Hydraulic fracturing (fracking) technology in gas or oil shale engineering is highly developed last decades, but the knowledge of the actual fracking process is mostly empirical and makes mechanicians and petroleum engineers wonder: why fracking works? (Bažant et al., 2014, “Why Fracking Works,” ASME J. Appl. Mech., 81(10), p. 101010) Two crucial issues should be considered in order to answer this question, which are fracture propagation condition and multiscale fracture network formation in shale. Multiple clusters of fractures initiate from the horizontal wellbore and several major fractures propagate simultaneously during one fracking stage. The simulation-based unitary fracking condition is proposed in this paper by extended finite element method (XFEM) to drive fracture clusters growing or arresting dominated by inlet fluid flux and stress intensity factors. However, there are millions of smeared fractures in the formation, which compose a multiscale fracture network beyond the computation capacity by XFEM. So, another simulation-based multiscale self-consistent fracture network model is proposed bridging the multiscale smeared fractures. The purpose of this work is to predict pressure on mouth of well or fluid flux in the wellbore based on the required minimum fracture spacing scale, reservoir pressure, and proppant size, as well as other given conditions. Examples are provided to verify the theoretic and numerical models.