In this paper I argue that empirical analysis is generally considered
to be easy and thus not worth teaching or explaining but that it is in
fact difficult and requires a place in our curricula.  I then discuss
how I planned to include coverage of this topic through lectures,
discussions and practical work and the approach that I took in the
Honours Analysis of Algorithms topic at the University of the
Witwatersrand in 2001.  I report on what I feel are the successes and
failures of the approach taken; discuss the results of a survey
conducted on the Honours class; and suggest some improvements for
2002.