Kryptos K4 attempt #257

It had been too long since I’d actually tried something constructive with solving Kryptos.  I tried something quick and dirty tonight and while it didn’t work, I’m proud of the effort.

Background:  I’m pretty sure there are two encryption methods on K4.  The first was some kind of substitution cipher followed by a transposition.  I don’t know either.

Algorithm:  Basically I thought a stream cipher would give the best chance at the flexibility I would need to fiddle with and the strength K4 appears to have.

Deconstruction:  If Y(mod26)=X+position with X being the numerical value of A-Z (1-26), what are the relative positions of the letters BERLIN that would give NYPVTT?

B     12, 38, 64, 90

E     20, 46, 72

R     24, 50, 76

L     10, 36, 62, 88

I     11, 37, 63, 89

N     6, 32, 58, 84

Test:  If the stream cipher was the method, there would be some kind of alignment in the available positions.

Result:  No pattern that I can see.

Rationale:  This would actually be something different from what I suspect has happened with K4.  For this to have worked, K4 would have to have been transposed first and then those new positions would be used for the substitution and then the transposition reversed.  It would be an interesting cipher method but I’m not sure how helpful it is as these things go.

It was still nice to actually try something instead of just waiting for someone else to figure K4 out.

I will name this cipher KPOP in lieu of some standard nomenclature that I’m sure someone will tell me if they know.

To Review – KPOP cipher: transpose plaintext, take numerical value of plaintext plus numerical value of position and add MOD26, reverse the transposition and this is the final ciphertext.

I’m not sure how you’d beat KPOP even if you knew the mechanism.  I think most of the supercomputers are mostly fixated on the big stream and block ciphers.  Why would anyone write code for stuff like this?  That’s probably how the Scheidt/Sanborn creation has made it this far.