How can 'nothing' make a difference?

A placebo is a sham, or fake, “medication” or “treatment” that, although inert, is designed to appear identical to an active actual or experimental treatment. Placebos can take the form of a “dummy” pill made of an inert substance like sugar or starch; a water or saline solution; or a wearable device like a patch, catheter or magnetic bracelet.
In any clinical trial, the question scientists must answer is not, “Does the treatment work?” but rather, “Does the treatment work better than a placebo?”
In order to be approved for use in humans by the U.S. Food and Drug Administration, clinical testing must prove that patient response to an experimental therapy is quantitatively or qualitatively better than response to a placebo.
Or, more specifically, the experimental treatment must outperform the placebo effect.
This phrase refers to any measurable or quantifiable response to an inert or fake treatment.
Clinical trial participants often respond to placebos in ways that mimic those taking the active treatment. Some receiving placebos even report expected (often unwanted) side effects.
A number of studies suggest that two primary factors contribute to the placebo effect: conscious factors such as expectations, and subconscious factors based on prior experiences.
A conscious expectation that a medication or treatment will work can prompt an individual to attribute any perceived symptom relief as proof of such, thus reinforcing and perhaps even strengthening expectations of a positive outcome.
Subconscious factors that influence the placebo effect may be based on an individual’s health care history, or even that of a friend or family member. Although a trial participant may not have a conscious expectation that a treatment will work, subconsciously he or she may equate the care and attention shown by researchers and physicians with a positive outcome, and attribute, correctly or incorrectly, any minor symptomatic relief to the active or placebo treatment he or she is receiving.
Placebo effect responses can be triggered by, or linked to:
In a placebo-controlled study, one group of trial participants receives an experimental treatment and a second group receives the placebo (fake) treatment. Such trials usually are blinded, meaning trial participants aren’t told (and presumably can’t figure out) whether they’re receiving the experimental treatment or a placebo.
In double-blind placebo-controlled trials — considered the “gold standard” for evaluating the efficacy of experimental therapies — neither the researchers nor the trial participants know who is receiving the experimental treatment or placebo. This is done to ensure that those conducting the study provide objective assessments, and that they don’t somehow unconsciously influence trial participants’ beliefs or expectations.
The purpose of the placebo group in any trial is to account for the placebo effect — that is, effects from treatment that do not depend on the treatment itself.
Without a placebo group to compare against, it is usually (but not always) impossible to know whether an experimental treatment provides any benefit.
So, ironically, in the end it’s the treatments with nothing in them that help get medications and treatments with that something special in them through clinical testing and into the hands of patients.