So I’m reading rec.humor.funny and this is today’s joke:
Subject: Anyone Have a Line on the Jesse Jackson Sex Tape? (Jacksons_Big_Hoe_Hopper@qylo.edu)
This newsgroup typically encodes dirty jokes in rot13, but this is clearly not rot13, nor any other simple letter substitution cypher. The only thing I can make of it is that the letter frequencies correspond pretty much to that of uncoded text: e is the most frequent, followed by t, n, i, o, and r. And j, k, q, x, and z all appear rarely. So, can someone help me decipher this?
I read an amazing book called the Code Book which i recommend to anyone. Just from a cursory glance i can say a few things - y and o stand for i and a, not necessarily in that order. Since there seems to be stability in the y’s and o’, and their are multiple instances with three letters in a row im wondering if it doesnt have garbage letters added, especially since the avg word seems too long. Anyway ill take a look.
You’ve also got some letters repeated in triplicate, though, such as “munyyy”, “eeeti”, and “looo” which pretty much eliminates any sort of direct letter-subsitution. There’s also a shortage of three-letter words, and on a casual glance, I don’t see any repeated words (other than the single-letter “words” “o” and “y”). Spacing and possibly punctuation are probably arbitrary. Considering that it’s got an English letter frequency, it’s probably some sort of rearrangement cipher, of which the simplest is a fencepost, where you read every nth letter. I’m not getting anything this way offhand, but I haven’t tried every value of n.
Its probably a multiple monoalphabetic substitution cypher. Since their are 4 single letter words (A, Y, I, O) there are probably two different cyphers that alternate one after another or randomnly. It is also strange that the two letters that one letter words must be were one of the four, that is, A and I. I wonder if there is no substitution followed by substitution etc. Then move on to two letter words. I don;t have the time right now, and it is just a joke about Jesse Jackson, but if you cant find the answer soon, ill come back and figure it out. It is trial and error. Remember that there are only a few two letter words, when you combine this with your understanding of the coding of one letter and then three letter words, the method will become clear. It just takes time. Frequency analysis probably wouldnt work cause there is probably, actually, almost definately more than one cipher alphabet. My guess is this could be decoded in an hour or less. There is the chance that the division of the letters into words is randomn, but my guess is for a joke board, it must remain a simple algorithm. Frankly, why in the nine hells does a joke board feel the need to encrpyt jokes??
ROT13 has been a long standing tradition on USENET. It’s not that it’s intended to be hard to decode - quite the reverse (rot13 is just a simple cipher that rotates the letters 13 places, built into most newsreaders). The idea was to protect the casual reader against inadvertantly seeing objectionable material. By taking the time to decode the text, you were demonstrating that you were aware that it might be offensive, and were choosing to read it anyway. It doesn’t work that well - people just decode the stuff, read it, and get offended because they didn’t expect it to be THAT offensive, or they expected it to just be dirty, not RACIST, and so on …
Unfortunately, Opus1 didn’t provide us with the entire usenet posting. Otherwise, we would have learned that this message has appeared in multiple newsgroups with different variations of the “encoded” text in each post. The original message is a solicitation for a sex tape and contains bogus headers. The obvious conclusion is that the “encoded” text is a method of avoiding various spam filtering routines - mainly the ones directed at filtering multiple identical posting to multiple, unrelated newsgroups. By using gibberish text, it will also avoid various kill file patterns. Not clever, but effective. If you have any doubts, search deja for: Jacksons_Big_Hoe_Hopper
As a side note, Rot-13 is only one of a family of simple text-rotation encryption methods, Rot(n), where n is the number of characters to move in the alphabet for the replacement. There’s a great utility at http://www.flywheel.org/rot/ which will decode all possible ROT variations at once.
As a straight exercise in decoding, this gibberish contains a number of clues that suggest it isn’t a simple monoalphabetic substitution, the triplicate characters, double initial characters, more variations of 2- and 3-letter words than you should expect in a simple message. It might have used a rotating substitution, but again, the triplicates give it away. It would simply be too complicated for such a low-brow Usenet posting. Anything that’s encoded on Usenet is generally meant to be easily decoded.
Okay, since rec.humor.funny is a moderated newsgroup, how the hell did it slip through? And, I did indeed provide the entire message. Go to rhf and look for yourself. If it’s supposed to be a solicitation for a sex tape, why isn’t there any URL or anything else that would let people actually order it? Not that I don’t believe you, but I’m still left scratching my head.
In the case of r.h.f, it might have been a case of the moderator being lazy. He probably removed the ad, but left the cryptic text in because it certainly looks like ROT-13. Or perhaps his moderation duties are performed by a script of some sort. In any case, all it takes to verify that this is not encoded text is a simple search of deja for the poster’s name.