castling into it

Nice mate.  I am not sure if there is a refutation to it earlier in the line as I haven't analysed it but it's aesthetically pleasing that's for sure! =]

the more I analyse it, the more i realise that without 15... 0-0, white has no concrete attack, even though he can stop castling rights.

Incredibly, after 13. Nd4!, white is substantially ahead. Black has no good way of dealing with the thread of Qf3 followed by Nd6+. Rybka recommends 13... Kd8 14. Nxf5 exf5 15. Rxf5 +1.50. If 13... Nxe3? 14. Qf3 Nxf1 15. Nd6+ Kd8 16. Nxf7+ Kc8  17. Nxh8 +2.30

However, you played 15. Rxf1?!, which is still winning, but not as good, since it allows a safe Kd8. The idea here for white is Qg3 followed by Qxg7. Black can't spend time defending that square because of the threat of Ng5. I dont' understand the nuances here, but Rybka thinks white is still well ahead here.

Of course 15... O-O?? was the critical blunder, as you identified. 16. Qh3 with the idea of Nf6+ was the right idea, but 16. Nf6+!! is a forced mate in 11. 16. Qh3 h6! 17. Nf6+ does not have nearly as much force, but it is still winning, as you can follow up with Ng4 with lots of threats. Rybka gives white a rook advantage in this line.

17. Rf3? gives black the chance to play Nc6, which he does. Although you dismiss 19... exf5, it is actually the correct move, as 19... exf5! 20. Qxf5 g6 21. Qh3 Kg7 22. Nh5+ gxh5 23. Rg3+ Qg5 24. Rxg5 hxg5 is OK for black. However, it is very difficult to see that, as it loses the queen. Black's response 19... Qc5?? is the final blunder, which loses the game. 20... exf5 hangs on a bit longer than 20... Qxe5?, but it won't stop white's crushing attack.

thank you for your help eebster! I missed 21...Kg7. A massive hole in my calculations, no doubt. What happens if 21...kg7 22. Nh5+ Kh8  ? I can't see a winning continuation  for white.

Hmm, you might be right. Looks like white's only move is Nf6+ and repetition ensues.

The line is probably still worth playing, though, as any other black responses lose, and the odds of him finding that draw are pretty low. However, Rybka's line (17. Nf6+! or even better 16. Nf6+!! with a mate in 11) would of course be better.

xevs, I pointed out 16. Nf6!! above, but it is actually a mate in 11, not 5. However, even after 16. Qc6, white still should win.

Eebster

Now I see the post. My bad :)

I don't have Rybka, however, I don't see any chance for Black after 16. Nf6 when at best the Queen has to be sacrificed to extend the game. I don't even see it taking 11 moves to finish the task... could you post that analysis for us in a game format?

Thanks

Heres a legendary example of castling 'into it' :

Yeah, sorry. I just didn't post it because it was a sideline. Here is some analysis I took from my computer of the position after 16. Nf6 (actually a mate in 10, as it turns out):

thanks alot man! I can't believed I missed the most flashy win in my calculations! nice spot.;