Hacker News Reader

Comment by jgrahamc | Hacker News Reader

In what sense is that a 'solution' to the cipher? It just looks like the sort of pattern spotting that the human brain is good at and doesn't seem to indicate any sort of systematic crack of Zodiac 340 at all. This sort of reasoning is typical of people who 'crack' ciphers by coming up with all sorts of complex explanations. I call this sort of 'crack' a 'Dan Brown'.

What's needed to crack Zodiac 340 is a systematic way in which The Zodiac enciphered the message. His Zodiac 408 had a clear and simple pattern.

My recent posts on the matter:

http://blog.jgc.org/2011/06/how-zodiac-enciphered-zodiac-408...

http://blog.jgc.org/2011/06/identification-of-homophone-sequ...

View on HN

Here's an image showing the redditor's claims: http://imgur.com/lQuYT.

There's really no logic behind this 'solution.' The 'i' 'c' part, for example, is certainly bogus, and the arrows are inconsistent. I'd bet money that if this is in fact a cipher text (and not a random arrangement of symbols) this decoding is not correct.

In order for a systematic solution to exist, the cipher must be systematic. Maybe Zodiac didn't make it mathematically or systematically.

Humans are pattern-loving creatures, maybe these Reddit users found a pattern that has escaped everyone else so far. Or maybe it's just gibberish.

Reminds me of the "Bible Code" folks, too.

Yeah its more of a starting point than a solution. The overall arrangement of the cipher text is pretty convincing. The words spotted not so much.

I wouldn't discount the Dan Brown approach though...it seems well suited for solving a puzzle by a guy who routinely misspells words & has a tenuous grasp on reality at best.

  The overall arrangement of the cipher text is pretty convincing.

Really? I don't get that bit. Can you explain it to me?

  has a tenuous grasp on reality at best.

In terms of cryptography in the late 1960s that could be done with paper and pencil I'd say that The Zodiac had a very good idea of what he was doing. In the Zodiac 408 cipher he carefully disguised the frequent letters with multiple symbols to make it hard but not impossible to break (see my blog posting above). His misspellings could easily be deliberate (as they mess with the letter frequencies). I wouldn't write him off, even if he did kill a large number of people.

Cipher arrangement: If you look at the middle row here: http://imgur.com/6MWFk The very first char in the row, the last and the middle fit perfectly into the zodiac sign drawn at the bottom. It could just be a coincidence, but it feels like the chars should be shifted around so that the whole thing makes a zodiac sign. All the funky chars being the sign & the text being in the white background.

Lots of the characters are also upside down, mirrored or both. Superimposing a zodiac sign on it would give us 4 quadrants that could be used to flip the characters back to their correct orientation.

As for the deliberate misspelling...maybe. I don't buy it though. All the wrong words are misspelled. e.g. http://upload.wikimedia.org/wikipedia/commons/f/f8/Zodiac_ci... He misspells paradise, but repeatedly uses the letters "kill". Thats all wrong for someone deliberately misspelling to mess with cryptographers. Admittedly I haven't looked at the ciphertext of the above URL, so maybe kill was suitably disguised there.

>I wouldn't write him off, even if he did kill a large number of people.

Yeah. You have to admit though: Anybody rambling on about killing to have slaves in paradise qualifies for "tenuous grasp on reality at best" category.

I didn't recognize that the cipher could be split into four quadrants (with the "+" symbol bisecting vertically and horizontally) before reading the comment; maybe he's cycling through the symbols based on the quadrant they're in? I noticed you wrote about substitution cipher, so maybe some relation to that (rather than cycle sequentially)? There were some interesting comments from the thread, although I agree with what you're saying that it's not systematically cracked if Zodiac was following the method he used in the other ciphers.

too late to edit my comment but an interesting analysis which does consider the quadrants: http://zodiackillersolved.blogspot.com/2008/12/340-cipher-so...

What's wrong with the work the FBI did?

http://upload.wikimedia.org/wikipedia/commons/7/77/FBIZodiac...

Look at pages 105-114. Seems pretty believable to me...

All the efforts look awful.

There's no analysis whatsoever of the number of similar-degrees-of-freedom different-meaning decipherments that would seem equally credible, or likelihood of deciphering a random text "successfully" given that many degrees of freedom.

Granted, what I'm complaining about isn't easy to compute. If I knew how to actually do it, I'd publish a paper.

Reply on news.ycombinator.com