Case File 063 · Investigation and Psychology

Why There Is No Single Reliable Sign of Deception

Why research has found no behaviour that reliably reveals a lie, and why nervous or evasive behaviour has many innocent explanations.

By Murder Book Editorial TeamPublished 16 September 2026Updated 16 September 2026

There is no single reliable sign of deception because no behaviour appears only when people lie. Looking away, fidgeting, pausing or seeming nervous can all happen when someone is telling the truth, and many liars show none of them. A major 2003 meta-analysis of 158 possible cues found that many behaviours had no link, or only a weak link, to lying. The differences that did appear were small averages across groups, not signals that can be read in one person. Later work suggests that even some of those estimates may be inflated. The hope for a clear tell is often compared to Pinocchio's nose: a visible sign that grows only when someone lies. Humans do not have one. That is why behaviour can be a reason to ask more questions, but never proof on its own.

Key takeaways

No behaviour has been found that reliably shows someone is lying.
Across 158 cues, most showed no link or only a weak link to deception.
The belief that liars avoid eye contact is common worldwide but poorly supported.
Nervousness has innocent causes, including fear of not being believed.
Some reported cue effects may be exaggerated by publication bias and small studies.

What is the Pinocchio's nose myth?

In the story, Pinocchio's nose grows every time he lies. It is the perfect lie detector: it appears only with lies, every time, and anyone can see it.

Many popular ideas about lying assume people have something similar, such as a glance to the side, a touch of the face or a shift in the chair. In 2019 the psychologist Timothy Luke used the image in the title of a review, "Lessons From Pinocchio", which argued that even the cue effects reported in the research literature may be much larger on paper than in reality.

A cue would need two properties to work like Pinocchio's nose:

It appears reliably when people lie.
It rarely appears when people tell the truth.

Research has not found a behaviour that meets both conditions.

What did the big meta-analysis of deception cues find?

In 2003, Bella DePaulo and colleagues published "Cues to deception" in Psychological Bulletin. They combined 1,338 estimates of 158 possible cues from earlier studies. Their summary was mixed:

On average, liars were somewhat less forthcoming and told less compelling stories.
They made a more negative impression and were rated as more tense.
Their stories included fewer ordinary imperfections and unusual details.
However, many behaviours showed no discernible link, or only a weak link, to deceit.

Context mattered. Cues were more pronounced when people were highly motivated to succeed, especially for reasons tied to their identity, and when lies were about wrongdoing.

It is important to read these findings correctly. They describe small differences in averages across many people and many studies. They do not mean that a tense person with a less detailed story is lying. Plenty of truthful people are tense, and plenty of true accounts are short.

Do liars really avoid eye contact?

It is the most widespread belief about lying. The Global Deception Research Team, a group of researchers working across 75 countries and 43 languages, asked people "How can you tell when people are lying?" The results, published in 2006, revealed a dominant belief across cultures: that liars avert their gaze.

The research does not back that belief up. The 2003 cue review found that many behaviours had little or no link to deception, and reviews since have repeated the point. The Annual Review of Psychology article by Aldert Vrij, Maria Hartwig and Pär Anders Granhag (2019) concluded that the nonverbal cues to deceit discovered so far are faint and unreliable, and that misbeliefs about behaviour and deception appear very hard to debunk.

There is also a gap between what people say and what they do. In a 2011 meta-analysis, Hartwig and Bond found that although people report using eye contact to judge lies, eye contact was only weakly related to the judgements they actually made. For more on how well people judge lies in practice, see Can You Actually Tell When Someone Is Lying?

Why can innocent people look deceptive?

Behaviours linked with lying in popular belief are often signs of stress, and stress has many causes. Being questioned about something serious is stressful in itself, whether or not you are guilty.

Research gives a clear example. In their 2006 meta-analysis of deception judgements, Bond and DePaulo found that people appear deceptive when they are motivated to be believed. Someone who desperately wants to be believed, perhaps because the consequences of doubt are serious, may come across as less credible, even when they are telling the truth.

Other innocent explanations include:

fear, embarrassment or grief;
cultural differences in eye contact, gesture and directness;
neurodivergence, anxiety or other conditions that affect how people present themselves;
fatigue, or discomfort with a formal setting;
protecting a secret unrelated to the matter being discussed;
ordinary memory gaps and uncertainty.

This is why the idea of behavioural leakage needs careful handling. A visible change in behaviour tells you that something is happening for that person. It does not tell you what.

Could the known cues be even weaker than reported?

Possibly. Luke's 2019 review in Perspectives on Psychological Science examined the deception-cue literature and ran simulations. It argued that many estimated effect sizes may be greatly inflated by publication bias (studies with striking results are more likely to be published), small numbers of estimates and low statistical power.

The review went further: given the state of the evidence, it may not be possible to tell whether a given reported cue is a real effect or a false positive. Luke warned about the hazards of relying on potentially illusory cues when assessing credibility in practical settings.

Not every researcher draws the same conclusion from this, but the direction is consistent. The more closely the evidence is examined, the weaker the case for a reliable behavioural sign.

Why do these myths matter in investigations?

If an investigator believes a specific behaviour signals lying, that belief can steer an interview. A 2025 scientific review paper on police-induced confessions by Saul Kassin and colleagues, published in Law and Human Behavior, addresses the practice of judging deception by observing a suspect's verbal and nonverbal behaviour. The authors put it bluntly: demeanour is not evidence. They conclude that such assessments do not offer a reliable basis for inferring deception, and are more likely to produce a presumption of guilt that can lead to more pressure and a greater risk of false confessions.

This is where the cue research connects to wider problems covered elsewhere in the Casebook, such as how confirmation bias can affect an investigation and why innocent people sometimes confess.

What works better than looking for a tell?

Hartwig and Bond's 2011 analysis concluded that lie detection is more readily improved by increasing the behavioural differences between liars and truth tellers than by teaching observers a list of cues. In other words, the focus in research has moved from watching to asking: interview approaches that gather information and allow it to be tested against what is already known.

Those approaches are covered in What Is Cognitive Load in an Interrogation? They are not perfect either, and none turns a single moment of behaviour into proof.

How does Murder Book treat behaviour?

Murder Book: The Art of Interrogation is an upcoming narrative interrogation game from Gungrounds in Zagreb, Croatia, inspired by real police footage and psychological interviews. Development screenshots show an in-game casebook listing behaviours such as Stonewalling, Stress Response, Avoidance, Non-Response, Submissive Gesture and Emotional Fading.

A list like that could look like a set of tells, but the research suggests reading each one differently. A Stress Response might come from guilt, or from fear of being disbelieved. Avoidance might hide a lie, or something the person is simply ashamed of. In Murder Book, spotting a behaviour is a reason to look closer, not proof: the question is how it fits with the statement, the evidence and the other possible explanations. There is no Pinocchio's nose to rely on, which is exactly what makes interrogation an interesting puzzle.

Murder Book is in development for PC (Steam) and mobile, with release currently planned for the end of Q2 2027.