We present a detailed evaluation of the expected rate of joint gravitational-wave and short gamma-ray burst (GRB) observations over the coming years. We begin by evaluating the improvement in distance sensitivity of the gravitational wave search that arises from using the GRB observation to restrict the time and sky location of the source. We argue that this gives a 25% increase in sensitivity when compared to an all-sky, all-time search, corresponding to more than doubling the number of detectable gravitational wave signals associated with GRBs. Using this, we present the expected rate of joint observations with the advanced LIGO and Virgo instruments, taking into account the expected evolution of the gravitational wave detector network. We show that in the early advanced gravitational wave detector observing runs, from 2015-2017, there is only a small chance of a joint observation. However, as the detectors approach their design sensitivities, there is a good chance of joint observations provided wide field GRB satellites, such as Fermi and the Interplanetary Network, continue operation. The rate will also depend critically upon the nature of the progenitor, with neutron star-black hole systems observable to greater distances than double neutron star systems. The relative rate of binary mergers and GRBs will depend upon the jet opening angle of GRBs. Consequently, joint observations, as well as accurate measurement of both the GRB rate and binary merger rates, will allow for an improved estimation of the opening angle of GRBs.