50 Chapter 2 Using the chi-square test against an encrypted text would allow certain inference to be made, but only where the contents, or the type of contents (random or of an expected distribution), of the text were known. For example, someone may use a program that encrypts files. By creating the null hypothesis that the text is completely random, and by reversing the encryption steps, a block cipher may emerge as the null hypothesis is disproved through the chi-square test. This would be done by reversing the encryption method and XORing against the bytes with a block created from the known text. At the point where the non-encrypted text matches the positioning of the encrypted text, chi-square would reveal that the output is not random and the block cipher would be revealed. Chi-squared ¼ . . . ð observed À expected Þ 2= ð expected Þ Observed would be the actual zero/one ratio produced by XORing the data streams together, and expected would be the randomness of zeroes and ones (50/50) expected in a body of pseudorandom text. Independent of having a portion of the text, a large body of encrypted text could be reverse encrypted using a block size of all zeroes; in this manner it may be possible to tease out a block cipher by searching for non random block sized strings. Modern encryption techniques generate many, many block cipher permutations that are layered against previous iterations