Click me if you can

Now you know how to trigger simple game actions such as rotating a cube. Since the cube is the only object in the scene, the user does not have any choice which spatial to interact with. But as soon as you have several objects in a scene, you need to find out what the user was looking at when he or she pressed an action key.

There are several possible targets of game actions: the target of navigational actions is typically the player character or vehicle—no explicit target selection is necessary. The target of a take action is, in contrast, one of many items lying around. The target of an attack can be a subset of enemy characters, while the target of a magic spell can be pretty much anything, depending on the spell: the player, an ally, an enemy, an item, or even the floor.

These seemingly different game actions are very similar from an implementation point of view: the player selects a target in the 3D scene, and then triggers an action on it. We call the process of identifying the target picking.

If the hot spot for clicks is the center of the screen, some games mark it with crosshairs. Alternatively, a game can offer a visible mouse cursor. In either case, you need to determine what 3D object the player was aiming at.