Recently, Keane et al. reported the discovery of a fading radio transient following FRB 150418, and interpreted it as the afterglow of the FRB. Williams \& Berger, on the other hand, suggested that the radio transient is analogous to a group of variable radio sources, so that it could be a coincident AGN flare in the observational beam of the FRB. A new observation with VLA showed a re-brightening, which is consistent with the AGN picture. Here, using the radio survey data of Ofek et al., we statistically examine the chance coincidence probability to produce an event like the FRB 150418 transient. We find that the probabilities to produce a variable radio transient with at least the same variability amplitude and signal-to-noise ratio as the FRB 150415 transient, without and with the VLA point, are \(P_1 \sim 6 \times 10^{-4}\) and \(P_1 \sim 2 \times 10^{-3}\), respectively. In addition, the chance probability to have a fading transient detected following a random time (FRB time) is less than \(P_2 \sim 10^{-2.9\pm 1.3}\). Putting these together and assuming that the number of radio sources within one Parkes beam is 16, the final chance coincidence of having an FRB 150418-like radio transient to be unrelated to the FRB is \(< 10^{-4.9\pm1.3}\) and \(< 10^{-4.4\pm1.3}\), respectively, without and with the VLA point. We conclude that the radio transient following FRB 150418 has a low probability being an unrelated AGN flare, and the possibility of being the afterglow of FRB 150418 is not ruled out.