Positional assessment at mid-game: Black has cleared the center without loss of a single center pawn of his own. This suggests a slight modification of the earlier plan to attack on the king side. Black will now attack on two fronts, both up the center and on the king side, for it will be difficult for White to defend on two fronts without getting wrong-footed. The fact that Black has won a pawn is not very significant (yet), for he has no passer; it might become significant later, after the heavy pieces have been traded off.

Black now proceeds to restrict White's maneuvering space as much as possible by advancing in the center.

While White makes a counter demonstration in the center.

Denying the White B g4. The h-pawn is poison: if 19. Bxh5, then e5-e4 is deadly, as the B cannot move and cannot be defended after ... g6.

Trying to hold the center. c2-c4 is no longer recommended, as it would allow ... d5-d4, giving Black his passer.

Black solidifies the K-side in preparation for R action in the center, according to plan: control the center, then attack on the k-side.

White controls the c1 - h6 diagonal, where Black is notably weak, and also makes way for the QR to come to the center.

I expected Ra1-d1, to which I was prepared to answer ... e4, and then ... c5 preparing for ... d4.

This effectively reduces Black's B to a pawn and relinquishes control over d4. On the other hand, it severely cramps White's position, and after ... c5, ... d4, White will be hard pressed to blockade Black's pawn storm in the center.

This is much better than retreating the White B immediately, which would cut off the QR.