A number of observables related to the $b to s l^+ l^-$ transition show deviations from their standard model predictions. A global fit to the current $brightarrow sl^+l^-$ data suggests several new physics solutions. Considering only one operator at a time and new physics only in the muon sector, it has been shown that the new physics scenarios (I) $C_9^{rm NP}<0$, (II) $C_{9}^{rm NP} = -C_{10}^{rm NP}$, (III) $C_9^{rm NP} = -C_9^{prime rm NP}$ can account for all data. In this paper, we develop a procedure to discriminate between these scenarios through a study of the branching ratio of $B_s to mu^+mu^-$ and the distribution of $Bto K^*mu^+mu^-$ decay in the azimuthal angle. The scenario II predicts a significantly lower value of $mathcal{B}(B_sto mu^+mu^-)$ and can be distinguished from the other two scenarios if the experimental uncertainty comes down by a factor of three. On the other hand, a precise measurement of the CP averaged angular observables $S_3$ and $S_9$ in high $q^2$ bin of $Bto K^*mu^+mu^-$ decay can uniquely discriminate between the other two scenarios. We define two azimuthal angle asymmetries, proportional to $S_3$ and to $S_9$ respectively, which can be measured with small statistical uncertainty.