It is very important to improve the electrical conductivities of polymer nanocomposites (PNCs) as this can widen their application. In this work, by employing a coarse-grained molecular dynamics simulation, we investigated the effect of the amphiphilic diblock copolymer (BCP)-mediated nanoparticle (NP) on the conductive probability of polymer nanocomposites (PNCs) in the quiescent state and under a shear field. The conductive probability of PNCs first increases and then decreases with increasing content of BCPs while, interestingly, it exhibits an N-type dependence on the A-Block-NP interaction. Furthermore, the conductive probability shows a non-monotonic dependence on the fraction of A block (fA) in the BCPs, which reaches the maximum value at moderate fA. Under the shear field, NPs self-assemble to form the sandwich-like structures in the matrix above a critical concentration of BCPs, which leads to the anisotropic conductive probability of PNCs. In addition, the sandwich-like structures of NPs will be broken down at a high shear rate, which reduces the difference of the directional conductive probabilities. Last, the mechanism of the formation of the sandwich-like structures of NPs is discussed. In summary, this work presents a simple method to control the conductive network formation, which can help to design PNCs with high electrical conductivity, and especially anisotropy.