In a competitive environment it is important to bring research project
s to an early successful conclusion. A stochastic model of a typical p
harmaceutical research project shows that it is often very desirable t
o allocate more scientists to a project than the number which is most
efficient in terms of progress per scientist-year.