The structure is the same save for the d-pawn, which can stay on -d2, or advance to -d3 or -d4 depending on the situation. The pawn on -b3 usually advances to -b4 in conjunction with pawns on -c4 and -a4.

Tempi are not that important in these positions. Also, White's flexibility with his d-pawn can be useful.

tony37 wrote on 01/07/13 at 14:13:38:


And yet it works quite well for white so long as white knows how to play the structures from the middlegame. I've personally been playing this line for a few years now, and I find it both fun and easy to play from the white side. Whenever I'm playing blitz on playchess and black opts for Nf6 and g6, I win almost every time playing Bxf6. In fact, going back through my last 25 games in this line, I won 23 and lost 2. 

Anyway, the formation I use is with the g3 line as mentioned by the OP. I've found it a very comfortable opening, although one of my losses was to a cheater. He played this strange idea of f5 and then Ba3, which I had never had any other human do before. I got absolutely crushed off the board, and when I plugged the game into Houdini afterwards, every move of his matched the computers choice, including the strange Ba3 move.

Uberdecker wrote on 12/30/12 at 16:26:59:


I would imagine this is the reference game. The problem with the early c5 is that it gives away the d5 square for the rest of the game, as Karpov demonstrates. It must be well enough known that only a handful of players have repeated Browne's move order over the last forty years. The popularity of Hedgehog set ups is such that most players prefer 3. .. e6 or the 3. .. Nc6. In the latter case, transpositions to Karpov-Browne can remain possible if Black later tries .. g6 .

I notice that ECO has treated 3. e4 as the main line; in one database I looked at (NIC online) it appeared slightly more popular than 3. Bxf6.

I was referring to this:  1. c4 c5 2. b3 Nf6 3. Bb2 g6 4. Bxf6, which I think is generally recognized as slightly better for White.

Uberdecker wrote on 12/30/12 at 15:29:50:


I'm not aware that 3. Bxf6 has been considered the main line in the Larsen case; I wonder if you might be thinking of the sort of thing where Black has played ...c5, like in a game Karpov-Browne.

As mentioned at http://www.chesspublishing.com/content/12/index.htm, GM Jobova has revived interest in 1 b3.

Amongst the games is the sequence 1. b3 Nf6 2. Bb2 g6 3. Bxf6 gxf6 4. c4 Bg7 5. Nc3 f5 6. Rc1 

Checking my personal database, I found that I had played this idea myself in the 1970s. Jobova continued with g3 which I imagine is better than the approach I would use with e3, Nf3 and Be2.

Larsen and Fischer had played 1. b3 in 1970 and later. The idea of taking on f6 is obvious enough, even if at first you only regard it as a blitz weapon.