played Kc7 then, and Bb7-Ba6 (instead of his ambitious weird a4-a3 plan, that can't really be good), how would Carlsen proceed then?

White can't play f3-Rg3-Rh3 since c4 will fall. And if he tries to protect c4 with Bd3 then Rd8 is harassing the bishop, and if white manages to push his B to e2/f1 then the pressure from Ng6 is gone, and black can get counter play on the d file.

Most especially <1971>, <jphamhock> and <CassandraThess>, who would beat the spots off him blindfold.

jphamlore pointed out a good example from the Alekhine-Euwe match on another page. Not sure this one really qualifies, but:

click for larger view

Tarrasch played 19.Kh1?? and Lasker missed ...fxe4 which won by force. But unfortunately for Tarrasch, that wasn't the last blunder in the game.

What was Carlsen's body language like whilst waiting for Anand to move? after he played 26.Kd2. He would have seen it was a blunder the moment his hand quit the King.

Apparently he so shocked he could not write down his move. Poor lad, he is not use to making to blunders. Now would have been a good time to see his poker face. We won't get many more chances .

It's a wonder Anand did not pick up a blunder-made vibe.

The mysterious 26...a4.

click for larger view

Like everyone else I was looking at the 26...Nxe5 and wondering what was wrong with it. (except those that had an engine pumping) Everything looks so good for Black. It cannot be a mere blunder and I dug in looking for the twist.

When 26...a4 was played.

click for larger view

Gosh he's seen it and is constructig a matiing net after Kd3 and Kxc4. 26...a4 covers the b3 square.

Black is threatening 1.Nxe5 2.Rxg8 Nxc4+ 3.Kd3 Rxg8 (3...Nb2+!) 4.Kxc3 Rd8. and Black is on for Ka6 and Bb5 mate.

click for larger view

Remember I had this in my mind - 'fantastic. The White King is stumped, it cannot move, it is just waiting for a check. this is going to be up there with the immortals'.

But then you see White can sac on c5 with Bxc5 and wreck it. It's all kack-handed, wrong move order, just simply wrong, it does not work. He had to play 26...Nxe5 and missed it.

Anand had been given a penalty kick and blasted it over the bar.

Then Carlsen played 27.Ke2 and it was gone. So I cannot figure what was going through Anand's mind with 26...a4. Possibly he can't.

How does it end? Carlsen could have rubbed it in and thus (in the words of Larsen) make one win count as two wins.

click for larger view

38...Ka6 39.Rxh6 Ka5 37.Rh8. Here, have a Rook.

click for larger view

Larsen's theory being to win in such a way derails an opponents thinking the next time you play. The thirst for revenge out weighs rational thought.

The only reason Anand could even attempt this is that he has a phenomenal memory, work ethic, and discipline. But he apparently has neither the resources nor the team he had in say 2010 versus Topalov.

The irony is that if Anand loses this match and if he puts in the physical training late 2015, he could go into Candidates 2016 again with a stripped down opening repertoire and win it.

To me bravely going out with the intention of coming back in defeat only on one's shield is simply playing the best of what one's got, and if the other guy is better, the other guy is better. It's not playing stuff one does not normally play with some vague hopes of attacks that rarely materialize.

<<keypusher>: One simple, probably dumb thing that occurs to me -- after 32....Ka7, does W have to take the rook? Can he just play 33.Bxa3?>

Carlsen knew it was blunder the moment he played 26.Kd2.

Professional chess players exhausted in game 6 after 5 games and two rest days? No game has gone over 50 moves.

Don't make excuses for them, It was a basic chess players double blunder one sees week in week out in open tournaments. In a way it gives us all hope. (though I am so disappointed for Anand. Carlsen is a brilliant chess player, these chances will round very rarely.)

After 26...♘xe5+ 28. ♖xg8 ♘xc4+ 29. ♔e2 ♖xg8 30. g3, Black is really up only one pawn (since the h-pawn is untenable), so despite his weak a and c pawns, White can probably draw if he can exchange his light bishop for Black's knight, leaving opposite colored bishops. As evidence that this position might be drawn, Houdini is unable to win this ending playing against itself; it ends up stuck in a drawn position (even though it still thinks is a Black advantage of something like 1.15). I'm not claiming that Black doesn't have a win, but it seems that the "proper" result is as likely to be a draw as a Black win.

However, it may be the case that Carlsen has actually managed to find the glitch in Anand's own wall.

The glitch is - get a position where you can press him, even if objectively there is nothing.

It turns out from game 6 (in both matches!) that Anand just doesn't like these positions, and he will do everything to exacerbate his own situation - sacrifice pawns needlessly, such as last year), or play unnecessary weakening moves such as ...h6, ...a4 and ...a3 this year.

There was probably nothing wrong with sitting in one place, moving the B or the K back and forth and waiting - Anand just doesn't like that.

So - establish even a slight pressuring position, then sit and wait for Anand to self-destruct.

Just why this works against him, isn't really clear. Evidence is, it does.

This kibitzer saw ...a3 coming before Anand played it. He also predicted the exact fashion in which Carlsen later exploited that error, and in detail.

There was nothing seriously wrong with Anand's position (even discounting the fact that he didn't find ...Nxe5 when it counted) before he went ...a4. This kibitzer would never in a million years have played this move in this kind of position, engines or not.

This kibitzer wouldn't get one whole point in a thousand tries against either champion or contender (the other day I beat a FIDE 2376 very nicely, in a Carlsenesque positional squeeze - but his position was way worse than Anand's today before he blundered a R) - it's just that I don't suffer from this particular glitch, and have no difficulty in sitting and waiting, and letting the other guy "prove it".

Can you be more specific pls, and give some concrete variations for this?

As I said, doubling rooks on the h file doesn't seem to work for white, in case his c4 pawn is not protected when the black B is on a6. And taking the h6 pawn with either the B or Rh5 while his R is still on g4, doesn't seem to work either, since white will have troubles defending both e5 and c4 once the black rooks are no longer stuck on the K side.

Firstly, many noted <Sally Simpson> Carlsen's body language after Kd2. Luckily for Carlsen, Anand was preoccupied and missed an opportunity to study the position closely. You, I and all Chess players and sportsmen/women understand the power of psychology in sports, war, contests, relationship battles etc etc. The power of psychology in the course of an event is incredibly telling!

Secondly, you write -probably- and continue to suggest that the "probability" of a draw was/is high if White exchanges his light colour Bishop for Black's Knight. It is very likely that had Anand seen 26.Nxe5 the resulting continuations and variations would have allowed him to foresee that an ending with light coloured Bishops would be drawn thereby attempting to avoid such an ending; probably not allowing the exchange of his Knight (unless of course attempting to prevent the exchange of Black's Knight results in a worse position for Black)

However - you saw me write about Anand's ...a3 and its consequences BEFORE Anand played that yesterday. How can you explain such a move from Anand? I wouldn't expect a player of my own level (2000 FIDE on a good day...) to play such a move in this kind of a position.

Your opponent has the 2B's and is pressuring you on one side of the board - now you stretch your own defenses and create a weakness, a liability, on the other side of the board.

Now do you see how our use of software ought to be exercised with caution? As Houdini is unable to win the resulting position after 26...Nxe5, whereas Deep Fritz judges the position as won for Black? (Well..even Houdini goes for -1.15 and is unable to win)