MNb wrote on 02/12/07 at 20:32:19:

Yes, 28.g5 looks like a great move.  I'm sure I underestimated it when I first looked (although, I'm also pretty sure the previously mentioned 28.Bxc4 also leads to a winning position).

One line (after 21.Rh6 Qf8 22.Qh3 Qg7 23.Re1 a5 24.f5! a4 25.f6 Qh8 26.Nc1 Bf8 27.Rh4 h6) is:

28.g5 h5? 29.Bxg6! fxg6 30.Qxe6+ Kh7 31.Rxh5+! gxh5 32.Qf5+ Kg8 33.f7+ Kg7 34.e6+ Nde5 35.Rxe5 and wins.  Of course, Black is not obligated to play 28...h5, but I think this illustrates the danger (for Black) inherent in this position.

Quote:

Haven't looked at this yet (only 27...h6 so far).  Maybe tonight.

Quote:

I've only looked at 22.Qe2 so far (which I found convincing enough to go with 21.f5 last move) but, as I mentioned earlier, I will certainly take a look at 22.Re1 as well if 21...Qg5 appears on the board.

In the line 21.Rh6 Qf8 22.Qh3 Qg7 23.Re1 a5 

I'm becoming convinced that 24.f5 is, in fact, quite strong.  After 24...a4 25.f6 Qh8 26.Nc1 Bf8 27.Rh4 h6 we get to the position which I noted that I wasn't particularly impressed with.

                          analysis position
http://www.france-echecs.com/diagramme/imgboard.phpfen=r1r2bkq/3n1p2/2b1pPpp/1p1...
Willempie wrote on 02/12/07 at 10:57:59:

I also considered 28.g5 but decided that 28.Bxc4 etc. looked more convincing.  I think 28.g5 is also a good idea, though.

Willempie wrote on 02/12/07 at 10:57:59:

23.f6 looks good here at first glance, but I haven't analyzed it.  So far I've only looked at 22.Qe2, which I'm quite happy with.  If 21...Qg5 appears on the board, I'll definitely consider 22.Re1 as well.  Just haven't looked at it yet.

21...Qg5 was my suggestion, although I didn't think it looked clearly better than 21...Ncxe5.  However, there are enough options to explore that maybe it could turn out to be the best try after all.  In any event, it certainly merits consideration and I'm glad to have put it on the table.

The reason I did not like 21.Rh6 too much, is that two pieces (Rh6 and Qh3) were used to bind one (Qh8). White will have to reorganize his pieces. That takes time, which Black can use to generate counterplay. But then again, I am not an IM. So probably I overrate temporary factors.
The reason I liked 21.f5 and did not mind giving up the e5 pawn were twofold. Snatching that pawn means opening the diagonal for Bd4. It also means opening the e-file. The combination of a rook on e7 and that beautiful bishop on d4 is lethal.

Willempie wrote on 02/11/07 at 16:58:59:

Yes, S_F made the mistake of postponing the essential bishop move until it was too late, and it's been all downhill from there.

OstapBender wrote on 02/11/07 at 02:02:06:


Yes, please forgive this oversight on my part. I read through several pages of this thread yesterday, and these words escaped my attention.

OstapBender wrote on 02/11/07 at 02:02:06:


Well, burying the black queen is certainly a major achievement. I suppose the only mitigating factor (aside from Black's queenside activity) is that a white rook (or queen, if we give up the exchange on h6) will have to remain on the h6-square in order to maintain the bind, and therefore may also have limited powers to influence events on the opposite side of the board. Still, White will always have the freedom to switch flanks at a suitable moment of his choosing. One final thought, looking ahead a bit further, is that if we reach an endgame then the black king is also entombed on the kingside, whereas the white one could play an important role (a king can usually be considered a 'four point piece' when it comes to the endgame).

Well, I look forward to seeing how the game develops! Cheers,
Andrew

I must say that this thread has made for fascinating reading, and it has been a great pleasure to play through the excellent analysis of so many contributors. I haven't analysed the current position in much detail, though I would like to comment on one of the earlier possibilities, namely 21.Rh6 (instead of 21.f5 as played in the game). To begin with, I must say that I would have instinctively preferred something other than 21.f5 (in principle I think a tempo is a worthwhile investment to safeguard that pawn), although that is purely a first impression and it could well be that analysis will prove 21.f5 to be the best move.

Anyway if I were white in this position, I would ideally be looking to attack on the kingside, while maintaining the e5-pawn if possible. There are many possible ways of doing this, but 21.Rh6 certainly looks sensible, and after 21...Qf8 22.Qh3 Qg7 23.Re1!? looks sensible, preparing f4-f5.
This has all been suggested by previous contributors, but the following seems to have been the last word on the subject:
OstapBender wrote on 01/30/07 at 05:35:02:


After 23...a5 24.f5 a4 25.f6 Qh8 I would say that the black position is close to strategically lost. The point (of which some people may be aware but which, until now, has not been clearly stated) is that the black queen is completely 'buried' on h8, with no means of escape. As long as the pawn on f6 remains in place, Black will have no way to free the queen without resorting to some other material sacrifices.
That means that White does not even need to decide the game with a kingside attack - he can simply play in the centre and queenside, safe in the knowledge that the black queen will not play a part in the game any time in the near future, if at all. Hell, it may even be possible to give up the knight on b3 and still be better (although 26.Nc1 must of course be correct).
Computers are a little slow to appreciate all this, as the positional problem of the queen is not something that will decide the game in the immediate future. 

Following something like (after 26.Nc1) 26...Bf8, it is not clear whether White should play 27.Rh4 or just leave the rook where is is. Now, the only complicating factor is that Black is rather active on the queenside. However I find it hard to believe that this should really be enough, given that his queen will not be able to participate in any attack. White can choose between 27.Rh4 or some kind of prophylactic move on the queenside (maybe eliminating the powerful knight with 27.Bxc4, or perhaps something else).
I don't have time to investigate this in full detail now, but anyway I hope this will help to shed a bit of light on this particular variation.
Cheers,
Andrew

I am glad you have learned from your game against Scholar: not postponing the essential break move until it is too late.  I am optimistic here.