How Well Do Friends Really Know Each Other?

Almost everyone who makes a quiz here expects their best friend to walk it. Almost nobody's best friend actually does. That gap is the whole point — and it's a lot more interesting than it is sad.

Everyone expects a 10. Nobody gets a 10.

There's a specific moment that happens when someone sends their quiz link to the one person they were sure would ace it. They wait. The score comes back. It's a 6. And there's this small, funny, slightly indignant beat of wait, we've known each other since we were fourteen.

It's tempting to read that as a failure — of the friendship, of the friend, of the quiz. I don't think it is. Ten multiple-choice questions is a bizarrely specific instrument. It doesn't measure whether someone would drive to pick you up at 2am. It measures whether a particular set of facts about you happens to be sitting in someone else's head, ready to be retrieved under pressure. Those are two very different things, and confusing them is how a fun quiz turns into a bad mood. So let's not confuse them.

Knowing someone isn't the same as being close to them

Closeness is about how a person makes you feel. Knowledge is about what you've stored. They correlate, sure — spend enough time with someone and facts accumulate by accident. But they're separate axes, and you can be maxed out on one while sitting at the middle of the other.

You can love someone for a decade and never learn their coffee order, because you've never been the one who ordered. You can know exactly how they'll react to bad news, what their face does when they're pretending something's fine, which topic to steer away from at dinner — and still have no idea what their favourite film is, because it has genuinely never come up. That's not a shallow friendship. That's a friendship with a different shape, one built out of moments instead of trivia.

The quiz only measures the trivia axis. It's honest about that if you let it be. A low score isn't a low friendship; it's a low overlap between what you know and what got asked.

What friends reliably get right

The surface layer holds up remarkably well, and it's worth understanding why. Friends nail the things you have said out loud, repeatedly, in front of them. Your loud favourites. The band you won't shut up about. The team. The comfort show you rewatch. The one food you're famous for refusing. Anything you've posted about is basically a freebie.

None of that is intimacy. It's repetition. A fact becomes knowable because you broadcast it, not because someone earned it. This is why the friend you see twice a year can score surprisingly well on your "favourites" questions — you've been publishing that information the whole time, and they were reading. The big obvious traits land for the same reason: if you're the loud one, everyone knows you're the loud one, including people who barely know you.

What friends reliably get wrong

Here's where it gets interesting. The misses cluster, and they cluster in a way that tells you something.

• Quiet preferences. The things you like but never announce, because announcing them would be weird. Nobody says "by the way, I prefer the aisle seat."
• Things you changed your mind about. Your friend is answering with the you from four years ago. They're not wrong, exactly — they're out of date, and you never issued a patch note.
• Your ick. Specific, vivid, deeply personal, and almost never volunteered unprompted.
• What you're actually proud of. People tend to be quiet about their real pride and loud about their fake modesty. The thing you'd pick is rarely the thing they'd guess.
• The version of you that picks the film when nobody's watching. The performed self chooses something defensible. The private self rewatches the same thing for the ninth time.

The common thread: none of this comes up in conversation. Not because you're hiding it, but because conversation is driven by what's happening right now, and "which of these four things would I grab in a fire" is not a thing that's happening right now. Your friends never got the data. You can't blame them for not having it.

Why a low score isn't an insult

If someone you love gets a 5, the honest reading is usually one of three things. They know an older version of you, one who was accurate at the time. They know a curated version — the one you bring to that particular friendship, which might be the funny one or the calm one or the work one. Or you simply never gave them the material, and they were guessing in good faith.

All three of those are fixable in about ten minutes of conversation, which is really what this thing is for. The best outcome of a quiz isn't the score. It's the message that comes after: wait, since when do you hate that? That's the game working exactly as intended. A 5/10 is a conversation starter with a scoreboard attached, not a verdict on anyone's character.

How to write a quiz that actually reveals something

Most quizzes are boring because they were written in ninety seconds. A little care makes an enormous difference. Some rules I'd stand behind:

• One unambiguous answer, always. If you'd hesitate between two options on your own question, your friends have no chance and the miss teaches nobody anything. "Favourite film" is often worse than "the film you've rewatched most" — the second has a real answer.
• Avoid anything deducible from your posts. If a stranger could get it from thirty seconds of scrolling, it's not measuring friendship. It's measuring literacy.
• Make the wrong options genuinely plausible. This is the one people skip, and it's the most important. If three options are obviously absurd, you've written a gift, not a question. Every wrong answer should be something a reasonable friend might actually believe about you.
• Mix easy wins with deep cuts. All-easy and everyone ties at 9. All-brutal and everyone ties at 3. You want spread, because spread is what makes a leaderboard mean anything.
• Never write a question you don't know your own answer to. If you're inventing your answer as you type, the question is fiction and the scoring is a coin flip.
• Resist making it all deep. This is the counterintuitive one. Ten earnest questions about your values are less revealing than one about which condiment you'd defend in court. The ridiculous questions are often the most diagnostic, because nobody has ever performed an opinion about condiments — so only people who actually pay attention will know.

If you're stuck, the question bank is a decent starting point — swap in what fits, then write two or three of your own. The custom ones are almost always where the good misses happen, because they're the only ones your friends couldn't have seen coming.

How to read the leaderboard

The thing that surprises people most is when the top scorer isn't the oldest friend. It's the coworker. It's the person you've known eight months. And the instinct is to feel weirdly guilty about it, like the leaderboard has ranked your relationships and got them wrong.

It hasn't ranked your relationships. It's ranked attention. Some people are simply collectors — they notice what you ordered, they remember what you said offhand in March, they file things. That's a real and lovely trait, and it's not the same as closeness. Meanwhile your oldest friend, the one who'd genuinely take a bullet, might be running on a beautifully detailed model of who you were in 2016.

So read the board as two overlapping things: who's been paying attention lately, and who has the most recent data. The winner usually has both. Your best friend usually has neither, and loves you anyway.

Go find out

The honest pitch for all this is small: you'll learn one or two genuinely surprising things about how you're seen, everyone will argue about question 7, and someone will text you at midnight to contest their score. That's a good evening. Make one, send it to the person you're sure will ace it, and enjoy being wrong.