When people encounter a problem, they want to get help from others. We develop a social mutual aid network model which is based on Calvó-Armengol's job contact model, and players choose the number of links to prevent them from falling in isolation. Players make links under social information which is like calculating an expected value. Although we assume players' strategies are symmetric, there are different degrees in the equilibrium, besides they make the same number of links. This paper tries to connect random networks and strategic networks. The player contacts with others randomly, and he chooses the number of links with strategy. Finally, we find that the tight networks may generate inefficient results.