Figure 4.5.10.6[Black to move]

You should have no trouble seeing the concept for Black here: White’s queen and rook are aligned on the e-file, and the queen is the rook’s only guard. This should be enough to cause you to play with a skewer; if Black could attack with his rook on the e-file—and he already has a rook there, ready to attack if his bishop moves out of the way—White would have to move his queen out of the way and allow White to play RxR. But would it turn a profit? When the middle piece is a queen rather than a king, of course there are many more ways the middle piece can leave its square while still going to the defense of the piece behind it. This isn’t necessarily a fatal problem. Its seriousness depends on the value of the skewering piece and its target. But here Black would be trying to use a rook to win a rook, and this only will be productive if the targeted rook is loose.

So Black has to worry about moves by White’s queen that will take it to safety and yet protect the rook. He studies the White queen’s lines and sees one way White might accomplish this: Qb1. Can Black do anything to prevent that move? Remember that Black’s skewer has the structure of a discovered attack, meaning that as he unveils it he has a bishop move to make. Perhaps he can use it to attack b1 and so make the square inaccessible to White’s queen. Indeed he can: he plays Bxa2, and now White’s queen has no squares where it both is safe and protects the rook. Black wins the rook next move—unless White prefers to sacrifice his queen and use his rook to recapture. (Black also has the option of Bf5, likewise attacking the b1 square; this time White’s response is QxR, again sacrificing his queen for a rook.)