Why we misread our friends

Play a game where you have to guess what your friends think, and a pattern shows up fast. The people you've known for ten years are not much easier to read than the ones you met last month. Sometimes they're harder.

That feels wrong, but it's one of the better-replicated ideas in social psychology. We tend to overestimate how well we know the people closest to us. Here are the main reasons, the studies behind them, and what you can do about it, whether you're playing a party game or just trying to understand someone better.

1. We assume others think like us

In 1977, Stanford psychologist Lee Ross and his colleagues asked students whether they would walk around campus for half an hour wearing a large sandwich board with a message on it, such as "Eat at Joe's". Some agreed and some refused. Then each student estimated how many other students would make the same choice.

The result: students who agreed thought most people would agree, and students who refused thought most people would refuse. Both groups saw their own choice as the normal one. Ross named this the false consensus effect, and it has been found again and again since, for opinions, habits and preferences of all kinds.

In a guessing game, false consensus is the biggest trap there is. If you think "obviously split the bill on a first date", you'll tend to assume your friends think so too. Your own answer leaks into your guess for everyone else.

A useful habit: when your gut says "they'd obviously pick the same as me", treat that as a reason to think again, not as evidence.

2. Closeness makes us lazy listeners

In 2011, a team led by Kenneth Savitsky at Williams College, with Nicholas Epley and Boaz Keysar at the University of Chicago, tested how well married couples understood each other. Partners sat back to back and tried to interpret ambiguous everyday phrases, each of which could mean several things. They also did the same task with strangers.

The couples expected to understand each other much better than they understood strangers. They didn't: their accuracy with their spouses and with strangers was, as Savitsky put it, "statistically identical". Follow-up experiments with friends found the same pattern. The researchers called it the closeness-communication bias: with people we know well, we feel understood, so we stop checking whether we actually are.

Epley summed it up as "an illusion of insight". With a stranger, you know you don't know them, so you pay attention. With a best friend, you think you already know.

3. Confidence grows faster than accuracy

A 1997 study by William Swann and Michael Gill at the University of Texas at Austin looked at dating couples and roommates. Participants predicted how their partner would describe themselves, and rated how confident they were in each prediction.

The longer people had been together, the more confident they were. But the length of the relationship had little to do with how accurate their predictions actually were. Time together made people feel like experts on each other without making them much better at the task.

If you've ever been completely sure what a friend would pick and watched them choose something else, this is the feeling those researchers measured.

4. We start from ourselves and don't move far enough

How do people imagine someone else's point of view? Research by Epley, Keysar, Van Boven and Gilovich (2004) suggests we start from our own perspective and then adjust, and we usually stop adjusting too early. They called this egocentric anchoring and adjustment. When people were rushed, their judgments stayed even closer to their own view.

That matters in any timed game. Under a ticking clock, you have less time to adjust away from your own answer, so your guesses drift towards what you would pick.

5. Sometimes the "obvious" answer wins for a reason

Not every bias works against you. In his 1960 book The Strategy of Conflict, economist Thomas Schelling asked people where and when they would try to meet a stranger in New York if they couldn't communicate in advance. Most chose the same answer: noon at the information booth at Grand Central Terminal.

Schelling called these shared, obvious answers focal points. When people have nothing else to go on, they gravitate to the option that stands out. In a guessing game, focal points are your fallback: for a player you know nothing about, the most famous name or the middle-of-the-road option is often the best bet. "Up to a point" is a focal point. So is Michael Jordan.

The skill is knowing when to use the focal point and when someone is the kind of person who'd deliberately avoid it.

How to read people better

None of these biases go away just because you know about them, but a few habits help:

1. Separate your answer from your guess. Decide what you'd pick, then deliberately ask "would they?". Treat the two as different questions.
2. Think about what they've done, not what they've said. A friend who says they're relaxed about money but always splits the bill to the cent has told you something.
3. Use recent evidence. People are more consistent within a topic than across topics. If someone picked "Tax them heavily" on billionaires, that's a strong signal for every money question in the same game.
4. Default to the focal point for people you don't know. With a stranger, the most popular or most moderate answer is usually the best guess.
5. Ask after the reveal. The closeness-communication research suggests we stop checking with people we know. A wrong guess is a good excuse to ask "wait, why that?" and actually update.

Try it yourself

All of this is easy to test. Pick a few questions with four possible answers (our 60 debate questions are a good start), have everyone answer privately, and guess each other's picks before revealing. Or play Mind Raider, which does the scoring for you: everyone picks, everyone guesses everyone, and you get 40 points for every friend you read right.

You might find you know your friends less well than you thought. That's not a bad thing. It's a reason to keep asking.