Correct.  Benefit of h7 square rather than h8, the King is less vulnerable to checks here plus able to keep a few more options open.

I thought it should have been played on the 14th move. :-)  Engines don't find this move very easily there, but once you force it, they realize it's a good move.  Engines still have a lot of trouble on both sides of this opening.

rybka was very reluctent to play a5,i had to force it.

looks like a very good try, my first 18-19 moves of analysis, a lot of it deleted when a crash happened in analysis, but i redid most of it; i considered it briefly i didn't really analyze it as black's moves were simple and forced to that point.  something i will take a look at.

I have a feeling you are about to see more pawns.

This what make the tour unique is that players and spectators can interact in the forum to have a more interesting game and make it more fun. This is a new norm for correspondence play.

Actually compared to last year you have talked less.

The engines scores are false in these positions, have had to sift these things through to beyond move 40 to determine the outcome of some lines.  Main advantage for me here is my attack will happen first, question is whether I can do enough carnage. High chances I'm doing fine although at the speed we've been playing, not really possible to be 100% sure, the positions ahead are as unclear as any I've ever seen.  I've got a good head start, having figured out a few moves back that Nb1 was indeed the best move, but still some positions remaining to check.  I'm at least 95% sure of my next move.  It's going to be the most important move of the game for me, so this is the point where I'll spend my time, one line at a time.  Will submit by Friday evening.

I won't be looking in on the forum again until the endgame, a few ideas I'm looking out for, in terms of breakthroughs and exchanges --- it's all very concrete from here.  The next few moves will probably be slow, then I expect they'll come quickly.  Estimate to be at least another 25 moves ahead.

+1

> thought 20.c4 was def gonna be played and seemed natural lol, I have no doubt Paul has his reasons

That's a key observation, and it's not at all simple to pass judgment on deep analysis.

If White had played c4, it was the move I expected, and I am not sure if it's value is better or worse, and from a human perspective it would be the only move I would expect, but looking at it with computers, perhaps not as Black's attack seems particularly imposing;  I had not intended to play Ng6, at least not immediately, and perhaps not at all - and my attack too seems a bit more direct, it's a different position.  I suppose one could argue on g6 the knight gets the away of Black's Kingside play, so that moves like Rg8 and Qe8-h5 are not possible.  There's a few tactical points too, e7 the knight can able c5, and Rxc7 does not occur with check with a knight on e7.  Plus Nd3 had some insidious points to it (and principally centralizes a piece, additionally some prophylactic and tactical points), and more or less forced me to play Ng6, bc Nd3 gives White the option of playing on the Kingside in some positions via g3-f4 (part of the reason of Bh6, to deter this), but Black gets a few additional options too; I may even be able to open up the position on the queenside eventually - moves like c5 (and Rxa5 if cb) are a bit more tenable than they were.

Am refraining to comment on my assessments of some positions and certainly not the current one, there's not much to say other than it is enormously complicated, and in many regards the machines have not dictated our play or assessed the positions with accuracy.  Still no exchanges yet!  The most intriguing KID structure position for me that I have ever played for sure principally due to the unusual development of the pieces, and  all the nuances - how it seems each move has to be dealt with in a different manner.  Plus this game is theoretically relevant to my repertoire, yet this is new ground being broken here (out of book on move 10!) ...  Paul also has shown some very fine quality in the depth of his ideas, and the placement of his pieces in a few positions, for instance, Qe2 of course, a farsighted move. I'm pleased to have matched him move-by-move so far.

ok, so i said in my last post that i would reserve my comments for the endgame in the case of concrete tactics that i expected... but paul's last few moves necessitated positional manuevering the last few moves, i thought most likely we would be immediately embarking on a tactical slugfest as seemed to be the case on an earlier c4... the lines that I had worked out I was doing at least okay and at worst not unfavorably unclear, i.e. quite a manageable semi-closed position with dynamic potential.

i guess now that it's not relevant --- to clarify 20.c4 bf6 and now if nd3 rg8, threatening h4 followed by qe8-h5

whereas 20.nd3 bf6?! g3! is a safe +/=, quite likely positionally winning for white after rg8 kh1
thus ng6 looked to me the most reasonable, preventing this g3 move as well as getting things going on the kside

[edit: on 19.c4 g4 nd2 [xf3 threatened] bf6  and now rg8 etc]

mostly though i've just been playing on intuition and one move at a time - i had no choice but to go for the race position, and there is no reason to look more than the next move - playing the pawn moves first is always right as long as the black kside pawn chain cannot be undermined, or white has some threat somewhere (like 15.Nb4 threatening Nc6 with an edge so Nb8) that has to be stopped  - so from there... i can push on to 19...g4.  i would go as far to say there is nothing to calculate for Black on every move until his 20th move if you know the king's indian position or play the modern defenses or if i want to not get simply rolled over on the queenside, and i am already familiar with this type of position, this was a bit different for me with qe2/bc8 instead of qd2/bb7 which i studied two weeks before this game ---  qd2 and bb7 misplaced instead of the ideal qe2 bc8, so it is similar.  so i don't need to consider much on my opponent's choices much, i just set up a mating attack and if i break thru on the kingside i win because i can add more forces than white can muster on that side;  i calculate more after my opponent's move... then i analyze concretely more deeply, bf6 and rg8 was my plan; but as i've indicated on nd3 this was not possible.  just have to make sure i don't allow white the slightest chance of counterplay.

my analysis for moves 21 to 23 is really quite gigantic i admit. intuitively though it's actually quite logical to see where to look, and the tactics should work out, i admit they're quite crazy. in an over the board setting any result is possible, although its much easier to play black.

I don't understand Nd3 here, either. I would like to hear from Paul why he played it.

my King i hope will be "fleet of foot".

i woke up to a crash,so had to start again.