Slightly less accurate is 29.e4, since this move limits the manoeuvrability of his own knight.

Black resigned.

Black’s pieces are harmoniously placed, but he has forgotten to develop his kingside, and his king has not castled. If one could just ‘correct’ the position by putting the bishop on e7, Black would have no problems, but it is White to move.

A blow on the most strongly-defended square in Black’s camp, which he is covering with five pieces, wins the game for White.
At the cost of a pawn, White opens lines, exploiting his advantage in development.

After 17...Qd7 18.Re1 Kf8 19.Qh5, the knight comes to f5 with decisive effect.

Black resigned.

In such structures, a great role is played by the pawn on a5, which holds back two black pawns, and stops him playing the strengthening move ...b7-b6. And with the c5-pawn unsupported, White played...

In whatever form the pawns on b4 and c5 are exchanged for each other (and ‘exchanged’ is the operative word, because if Black captures on b4, his pawn will not survive there long), Black is guaranteed a backward pawn on the b-file. On the other hand, White’s pawn pair will represent a real threat, because if the b7-pawn falls, they will be connected passed pawns. It is important too that the Bg6 is just a spectator to what is going on. A knight on d7 would be a much more useful piece, taking part in the blockade of the white pawns.

The b1-square is attacked, so White uses a different square.

The connected pawns are impossible to block, so White is in no hurry, and makes a useful prophylactic move first.