Figure 6.2.4.6[White to move]

Black’s castled position is intact, but observe that White has three pieces trained on f7, an underdefended square next to Black’s king. It ought to be possible for White to translate this coordinated firepower into at least some material gain; but how? The natural way to start is by opening a line, so White starts by removing the f-pawn with check: 1. Bxf7+. If nothing else, this begins to set up Greco’s mate by attacking g8. Black has two choices: take White’s bishop or move his king to h8.

(a) If he plays RxB, White takes advantage of the multiple pieces he has focused on the square by playing 2. QxR+; Black’s king is forced to h8, and now White has a simple back rank mate with Qe8 (Black can interpose his bishop at f8, but it’s futile).

(b) If Black replies to Bxf7 by playing his king to h8, half of Greco’s mate is in place. And White has a knight he can use to force open the h-file with the standard 2. Ng6+, but then how will he take advantage? With his queen, of course: after Black plays h7xN, White has 3. Qh3#. It's important to remember that despite being way over on b3, the queen is ready to participate in an attack down the h-file on a moment’s notice.

Notice that both variations are further examples of a queen alternating between diagonal and vertical roles, a useful possibility to master. This position also illustrates another point you want to learn by studying combinations and mating patterns: how to sense when your accumulation of coordinated force against a point has become lopsided enough that it is likely to yield material, even if you have not yet worked out the details. At the outset White has three pieces directed against the Black king's position, and Black doesn't have comparable defensive resources. This should cause you to search avidly for a sacrifice and combination.