Average case analysis is a method used in algorithm analysis to determine the expected performance or resource usage of an algorithm under a probabilistic distribution of all possible inputs. It provides a more realistic measure of an algorithm's efficiency compared to worst-case analysis, especially when certain inputs are more likely than others.