OK, you force me to take out the "big gun"! The last variation i showed was the one that made me stay away from the Marshall Gambit when i studied it from Brunello's and Vigorito's books. The one i'll present you now is what i believe to be a serious try for the advantage for White. Gustafsson misses a move order trick, so don't go to him for help. 



I know, i know....White is better but does he have realistic winning chances? I think yes, he does. 

Shall we go now to another variation? (which i may be able to spell it correctly this time?) 

PS
@Tony, Stefan: Thank you for saving my time by showing a good way against this Steiner line!

That's right. The whole question is how to win as White,which appears to be difficult with expert defense.

OK, i have a problem with the MG. I was corrected by Markovic not long ago again! 

For starters, look at this line that i found while reading Vigorito's great "Understanding the Marschall Attack". I think that this is the traditional Main Line, right? I'd never like to be in Black's shoes. Where are his winning chances? Objectively it must be a draw, but i assure you it is not an easy one. Some time ago i played lots of training games in this line with Black and i drew about 65-60% and lost 35-40% of those games. We can look at Black's modern alternatives if you have nothing to add here.



Also 20.Qf1 seems promising to me. I really cannot decide which move is more unpleasant for Black

As an American, I grant Ametanoitos special dispensation to spell it anyway he wants.

Let's do it. 

As far as refuting the Steiner goes, I propose we base it off of Nakamura's convincing win over Short in London 2010, utilizing the under-promoted 11.g3. The only database I have access to at work has White scoring 100% after 11.g3 in a small number of games. The Naka game is below:



Black can also try 13...Qd6, though I'd be confident handling White after 14.Qd3 Qxc6 15.Nd2, when Black can't avoid losing the f3-pawn, with what looks to be a much improved version of normal Marshall attacks to me - the structure is the same, but Black has way less pressure. My computer thinks 15...Bf8 16.Bxf6 Qxf6 17.Bd5 Rad8 18.Bxf3 Rxe1+ 19.Rxe1 Be6 20.Qe3 is best play, but this looks strictly better for White to me, somewhere between += and ± to me.

I had written for ChessCafe on the Steiner Variation of the Marshall Attack, where Black sacrifices a second pawn and did well. I am not aware that somebody had refuted it. These two pdf documents are named kaiss11.pdf and kaiss12.pdf and can easily be found on the web. It would be interesting for me if somebody could find a plus for White.

Bibs wrote on 07/27/12 at 00:26:20:

My main point was that there is nobody here to do the work on the Spanish main lines. It is less surprising that people help to study 3...f5 or 3...Bc5, perhaps they play it with Black. But most amateurs (and their opponents!) avoid the main lines. As I said, my personal focus is on 4...Nf6 5.Bxc6 which gives White a plus. I don't think that other moves are better. But if you, Bibs, step forward to do the heavy-lifting, it is of course highly welcome, and I won't get in the way.