Zodiac killer's unsolved code "behaves like fake ciphers" and therefore probably a ruse


#1

Originally published at: https://boingboing.net/2018/03/25/zodiac-killers-unsolved-code.html


#2

Is it possible it’s a one time pad? That would not be “cheating” per se but it would be uncrackable.


#3


#4

Yeah, but that would require a bit more cryptographic knowledge than would be plausible for a typical serial killer of the 1960s/1970s.


#5

And I thought he was an honest serial killer. :sleepy:


#6

It’s a word search


#7

The link to the blog post does not work: there’s a trailing / that breaks it. This link works:

http://tsjuzek.com/blog/z340.html

(Don’t feel bad, the link on the blog’s main page is broken too).


#8

How so? One time pads were commonly used in WWII, and their nature and use would require no more cryptographic knowledge than would be available in a library book on the subject (such as the books that went missing from the SF library system at the time).

It’s not a one time pad issue, though. The reason a one time pad is challenging to crack is that there is no primer embedded in the message to give the reader the information they need to decrypt, because the sender and reader already both have it, and use it only once, for that message. Messages with the primer embedded are easier because you just have to find the primer. A series of messages using the same primer are easier because the statistical character analysis is more robust. But the ciphers used with a one time pad are not inherently more complex or challenging that those without. The one time made is just about the method the recipient gets the info needed to decrypt, nothing about the algorithm of encryption.

The reason the author believes this message can’t be deciphered and is fake is not that it’s too difficult to crack. It’s that it shouldn’t be too difficult to crack, but it is, because the characters are too random to be real.


#9

turns out he’s a serial liar too. What a character defect!


#10

image


#11

I agree that this looks like some very detailed statistical analysis, and if the analysis is robust, the arguments are compelling.

But wow, that graph is a disaster. It could be a ‘what not to do’ case study in graphic design and science communication.

The legend is too small to read easily, even though there is plenty of room for it to be larger. The legend actively detracts from the ability of the reader to understand the graph, because the legend implies that the colours of the datapoints are what differentiates various ciphers in the analysis, but it is actually different shapes that are used for different groups. The legend implies a single colour for true ciphers and a single colour for fake, but each category is in fact unnecessarily represented by a range of colours — while those colours are related within each group (greyish and blue-to-teal), the outliers in each group look nothing like the point in the legend. At least three of the true cipher points are so desaturated that they more closely resemble the colours of the fake cipher points that the other true cipher points. The different colours within each group are pointless — they convey no information, they only make the graph more confusing and difficult to read. The z408 circle colour is indistinguishable from that of some of the true cipher points for a colour blind person.

Each different group should have a single unique shape and a single unique colour, the colours should all be mutually distinguishable for a colour blind person, and both colour and shape should be accurately described in a larger, more legible legend.


#12

Actually, it’s about ethics in serial killer cryptography.


#13

Books on cryptography tended to be rather thin on actual details until recent years when large numbers of people outside the intelligence community needed to understand it for practical reasons. And large parts of things that you’d think would be public knowledge about WWII cryptography by then weren’t yet.


#14


#15

Perhaps not a one time pad, which seems technically unlikely, but I wonder he made some stupid trick to “decoding” that permitted him the thrill of not cheating but which he knew randomized it.

For example: retain the unencrypted original, and make a “key” which would confirm the arbitrary relationship between original and the otherwise random cipher if all were discovered.

Or: say the first half of your text is a message, and the second half of your text is random. You encrypt the whole text in some way that you know won’t affect its randomness. Then you send out the half which isn’t the message.

This strikes me as a very serial-killer ish way of telling yourself you aren’t cheating.


#16

Ok, but does it require more library using skills than was common for serial killers in the 60s/70s? Especially for serial killers with something to prove about their Intelligence?


#17

That’s a nice theory Rob, but I think it’s obvious that a team of NSA cryptographers cracked it and alerted their superiors long ago. Unfortunately, the Deep State (led by Michael Ignatieff) hid the truth because they hoped that Ted Cruz would overtake Trump in the Republican primaries. When this did not occur, they knew they had to erase any evidence that they knew the killer’s identity and did not pass it on to law enforcement - the public would not tolerate another instance of the deep state standing idly by. So they were told that they would meet with SFPD to discuss their findings, but first would be given tickets to a free concert in… Las Vegas.

Your theory is interesting but I have three (3) citations and yours has none so… thanks for playing. :wink:


#18

If you’re wondering why the cracked Zodiak cipher is way past the rest of the “true” ones…

It looks like he wasn’t really that good at crypto. After the first one its hard to believe he’d hit a home run that would fall in that lower left corner like that.


#19

I assume that the colour range conveys information along some other axis, but the key for that has been excluded, either during the C&P for this blog, or in the original. But your point is still valid - each group should have a unique shape and a unique colour range.

Also; thanks for remembering the mighty 10% :+1:


#20

While this is a fascinating analysis, I decided long ago that I don’t care what serial killers think; they do not deserve the public’s attention. They should be forgotten in whatever prison they land in; their writings and manifestos burned unread; their ashes interred in graves unmarked.

No one should ever think that murder is a valid path to the public eye.