Comparison between microlensing on quasars observed and simulated makes it possible to draw conclusions about physical parameters such as source size, quasar brightness profile, mass distribution of the lens, abundance and mass of microlenses, and relative transverse velocity between source and lens. By simulating magnification maps for different values of the parameters we intend to calculate the probability that a variable takes a particular value. One of the problems faced by this statistical study is that experimental errors and sources of variability other than microlensing can significantly affect data and results. To minimize this problem, we reduce the phenomenon of microlensing to a series of discrete events, the caustic crossings. If the source size is small (X-ray-emitting source), then each caustic crossing appears well resolved, would be of great amplitude (unaffected by measurement errors) and would be difficult to confuse with other types of variability. From the simulations we have calculated the probability that we are faced with a given distribution of mass in stars from the number of observed caustic crossings. We apply this study to the case of the quasar image D QSO 2237 + 0305.