What is Dunbar's number?
The name refers to a proposed group-size estimate around 150. It is not a tested personal ceiling or a required number of friends. Hill & Dunbar (2003) studied Christmas-card networks in a Western sample. Their observations do not validate this calculator or establish how many supportive relationships you personally can maintain.
Zhou et al. (2005) combined published group-size data and examined approximate hierarchical patterns. The paper notes differences in network definitions, sample sizes and individual variation. Group patterns are not personal quotas or evidence for our input multipliers.
How this editorial model works
This is an unvalidated editorial planning estimate. The six starting values are 5, 15, 50, 150, 500 and 1,500. The age, work, self-rating, children, partnership, move and commitment factors are product choices. They were not fitted to a validation sample, and the cited papers do not supply these exact multipliers.
model circle[i] = round(baseline[i] × ageFactor × workFactor × selfRatingFactor[i] × childrenFactor[i] × partnershipFactor × moveFactor × commitmentFactor)The self-rating effect is dampened for the first two circles and amplified for the last two. The children factor applies from the third circle onward. Only the circle starting at 150 has its multiplier clamped between 0.28 and 1.50; its rounded result is the headline model value. These bounds prevent product outliers, not unhealthy networks.
A comparison notice appears when at least two of five input categories apply — the work category 41–60 or 61+, a move category within a year, self-rating 3 or below, children aged 0–5, or the 65+ age category — and the headline value is below 110. Those are editorial thresholds, not validated risk factors, a diagnosis or proof that you should have five close friends.
What the circles mean here
The inherited labels are retained for compatibility. In this tool they are model labels, not classifications of your actual contacts, guarantees of support or measured closeness. The circle values are cumulative examples; do not add them to count distinct people.
- Support Clique (baseline 5): An editorial innermost-circle example; you decide who belongs there.
- Sympathy Group (baseline 15): A nearby-circle example.
- Close Friends (baseline 50): A wider-circle example.
- Good Friends (baseline 150): The circle used for the headline model value.
- Meaningful Contacts (baseline 500): A broad-contact example.
- Acquaintances (baseline 1,500): A still wider-contact example.
Use the estimate for reflection
You can change inputs to see how the model behaves. A higher value does not show better social ability; a lower one does not diagnose overload, loneliness or an unhealthy network. The calculation cannot predict recovery after a move, assign a partner a mandatory place or prescribe contact frequency.
Optional next steps: write down people you want to contact, consider your actual available time and ask what contact each person prefers. There is no need to fill every circle, prune your contacts to the output or wait for a fixed recovery period.
Answers used for calculation stay in the browser, but calculating also places their encoded values in the page URL. Encoding is not encryption. Anyone with the link can read them, and URLs may appear in browser or server logs. The email form is a separate launch-list subscription.
Join the launch list: one email when it opens. Platforms, plans and capabilities need their own approval; signup creates no app access or purchase.