Why Who You Know Matters as Much as What You Know
A network is a collection of nodes (entities) connected by edges (relationships). The mathematics of networks — graph theory, developed by Euler in 1736 to solve the Königsberg bridge problem — seems abstract. But the pattern it describes governs how disease spreads, how information propagates, how social influence flows, how financial contagion occurs, and why some ideas go viral while structurally similar ideas die quietly.
The central insight of network theory is that the structure of connections matters as much as the properties of individual nodes. A city with great roads to nowhere is not as useful as a city with good-enough roads to everywhere. A person with 10 strategically placed contacts may have more influence than one with 100 poorly connected ones.
Core metrics
Degree is the number of edges a node has — how many direct connections. High-degree nodes are "hubs." In a social network, a hub is a person with many followers or friends. In a protein interaction network, a hub protein participates in many interactions.
Betweenness centrality measures how often a node sits on the shortest path between two other nodes. A node with high betweenness is a bridge — many connections between distant parts of the network route through it. Remove it, and the network fragments. This is the "broker" position: controlling information flow between groups that don't otherwise connect.
Clustering coefficient measures how interconnected a node's neighbors are. High clustering means the people you know mostly know each other — you're in a tight clique. Low clustering means your connections span different groups that don't connect to each other — a structurally advantageous position (Ronald Burt's "structural holes").
Path length is the average number of steps to get from any node to any other. The famous "six degrees of separation" (Milgram's small world experiment) found that the average path length between any two humans is approximately six. The internet has dramatically shortened this.
Scale-free networks
Many real networks — the web, citation networks, protein interaction networks, airline routes, social networks — follow a power-law degree distribution: most nodes have few connections, and a very small number of nodes have enormous numbers of connections. Barabási and Albert showed in 1999 that this "scale-free" structure arises naturally when networks grow through preferential attachment — new nodes are more likely to connect to already-well-connected nodes ("the rich get richer").
Scale-free networks have two striking properties. They are robust to random failure: removing random nodes rarely disrupts the network badly, because most nodes are low-degree and their removal has little effect. But they are highly vulnerable to targeted attack: if you identify and remove the hubs, the network rapidly fragments. This property affects everything from internet resilience (surprisingly robust to random router failure, vulnerable to targeted attack on major routing nodes) to epidemic control (random vaccination is inefficient; vaccinating hubs disproportionately rapidly reduces transmission).
Weak ties and structural holes
Mark Granovetter's 1973 paper "The Strength of Weak Ties" is one of the most cited papers in sociology. The finding: your close friends are likely to share your information and opportunities, because you all know the same people. Your acquaintances (weak ties) — people you know but don't interact with often — are more likely to bring genuinely new information, because they are embedded in different social clusters. Weak ties bridge structural holes.
The practical implication for career and innovation: most job opportunities are found through acquaintances, not close friends. Most cross-disciplinary innovations come from people who bridge communities that rarely interact. The most valuable network position is not maximum degree but maximum bridge — spanning communities that don't otherwise communicate.
Cascades and contagion
Network structure determines whether a contagion (disease, idea, financial crisis, social norm) will spread globally or die locally. The basic reproduction number R₀ — the average number of secondary cases caused by one case — is a network property as much as a pathogen property. In a highly connected network, even a weakly contagious pathogen can achieve pandemic spread. In a fragmented network, a highly contagious pathogen may be contained.
This is why superspreaders matter disproportionately in epidemics: a single high-degree node infected early can expose far more secondary cases than the average node, dramatically increasing effective R₀. Contact tracing is effective because it exploits network structure — tracing back to the source of an infection maps the actual transmission network rather than assuming random mixing.
Network effects and lock-in
A product or platform exhibits network effects when its value increases as more people use it. This is the network theory of value creation. Metcalfe's Law (the value of a network scales with the square of its users) implies that network effects are positive feedback loops — once a platform achieves critical mass, it becomes progressively harder to displace because the value it delivers to any individual user is a function of the entire user base, not just the product's intrinsic quality.
Quick answers
What is Network Theory?
Network theory reveals how the structure of connections — not just the quality of individual nodes — determines what flows, spreads, and survives.
Where does this concept come from?
The concept originates with Leonhard Euler (1736, Seven Bridges of Königsberg); modern network science: Duncan Watts & Steven Strogatz (1998).
