A magician is observed going through the familiar motions of making an object disappear, but there’s a catch: there was never an object to begin with.
Yet roughly one-third of participants in a recent experiment reported seeing it. Many described an object consistent with those they had seen in earlier tricks, apparently imagining something the magician’s hands never really held.
The experiment, known as the Phantom Vanish Trick, is among 55 empirical studies examined in a new review exploring how illusionists exploit attention, perception, reasoning and memory. Together, the findings offer a revealing look at how the brain turns sensory information into the reality we experience.
Published in Frontiers in Psychology, the review by Miguel De Lucas and Jon Andoni Duñabeitia of Universidad Nebrija in Madrid brings together empirical research spanning 1995 to 2026. The strongest evidence shows that looking directly at something does not necessarily mean we consciously see it.
“A magic trick is not merely an entertainment device; it is a carefully designed cognitive event in which attention, perception, expectation, memory, causal reasoning, and emotional appraisal are manipulated to produce the experience of impossibility,” researchers write. “Misdirection operates in part through covert attention and cannot be explained solely by where spectators are looking.”
The familiar explanation for misdirection is that a magician makes an audience look elsewhere while carrying out the secret move. But eye-tracking experiments suggest that the relationship between looking and seeing is much more complicated.
In research involving a disappearing lighter, observers could fail to notice the lighter falling even when their eyes were directed toward, or near, the relevant area. Their eyes had access to the event. Their conscious awareness did not necessarily follow.
Researchers call the allocation of attention without a corresponding eye movement “covert attention.” A spectator may be looking at one part of a scene while devoting mental resources to something else. Magic provides a controlled way to separate those processes and examine where awareness breaks down.
The timing of a performance can matter as much as its distractions. A study of spontaneous blinking found that viewers’ blinks were highly synchronized and tended to cluster around a magician’s secret actions rather than the apparent magical climax.
The researchers suggested that the accompanying relaxation in attention could help conceal the method. An audience watching the same routine may therefore share brief lapses at particularly useful moments for the performer.
Other experiments show how someone’s expectations can influence what they perceive.
In the Vanishing Ball Illusion, a magician tosses a ball into the air, then pretends to repeat the throw while keeping it hidden. Spectators may perceive the ball rising and disappearing, although it never left the hand.
Another study found that viewers did not even need to see a genuine throw first. Their existing knowledge of how objects move appeared sufficient to support the illusion. For example, experiments involving a coin revealed that appropriate sound accompanying the fake throw increased the probability that viewers would experience the visual effect.
The Phantom Vanish Trick pushes the question of “seeing is believing” even further by removing the object altogether. After watching videos in which a magician made real objects disappear, participants saw the performer mime a similar routine with empty hands. The movements suggested that something was being handled and then vanished. Yet, no object was ever present.
Around one-third of participants reported seeing an object during that final performance. Many described something consistent with objects shown in the preceding tricks, suggesting that their expectations helped fill in what was missing. The videos were silent, ruling out spoken suggestions from the magician as an explanation for participants’ misperception.
Researchers interpret those findings as evidence that expectations can lead people to perceive something that is not there. After watching earlier tricks, some viewers apparently expected another object so strongly that they reported seeing one in the magician’s empty hands.
Interestingly, learning a magic trick’s secret doesn’t necessarily clear things up.
Some tricks exploit “amodal completion,” the visual system’s tendency to fill in hidden portions of objects. Researchers highlighted past studies showing that illusions relying on amodal completion could persist even after spectators understand the method.
Essentially, even when spectators know how a trick works, they still may see the illusion. Researchers can use these types of tricks to study why understanding a deception does not always change what we see.
Even when people successfully spot a deception, they may still be mistaken about how it occurred. In one study, participants believed they had looked directly at the critical area when detecting a trick. Yet, eye tracking showed that most successful detections occurred through peripheral vision.
Brain-imaging work gives a preliminary glimpse of what happens when an event appears impossible. Studies revealed that neural networks associated with conflict recognition, prediction, and violated expectations play key roles.
In one imaging study, researchers found increased activity in the caudate nucleus when they analyzed viewers’ responses across the full magic trick. They suggested that this activity reflected the expectations spectators built as they watched the magician’s actions unfold.
Earlier research offers a possible explanation. In a 2011 experiment, the caudate became active when viewers watched a dancer deviate from movements they had learned to expect. The finding suggested that this brain region helps register when an observed action breaks an established prediction, even without reward or punishment.
During a magic trick, spectators likewise anticipate familiar outcomes—for example, a coin should remain inside a closed hand. The imaging results are consistent with the caudate helping track those expectations, although they do not establish that its activity represents an attempt to solve the deception.
However, of the studies examined, only five fell within the review’s neural category. Researchers caution that the evidence cannot yet establish a definitive model of how the brain processes magic.
The wider literature similarly has its limits. The median study included only 50 participants, almost half had fewer than 50, and the research was geographically concentrated. The researchers charted the evidence in a scoping review rather than pooling results in a meta-analysis.
Nevertheless, researchers say magic offers a way to compare a known sequence of physical actions with how someone perceives the experience. A lighter falls, a coin stays hidden, or an object is never introduced. The spectator’s perception of what occurred can be widely different from the underlying reality.
For scientists studying perception, magic tricks offer an exceptional opportunity to investigate how the brain arrives at a convincing experience of events that did not happen as we perceive them.
“In this sense, magic is not merely an object of scientific study,” researchers write. “It can also serve as a methodological instrument for investigating cognitive processes that are difficult to examine under ecologically valid conditions.”
The study, “Magic and cognition: a scoping review of empirical evidence on cognitive and neural processes,” appeared in Frontiers in Psychology.
Tim McMillan is a retired law enforcement executive, investigative reporter and co-founder of The Debrief. His writing typically focuses on defense, national security, the Intelligence Community and topics related to psychology. You can follow Tim on Twitter: @LtTimMcMillan. Tim can be reached by email: tim@thedebrief.org or through encrypted email: LtTimMcMillan@protonmail.com