QuizmasterJ recently posted about Kryptos here.

I’ve been pretty stuck on BERLIN = NYPVTT.  At the time, I didn’t really think Sanborn had given us much with his decoding charts.  I’m sluggishly thinking of maybe revisiting them and the idea of directionality.  He likes matrices so much, it must sort of appeal to some part of his mental modeling of the codes for Kryptos.  Is it really as simple as that he switched the ordering of the ciphertext for K4?  Could it be that easy?  Or still ridiculously hard?

Anyways, go visit quizmasterj and support a burgeoning blogger.  Then go solve Kryptos.