In Defense of the Random Selection of Research Projects