Kramnik, commenting on his game against Nakamura, said that his revelation (after studying the position most of the night) was that White doesn't have to take on e4, and thereby got a great game.

Would you mind providing the moves before 7...Ne4?

Thanks!

We have discussed this option in the past and Black is far from equal here.

Dean wrote on 03/06/10 at 12:11:43:


No. A reasonable definition of "=" is that each side has equal chances, objectively based on existing positional and tactical elements.

If Joe-Patzer can score 50% against Jane-Patzer who gives a flying cupcake?

Quote:


A line having more nuances isn't what determines equality, or advantages.

Quote:


Much less frequent so long as the computer is used for its intended purpose - blunder checking, and offering the occasionally interesting continuation. Especially when a GM is writing about their area of expertise.

Quote:


First a basic point is that "a patzer" in our context means nothing else than someone with significantly lower ELO than a GM. This patzer could have very low or very high skill in chess analysis with computer. The patzer may be dull or having a very high IQ, having no chess talent or having as much chess talent as any GM.

Therefore a patzer(or anyone else) who has developed high skill in centaur analysis will of course not just go with Rybka's simple evaluation.

I think the only reasonable definition of  "=" is that you score 50% against an equally strong opponent.

While this is based on knowledge of the particular position the evaluation will always be partly subjective.

But the critical question is if not white needs to know as many variations after 13...Bd7!
For example, does black or white need to know tricks as 14. Nd2 Nxd4! 15. Bxd4 Ba4!

Quote:


Well... hope that is not how they raise children from where you come from...?

At least Rybka's middlegame evaluation is probably more accurate than you think. It has been calibrated by extensive tests of what scores well in a large set of actual games.

Of course centaur analysis is a team work where you learn more and more about both the weaknesses/strengths of the computer and yourself.

Quote:


Compare with following sentence:

If you've ever developed a very thorough repertoire then you know how often humans/IM:s/GM:s/theory books are flatly wrong in their evaluations, and worse yet fail to often grasp the right plans.

Dean wrote on 03/05/10 at 23:20:16:



If you really believe this, then the only reason I can see for buying any opening book is to see what lines patzers and GMs who don't have your computer acumen will play.

Dean wrote on 03/04/10 at 21:21:41:



Let me start with the highlighted portion of the quote:  Avrukh has shown in his first volume that he does use strong chess engines in his analysis.  I fully expect that he did the same again here, though I haven't bought my copy yet to be sure.

Avrukh can be accused of many things, but "lazy-following" another author doesn't seem to be one of them.  From what little I've seen, I agree completely with BPaulsen's sentiment that "equality" is based on silicon analysis rather than Avrukh's judgement.   

Until proven otherwise, I will trust Avrukh's assessment.

BPaulsen wrote on 03/05/10 at 02:31:40:


Wrong. Rybka claims += after 14.b3, after which I concluded that a serious evaluation of this position is impossible without a closer look, something I don't have time for now.