# Network centralization > Centralization measures how much a network is organised around one person. Freeman's degree centralization, how to read it, and a worked example. Source: https://netgraf.isik.co/learn/network-centralization Updated: 2026-09-24 Network centralization measures how much a whole network is organised around a single person. It compares the most central person with everyone else: a star, where one hub touches everyone, scores 1, and a network where everyone has the same number of ties scores 0. ## How it's calculated ``` centralization = sum over v of (max degree - degree(v)) / ((n - 1)(n - 2)) ``` Freeman's degree centralization for an undirected network. The bottom line is the value for a star of the same size, the most centralised shape possible. Linton Freeman defined centralization in 1978 as the network-level partner of each centrality measure: there is a betweenness centralization and a closeness centralization too. Degree centralization is the one most often reported. ## Reading it Netgraf reads the number out rather than leaving you with a decimal: below 0.2 is spread fairly evenly, up to 0.45 moderately hub-driven, up to 0.7 strongly hub-driven, and above that almost a single hub. A highly centralised team is fast when the hub is available and stuck when they are not. ## In Netgraf The **Who carries it** card in the Insights panel gives the centralization and the share of all ties held by the busiest tenth of people, and the one-line summary at the top of the panel opens with it. ## Questions ### Is a centralized network bad? Not necessarily. Centralized networks coordinate quickly on simple tasks. They are fragile when the hub leaves or is overloaded, and they tend to do worse on complex problems that need many people to share information directly. ## Example maps - https://netgraf.isik.co/m/zachary-karate-club ## References - Freeman, L. C. (1978). Centrality in social networks: Conceptual clarification. Social Networks, 1(3), 215-239. https://doi.org/10.1016/0378-8733(78)90021-7