There are numerous algorithms which efficiently solve the problem of finding occurrences of a pattern in a text. Well-known algorithms which do not preprocess the text include Knuth-Morris-Pratt and Boyer-Moore (and variations). These algorithms naturally abstract to a comparison based model of computation. We will discuss a lower bound on the number of comparisons for this class of algorithms derived by Galil and Giancarlo (1990) and refined by Cole, Hariharan, Patterson, and Zwick (1993).