I want to start with the data, because the data is what stopped me in my tracks.
Researchers showed participants recordings of magic performances. Same tricks. Same technical execution. Same presentation style. The only variable was the name attached to the performer — whether participants were told the performance was by a male magician or a female magician.
The results: audiences rated the same trick as less impressive when they believed it was performed by a woman.
Not dramatically less impressive. Not laughably less impressive. But measurably, statistically significantly less impressive. The bias was consistent, reproducible, and — here is the critical part — completely unconscious. Participants did not know they were rating performances differently based on the performer’s gender. They believed their ratings were based on the quality of the performance itself.
When I encountered this finding in the research collected by Gustav Kuhn and Alice Pailhes, I sat with it for a long time. It made me uncomfortable in ways that I think are important to articulate, because the discomfort is part of the point.
The Study Design
Let me describe what the researchers actually did, because the methodology matters.
The key design feature was control. In the most rigorous versions of these studies, participants watched identical performances — literally the same video. The only thing that changed was the information provided about the performer. In some conditions, the performer was described with a male name. In others, a female name. The participants did not see the performer’s face or body in a way that would allow them to determine gender independently; they relied on the name provided.
This is important because it eliminates every variable except the name. Same technique. Same presentation. Same camera angle. Same everything. The only thing the participants had to differentiate the conditions was a name that implied gender.
And still, the ratings differed.
The effect was not enormous in absolute terms — we are not talking about a chasm between male and female ratings. But in social science research, the fact that it was consistent and statistically significant is what matters. It means the bias is real, it is replicable, and it operates across enough of the population to represent a genuine pattern rather than noise.
Why It Happens
The bias has roots in what psychologists call schema-based expectation. Our brains maintain mental models — schemas — for what different categories of people are like and what they are capable of. These schemas are built from a lifetime of cultural exposure: media, personal experience, societal narratives, and the accumulated weight of everything we have ever seen and heard about a given group.
The schema for “magician” is overwhelmingly male. Magic has been, historically and currently, a male-dominated field. The vast majority of famous magicians are men. The vast majority of magic on television features men. The vast majority of magic instruction is taught by men. When someone says “think of a magician,” the image that comes to mind for most people is a man.
This is not a statement about ability. It is a statement about cultural exposure. The schema is built from what we have seen, not from what is possible.
When a performance is attributed to someone who matches the schema (male magician), the performance is evaluated through the lens of congruent expectations. The brain says: “this fits.” The performance is processed fluently — smoothly, without friction.
When the same performance is attributed to someone who does not match the schema (female magician), the brain encounters a mild incongruity. The performance does not match the default expectation. This incongruity creates a subtle processing friction — a form of disfluency, as we discussed in the previous post. And that friction, operating below conscious awareness, slightly reduces the perceived quality of the performance.
The participant does not think, “This is less impressive because it was done by a woman.” They think, “This is less impressive.” The gender inference happens upstream, in the schema-matching process, before the conscious evaluation begins. By the time the participant is aware of their judgment, the bias has already done its work and is invisible.
The Orchestra Audition Parallel
There is a famous study from the classical music world that illustrates the same phenomenon in a different domain.
Major orchestras historically held auditions where the committee could see the performer. Under these conditions, women were significantly less likely to advance to later rounds. When orchestras introduced blind auditions — with the performer hidden behind a screen — the rate of women advancing increased by fifty percent.
Same performances. Same music. Same technical skill. The only change was whether the evaluators could see the performer’s gender. And that single variable shifted outcomes dramatically.
The orchestra study is often cited as one of the clearest demonstrations of unconscious gender bias in performance evaluation. The magic research extends this finding into a new domain and confirms that the bias operates even when the evaluator is not a trained professional but an ordinary audience member.
My Own Reckoning
Reading this research forced me to examine my own biases, and the examination was not comfortable.
I am a man in a male-dominated field. I benefit from the schema. When I walk up to a table at a corporate event in Austria and introduce myself as someone who does magic and mentalism, the audience’s default schema activates in my favor. They expect a man. They see a man. The congruence creates fluency. My performance starts with a small tailwind that I did nothing to earn.
A woman walking up to the same table with the same introduction faces a small headwind. Not because the audience is hostile. Not because they are consciously prejudiced. But because the schema does not match, and the mismatch creates friction, and friction lowers evaluations.
I want to be clear: this is not about individual audience members being sexist. This is about a statistical pattern across populations. Any individual audience member might rate a female performer higher than a male performer. But across hundreds of evaluations, the pattern tilts. And the tilt is enough to create real, measurable disadvantage.
When Adam Wilber and I were building Vulpine Creations, we had many conversations about who the magic community includes and who it leaves out. Magic has a diversity problem — not just gender, but age, ethnicity, and cultural background. The research on gender bias in magic evaluation suggests that part of the problem is not just access and representation, but the psychological machinery of evaluation itself. Even if more women entered magic, they would face a systematic evaluation disadvantage that their male peers do not face.
Understanding this does not fix it. But understanding it is the first step.
The Implications for How We Watch
Here is what I find most troubling about the gender bias research: it is unconscious, which means it is resistant to good intentions.
People who scored high on measures of egalitarian attitudes — people who explicitly believe that men and women are equally capable — still showed the bias. The bias does not live in beliefs. It lives in schemas. And schemas are updated slowly, through accumulated experience, not through intellectual commitments.
This means that you cannot eliminate the bias by deciding to be fair. You can only reduce it by changing the inputs that build the schema. More exposure to female magicians — on television, in live performance, in instructional materials, in the cultural conversation about magic — will gradually update the schema until “magician” no longer defaults to “male.”
But that update is slow. Schemas built over decades do not change overnight. In the meantime, the bias operates in every audience, at every show, affecting every evaluation.
What Performers Can Do
If you are a female magician or mentalist reading this, the research has a discouraging headline but a more nuanced story underneath.
The bias is real, but it is not insurmountable. The studies show a statistical tilt, not a deterministic outcome. Individual performance quality still matters enormously. An excellent female performer will be rated higher than a mediocre male performer. The bias shifts the baseline, but it does not override quality.
Moreover, the framing research we discussed in the previous post offers a partial counterweight. Accolades and credentials significantly increase audience evaluations. A female performer whose introduction includes specific, credible accomplishments benefits from the accolades effect, which can partially offset the schema-based penalty.
The most effective accolades, according to the research, are television appearances and specific skill claims (“one of only four people who can perform this”). For female performers navigating gender bias, building and prominently displaying credentials may be not just a marketing strategy but a psychological necessity — a way to update the audience’s schema in real time, before the performance begins.
But I want to be honest: suggesting that women need to work harder to prove themselves is not a solution. It is a coping strategy for an unjust situation. The actual solution requires changing the schema, which requires changing the culture, which requires changing who gets opportunities, platforms, and visibility.
What Male Performers Should Do
I want to address this directly because I think male performers have a responsibility here that is easy to avoid.
First, acknowledge the bias. It exists. The data is clear. Dismissing it as “just one study” or “not that big” is a form of willful ignorance that benefits the people who are already benefiting from the bias.
Second, amplify female performers. Share their work. Recommend them for gigs. Include them in shows and events. Every time a female performer is seen and appreciated, it chips away at the schema that says “magician equals male.” The schema updates through exposure, and exposure requires opportunities.
Third, examine your own evaluations. When you watch a female performer and find yourself thinking “she’s good for a woman,” catch yourself. The qualifier reveals the bias. She is good, period. Or she is not good, period. The gender should be irrelevant to the evaluation, and if it is not, the problem is in the evaluator, not the performer.
Fourth, talk about it. The research shows that awareness alone does not eliminate unconscious bias. But it does create the conditions for change. When the magic community talks openly about gender bias — acknowledges it, studies it, works to counter it — the culture shifts, slowly. Silence preserves the status quo.
The Broader Pattern
Gender bias in magic evaluation is part of a larger pattern in how humans evaluate performance across domains.
Studies show similar biases in evaluations of musicians, athletes, academics, business leaders, and artists. In every domain where the default schema is male, female practitioners face a small but consistent evaluation penalty. In domains where the default schema is female (nursing, teaching, social work), male practitioners sometimes face a similar penalty, though the research is less extensive.
The common thread is schema congruence. Performance is evaluated more favorably when the performer matches the audience’s default expectation for that category. And default expectations are built from cultural exposure, which reflects historical patterns of access and representation.
Magic is not uniquely biased. But it is not exempt from bias either. And given that magic prides itself on challenging assumptions — on making people question what they believe to be possible — there is a particular irony in the magic community’s slowness to question its own assumptions about who belongs on stage.
Why I Wrote This Post
I debated whether to include this topic in the blog. It is not about technique. It is not about showmanship. It is not about the craft of creating impossible moments. It is about something less comfortable: the way our brains evaluate the people who create those moments.
I decided to include it for two reasons.
First, because the research is part of the science of magic. This blog series is about what cognitive science tells us about how magic works. And how magic works includes how audiences evaluate performers, which includes the biases that shape those evaluations. Leaving out gender bias because it is uncomfortable would be intellectually dishonest.
Second, because I believe the magic community — a community I joined as an outsider, as an adult, with fresh eyes — can be better than its history suggests. Magic is an art form built on impossible things. A more inclusive community is not impossible. It just requires the same thing that any good effect requires: deliberate design, honest evaluation, and the courage to change what is not working.
The data says that audiences rate the same trick differently based on the performer’s gender. The data says the bias is unconscious. The data says it is significant.
What we do with that data is not a scientific question. It is a moral one.
And I think the answer is clear: we work to change it. Not because the data tells us to. But because the kind of community we want to belong to — the kind of art form we want magic to be — demands it.
Same trick. Different name. Different rating. That is not magic. That is bias. And bias, unlike magic, should not be left unexplained.