Matemax wrote on 11/14/08 at 11:53:36:


I was not at all impressed by my computer's endorsement of 18...fxe4.  It's precisely in this sort of position you distrust silicon advice the most.  But yes, 19.Nxe4 Bxd5 (perhaps 19...Nbxd5 is better -- I'm not sure) 20.g5 and now Black must certainly play 20...hxg5 (my computer does not like this move), since to allow gxh6 is disastrous: White in most lines after gxh6,gxh6 can just go ahead and play Nfg5, with an attack that far outweighs the value of the sacrificed piece, should Black take it.  No, the only hope is 20...hxg5 after which 21.Nfxg5 is met with 21...Nf6 and 21.Bxg5 with 21...Qb8.  In both cases White's attack is strong, but perhaps Black can defend.  At least he has that nice little g-pawn to hide behind.  I don't want to analyze much more (or hear much more from anyone else), because I have a game in progress.  I will report back when's it has progressed enough.

I notice that in the above-referenced game in the HIARCS forum, 17...c4 has just been played.  Wisely, no doubt.

I do not like 20...b4.
Instead maybe deserves attention the following:
1) 20...hg5 21.Neg5 (21.ab5 ab5 22.Ra8 Ba8 23.Neg5 N5f6 keeps the balance) 21...Re1 22.Ne1 (22.Qe1 Qe7 followed by N7f6 keeps black in the game) 22...Qe8 and black will fnd a defence against white's attack leaving with a pawn up. Of course the defence will be not so easy. I checked the position and seems Black survives.
2) 20...N7b6? 21.Qc2 b4 22.Rae3 is good for white.
Also 19...Nb6 is interesteing for black, which i think also keeps the balance.

Hi Markovich, we have a game which is following a similar path over at http://hiarcs.net/forums/viewtopic.php?t=1807

Regards

tafl wrote on 11/12/08 at 20:42:58:


Ah, I missed that.  Thanks.  But my comment about Chesspub remains true.

I complained before about the dearth of this, and it has been somewhat remedied.  However I was recently a little shocked to see that the highly theoretical 17...f5 variation has not been treated at all.  The PDF file remarks after 16...Nd7 that "Black's last move introduces two different plans," and then proceeds to consider only 17...c5. 

Yet there have been some major theoretical advances in 17...f5, or so it seems to me.  Indeed I was quite shocked in a recent CC game to encounter 18.g4!, a move that isn't entirely new but seems to be the coming thing and has scored very well indeed for White.  (Irritatingly, it is also not considered in J&J's Zaitsev book.)  There is a fascinating game where Ivanchuk tried 18...fxg5 19.hxg5 Ne5 but was brilliantly defeated by his opponent, I forget whom.  And in some more recent games Black tried to keep the position blocked by playing 18...f4 but without any success at all in the games I could find.

It would be mighty nice if Victor would sometime soon remedy what seems to be a pretty big gap in Chesspub's Spanish coverage.  Frankly concerning 18.g4! I am prepared to believe that it refutes 17...f5.  In no game that I could find was 18...fxe4 played, and you can see why not when you look at how well White's pieces are poised to attack.  In my game, still in progress, I nevertheless dared to do that, and I am hoping against hope to be able to defend (no help, please).  The game has advanced a few moves beyond the ones I quote, and so far I am still alive, but very much in fear of the outcome.