At the beginning of 2015, in an inductive investigation of four entrepreneurship-related data sets (N=12,000+), Crawford, Aguinis, Lichtenstein, Davidsson, and McKelvey discovered power law distributions in all resource-, cognition-, action-, and environmental-based input variables, as well as all revenue-, employee-, and growth-based outcome variables. Given the large number of outliers in these distributions, the findings call for new theory and method to identify the mechanism that generates all observations. Grounded in a complexity science paradigm, I propose the mechanism to be self-organized criticality—where thresholds in the distribution have the potential to produce large, nonlinear, and cascading effects—and empirically demonstrate that outliers are most likely to emerge when venture endowments and expectations for future growth are above critical values.