0
0
You know the moment. Someone at the table says something slightly off — a pause, a look, a word that lands wrong — and you sit up.
It’s him. It’s been him the whole time.
You are so certain that you say it out loud, to nobody, in an empty room.
Then the votes come in, the wrong person is sent home in tears, and you find out you were not even close.
If you have ever wondered how to spot a liar, that gap — between how sure you felt and how wrong you were — is the entire problem, sitting in one moment.
This is the engine of The Traitors, and it is about to run again. A new celebrity series is coming to BBC One and iPlayer this autumn, ten episodes, with Claudia Winkleman back in the coat. The format has already been remade in Britain, the United States, Australia and well beyond, and it works the same way everywhere. A group of strangers. A few secret traitors hidden among them. Every day, everyone sits in a circle and votes someone out based on almost nothing.
And here is the strange part. Nearly everyone watching believes they would be good at it.
Why the Game Feels So Winnable From the Sofa
The show is built to make you feel clever. You get information the players do not have. You see the traitors meeting at night. So when someone at the breakfast table stumbles over a sentence, you are not really reading them — you are matching them against an answer you already have.
Take the answer away and the whole thing collapses. That is why players who seemed obvious to you on screen sit undetected for weeks in the room.
But there is something more interesting going on than a television trick. The game asks people to do a specific thing: look at a face, listen to a voice, and decide whether that person is lying. Researchers have been testing exactly that for decades, and the results are not what the show trains you to expect.
How to Spot a Liar: What the Research Actually Found
In 2006, two researchers, Charles Bond and Bella DePaulo, pooled together the whole field. Their review in Personality and Social Psychology Review gathered 206 documents covering 24,483 people who had been asked to judge whether someone was telling the truth.
Across all of it, people got it right 54% of the time.
A coin gets 50%. So after decades of study, the average person is barely better than a coin toss at telling truth from lies.
That figure gets quoted a lot. But it hides the finding that actually matters, and almost nobody repeats this part.
Split the 54% in two. When people were shown a lie, they correctly called it a lie 47% of the time. When they were shown the truth, they correctly called it the truth 61% of the time.
Read those two numbers again. On truths, people beat the coin. On lies — the one thing the game is actually asking you to catch — people did worse than the coin.
We are not mediocre lie detectors. We are decent honesty detectors who are below chance at the other half of the job.
Why We Feel So Sure Anyway
Researchers have a name for the reason: truth bias. Left to ourselves, we mostly assume people are telling the truth. Most of the time, that assumption is correct, because most of the people around us on most days are not lying.
So our confidence is built on a real track record — just not the one we think. We are right about people constantly. We are right about the honest ones. And that feeling of being a good judge of character comes along for the ride when we finally meet someone who is lying.
It is worth being careful with the 54% figure, though, and the researchers who work on this are careful with it themselves. A 2025 re-examination by Timothy Levine and Kim Serota argues that headline accuracy numbers like this one need to be read against how often lies actually occur in a given situation, because the same tendency that makes people accurate about truths makes them inaccurate about lies. The single number is real. It is also doing a lot of hiding.
Which lands somewhere uncomfortable. The Traitors hands its players the one task the evidence says people are worst at, then asks them to act on it, publicly, with someone’s place in the game at stake. No wonder the circle so often gets it wrong.
What Tends to Work Better Than Reading a Face
The research points away from demeanour. Squirming, stammering, looking away, touching the face — these read as guilt and mostly are not. Nervous people look nervous. Honest people under suspicion look exactly like liars, because being accused is genuinely stressful.
What holds up better is duller and slower:
- Go to what can be checked, not what can be felt. A story either fits with something else you know, or it does not. A feeling about someone’s eyes does not.
- Let people talk longer. More detail gives you more that can be compared against everything else. Short interrogations mostly produce short answers and confident guesses.
- Notice when you decided. If you settled on someone in the first minute, you were not gathering evidence after that. You were collecting reasons.
That last one travels well past the television. Most of us have quietly decided about someone — a colleague, a relative, a friend who went distant — and then spent months reading everything they did through that decision. If you have ever wondered why you always seem to be the one reaching out, the research on what silence usually means is a useful corrective to the story we tend to tell ourselves.
If you want a rough read on how well you actually track other people, our free emotional intelligence assessment takes a few minutes and is more honest than the confidence you feel while shouting at a television.
The Part the Game Cannot Show You
In The Traitors, accusation is the only move available. You cannot ask someone a real question. You cannot wait. You cannot decide the relationship matters more than being right. The format removes every option except pointing across the table, which is exactly what makes it good television and exactly what makes it nothing like your life.
Because in your life, suspicion has a cost the game never charges. Every hour spent deciding who cannot be trusted is an hour not spent being someone worth trusting. And we have become fluent at naming what other people are doing to us — a whole vocabulary for it, which is part of why so many people have gone off therapy-speak — while getting no better at the older, harder question.
There is an account of a dinner, told and retold for a very long time, that turns this exactly inside out.
A man sits down to a last meal with twelve close friends and tells them, plainly, that one of the people at that table is going to betray him.
Now picture the circle. Picture what The Traitors would do with that sentence. The accusations, the alliances, the eyes going round the room looking for the flicker.
None of that happens. Not one of the twelve points at anyone else. Every single one of them turns the question inward and asks the same four words: Lord, is it I?
Not who is it. Not I knew it. Just — could it be me?
And the detail that keeps the story from being tidy: the one who had already arranged it asked it too. The question does not sort anybody. It just leaves twelve people looking at themselves instead of at each other, which may be the only honest thing a room full of people can do with a suspicion.
The Question Worth Taking Off the Sofa
You will not get better at spotting liars. Sixty years of research is fairly settled on that, and the small edge available goes to people who check things rather than people who read faces.
But there is a second question sitting underneath the first one, and it is the one you can actually answer.
Next time you are in a room you do not fully trust — a workplace, a family group chat, a friendship that has gone quiet — you already know the game version of the question. Which of them is it.
Try the older one instead. It is shorter, it is harder, and unlike the other one, you are the only person alive who can answer it honestly.
Then go back to the television, where the shouting is free and nobody has to live with the vote.
A Question Worth Discussing
Does watching a show built entirely on suspicion make people more suspicious in real life — or does it get it out of our system safely? Tell us where you land in the comments; people genuinely disagree about this one.
Share This
- People catch lies 47% of the time. A coin does better. We’re not bad lie detectors — we’re good honesty detectors doing the wrong job. Explains a lot about shouting at The Traitors.
- The best-known number on lie detection is 54% accuracy. Split it and it gets worse: 61% on truths, 47% on lies. We’re below a coin toss at the only half that matters.
- At the most famous betrayal dinner ever told, one man says “one of you will betray me” — and not a single person at that table points at anyone else. Every one of them asks whether it’s them. Try running that in a group chat.
Questions People Ask
How can you tell if someone is lying?
Not reliably from their behaviour. Research on deception detection has found that common signals people trust — fidgeting, avoiding eye contact, stammering, touching the face — do not separate liars from honest people, because being under suspicion is stressful for anyone. What performs better is checking a claim against outside information, letting a person give a longer and more detailed account so there is more that can be compared, and noticing how early a judgment was made. Someone who decided in the first minute spends the rest of the conversation collecting reasons rather than evidence.
Are people actually good at detecting lies?
No. A 2006 review by Charles Bond and Bella DePaulo in Personality and Social Psychology Review pooled 206 documents covering 24,483 people asked to judge truthfulness, and found an average accuracy of 54% — barely above the 50% a coin would produce. The average is also uneven: people correctly identified 61% of truths but only 47% of lies, meaning accuracy on lies specifically fell below chance.
What is truth bias?
Truth bias is the tendency to assume that what someone says is true. It is why people score better at recognising honesty than at catching lies. In everyday life the bias is usually well-founded, since most statements most people hear are not lies, and it produces a real record of being right about people. That record is what makes many people feel like good judges of character even though their accuracy at detecting an actual lie is close to chance. A 2025 re-examination by Timothy Levine and Kim Serota argues that overall accuracy figures should be read against how frequently lies actually occur in a given setting, because truth bias raises accuracy on truths while lowering it on lies.
When does the new series of The Traitors start?
A new celebrity series of The Traitors is scheduled for BBC One and BBC iPlayer in autumn 2026, running to ten episodes and presented by Claudia Winkleman. An exact premiere date had not been announced at the time of writing. Earlier series of the UK version have also been made available to viewers outside Britain through streaming platforms, and separate versions of the format are produced in the United States, Australia and other countries.
Why is The Traitors so addictive to watch?
Viewers are given information the players do not have, including seeing which contestants are secretly traitors. Every hesitation or awkward answer on screen is then matched against an answer the viewer already holds, which feels like skilled deduction rather than hindsight. The format also removes every response except public accusation — players cannot wait, investigate, or set the question aside — so pressure builds continuously and every episode ends in a decision someone will get wrong.
