Almost all outcome variables in entrepreneurship show right-skewed long-tail distributions, rather than bell-shaped normal distributions that are commonly assumed in entrepreneurship research. However, it is still unclear how the long-tail distributions are generated in the entrepreneurial process. Hence, this study aims to explain the generative process of the distributions and the extreme outcomes on the right tails of the distributions by reproducing the empirical reality using agent-based modeling and simulation (ABMS). We also discern the roles of luck, skill, and environment in the entrepreneurial process by the simulation approach. The simulation results show that the long-tail distributions and extreme outcomes are reproduced by the mechanisms of preferential attachment and multiplicative effect, even without any individual or environmental variations. These results support the essential role of luck to generate the stylized fact of entrepreneurship.