We analyse the concept of strong overgeneral rules, the Achilles' heel of traditional Michigan-style learning classifier systems, using both the traditional strength-based and newer accuracy-based approaches to rule fitness. We argue that different definitions of overgenerality are needed to match the goals of the two approaches, present minimal conditions and environments which will support strong overgeneral rules, demonstrate their dependence on the reward function, and give some indication of what kind of reward functions will avoid them. Finally, we distinguish fit overgeneral rules, show how strength and accuracy-based fitness differ in their response to fit overgenerals and conclude by considering possible extensions to this work.
[This work has been subsumed by "Strength or Accuracy: Credit Assignment in Learning Classifier Systems" PhD Thesis, 2002. School of Computer Science. University of Birmingham. Birmingham, U.K. http://www.cs.bris.ac.uk/~kovacs/author.directory/thesis/thesis.html]