It isn't a trick. It's why pharmaceutical trials have a placebo control, not a nothing-control. There's been studies done on the effect of different placebos:

It isn't a trick. It's why pharmaceutical trials have a placebo control, not a nothing-control. There's been studies done on the effect of different placebos: bigger pills, with bigger markings, administered by professionals in professional settings etc work better. My pet theory is that, since almost all medicine merely assists the body's own healing (e.g. a plaster cast doesn't heal the bone), the belief of safety and being looked after is enough for your body to switch resources away from dealing with a threat towards healing. (You can't afford to heal while your attacker is still around.) A placebo signals safety.