Network density

How density is calculated — and what counts as high

For a network of n people, the maximum number of undirected ties is n(n−1)/2. Density is simply existing ties divided by that maximum, giving a value between 0 (nobody knows anybody) and 1 (a complete clique). For ego networks — the network around one person, which is what matters for personal relationship management — you typically exclude yourself and ask what fraction of your contacts know each other. Two regularities are worth knowing. First, density falls mechanically as networks grow: in a village of 50 everyone can know everyone, in a city of 50,000 they cannot, so comparing densities across very different network sizes is misleading. Second, real ego networks are 'lumpy' — density is high within clusters (your family is probably a near-clique) and near zero between them. That lumpiness is why a single average can hide the most important structural facts, and why analysts usually look at density per cluster plus the bridges between clusters.

Coleman vs. Burt: the closure–brokerage trade-off

The deepest debate in network sociology runs straight through density. Coleman's closure argument (1988) says dense networks create social capital: when your contacts know each other, obligations are enforceable, information about trustworthiness circulates, and cooperation becomes safe. Ronald Burt's structural-holes program says the opposite: his studies of managers found that compensation, promotions and good ideas concentrate among people whose networks are sparse — rich in connections to groups that do not otherwise touch. The resolution most scholars accept is that the two describe different goods. Closure produces trust, support and safety; brokerage produces information, opportunity and innovation. Which you need depends on the task: raising children and surviving crises reward closure, while finding jobs, customers and ideas rewards brokerage. For an individual the implication is architectural — keep your inner circle dense and your outer network deliberately spread across worlds, rather than letting every layer drift toward the same comfortable cluster.

Reading the density of your own network

A rough density audit takes twenty minutes and no software. List your 30 most important people, then for each pair you can think of, mark whether they know each other. You will see the clusters immediately — and, more usefully, the gaps. Questions worth asking: Does my emotional support all come from one dense cluster (a risk if that cluster is also my employer)? Which two clusters would benefit from an introduction I could make? Who is my only link into an entire world, and when did I last talk to them? The audit stays alive rather than one-off if your contact notes record the circles people belong to and who introduced whom: then your dense cores and your fragile single-thread bridges remain visible at a glance. Whether Endearist, our own app in development, will offer such a relationship map is not decided; its planned core is private per-person notes stored on your device.

Local-first vault, fully offline: not publicly available yet. Status: planned for v1.