When trying to solve a Kryptos solution, there’s really only one main assumed prediction, that the method you are using will solve the fourth part of Kryptos.

Nevertheless, it’s possible to take my hypotheses and try to make some reasoned predictions about each.

“Kryptos” is a plaintext artifact in the ciphertext
This one is simply right or wrong.  I’ve attempted in several different ways to avoid the knowledge that there are several ways of actually spelling the word Kryptos in the text of K4.  I am predicting that it is the most likely, the digraphs on the right side of the ciphertext with a “T” from the left of row 3 of K4.  Is this certain?  Not by any statistical means, it is rather an acknowledgement of the fact that I’m hoping it won’t disrupt the translation of the rest too badly to allow partial recovery.  If I can even recover a 10+ letter fragment, it may be possible to go back and re-choose the letters used to remove “kryptos”.  Why am I convinced that we need to remove it?  It’s too coincidental that it can be easily noted.  For what appears to be a secure cryptosystem for K4, the difficulty of coordinating plaintext and ciphering to allow the specific keyword “kryptos” to appear in the ciphertext would be significantly difficult.

Transposition has scrambled the digraphs
It is possible that  the message was transposed first and then substituted.  In this way, the ciphertext would hold little trace of the plaintext and would likely be very random and the frequency analysis would be nearly flat instead of muted.  This is a hope and reasoning but not a certainty.  It makes a stronger cryptosystem to fractionate and then diffuse so I will choose this as the most likely method that would have been used.  If I fail miserably, this may be an area of alteration to revamp my methods.  As for the type of transposition, it would not need to be exotic – a simple, standard transposition capable of splitting letter pairs would suffice.

A digraphic substitution ciphered the plaintext
Why digraphic?  As I’ve already shown, neither transposition alone, simple substitution or polyalphabetic substitution would be enough to ensure K4’s security.  Many of the known digraphic methods involve matrix enciphering and given Sanborn’s obvious predilections, these are almost certainly involved in Kryptos in some manner.  By itself, a cipher of this kind is still vulnerable so it would need to be paired with another method.  It is doubtful an unknown or strange digraphic system was used.

I predict that this will work.

What’s the point otherwise?