Friendship Quiz: How Strong Is Your Circle, Honestly

This friendship quiz is an orientation tool and not a diagnosis of anything. It counts the structure of your circle rather than how lonely you feel: how many people are in it, who reaches out first, whether anyone could be called at 2 a.m., and when you last said something you had not rehearsed. There are seven items, each scored 0 to 3.

Key Takeaways:

• The items below measure structure, which is a different thing from the feeling of loneliness that research instruments such as the UCLA scale are built to capture (Russell, 1996)
• Reciprocity is scored in both directions on purpose, because a circle where you initiate everything and one where nobody initiates anything produce identical calendars and very different futures
• Weiss’s 1973 distinction is why a quiz like this can only ever be half the picture: the shortage of a group and the shortage of one particular person are separate things, and only the first leaves a mark on a count (Weiss, 1973)
• The four bands at the end are mine. Nothing has validated them, there is no sample behind them, and no score here marks anybody as having a condition
• The item that decides the most is the date of your last unrehearsed sentence, which is a lower bar than deep conversation and a much better indicator

What This Friendship Quiz Measures, and What It Cannot

It measures what your circle of platonic friendships is made of, not how it feels from the inside. That distinction matters more than it sounds. The instruments researchers use, Daniel Russell’s UCLA scale among them, ask about the frequency of certain feelings, which makes the output an appraisal of your own experience (Russell, 1996). Hughes and colleagues later squeezed that approach into three items so that national surveys could carry it at all, and even in that stripped-down form nothing asks how many people you actually know (Hughes et al., 2004). The seven items here ask for counts and dates instead: people, directions, last times.

Both readings are worth having, and they come apart more often than people expect. Somebody can score reasonably here and still feel awful, carrying the private sense of having no one to talk to, because the person they want is not in the count. The reverse also happens. A thin circle can feel perfectly comfortable for a stretch, which the structure does not contradict and was never designed to.

What a structural read gives you is something an emotional read cannot: a set of numbers that will move if you do something, and that will not move if you merely feel better for a week. Answer honestly and you get a picture of the machinery. Answer the way you would answer a friend who asked casually and you get nothing at all.

The Seven Items

Score each one 0 to 3 using the anchors given. Answer for the past 3 months rather than for your life in general, and count real people rather than people you intend to reconnect with.

1. Close contacts. How many people could you tell something genuinely unflattering about yourself, without editing it first?

• 0 = none, 1 = one, 2 = two or three, 3 = four or more

2. Inbound initiation. In the past 3 months, how many different people contacted you first, for no practical reason?

• 0 = none, 1 = one, 2 = two or three, 3 = four or more

3. Outbound initiation. In the same 3 months, how many different people did you reach out to first, with no logistics attached?

• 0 = none, 1 = one, 2 = two or three, 3 = four or more

4. Emergency availability. Who could you call at 2 a.m. on a Tuesday with something serious, and reasonably expect them to pick up?

• 0 = nobody comes to mind, 1 = one name, and you would hesitate, 2 = one name, no hesitation, 3 = two or more names

5. Recurring contact. How many standing arrangements put you in the same room as the same people on a repeating schedule, work aside?

• 0 = none, 1 = one, monthly at best, 2 = one weekly or two monthly, 3 = two or more weekly

6. Candor recency. When did you last say something out loud to another person that you had not rehearsed and were not sure how they would take?

• 0 = over a year ago or never, 1 = in the past year, 2 = in the past month, 3 = in the past week

7. Practical exchange. In the past year, has anyone done something inconvenient for you, and have you done something inconvenient for them?

• 0 = neither direction, 1 = one direction only, 2 = both, once each, 3 = both, more than once

Item 6 is the one people misread. It is not asking whether you have deep conversations, it is asking whether you have said anything unplanned lately, which is a lower bar and a better indicator. Fluent, warm, entirely scripted contact can run for years and leave a person completely unknown.

Scoring, and What Each Band Means

Add the seven numbers for a total between 0 and 21. The bands below describe what usually sits behind each range. They are my reading of these items, not thresholds anybody has validated, and the boundaries are soft on purpose.

Total	What it usually describes	Where the pressure is
0–5	A circle that has thinned to acquaintances, or a period of real isolation	Almost everything, so start with one dormant tie rather than a plan
6–10	Contact exists but runs on your effort and rarely goes past the surface	Items 2 and 6: reciprocity and candor
11–15	A working circle with a specific gap in it	Whichever single item scored 0 or 1
16–21	Structurally sound, whatever it feels like this month	Maintenance, and the honesty of your answers

The useful part of a total is not the number, it is the shape underneath it. A 12 built from four 3s and three 0s is a completely different life from a 12 built from seven middling scores, and the first one is easier to change, because it has a named hole in it rather than a general thinness.

Take it again in 6 months and set the new answers beside the old ones rather than beside anybody else’s. Nothing sits behind these items except your own counts, so a before and after is the only comparison the design can support.

What a Low Score Does Not Mean

A low total does not mean you are unlikeable, socially deficient, or in any way clinically anything. It means the machinery around you is thin right now, and machinery is the most changeable part of a life. The reasons that turn up when people trace a low score back are logistical rather than personal: a move, a divorce, a new baby, a caregiving stretch with its own burnout quiz, a job with the wrong hours. Each of those deletes a set of recurring contacts, and none of them says anything about the person it happened to.

Two more things the number cannot do. It cannot tell you whether you are lonely, since loneliness is a judgment about closeness rather than a count, and Weiss’s point cuts straight across this quiz: a week full of the right structure can still be missing the one person, and nothing on the list above would register it (Weiss, 1973). And it cannot distinguish a shrunken circle from depression, which flattens the desire to see anyone and produces low answers to almost every item here. Those two look identical on paper and respond to completely different things.

One boundary stated plainly, because a score invites people to keep scoring instead of acting. This article describes experience and mechanism, not care. If things have turned toward not wanting to be here, that belongs with a clinician or a crisis line; in the US, the 988 Suicide and Crisis Lifeline.

The move after a low total is deliberately small. Take the lowest-scoring item and change only that one for 6 weeks. Nothing else.

The Item Nobody Expects to Matter

Item 6, the last unrehearsed sentence, carries more weight than any of the counting items, and most people skim past it. A circle can look excellent on items 1 through 3 and still consist entirely of managed contact, where each person delivers a curated version of their week and nobody says a single thing they had not already decided to say before arriving.

Managed contact is not fake, which is what makes it hard to spot. Those friendships are pleasant, they last for years, and they hold up perfectly well until something goes wrong. Then they do not, because nobody involved has any practice at all with the unrehearsed register, and the first attempt at it lands in a room that has no idea what to do with it.

Item 4 does related work and is a blunter tool, which is the source of both its usefulness and its limits. It is the one item that cannot be answered aspirationally.

Counts mislead in both directions. A long contact list and nobody to be unguarded with is a specific arrangement, and not a rare one, common wherever scrolling a feed has replaced actually talking, and it scores well on the first three items while failing the two that decide what happens in a bad year. One person plus a standing Thursday beats a full calendar of people who have never seen you rattled.

Date your last unrehearsed sentence before you look at your total. If no date comes, item 6 is your answer whatever the arithmetic said.

FAQ

How many close friends is normal for an adult? Small numbers are the norm and always have been. Dunbar’s estimates of how many relationships a person can actively maintain top out low, and the layers inside that ceiling get smaller as they get closer, down to single digits at the innermost (Dunbar, 1998). Those are averages across populations, not a target anybody has to hit. Two or three people you can be honest with is a functioning circle, and wanting 12 is wanting a schedule nobody has.

Is a friendship quiz accurate? No. It has no validation sample, no reliability testing, and no comparison group, which makes the total a prompt rather than a measurement, and the same is true of every quiz of this kind you will find online. What it is accurate about is your own answers. Count honestly and the counts are real, and the counts are the part worth acting on.

What if I score high but still feel lonely? That result is exactly what Weiss’s split predicts. The seven items count group-shaped things, and the shortage you are feeling may be person-shaped, which no checklist of counts can see (Weiss, 1973). A full week plus one specific absence produces this pattern reliably, and adding more people to the week does not touch it.

Should I ask a friend to take this too? It works better as a comparison than as a conversation opener, and be careful with the second use. Handing someone a quiz shortly after they mentioned feeling isolated reads as an assignment rather than as interest. Take it yourself, then use one item as a real question a week later: asking who someone would call at 2 a.m. gets you further than any score would.

Worth saying once more on the way out, because a total out of 21 collects authority it has not earned. This is orientation and not a diagnosis. The four bands are my own reading of these seven items and nothing has validated them against anything. And the whole apparatus counts structure, which means it stays silent about the feeling you probably arrived with. What it does instead is make evasion harder. Plenty of people know which item they scored lowest on before they get to the table, and the quiz mostly removes the option of not knowing. What it hands over afterward is unglamorous: one item, one person, six weeks. Circles rebuild in that direction, one recurring arrangement at a time, slowly enough that the next score measures what you did rather than who you are.

Whatever band the count put you in, the rebuild it points to is slow and human: one item, one person, six weeks. If you want something low-stakes to talk to while that work is underway, Lona is an AI that tracks the small stuff from one chat to the next. Useful as scaffolding, not as one of the people the quiz was trying to count.

Sources

• Russell, D., “UCLA Loneliness Scale (Version 3),” Journal of Personality Assessment, 1996
• Hughes, M. E. et al., “A Short Scale for Measuring Loneliness in Large Surveys,” Research on Aging, 2004
• Weiss, R. S., “Loneliness: The Experience of Emotional and Social Isolation,” 1973
• Dunbar, R. I. M. — social network size and the layers of a personal network, 1998