Figure 5.2.5.8[Black to move]

Black has a possible capture in QxR but it won't work now because of the protection provided by White’s other rook on c1. Black considers RxB+, asking whether it would either draw the c1 rook away from guard duty or win the bishop for free. Sure enough, it requires White to play RxR, after which Black plays QxR—recovering his rook and netting a piece.

By now that much is easy. Yet there is a still stronger sequence; you find it by playing with the alternative move orders available to achieve your aims. This time start with QxRc3. Now White plays RxQ; Black replies RxBd1+; and since Black’s rook cannot be captured White must interpose his queen on f1. Black plays RxQ, winning back his queen. After White recaptures KxR, what is the net result? Black’s gain is the same: he has won a queen, a rook, and a bishop for a queen and a rook. The difference is that here Black has traded queens as well as winning a piece—a favorable adjustment, since the advantage of an extra piece becomes more meaningful as the number of pieces on the board decreases. That is why the usual goal once you obtain a material advantage is to trade away pieces, simplifying the position to make your advantage loom larger. Better still if you can simplify pieces away during the course of your tactical sequence.

The larger point is to remember that there are two ways to proceed if you find an overworked enemy piece. You can, at least in principle, go after either of the targets it protects (whether pieces or squares). Sometimes the right move order is obvious; sometimes it isn’t. Here it might seem obvious since the most familiar move order does win a piece, but there is an advantage to taking a second look at the counterintuitive second sequence that leads with a more valuable piece—the queen. Notice, by the way, that if White’s king had a flight square (e.g., if his h2 pawn were on h3), the second sequence wouldn’t work: White could reply to the rook check at d1 by moving his king rather than by interposing his queen. But the first sequence still would be productive.