1.d4 Nf6 2.Nf3 e6 3.c4 c5 4.d5 b5 5.de6 fe6 6.cb5 d5 7.g3 a6 8.ba6 Bd6 9.Bg2 0-0 10.0-0 Nc6 11.b3 Ba6 12.Bb2 Qe8! 13.Nbd2 Rd8 14.Re1 Ng4?! 15.h3 Nf2! 16.Kf2 Nd4! 17.g4 h5 18.g5! Qf7 19.e3? Nf3 20.Nf3 Bh2!! 21.Bf1 Qc7! 22.Ba6 Qg3 23.Ke2 Qg2 24.Kd3 c4! 25.Kc3 Rf3 26.Qd4 Rf2 (and Black is winning) 27.Qb6 Be5 28.Kb4 Rb8 29.Bb7 Bb2 30.Qe6 Kh8 31.Kc5 Ba1 32.Bd5 Qg5 33.Ra1 c3 34.h4 Qh4 35.e4 c2 36.Rc1 Qg5 37.Qh3 Qc1 0-1 [Nisipeanu]
I think that a player (maybe at the exception of Mister Huebner) cannot analyse all possible moves, and it's obvious that 17.g4 (game) or 17.Nf1 are more flexible, the Nd4 can be exchange later or repulse by e3.

I want to thank everyone for their input (especially MNb, as now I'll have to go back and look at that transpositional opportunity and see if 18...e5 is better than 18...Qg6).

Also, and with all due respect to Smyslov_Fan, I too think the position is FAR to complicated to be evaluated on the basis of general principles and pedagogy.  I'm aware that I didn't supply any "analysis" of my own, but I didn't really have anything to offer as I couldn't really come up with anything clear.  I think that when one side is up an entire extra piece and there are huge imbalances (center vs. queenside pawns), dismissing a move on general principles is silly ("Black's center pawns get to advance with a gain of tempo.  That's my guess as to why it wasn't even considered").  Of course, we all use general principles to help our decision making over the board, but after the game, it's concrete analysis to get to the truth of the position!

Having said that, I am again aware that I didn't contribute any of my own...

Smyslov_Fan wrote on 08/21/07 at 15:17:49:


I am aware of that. What was to be gained by 17.Bxd4 was a) the removal of a centralized piece and b)the long term advantage of two connected passed pawns.
At the other hand I cannot imagine, that very strong players like Georgiev and Nisipeanu would reject 17.Bxd4 without reason. And yes, the instinct of two GM's is usually more reliable than silicon analysis.
Still, with my poor understanding, I don't get it (happens quite often). After all that knight on d4 only slightly later removes an important defender. So I will look at Ostap's line, look if I see the point.

Later edit:
Got it. It is in fact in the notes, but via a transposition: 17.Nf1 Qg6! 18.Bxd4 cxd4 19.Qxd4 e5 (variation b2). It appears, that 17.Bxd4 cxd4 18.Nf1 Qg6 and 17.Nf1 Qg6 are of the same value. Of course I should have looked at Nisipeanu's analysis at chesscafe in the first place.  
Still I find Nisipeanu's analysis not entirely convincing. He thinks variation b2221 (26...exf3 27.Bxf3 Bd6 28.Rc6) good for White and variation b2222 (26...Bd6 27.Rc6) only slightly better for White. Huh? 26...Bd6 27.Rc6 Bb5? loses to 28.Rxd6 Rxd6 29.Rxe4, so it has to be 26...Bd6 27.Rc6 Bb7 and indeed it is hard to decide where the rook has to go. I suspect Nisipeanu simply thought his analysis enough at this stage.

Yes, this is clear and logical.  Nice display of pedagogical skill, S_F!

Still, I wonder if 17.Bxd4 would be rejected so quickly by Georgiev (or by Nisipeanu when contemplating the sacrifice 15...Nxf2)?  I'm sure GM's simply don't calculate certain lines that are not positionally well motivated.  On the other hand, at a critical juncture of the game (such as the position where Black decided to play 15...Nxf2) my impression is that everything comes down to concrete analysis - i.e., 17.Bxd4 was probably analyzed OTB and rejected for tactical reasons.  I suppose I'll just have to wonder.  Not having anything close to the positional judgement of a GM, I guess I may never know...

Thanks for shedding some light on the decision-making process, though.

After 17.Bxd4 cxd4 18.g4 (maybe a mistake here) 18...e5 (the disregarded line I gave above) I think White is probably lost.  Here’s a sample continuation:

19.Kg1 e4 20.Nxd4 (what else?) 20…Qe5 21.Nf1 Rxf1+ 22.Rxf1 Qh2+ 23.Kf2 e3+ 24.Kxe3 [24.Kf3?? Qg3#] 24...Qxg2 25.a3 Bc5+-

Find an improvement for White (maybe White's 18th) or you have no argument in favor of 17.Bxd4

After 16.Nd4 (diagrammed position above), Black has serious compensation for the sacrificed piece but according to the analysis given by Nisipeanu White is still OK (maybe even slightly better).

While I agree that you can argue that 17.Bxd4 deserved mention, I strongly suspect that analysis will show that this move is inferior to the ones mentioned (17.Nf1 and the game move 17.g4) — perhaps even losing.  I also think, BTW, that the position is just a tad too complicated to be assessed by some facile positional observation (e.g., mobile center, queenside passers, etc.).

To me the position after 17.Bxd4 cxd4 is clear as mud as well. Sure Black has full control of the centre, but if White can stabilize it and also protect the weaknesses on his kingside (Nf1), won't the passed pawns play a role?

Thanks for the diagram, Ostap!

This seems like a clear case of the central pawns being worth more than a piece. I don't think you need deep computer analysis to realise that the pawns are going to overrun White's position after White gets rid of his main dark-squared defender.

In the game Georgiev – Nisipeanu, Fuegen 2006, (a very pretty game), can someone pleas tell me why White didn't just play 17.Bxd4?  I don't see how Black's attack should continue after this move, and it isn't mentioned at all in the article found here:
http://www.chesscafe.com/informant/informant.htm

I feel like I must be missing something, because Black's play is roundly praised and 15...Nxf2 is given an '!'.  Perhaps something like 17.Bxd4 cd 18.g4 (maybe asking too much) Qf7 19.Kg1 e5 gives Black play for the piece, but I don't understand why 17.Bxd4 isn't even mentioned.