Caro-Kann fantasy variation novelty

I have found what I believe to be a novelty in the Caro-Kann fantasy variation.

Could someone who plays the fantasy variation give an opinion please?

This is a recent game I played in this line. However, after Nc3, I chickened out with e6.

Hard to tell what to make of your variation.

The databases show a number of fairly recent matchups using your line up through the knight's return to g8.

I feel like it's tough to give a fair evaluation here, because the variation shown relies pretty heavily on trading off the light squared bishops, and giving yourself that nice f5, light-squared post for the knight, once the opposing bishop isn't around to harass him.

But what is white's bishop doing on d3 to begin with?  He doesn't seem to be serving any purpose there, just development for development's sake.  It's a lot more common in the fantasy variation to delay development of white's KB until its strategic relevance becomes clear.

Standard attacking development, a la Be3, Qd2, etc., and you're a tempo or two down and looking down the barrel of a loaded gun.

Why would black play ...Nf6 if intending to return the knight to g8 in case of e5?

The variation I play apparently has a great score for white in the DB...
http://www.chess.com/explorer/index.html?id=5896494&ply=17&black=0

1.Forcing e5 at the cost of a tempo is fine because the position is closed, plus f3 hardly counts as a constructive tempo.

2. The knight is actually best placed on g8, because it is a flexible transition to h6, f5, e7, or c6.

3. "Standard attacking development" is pretty well met:

Thanks for the thoughtful analysis! I'll try to come up with an improvement for black in the next couple of days. Perhaps Qb6 is wrong, seeing as it doesn't threaten anything and sets up a fork on a4. Hmmm....

White didn't play the obligatory prophylactic Kb1 in the given variation.  After which, the black attack amounts to little.  Sorry that where I am now is a little bandwidth-sparse, so I'm not going to plod through a graphic, but after Kb1...Nd7, Ne2...c4, Ng3 drives back the bishop while freeing the white's same, and Bb5, after which black's king is stranded in the middle for the remainder of the game, white is fully developed, and half of black's army lies motionless on the kingside.  This is at least +/=.

The bigger issue, I think, is that white is by no means under compulsion to allow so much counterplay on the Q-side.  After Be3...h4, white comes c4 and black's counterplay is squashed, while at the same time, Bf5 ideas leave b7 hopelessly weak to a suddenly mobile white queen.  I'm just fiddling around on my brain board, but I don't see an easy path to equality for black here.

The move order last time was innaccurate, due to c4, which, although only +/= after dxc4, is troublesome. A better move order:

My only opinion is that playing your novelty, e6 has to be defended. Else, white could cramp all of black's position together and with Qh5 check start an attack, all from one pawn sac

The knight is only well placed on g8 after e4-e5. Before that black should just develop normally. Anyway black will be willing to grab the e-pawn, especially in the variation (that I think is much more promising for white than e5) Nc3 Qa5!?. Next, you have the misconception that in the fantasy white's only aim if black plays solidly (I'll use the e6 line as an example) is e5 and f4. This is just a silly plan. White intends to develop normally with Nc3 and Be3, not waste time turning the game into a french that is good for black.

According to Gallagher's Starting out: the Caro-Kann, Gallagher (a fan of the fantasy) notes the main line after 3...e6:

4.Nc3 Bb4 5. Bf4 Nf6 Qd3 with the comment:

"Black was trying to tempt white into playing e5, but this advance doesn't fit in with f2-f3..."

Also, f4 doesn't make sense even after 3...Nf6:

Lastly, f3 isn't a gambit. It is sharp and aims for open positions like a gambit, and their are gambits you can play within it, but I feel like dxe4 fxe4 just rids white of the problems of f3. Plus, you can't even grab the e-pawn with anything. What you might be suggesting is that white just gambit the e-pawn after Nf6, which is more sensible. For example:

I'm not so sure.  The problem is that 21...Bxb2 loses instantly to 22. Qd6++

"Black's attack is extremely strong in this variation, and his king is completely safe in the center."

I'm not so sure.  The problem is that 21...Bxb2 loses instantly to 22. Qd6++

Here's the thing about 4...Ng8. Black has lost two tempi. White has already used one to make the move f3, which isn't that great - though one variation above showed that it can be useful. But white has the other tempo in-hand, in that it's white's turn.

Usually the move order is 1.e4 c6 2.d4 d5 3.e5, and it's black's turn. So from that position, white "has" to play f3, but then gets to move again. So going into the whole discussion, we should be extremely wary of this idea for black.

The two setups that I always think about for white are the Nc3/g4 setup and the Nf3/Be2/o-o setup, championed by Nigel Short. Since white has already played f3, it makes a whole lot of sense to play the Nc3/g4 line. Note that Bd3 is maybe not best, since it sort of looks like black can play as the OP showed, to force Bf5. But in case of 5.Nc3, black still must prepare Bf5 via h5, because Bf5 right away runs into g4. The next questions for me are whether white should play Bd3 or Be2, and also after Bd3 Nh6 Nge2 Bf5, how good is Bxh6 for white?

I just don't trust this line.

Against people who know some theory youll be faced with this instead of that quite inferior line you gave for black: