Where does Black's LSB belong in the Accelerated Dragon?

In accelerated dragon games, Black usually develops his LSB in two ways:

Like this

Or like this

What is the idea behind the different bishop placement? Is it just a matter of taste? In my opinion, it makes more sense on c6 as it supports the b5 break.

Be6 is based on the idea of a6-b5 sometimes prepared by Qa5 and Rfc8. In some cases b5 can be played even at the cost of a pawn when it gives the right kind of Benkö-style position. Sometimes White tries to use the position of the bishop by playing f4-f5 but he's not forced to; more often he goes for the standard centralization of major pieces followed by a timely Nd5.

Bd7-c6 is a complete different idea based on the domination of dark squares. Black provokes the weakening f3 (Bd3 is also possible but creates a slight piece disharmony), then plays a5 to avoid b4-b5 plans by White embarrassing the bishop, and later Nd7-c5 and Qb6. Of course White tries to avoid the trade of dark-squared bishops, unlike in a regular dragon.

There are some move-orders designed to take on d4 even before Be3 appears, like 1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 g6 5. c4 Nf6 6. Nc3 Nxd4 7. Qxd4 d6 followed by Bg7 and 0-0 and later the customary developing scheme Be6-Qa5-Rfc8-a6-b5.

I'll pretty much duplicate Nerwal's excellent explanation. Though I'm not up on latest computerized theory, in general I also found that it looks like the ...Bd7 then ...Bc6 plan and ...Be6 plan  both cover the important d5 spot in the Maroczy Bind which often allows the e-pawn to sit back for awhile supporting d6.

Then...Bd7...Bc6 also pressures the e4 pawn which might be awkard to defend after...Nd7 ...Nc5 so it might force White f3 which could hinder any Center/kside f4 push. But it allows White pretty good leeway on Qside especially in trying to control the c-file which often wasn't very good for Black.

On the other hand, ...Be6 plays the other way. Pressuring c4 looking for b3 which usually allows a Black thematic Qside build on the c-file and ...Qa5. But maybe White's sharpest is f4 at some point with Center/kside initiative and with Black backing off ...Bd7 though I've seen some where White diffuses Black Qside counter with no castle, Qd2, c-file control with exchanges ending up with 2 B's with maybe better King plus.

For an extreme ...Bd7...Bc6 plan we'd sometimes try something like, a probably dubious line in today's world: 1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 g6 5. c4 Bg7 6. Be3 Nf6 7. Nc3 Ng4 8. Qxg4 Nxd4 9. Qd1 Ne6 10. Qd2 d6 11. Be2 Bd7 12. O-O O-O 13. Rac1 Bc6 14. Rfd1 Nc5...

@Nerwal I'll quote Nckchrls: "excellent explanation!" Thanks a lot. I'd be interested to know how you have such knowledge: is it in a book you read or did you arrive to it by analysing games?

@Nckchrls thanks a lot! Why is the line you mentioned dubios? Is it because White can quickly post a knight on d5 or maybe play for a quick b4?

Regarding the plan with Be6 there is some material about it in Shereshevsky's "Mastering the endgame" volume II.

Bd7-c6 I played this as Black several times and there are hundreds of master games in this line to look for. I remember an old chessbase video by Felgaer where he explained his game against Judit Polgar which ended in a draw. More recently there is this interesting video : https://www.youtube.com/watch?v=DtSgIMZ8dmc ; the subsequent plan with e6 and Be5 looks dumb at first but actually it's not easy for white to handle it and for once Black gets active play in the Maroczy Bind.

Thanks a lot! I think there was a similar game (with the plan ...e6 and ...Be5) in "Chess Structures: a GM guide". I'd recommend that book to anyone, as I think it's worth its weight in gold.

Edit: it was Pavlovic-Hoyos USA 2012

Pfren, could you provide a link to those games or suggest where I could find them? I'd be grateful.

With an ...Ng4, Qxg4 idea, thanks for the ...b6 and ...Bb7 idea. Probably is an improvement.

The problems in the ...Bd7 ...Bc6 line usually came as ...Ne6, ...d6 , ...Bd7, ...Bc6 often took too long once somebody was prepared. The position could end up cluttered without really getting the pressure desired anyway. Later analysis usually showed, Black typically had to throw in ...a5 to hinder White's b4 which added another move.

Even then though it seemed holdable but awkward. I'm guessing a computer might easily punish the tempo costs nowdays, however.

I'll definitely check out the ...b6 ...Bb7 take. Could make the whole idea more comfortable again.