Neric wrote on 11/25/10 at 14:41:47:
His proof probably consists of some games where a 2600-player beat some 2400-player, because Jobava vs. NN is pretty much the most recent reference.
Can you tell me which of the following you class as Jobava vs 'NN' - or is it all of them?:
Jobava,Ba[2556]-Kovacevic,A[2553] Leon 2001 (1-0)
Jobava,Ba[2614]-Alekseev,E[2634] Moscow 2006 (1-0)
Jobaba,Ba[2669]-Pogorelov,R[2451], Budva 2009 (1-0)
Jobaba,Ba[2669]-Inarkiev,E[2656], Budva 2009 (1/2-1/2)
Jobava,Ba[2695]-Timofeev,A[2655], Rijeka 2010 (1/2-1/2)
I am curious about how an unbeaten 4/5 vs 2590 rating average is dismissed. Or did you miss some of these games due to the move order? I may have missed more. Frankly if the line is capable of that, I wouldn't be so quick to claim it's totally equal - especially since Jobava only duffed up someone 200 points below him once.
I'd like to note that Alekseev won the Russian Championship later on in 2006. You can't call him 'NN', at least.
BTW, IM John Watson - generally considered to know a thing or two about the English - is quite positive about white's chances in Mastering The Chess Openings Volume 3, and basically implies that theory is in white's favour using this Bg5/Qc1 stuff. Marin isn't a lone voice in the wilderness. Watson bases his analysis on Jobava-Alekseev, and having flicked through Marin's work I know he was looking at Jobava's games.
I'd also be obliged if you could explain exactly how black transposes to Bf4/Bd2 by avoiding ...h6. Watson's explanation as to why black flicks it in is that after 7...0-0 8 0-0 d5 9 Rb1 will probably end up transposing after a later ...h6, and white has extra options like 9 Qc1 and 9 Qa4 so black avoids these.
Finally, Marin actually thinks 6 d4!? is quite promising for white but he didn't think it fitted well with the repertoire.
Oh and about specific analysis...
Neric wrote on 11/25/10 at 14:41:47:
Let's take a look at the line: 1.Nf3 d5 2.g3 Nf6 3.Bg2 c6 4.c4 dxc4 5.0-0 b5 6.a4 Bb7 7.b3 cxb3 8.Qxb3 Nbd7 9.Ba3 a6 10.Rc1 Qb6. Here Marin gives analysis by someone else (obviously he doesn't want to take responsibility) that goes 11.Ng5 e6 12.Bxf8 Rxf8 13.Nc3 with "compensation". Amazingly the totally obvious 13...h6 14.Nge4 Nxe4 15.Nxe4 c5 equalizes on the spot, so much about that...
First, "with compensation" actually doesn't mean "with an advantage" - equality is not contradictory. The with compensations sign even has a "=" in it. Ever wonder why?
In the meantime, I'll thank Stockfish for the analysis, since it is the only engine I have claiming this "instant equality" - tossing off that "obvious", "instantly equalizing" line seems to be the provence of computers. After 15...c5, it seems to think 16 Qb2 f5 17 Nxc5 Nxc5 18 Bxb7 Nxb7 19 Qxg7 Rf7 20 Qh8+ Rf8 21 Qg7 Rf7 22 Qh8+ is best play and hence the 0.00 evaluation.
A couple of my other engines jump on the chance to play 17 a5! Qc7 18 Nxc5 Nxc5 19 Bxb7 Qxb7 20 Rxc5 Kf7
- which, whatever it may be, is not "instant equality". Black has some long term chances thanks to the protected passed pawn on b5, which is about his only asset, but white has a strong iniative and black's king is having to walk. I'm not sure white is better, and I can't find anything absolutely clearcut, but black has to be extremely careful. Certainly the engines don't seem to agree on the evaluation here.
As for the part I've bolded above - it's fairly common in academic circles and books to name someone whose analysis you've taken. Otherwise, you know, you get accused of theft, or not given credit. Your interpretation only makes sense if you are desperately trying to make someone look bad, whereas it actually shows Marin to be conscientious.