Statistical code crackingapproaches, what do they work for? In what ways can they differentiate or not?

For instance, say you have a string of code that is built as follows:


GH%GVrddcdejcxˇ‰ÎÍÍÓ‰®!2221





and 1) Not all parts of the code are relevant


2)different parts of the code mean different things


3)an unknown variable is which parts of the code mean what, and this can vary based on the code


4)The code is combined with additional information in such a way that some information is not identifiable until the code is modified by identifying and removing other pieces of the code


5)For each piece of identifiable information in the code, it is possible for parts of the information to be considered affirmative if the data fits within a range (i.e. B-N; *-4 etc.)

Statistical code crackingapproaches, what do they work for? In what ways can they differentiate or not?
Download Simon Singh's "The Code Book" for free at his web site





http://www.simonsingh.net/Shop_-_Crypto_...





...You can read all about it.