>>12508951While superposition can never be directly observed, in order to explain observed results in double slit experiment for example we must think of the photons of having been in a superposition before observation. They arrive at the screen as discrete points, but in order to explain the resulting interference pattern they must be thought as having gone through the slits as waves. If it was just a normal 50/50 chance, where we just don't know through which slit the photon went each time, the photons would always arrive at the screen directly behind either slit - which does in fact happen *if* they are detected at the slits, which forces them to take a definite location.
