This looks a bit like a mirror image of the previous position: the pin Black imposes on White’s knight is structurally the same. Can White use the same idea to escape? Perhaps Bxf7+, requiring Kxf7 (if the pawn’s loss is to be avenged), allows White to then play Ng5+ and break the pin. But not quite. It worked last time because the knight was safe on the square from which it gave check, and because the pinning bishop was loose. Here the conditions are different. After 1. Bxf7+, KxB; 2. Ng5+, Black is not required to retreat his king and give White a free move. He can capture the checking piece with his queen, QxN. Now that charming discovered attack from last time, culminating with the queen taking the bishop that used to impose a pin, no longer looks so great for White, because Black’s bishop suddenly has protection from the queen on g5. True, White can play BxQ (the knight was not loose when it gave the check), but then Black can also play BxQ. After White takes the bishop with RxB, he still hasn't made up for the bishop he sacrificed at the beginning to smoke out Black’s king. The point is the importance of not playing patterns like the previous one as mechanical gimmicks. Treat them as ideas, and carefully study each side’s best replies. (White’s best move is simply Qb3.)