I don't really think that the fantasy variation is all that good for white. I've been successful when I've played against it. That said, if you want to play the Caro-Kann, you must know some of the theory of the Fantasy variation. Otherwise, you will get run over.

I'd personally suggest  3... e6, but if you're not comfortable with French-style positions, you might want to try something else. 3... e5 looks a little risky to me. White seems to be doing quite well after 4 dxe5 Bc5 5 Nc3 Qb6 6 Na4. I personally like 3... g6 much better than e7-e5, but that's just me.

Nietzsche wrote on 04/20/09 at 19:24:44:


3.f3 is not about proving an advantage; it's about dynamic imbalance.
After 3.Nc3 dxe4 4.Nxe4 and 3.exd5 cxd5 4.c4 Nf6 5.Nc3 Black also just has to pick one; White needs to know all of them and probably much deeper than the lines after 3.f3.
Nobody claims that 3.f3 is theoretically better than or even equal to 3.Nc3 or 3.exd5 cxd5 4.c4.

Stefan Buecker wrote on 04/18/09 at 17:53:08:


Which issue do you mean? I remember an article may be in year 2004 of J.Konikowski about the fantasy variation in the magazine 'Rochade'.

After 3...e5 i had analyzed the following variation:

4.dxe5 Bc5 5.Nc3 Qb6 (5...d4 6.Nce2 Qb6 7.Qd3 Nd7 8.f4) 6.Na4 Bf2+ (6...Qa5+ 7.c3) 7.Ke2 Qd4 8.Qxd4 Bxd4 9.f4 b5 (9...dxe4 10.c3) 10.Nc3 b4 11.Na4 dxe4 (11...Bg4+ 12.Kd3) 12.Le3 and White has an advantage. 
If White does not want to take a risk, he can play 4.Nc3 dxe4 5.dxe5 Qxd1+ 6.Nxd1 exf3 (6...Nd7 7.f4) 9.Nxf3.

I'm with several of the previous posters. As someone who has played this as white for most of my career, 3...e5 is the move I've always been worried about, and I have a pretty poor record against it. On the other hand, I have scored quite well against 3...e6, even with the dubious gambit 4.Be3 as well as the main moves.

The line is objectively fine for white, but saying that, it's objectively fine for black as well. Leads to a slightly different type of position to a normal C-K where both sides can play for the win.

The line 1.e4 c6 2.d4 d5 3.f3 used to be the "Classical Gambit" or "Tartakower Attack" (Tartakower himself mentioned the latter in his game collection: "some theoreticians call it ..."). As far as I know, Curt von Bardeleben invented 3.f3 and published a theoretical article in Deutsche Schachzeitung in the 1880s, but to name it after Tartakower, who played 3.f3 against strong opponents, is also an excellent choice. 

Now many authors prefer "Fantasy Variation", maybe because it sounds more impressive. But what was the origin of that name? In Understanding the Caro-Kann Defense (1980), Andy Soltis titled the section on 3.f3 "Fantasy Variation". Does anybody know of an earlier use of the name?

You can find an extensive coverage of that variation of the book Starting out the Caro-Can by Gallagher.  That line is a favorite of his and on that book he present it with very "spice-tricky" details that black have to play accurate to avoid a quick crash. 

By me, the best for black is to play e6 to french like position -equal .

I agree with Willempie, and I think that 3...e5 is the best answer.

It's a good sideline try for advantage. 3..e6 is Black''s best and leads to French like positions that may not suit a CK players. 3..e5,3. ..g6, 3...xe4 all give White comfortable play