When someone lies to you, how likely are you to catch it, how much better than a coin flip do most people actually do, and what does a former Secret Service agent who trained interviewers say about reading a face? The odds you catch a lie are about 1 in 2.1 (our arithmetic): people judged 47% of lies to be lies in a 2006 review by Charles F. Bond Jr. and Bella M. DePaulo of 206 documents and 24,483 judges. Across lies and truths together, people were right 54% of the time, about 1 in 1.85 (our arithmetic), a little better than a coin flip, and they leaned toward belief, judging 61% of truths truthful. For how often people lie in the first place, see our look at lying.
Book of Odds spoke with Paul Kelly in 2010; the interview was online by June 3, 2010.
Key numbers
| Figure | Source and year |
|---|---|
| 54% of lie or truth judgments correct | Bond and DePaulo meta-analysis, 2006 |
| 47% of lies judged deceptive | Bond and DePaulo meta-analysis, 2006 |
| 61% of truths judged truthful | Bond and DePaulo meta-analysis, 2006 |
| 24,483 judges in 206 documents | Bond and DePaulo meta-analysis, 2006 |
Who Paul Kelly is
Paul Kelly is a former Secret Service Special Agent, who served as a course director and senior instructor at the US Secret Service Training Academy, has worked with the Department of State’s International Law Enforcement Academy and its Anti-Terrorism Assistance Program, and taught at the NSA’s National Cryptologic School.
Mr. Kelly was, at the time of the interview, an instructor with the Paul Ekman Group, which then trained law enforcement, doctors, therapists, teachers, and corporate negotiators in the interviewing methodology pioneered by Dr. Ekman, whose work inspired the popular Fox TV drama Lie to Me, which was about to start its third season in June 2010.
In the show, Tim Roth played Dr. Cal Lightman, a tough-minded investigator who solves cases using his skills in sifting truth from lies. The stories are fictitious, but the techniques come right out of the playbook Kelly used as a trainer.
The figures in Kelly’s answers, including his 54% average, the seven facial expressions, the length of a micro-expression and the more than 1,000 TSA staff his group trained, are his, from 2010, and are not checked here; the box above gives the published figures. His remark that polygraphs are not admissible in court is a 2010 general statement, not legal advice; US rules vary.
The interview
Book of Odds: I went to the “Lie to Me” website and tried the “Lightman” tests to test my own skills at detecting truth and lies, and I didn’t do too well.
Paul Kelly:
Well, welcome to the group! The vast majority of people don’t do too well, and in fact most people think they’re much better at detecting deception than they actually are. The average person is accurate 54% of the time in discerning truth or deception.
Book of Odds: Is the ability to detect deceit something that can be learned?
Paul Kelly:
I think part of that is life experience. It generally comes out that even with the people who do much better than average; it’s not an innate ability.
Book of Odds: You’re a former Secret Service agent. Dr. Maureen O’Sullivan, who has worked closely with Dr. Ekman, wrote that Secret Service agents were one of the high-performing groups. You, in fact, are one of the people Dr. O’Sullivan refers to as “truth wizards.”
Paul Kelly:
[People in] law enforcement tend to do better than the average citizen, and with the groups that they have tested, Secret Service agents came out higher than other groups. Why? Secret Service is trained to actively listen and actively observe. Being very proactive rather than reactive is the cornerstone of the Secret Service methodology.
Book of Odds: What are “micro-expressions” and how do they fit into the training?
Paul Kelly:
Dr. Ekman has identified seven different facial expressions that people show as unconscious “micro-expressions” that last for just the blink of an eye: happiness, sadness, anger, surprise, fear, disgust, and contempt. Now, generally you’re not going to confuse happy and sad. Sometimes people will confuse fear and surprise.
Book of Odds: What about disgust and contempt?
Paul Kelly:
It’s funny you mention those, because of those seven emotions, the ones people get least are disgust and contempt. And yet they can be a serious indicator to an interviewer. But you’ll never miss contempt once you’re trained to recognize it! In fact it’s amazing, once you start being aware of disgust and contempt, you see them a lot more. More than you might want to.
The other one that’s important in law enforcement is anger.
To be a good interviewer, number one, you have to learn to listen, to be an active listener. Number two is to actively observe what’s going on in terms of the non-verbal. Sherlock Holmes said, “I have trained myself to notice what I see.” That’s the first step to a good interview.
Book of Odds: What’s the difference between macros and micros, other than duration?
Paul Kelly:
Macro-expressions are what people want you to see, whether or not it’s what they’re actually feeling. Many cultures will mask their true emotions, with a smile for example—the protocol smile vs. the sincere smile.
The interesting things about these micro-expressions are, first, that they’re cross-cultural, and second, that they’re involuntary, as opposed to a macro that you can fake. Micros, however, you don’t fake. They just happen. Facial micro-expressions are an invaluable tool to anybody in any arena of social dynamics, whether you’re a police officer or a therapist or a social worker or a negotiator.
Micros are very short in duration, anywhere from a tenth to a thirtieth of a second, the blink of an eye. What we’re looking for is, first of all, is to determine if it is consistent with whatever else we’re hearing or seeing from the person. If it is not consistent, that’s what Ekman calls a “hot spot.”
Book of Odds: An indicator that the person is likely engaging in deception?
Paul Kelly:
Not necessarily. An indicator that there is an inconsistency; it may or may not be an indicator of deception. As an interviewer, when you see or hear a hot spot, you’re probably going to want to follow up on it. But micros, hot spots, themselves are not proof positive that somebody is lying to you.
Normally in a micro you’re seeing either a conscious suppression of emotion, or an unconscious repression of emotion. That’s why a lot of times you’ll only see a partial, what we call a “mini” of a micro. That’s at more of an advanced level.
Book of Odds: What kind of feedback do you get about the training?
Paul Kelly:
It’s very seldom that we do not get a positive response to the training. And that applies to corporate folks too, who really benefit. There are very few people who do not improve in a matter of just a couple of hours. A lot of it is awareness.
The first step is getting to recognize that one of these micros has occurred, and you can see some examples on the Lie to Me website and at Dr. Ekman’s website.
Book of Odds: What about polygraphs?
Paul Kelly:
There are still federal agencies that require polygraphs as part of the application process. But they’re still not admissible in court. It comes down to the operator, and the planning of the session, including the formation of the questions, all of which are usually closed questions that can only be answered by a yes or a no. That’s where cognition comes into play. On a polygraph, they will tell you the questions in advance, they have to—they want to—because right away when you hear the question, you’re asking yourself, “When he asks me this, do I really want to say yes, or do I want to say no?”
Both the polygraph and the good interviewer are relying on picking up this leakage that occurs because of increased stress or anxiety, because the subject is going to have to lie. And even the sociopath doesn’t want to get caught. They might not have a sense of guilt but they still don’t want to get caught.
Book of Odds: How is this training useful in the corporate world?
Paul Kelly:
Even though they have a much lower burden of proof than law enforcement or criminal justice, corporate negotiators still want to know if the person across the table is being honest.
A lot of times people will say, Oh, I’ve got a bad feeling about that guy, or my gut tells me, or call it intuition. I think a lot of times people are picking up on these verbal and non-verbal indicators; they just haven’t been trained to be able to articulate what they’re picking up on. That’s why some people do better than others. That goes back to their life experience.
Book of Odds: Is there a danger that these techniques could be overused, or inappropriately used?
Paul Kelly:
This gets to another key point in Ekman’s interviewing methodology: the moment you see something, a micro for example, you don’t want to jump to a conclusion like “he’s showing fear, that means he’s guilty.” He might be showing fear because he’s talking to a detective. You have to come up with an alternative hypothesis. Even if I think it’s because he did it, what else could it be? By forcing yourself to keep an open mind, you’re keeping your objectivity, which will increase the validity of your assessment.
With the TV show, if you read Ekman’s comments on the episodes, “The Truth behind Lie to Me,” he does a very good job of separating the science from the Hollywood. Lightman many times is abrupt, obnoxious, makes a quick rush to judgment, and these are not really the traits of an effective interviewer. However, I think much of the science that is incorporated is based on stuff that’s happened. There is a precedent or an example for making a teaching point of whatever the circumstance was that you see in the micro or the body language, the shoulder shrug…
But it’s important to note that there is no Pinocchio clue. And unfortunately there are people out there who are supposed to be experts in the field who will tell you that there are single displays which conclusively indicate that someone is lying. A lot of people will get up there and say, “If they won’t look me in the eye, that means they’re lying,” or “If they rub their nose, that means they’re lying,” or “If his lips are moving, he’s lying.” It’s simply not the case.
Book of Odds: You mentioned that once you’re attuned to this, you can’t turn it off. Couldn’t that make maintaining objectivity even more complicated in your personal relationships?
Paul Kelly:
When trust, friendship, or romance gets involved, objectivity often takes a hit. How many times have you heard in a trial, or on a TV show, when a parent says something like, “My son’s a good boy, he’d never do anything like that”? Or “My spouse would never cheat on me.” A lot of times people don’t want to know the truth. Or as Jack Nicholson said in A Few Good Men, “You can’t handle the truth.” Some people don’t want to know the truth and would rather be an ostrich.
Book of Odds: You said micro-expressions are cross-cultural. But there are other indicators that are not universal, right?
Paul Kelly:
For example, there are some cultures where if you look an authority figure in the eye, even in a business environment, it’s an insult, seen as a challenge to their authority. In some cultures, if you’re a policeman interviewing someone, they would never look you in the eye. Whereas in our culture we consider it polite, and showing sincerity, to look someone in the eye when you’re talking to them.
On your site you have an article on Tiger Woods, and it shows a great photo of somebody with their hand behind their back and their fingers crossed. In the interviewing world we call that a cultural emblem, and cultural emblems can tell you an awful lot. Because they’re culture-specific you have to be careful how you use them. In our culture a thumbs-up is good, in others it’s an insult. V is for victory when your palm is facing the other person, but if you turn your wrist around and show those same two fingers with the palm facing you, that’s “giving the finger” in the UK.
I also read your article on the No Fly List. The Ekman Group trained over 1,000 TSA people for airport behavior observation. The consequences of a mistake vary depending on the arena. With the TSA, while the odds of them encountering a terrorist might be pretty low, what’s the consequence if they’re wrong? A plane could go down, with hundreds of people killed.
Ekman talks about the “Othello error.” In Shakespeare’s play, Othello makes both of the two fundamental mistakes an interviewer can make. He believes the liar, Iago; and when he confronts Desdemona he refuses to recognize the truth—he sees fear and thinks it’s because she’s guilty, but it’s actually because she knows he’s not going to believe her. That’s another cornerstone of Ekman’s training: an experienced interviewer should place as much emphasis on recognizing the truth as on detecting deception, and unfortunately too often that’s not the case. But if you can recognize the truth, that’s going to help you separate the wheat from the chaff.
I should mention that at least the basics are easy to learn, and you can buy an online subscription that will let you access Ekman’s software. But then you have to decide what you do with this. Once you start to be more attuned to non-verbal indicators, especially the micros, you start seeing it all the time. The question then is, especially in a social setting, where do you want to go with this?
The short version
In a 2006 review of studies with 24,483 judges, people caught 47% of lies and were right 54% of the time overall, a little better than a coin flip. In 2010, Paul Kelly answered Book of Odds’ questions on micro-expressions, polygraphs and interviewing; his answers appear above as he gave them.
Last checked: 2026-10-22, against the Bond and DePaulo 2006 meta-analysis (PubMed 16859438 and the Crossref record). This article was machine-checked against the sources above and has not been expert-reviewed. Spotted an error or a newer public figure? Please tell us on our contact page with the article link and we will review it.
Sources
- S1 Charles F. Bond Jr. and Bella M. DePaulo. Accuracy of Deception Judgments. Personality and Social Psychology Review 10(3), 214 to 234, 2006. https://doi.org/10.1207/s15327957pspr1003_2 Abstract: https://pubmed.ncbi.nlm.nih.gov/16859438/ Accessed Oct 22, 2026.
- S2 Book of Odds. Interview with Paul Kelly, as published in 2010 (Wayback Machine capture of June 3, 2010).
