We compute and analyze the probability that a particular random variable will assume the smallest value among a set of random variables. This work was motivated by wanting to predict the "winner" in R&D and patent "races". Some general results and comparative statistics are provided. For symmetric distributions we derive bounds on the probabilities of interest. We compute the probabilities of who will be the smallest (fastest) for normal, lognormal, Weibull, Pareto, binomial and PH-distributions. We also analyze several models of multivariate exponential distributions.
|Number of pages||6|
|Journal||IIE Transactions (Institute of Industrial Engineers)|
|State||Published - Jun 2002|