When I have a question about a GM move, I will look at the board and try to find out why. In this case, I have no idea other than he wants to push a pawn to a4 but he can't afford the b4 response. So Bc5 kicks the f8 rook with tempo, allowing a4 for free. But we see black pushes b4 anyway after using a simple tactic. I prefer Rfe1 because the pawn on e4 might need some help someday. It's a stock move for me.
So I asked Stockfish. For quite a while it suggested Bc5 too. Then it switched to Rfd1. The scores it assigned were within a few centipawns of Bc5. We see this move starts eyeing ownership of an open file and lines up on the black Queen. We can see simple tactics start happening. After (say) Qe2, then Bxb5 is possible because of the discovered attack on the Q.
Now that I know this, I can choose other moves and see the response. This will tell me what GM Efimenko was seeing.
(What I can see as I examine moves is that Black wants to play Be6 and we need to be ready for Qd7 or Qc8 threatening the pawn on h3. Be6 is such a nice developing move it is happening regardless of the variations Stockfish is flashing at me.)
My move (12. Rfe1) yields a 25 centipawn loss, and the next 4 moves for White are passive pawn moves and rook shuffles. I surmise that while not bad, this move cedes initiative.
How about 12. b4 to stop that b5 pawn? This move breaks the Queenside open, both bishops to slice into that area, and the rook on f1 can slide over too. But after 12... a6 Black doesn't have too many problems. Stockfish says this move is also loses about 25 centipawns. I think it gives White a more active game than 12. Rfe1 and if that's the sort of thing you like, it could be ok. One nice thing about this exercise is that I normally would not have considered b4 at all. And now I found a not-terrible move on my own.
For my last try, I see a lot of the variations White's knight on c3 gets kicked or is moved to e2 and black's Be6 move cannot be stopped. What if we accept that and move the Knight first? After 12. Ne2 Stockfish does indeed start a queenside attack, losing maybe 10 centipawns from the preferred move. Stockfish attacks the backward and weak c pawn with Qc3 which cannot be defended adequately. Black drops the pawn on a7 but White cannot hold onto it. Rolicking and dynamic.
[fen "r1bq1rk1/p4pbp/2p2np1/1p2p3/4P3/2NBB2P/PPPQ1PP1/R4RK1 w - - 0 1"]
1. Ne2 Be6 2.Qc3 Rc8 3.Bxa7 c5 4.Bxc5 Nd7 5.b4 Nxc5 6.bxc5 Qe7 7.Bxb5 Rxc5 8.Qb4 Rb8 9.a4 Qc7 10.Rfd1 Rxc2
Wow, that really is a lot of "engine" moves. Black has some monster rooks and queen for the pawn.
I probably would have been lame and defended the c6 pawn. For example, my b-player response might have been 1... Qc7 to defend a7 and c6... While not a bad move, it gives 10 centipawns away (+46 after my move) but avoids the razor-sharp engine moves (as if I could find them).
My next b-player move is the fugly Bb7. I would actually not play that because it turns a good bishop into a big pawn. But it does connect the rook and queen. Stockfish thinks it is better than my queen move. This move encourages White to weaken his Kingside with 'f3' to blunt the new pressure against the pawn on e4. In a way, this move refutes 12. Ne2. In retrospect, Ne2 does indeed remove a defender of e4... Who knew, lol!
