A polygraph examiner and a bizarre magic performer both watch faces for a living. That’s about where the resemblance ends. I’ve spent years reading books on interrogation technique, the Reid method, statement analysis, the whole shelf, and I keep running into the same problem. The room is wrong. The stakes are wrong. The subject is wrong. Everything that makes those techniques work depends on conditions a mentalist almost never has.
I’m not dismissing the material. Some of it transfers beautifully, once you strip away the cop-show language and look at what’s actually being measured.
Most performers borrow the vocabulary anyway because it sounds authoritative. “Tells.” “Leakage.” “Baseline.” Words lifted from a world of locked rooms and criminal stakes, then pasted onto a volunteer who’s had two drinks and just wants to see something impossible happen. That mismatch is worth taking apart piece by piece.
The Interrogation Room Assumes a Liar Under Threat
Paul Ekman’s original microexpression research, and most of what came after it, was built around a specific scenario. Someone has done something, they know it, and they’re being asked direct questions with consequences attached to a wrong answer. The fear of being caught produces a measurable physiological signature. A flash of contempt at the corner of the mouth. A flinch around the eyes lasting less than half a second. These are real. I don’t dispute the data.
But notice what has to be true for that signal to appear at all. The person has to be concealing something specific. They have to believe getting caught carries a cost. And they have to be suppressing a genuine emotional response in real time.
Your volunteer at the table has none of that.
She isn’t hiding a crime. She’s holding a playing card, or thinking of a childhood memory, or gripping a sealed envelope you asked her to focus on. There’s no threat, no guilt, no suppressed emotion fighting its way to the surface. Applying interrogation-grade reading to that moment is like bringing a stethoscope to listen for an engine that was never running.
What Actually Transfers: Baseline, Not Betrayal
Here’s the part worth keeping. Baselining is real, and it’s useful, just not for the reason interrogators use it.
In a police interview, baseline means establishing how someone behaves when telling the truth about something low-stakes, “what’s your name,” “where do you live”, so you can later notice deviation when the real questions start. In a séance or a sealed prediction routine, I use it for something quieter. I want to know what this particular person looks like at rest, before anything charged has been said, so I can notice the smallest shift once it has.
That shift is rarely dramatic. It’s a change in breathing rhythm. A hand that stops moving instead of one that starts. Weight transferring off one hip. These aren’t tells in the interrogation sense, evidence of concealment, they’re evidence of attention. Something just got heavier in the room, and the body register it before the person consciously decides to hide or reveal anything.
I tested a version of this at a private event near Louisville a while back, watching a small group of volunteers as I named locations on a hand-drawn map, waiting to see which name made someone’s shoulders drop half an inch. Nobody told me anything. Nobody needed to.
My hands stay still while I’m watching for this. Moving draws the eye away from theirs, and the whole read depends on nobody knowing they’re being read.
The Seven Universal Expressions Are the Wrong Tool for This Job
Ekman’s seven universal expressions, anger, disgust, fear, happiness, sadness, surprise, contempt, get quoted in mentalism circles like scripture. Learn to spot a flash of disgust and you’ll know when someone’s lying, or so the pitch goes.
This is where I get skeptical, and I stay skeptical.
Those seven categories were built to classify emotion, not to detect deception. Ekman himself has said as much in his later writing, the presence of an emotion tells you nothing about its cause. A flash of fear on your volunteer’s face might mean she’s thinking about something painful. It might mean she’s afraid of looking foolish in front of her husband. It might mean absolutely nothing, a random neural firing that has nothing to do with the number she’s thinking of.
Trying to read specific content, which card, which name, which memory, off a universal emotion category is a category error. Emotions don’t carry that kind of resolution. It’s the equivalent of using a thermometer to find out what someone’s thinking about. You’ll get a real number. It won’t tell you what you want to know.
This is the point where most guides get it backwards.
They teach you to hunt for named emotions when you should be hunting for change.
What Bizarre Magic Actually Needs From the Body
Strip the interrogation framework down to its actually useful parts and you get three things worth building a technique around.
- Timing over content. You don’t need to know what emotion someone’s body is showing. You need to know when it changed, relative to what you just said or did.
- Asymmetry as a flag, not a verdict. A one-sided mouth movement, one shoulder rising higher than the other, these are worth noting because they’re rarer than symmetric movement, not because they mean deception. Rarity earns attention. It doesn’t earn a conclusion.
- Suppression, not expression. Interrogators look for a flashed emotion. I look for a stopped one. Someone catching themselves mid-reaction, a laugh cut short, a breath held instead of released, tells me more than any named expression ever has. That’s a person managing themselves in front of an audience, and a bizarre magic performer’s whole job is noticing what people manage.
None of this requires the framework of guilt and concealment. It requires patience, and a willingness to watch instead of narrate.
I don’t necessarily believe you can read a specific number off a stranger’s face with anything close to reliability. I do have a strong conviction that you can read the moment a stranger stops thinking about nothing and starts thinking about something.
That gap, between nothing and something, is where the whole performance lives.
Cold Reading Fills the Space Interrogation Leaves Empty
If microexpression work only tells you when, cold reading is what tells you toward what. This is the piece most performers skip because it feels less scientific, less “trained by the FBI,” more like an old carnival trick. That instinct is exactly backwards.
Cold reading, done properly, is a feedback loop. You make a statement broad enough to likely land, watch the body’s timing response to it, then narrow based on what you saw. The microexpression training tells you the volunteer’s shoulders dropped when you mentioned a body of water. Cold reading is what lets you follow that thread, “near water, but not on it, somewhere you could hear it,” and keep narrowing until the room goes quiet in the specific way rooms go quiet when you’ve actually landed somewhere true.
Used together, body reading gives you the signal and cold reading gives you the vocabulary to act on it out loud, in real time, without ever naming what you’re doing.
Where This Gets Genuinely Unnerving
The part I still haven’t fully made peace with, after years of using this stuff on stage, isn’t the technique. It’s how little the volunteer knows they’re transmitting.
People believe their faces are private property, that a poker face is achievable with enough will. My experience says otherwise. The body leaks intention constantly, not through some elaborate deception signature, but through simple biological economy, suppression takes more effort than expression, and effort shows.
I’ve watched a grown man’s jaw tighten over a name he hadn’t said out loud in years.
He never confirmed it. He didn’t have to.
Whether that’s a “tell” in the interrogation sense or just a nervous system doing what nervous systems do under quiet pressure, I genuinely don’t know. I’m not sure it matters. What matters is that the reading was true, the room felt it, and nobody in that room could explain afterward exactly how I knew.
That unexplainable residue is the whole business we’re in.
Building This Into Your Own Work
If you’re going to take anything from this into your own act, take it in this order. Learn to baseline a stranger in under thirty seconds of small talk, before any charged question gets asked. Practice noticing timing rather than naming emotion, when did the shift happen, not what was the shift. Then pair that timing data with cold reading language so you have somewhere to put what you noticed.
Skip the parts borrowed wholesale from interrogation rooms. Your volunteer isn’t a suspect, and treating her like one, hunting for guilt signatures on a face that has nothing to hide, will make your reads worse, not better. You’ll see things that aren’t there because you’re looking for the wrong category of thing entirely.
So yes, the vocabulary is worth knowing. Just don’t mistake borrowed vocabulary for borrowed truth.
If you want the physical pieces that make this kind of work land at the table, the sealed items, the props built for exactly this quiet style of reading, take a look through what we carry.
