b) In the process of tting a distribution, it is common that after the last step of the tting, one is not convinced by the t and reconsiders which distribution is actually relevant. For example, exponential distributions were chosen as a t but actually all ts are poor because the tail is heavier than any exponential distribution known/tested. Explain your strategy for dealing with the unsatisfactory t, in terms of which class of distribution you would select, how would you choose it, and why