Black at present does not attack anything, but he sees the alignment of White’s bishop and queen and sees as well that the queen is the bishop’s only guard. Exploration of a skewer is indicated with 1. …Bd5, where Black’s bishop takes protection from its rook. The problem is that White can just move his queen over to g4, out of harm’s way and yet still able to guard the bishop on f3. But now rethink the challenge as it then would appear: we still have White’s bishop under attack, so we no longer are looking at a skewer so much as we are looking for a way to remove the bishop’s guard—the queen—so that BxB will become profitable for Black. Be persistent in looking for resources you can use to attack the queen again; remember your pawns; play 2. …h7-h5. The queen must move, and this time it will have to venture to g5, out of reach of the bishop. Black takes it next move.

Instead of h7-h5, Black could have chased off the queen with f7-f5. Why not that move instead? Because you must be wary of pawn moves that open lines to your king; moves of the f-pawns off the seventh rank are frequent examples. When you consider so moving a pawn, imagine ways your opponent could put a bishop or queen on that diagonal and give check. Here White would be able to reply BxB+; after Black spends a move on RxB (the priority of check—that’s the point), White would move his queen to safety, and the two sides would have done nothing but exchange bishops.

Consider another variation. Suppose White replies to the pawn push h7-h5 by moving his queen to f4. Black plays 3. …QxQ; 4. e3xQ, BxB and wins a piece. But now imagine White’s rook starting on c3 instead of c1. Would it change your analysis? It should; for then after 4. e3xQ, the bishop on f3 no longer is loose: a line has been cleared, causing the bishop to become guarded by White’s rook. This variation would have saved the day for White and prevented the sequence from succeeding for Black. It's worth playing this variation through in your mind’s eye until it becomes clear, as the key point—the movement of the e3 pawn, and its unexpected consequences—is a bit subtle.

All right; if both sides see the original diagrammed position correctly, what really happens? Black starts with 1. …Bd5, and White replies right away with 2. QxB, RxQ; 3. BxR. In this way White gets back a rook and bishop for his queen, a better deal than losing a piece. But don’t stop there; ask what happens next. White’s bishop would be left loose on d5, so Black looks right away for forks with his queen. 3. …Qe5 attacks the bishop and the loose pawn on b2 at the same time; after White defends with Rd1, Black picks up the pawn to make the outing more worthwhile.

This final result illustrates the role that tactics play in good chess: the really nifty sequence you see often doesn’t get played out, because once it begins your opponent sees the rest of it coming and makes a lesser sacrifice to avoid worse trouble. Thus the most beautiful combinations may never get executed—but seeing them still is the key to winning, because the threats they create force sacrifices that still result in decisive gains.