So a lot of ideas have been put out there for the fourth part of Kryptos and I’ve personally considered quite a few different possibilities for the ciphering system for K4.  At this point, as of March 28th, 2012; the most likely decryption in my mind is reversing a superencipherment of a keyword (additive or subtractive) followed by reversing a transposition.  It’s simple to do and at the same time, probably enough to keep us stumped this long.

How could it possibly be that simple?  #1, Scheidt knew Sanborn was a neophyte with enciphering, #2, it’s a classic move from WWII methods to encrypt the message with a simple additive and then do a different cipher method.  Frequency analysis among other things helped both sides unravel the messages, #3, it’s too logical for most Kryptos fans to consider.

How it works is basically changing the message into numbers, changing the keyword into numbers and then adding them (you can also subtract them) before you run them through some other substitution/transposition algorithm.

I’ve been working on trying to reverse it.  The idea is that by adding or subtracting keywords to the K4 text, we can change the letter frequencies.  If, for example, I subtract the digits 3-15-4-5 (CODE) from the digits of K4 (mod26) then analyze the letter frequencies I should be able to estimate how likely the result is to being plain English.

Why transposition?  Because the first half is substitution and we know the first part of the second half is transposition.

How do you know he didn’t do the superencipherment first then the transposition?  I don’t but if that’s how he did it then this method of trying to untangle the message won’t work.



Subtracting KRYPTOS.