Message Starts

The first trigram value in every message is different, followed by large shared sections. The values of these initial trigrams are all over the place. The most common theory is that they could be numerical indices for the messages, like 1-9 to label each message, or possibly A-I if numbers are not in the plaintext alphabet. Another possibility is that they are not part of the message as a normal character, and are instead something like a checksum, or the last character of the message moved to the front, or something similar.

This feature has been demonstrated to be possible to reproduce with deck ciphers, with the internal state of each message after the first character being slightly different, but similar enough that shared sections are still possible. This is also possible with other mechanisms like CTAK, but only if the initial state of each message is different, and is just finetuned to make the first letter resync the state in every message.

There are also some interesting observations about the base 5 values of these first trigrams, which might be a hint for some reason that base 5 was used, but we haven't figured anything out yet.