Branching Process

A branching process is a stochastic model of population growth in which each individual produces a random number of offspring according to a fixed probability distribution, then dies. Introduced by Francis Galton and Henry Watson in 1874 to study the extinction of family names, it has become a fundamental tool in probability theory, epidemiology, and the analysis of cascading failures.\n\nThe critical insight is the extinction theorem: if the mean number of offspring is ≤ 1, the population dies out with probability one; if the mean exceeds 1, there is positive probability of infinite survival. This sharp threshold mirrors the percolation threshold in network science: below criticality, clusters are small and finite; above it, a giant component may emerge. The connection is precise — component size distributions in the Erdős–Rényi model near p = 1/n are described by a Galton-Watson branching process.\n\nThe branching process is the simplest model in which randomness at the microscale produces a sharp macroscopic threshold between extinction and explosion. That this threshold reappears in percolation, epidemics, and network connectivity is evidence these systems share a universal critical structure — one we understand precisely because the branching process is simple enough to solve completely.\n\n\n