This is Black's standard approach to the Modern Variation. An interesting sideline is 5....Qe7?! after which White does best to play 6.O-O Nxe5?! (otherwise the Knight will have to retreat to g8) 7.Nxe5 Qxe5 8.Re1 Ne4 when Tim Harding recommends 9.f3! d5 10.Nd2! f5 (10....Bd6 11.Nxe4 Qxh2+ 12.Kf2 dxe4 13.Rxe4+) 11.Bxd5! Bd6 12.Nxe4 Qxh2+ 13.Kf1 and Black's King is in greater danger than White's. Black might also try 5....Ng4 6.O-O d6 7.exd6 Bxd6 (7....Qxd6 8.h3! Nf6 9.Ng5 Nd8 10.Re1+ Be7 11.Qe2 Bf5 12.Bf4 Qd7 13.Nc3! Estrin) 8.Re1+ Kf8 (8....Be7 9.Bg5 O-O 10.Bxe7 Nxe7 11.Qxd4 +=) 9.Na3 and White has the better of it despite his temporary pawn deficit. And on 5....Ne5 White can play the interesting gambit 6.Bd5!? Nc5 (or 6....Bb4+ 7.c3!? dxc3 8.O-O!) 7.c3! dxc3 8.Nxc3 Be7 9.Be3 O-O 10.Qe2 followed by O-O-O with a game that any Urusov fan should love.
Torre writes that this move is "intentionally trying to provoke White into an attack on Black's King, because in this case neither player's interest was served by a draw." Perhaps Black could more safely play 14....f5 to prevent White's pawns from advancing together, but White might then try 15.g4!?
Torre refuses to delay his attack or allow Black to play ....Bg5 to reduce forces. The exchange sacrifice must be accepted or the Rook will move powerfully to g3.
Torre notes that 20....d3!? would have been an interesting idea here and would certainly have distracted White from his attack. An attack on the wing is always best countered in the center....
Torre notes that 22....Qd5 is refuted by 23.Bf6!
Torre writes "This doubtful move nullifies all of White's previous effort. Correct was 31.Qg4+! Kd5 32.Qf3+ Re4 33.c4+ dxc3 34.Rd1+ Kc6 35.Be5 Qg4 36.Rd6+ winning."
Torre writes, "An error. Indicated was 34....Qxc7 35.Rxc5+ Kxc5 36.Qxc7+ Kb4 and Black obtains at least a draw." Now the game comes to a sudden conclusion.
On 37....Kb4 38.Qa5 mate.