Yeah, the idea was to discuss 3...Nc6!? in general and I just threw out a couple of lines. There are many transpositions to more known lines, but the early deviations could actually be beneficial for Black as long as White doesn't know how to respond. I think some players might be surprised by a move like 7 ...Bxc3!? because it goes against the first instinct that the bishop is something Black would cherish. However, the statistics, even if the numbers are pretty unimportant (small sample number is one reason), still point to this being a fairly ok line for Black. Maybe that is just because it is a slightly odd move order.

I think I would play a Botvinnik setup if faced with this OTB, but I feel there must be more critical lines that White could try, as your last example. An early e3 and d4 feels like that, but from what I can see in games already played, Black might be fine against this, provided he is ok with a Hedgehog with the knight on c6 (which of course is put there quite early...). I generally don't like that myself, but prefer to keep that issue flexible. In this case, with White already commited to a fianchetto, it isn't too bad I think. Food for thought.

Eric, yes, I have played an early e3 in this line when trying it out at Playchess. And yes, the idea is of course to go 8. d4. I also noted e.g. Ljubojevic-Betschart, 1988. Although not against a very strong player, still interesting.

I didn't have so much time to go through all alternatives yesterday and just wanted to post some ideas from what appears to be the most common moves.

The reason I ask about this, though, is that there aren't too many games with this move order. However, I have found 4 games by De Firmian that started with 1. c4 c5 2. g3 b6 3. Bg2 Nc6!?. And in total, there are 6 games between players rated over 2400, with 0 White wins, 4 draws and 2 Black wins. Bent Larsen has played it with Black and White once and drew both times.

I think it ought to be mentioned somewhere (maybe The Dynamic English, I haven't checked yet and just came to think of it). Anyway, Marin's repertoire is based on mainly the Botvinnik setup. Or in case of Black trying to reach the Hedgehog, he only gives ...b6, and then ...e6 before Black plays ...Nc6, giving White time to play for d4 without exchanging the light-squared bishops.

I can't find a thread on 1. c4 c5 2. g3 b6 3. Bg2 Nc6!? (but I have a vague memory of it being mentioned here at Chesspub) and this is not mentioned in any of Marin's books as far as I can tell. I have actually met this exact move order quite a few times and it is a bit annoying.

It might transpose into some of the main lines, but Black and White also have independent options of course. Like 1. c4 c5 2. g3 b6 3. Bg2 Nc6!? 4. Nf3 Bb7 5. 0-0 g6 6. Nc3 Bg7. This has transposed into Kosten-Bauer, 2004. White might go 7. e3 here, and after 7... Nf6, White is probably better after 8. d4! However, Black has 7 ...Bxc3!?, a move played a few times. After e.g. 8. bxc3 d6 9. d3 Qd7 we are now in a variant that has been played 12 times as far as I can tell, with nothing special for White.

Has anyone studied this more in detail? Any improvements for White? Maybe go into a Botvinnik set-up with Nc3, e4 and d3, followed by Nge2? Or just a Hedgehog type-position with Nf3 and d4 instead of e3?