When queens, for instance, have been exchanged on the board, the player who initiated the exchange may at once rescue his or her imprisoned queen and drop it. The other player can then similarly drop a queen, but may find this insufficient compensation. Dropping first, particularly with a queen, can easily win the game.
An often grave disadvantage of rescuing a man is that you do have to parachute the rescued man at once. In contrast, your opponent is now able to add a released prisoner to a store of paratroopers which may be growing alarmingly large. Experts will be wary of this danger, the result sometimes being that neither is willing to initiate an exchange of prisoners until one of them launches a big attack, rescuing and dropping several men in swift succession.
Much of the time, an imprisoned enemy should be considered nearly as threatening as an enemy already waiting to be parachuted. Remember, after an exchange of prisoners, the rescued prisoner is parachuted immediately. Think of the men in your prison not only as useful "cash" for "buying" your men back from prison at a time of your choice, but also as bombs liable to explode at a time chosen by your opponent.
The situation is not as in "Shogi", where capturing a knight, for instance, means you at once have a new knight to drop, a knight that has "changed sides". If you want a knight to drop, you must get your opponent to capture one of your own knights, after which you must rescue it. A man can sometimes be sacrificed simply so that it can then be rescued and dropped. You can often virtually force your opponent to capture a piece. Sometimes you can force it absolutely (as your opponent's sole way of escape from a check).
Correspondingly, there is a possibly grave danger in capturing anything: it gives your opponent the chance of rescuing and then dropping it. A man you capture might perhaps return to mate you immediately. You have to be careful even about capturing undefended pawns.
Often the best way of stopping your opponent parachuting somewhere is to parachute there yourself.
Knight drops can be extremely powerful, while bishop drops tend to be less useful. So capturing an enemy bishop at the cost of a knight of yours, a knight which you can then rescue and drop, is often a good idea, while capturing a knight at the cost of a bishop is often a bad one.
A rook may be little or no more useful than a knight as a droppable piece, or even as a piece on the board (since the game will never reach a stage comparable to the endgame in western chess, in which rooks become very powerful). But of course the sheer fact that the rook is "officially worth more" can in practice mean that it really is worth more to you as a prisoner. Knights cannot be used to rescue rooks, whereas rooks can be used to rescue knights.
An imprisoned queen may not be very useful to the owner of the prison. Yes, it can be used for rescuing absolutely any man, but only through giving the opponent a droppable queen. This is one reason why it may sometimes be good to sacrifice a queen to win a knight and a bishop.
Naturally, even a pawn can be very powerful if dropping it leads to a mate. Rescuing a pawn in exchange for releasing an enemy knight could often make sense. So, sometimes, could rescuing a knight in exchange for releasing an enemy queen.
Dropped pawns can be specially useful when placed so that they threaten to promote soon. Also for double attacks ("forks"), or for disordering a king's defenders. An exposed king may even have a pawn dropped in front of it just so that the king will be forced to capture the pawn, becoming still more exposed. Dropping two or more pawns in swift succession can be devastating.
In one game, White dropped a pawn onto the seventh rank, forking two rooks and threatening to promote by capturing one or other of them. Black replied by rescuing a man, paying for the rescue by releasing the only white piece which could be used for the promotion; the rook capture was then impossible.