Have you read the article we're commenting on? Obviously mathematicians know that n! * k, with k the length of the code, would work. The idea is to superpose the end of a code with the start of the next, so you don't have to test nearly as many combinations.