shams, i'm fully aware that riverbeast - and kingscrusher for that sake - are strong players, and stronger than me. and i didn't think that he was unaware of the current state of super-gm chess. but it seemed (and seems) that his enthusiasm for the success of the king's indian defence made him temporarily "forget" that modern chess relies on concrete lines and moves a lot more than in the days of dogmas and "sound" positions.

equally important: generally i agree with the gist of what riverbeast and kingscrusher say regarding computer evals and numerous chess fans' abuse of them (but as badknight points out, they are typically at a loss for other ways to make up an opinion about a game or position, due to limited understanding), and i agree regarding the specific problems that engines have with the kid position types, <but> regarding this specific game i think that concrete analysis will reveal that black was indeed clearly lost, objectively speaking, <and> i think that it didn't require some ultra-computerish lines or variations to win the game for white. neither do i agree that this position after black's 21st move -

click for larger view

is "easy" to play for black and "hard" to play for white. i feel that this is to rewrite the story of the game retroactively, based on anand's failure to play well from the white side of this. (and there's nothing that indicates that anand is in tip top shape for the time being.)

as i pointed out (on the nakamura page, i think), my feeling's that anand played too superficially, relying on general knowledge instead of concrete calculation, when he decided on and played 22. Kh1? - which gave away most of the <objective> advantage in my opinion. i know hesam7 claims that white's still objectively winning until 29. Nc4? but after 22. Kh1? i think most of what riverbeast, kingscrusher and naka claim about the computer evals of kid positions is again true, regarding the concrete position at hand.

naka found the very good concrete move 22... Bf8! which is, amongst others, based on one of the known attacking patterns for black here, in case of 23. Bxf8?? (when Nxe4, Qh4, Bxh3, etc. decides). 22. Kh1 loaded this gun for black - but there was no concrete reason to play Kh1 yet. it was a schematical & superficial move when the position required something else, and anand must have failed to consider black's concrete response 22... Bf8!

assume white plays 22. Qd2!? instead and consider what happens if black again tries 22... Bf8?! (now it's kind of pointless, but for illustration):

click for larger view

the only real change compared to the game is that white now can solve the issue of black's immediate threat (capturing the white dsb) by playing 23. Bxf8! since black here lacks a vital tempo due to 23... Nxe4? 24. fxe4 Qh4 25. h3 Bxh3 26. Qxh3 <not> being with check. (well, there's one more change, too - a white queen on d2 or c2 helps protecting against mate on h2 and f2 in the Qh4-h3-Bxh3-Qxh3 attacking lines - which is why white often puts a rook or a queen on the 2nd row in these positions).

this isn't about "computer perfect play" but about anand missing a single and not too surprising move from black; once Bf8 had been played anand must have realized that the position just had become a little awkward - unless he really wanted to follow up Kh1 with d6 to begin with. the play that followed showed that white lost much of his pieces dynamic potential after the committal d6-move.

according to himself it was

1) get anand out of his comfort zone

2) play "untheoretical" (the h6-move) to further confuse matters

3) attack white's king (14... f4 and so one, playing the typical attack despite being a tempo down on the standard position).

i don't claim it should've paid off - as i think should be obvious from my other comments around the game - but i think the above sums it up rather well.

<Nakamura is doing something unbelievable. He is entering a position, which is even considered to be dangerous for Black with an extra tempo for them (h6 is not really a great help and I guess soon h6-h5 move will be made)>

nakamura knew he did this - it was his plan, no matter if you (or shipov or whoever) think it was crazy or a mistake. objectively it probably wasn't a good move, but it still was his "concept" for the game - if we're going to take nakamura's words on face value.

And on topic - I believe that experienced GM ( preferably with lots of XP on both sides of the KID ) armed with strong computer and an hour or two of work on critical positions would give us excellent picture of what happened objectively.

"objectively, for players that always offer and accept draws after one half-move, chess is drawn."

Just to take future world champion Alekhine as an example, he seemed to have instinctively opted to play g3 and fianchetto his king bishop and would exchange his d5 pawn whenever black pushed his e-pawn, thus opening the center. This made it all but impossible for black to do a kingside pawn storm. (Until today the g3 and Bg2 idea is employed by many white players in order to attenuate any potential black pawnstorm.)

Alekhine vs G A Thomas, 1923

Alekhine vs J H Morrison, 1923

Kasparov was the greatest KID practitioner in the 1980s to 1990s. Here is a typical Kasparov KID win following the model described above.

Ljubojevic vs Kasparov, 1993

Kasparov quit the KID after Kramnik administered some beatings to him in this opening. Notice though that in Kramnik's KID victories, he never allows black unimpeded kingside play; instead he partially opened the center and went for strong central and kingside activity himself.

Kramnik vs Kasparov, 1994

Kramnik vs Kasparov, 1994

Kramnik vs Kasparov, 1997

These are the crucial games that may have caused Kasparov to quit the KID.

Thus you have two world champions, one in the early and the other in the modern era of the KID, both avoiding a totally closed pawn center; and playing for activity in the center and kingside.

Your post is really good!. Can I use it as a basis for a video?! Also note Kramnik even beat KID exponent John Nunn who was deeply theoretical.

I lost two KID's in my recent Classic unfortunately but they were both fun games to a point.

As a consequence of the pain of these couple of losses, I revised my d5 structure plans recently with this video which may be of interest to KID enthusiasts:

http://www.youtube.com/watch?v=PORH...

What you say intuitively about the fianchetto vs the Kings Indian is very important - Avrukh took the Fianchetto idea and basically did his opening books out of them, playing the Fianchetto against everything. Avukh's volumes seem to receive a lot of praise and even though I did buy them, I haven't had time to read them.

http://www.amazon.co.uk/1-d4-v-1-Gr...

Funny enough also in my two recent KID losses - the opponent did have a defensive fiancheto bishop. In one game though, there was a side effect of playing f3, that I should have picked up on more to save some valuable tempi. Also if White is using more time for prophylaxis with a fianchettoed bishop, sometimes black has alternative plans like play on the Queenside or even to undermine the center. The fianchetto is after all not looking after the f1-a6 diagonal as much, and therefore there is different potential effectiveness of these plans as well as the more direct K-side attack which is less likely to work.

I think it was Reti who popularized the KID in the 1920s. Reviewing his treatment of the opening, and the reactions of his esteemed opponents, is fascinating, and shows clearly that the pre-WW2 masters were enthusiastically pursuing theoretical opening debates.

Reti's first two KID games in CG database were no less than against Akiba Rubinstein.

Rubinstein vs Reti, 1919

Note that Rubinstein instinctively essays the g3 and Bg2 fianchetto (Rubinstein as you say was fond of the g2 fianchetto), and then exchanges center pawns to open up the d-file. Rubinstein's treatment renders a Kingside pawnstorm impossible, and Reti typically goes for active Queenside play. Rubinstein then was regarded as one of the strongest players in the world, and Reti must have been encouraged with his drawing the game.

Reti's second game with the KID against Rubinstein in 1919 resulted in a loss for him.

Rubinstein vs Reti, 1919

He exchanged his LS bishop for the knight on f3, and tried the f5 push, but Rubinstein typically countered in the center. Nevertheless IMO Reti won the opening battle, winning a pawn. It was Rubinstein's endgame brilliance that overwhelmed him.

Given his generally good result in the opening stage of the game, Reti tries again against Rubinstein.

Rubinstein vs Reti, 1922

Again Rubinstein essays the g2 fianchetto and this time is induced to close the center. Reti then embarks on the f5 pawn push; but we can see here the attenuating effect of the white fianchetto on black's planned kingside attack. Strong Queenside counter-play by white forces black to go for a perpetual.

In the following 1922 games involving Reti, white attempts to counter the black f5 pawn push with g4. Note that Reti himself as white this time played the g2 fianchetto.

Alekhine vs Reti, 1922

Saemisch vs Reti, 1922

Reti vs Yates, 1922

In his next two games, Reti played c5, transposing the game into a Benoni-like structure. He obtained a strong initiative in these games- and probably deemed them an opening success.

Gruenfeld vs Reti, 1923

Marshall vs Reti, 1924

The latter looks distinctively Benko- gambit like.

Reti also tried the KID against the formidable Alekhine in New York 1924,

Alekhine vs Reti, 1924

but Alekhine does the g2 fianchetto and completely opens up the center. This game must have discouraged Reti because he abstained from the KID for two years. By 1926, he must have found improvements and assayed it once again vs Alekhine.

Alekhine vs Reti, 1926

AAA responded again with the g2 Fianchetto, and outplayed Reti in a nice endgame.

Reti then quit playing the KID against top masters. He had waged a theoretical battle against two of the best. First against Rubinstein, although with a losing score, his KID could be deemed an opening success. Second against Alekhine, Reti must have regarded his KID a failure.

Reti did play it against a weaker master, Landau, who used AAA's idea, but Landau was not of AAA's caliber. Reti almost routinely won.

S Landau vs Reti, 1927

Ironically Reti in turn tried Alekhine's idea himself as white against another KID practitioner Yates; and lost:

When WWI broke out many European players (Alekhine, Janowski, Reti,, Vidmar, Bogolyubov, etc.) happened to be in Germany and were detained. They played chess day and night during the war years and mainly studied the Indian Defense now known as the KID which Nimzovich was said to have invented. These players were so successful with the defense after the war that Reti studied it for use with White and came up with the Reti Opening.

Related to Anand's caution OTB it may have been warranted, Nakamura's Kside attack can be quite dangerous.

An lesson example I use is