Zodiac Discussion Forum

Let’s Crack Zodiac …
 
Notifications
Clear all

Let's Crack Zodiac – Episode 1 – Graysmith

28 Posts
16 Users
0 Reactions
8,391 Views
doranchak
(@doranchak)
Posts: 2614
Member Admin
Topic starter
 

Quarantine inspired me to create a new video series on Youtube.
It’s called "Let’s Crack Zodiac".
True crime meets classical cryptography.

Episode 1: Graysmith
Almost 35 years ago, Robert Graysmith published his blockbuster book, "Zodiac", and claimed to solve a famous unsolved Zodiac Killer cipher. But did he really solve it? Let’s break it down.

https://youtu.be/N_Oh4snhF70

https://www.youtube.com/watch?v=N_Oh4snhF70

http://zodiackillerciphers.com

 
Posted : April 12, 2020 4:37 pm
(@sincos)
Posts: 11
Active Member
 

Great video! Will wait for next episode.

Maybe an episode of the solved 408?

 
Posted : April 12, 2020 8:40 pm
(@tlaz444)
Posts: 13
Active Member
 

This was great, thanks Doranchak!
Looking forward to more.

 
Posted : April 12, 2020 8:53 pm
(@mr-lowe)
Posts: 1197
Noble Member
 

well presented

 
Posted : April 13, 2020 3:07 am
(@largo)
Posts: 454
Honorable Member
 

Quarantine inspired me to create a new video series on Youtube.
It’s called "Let’s Crack Zodiac".
True crime meets classical cryptography.

Great idea!

Really well presented and well explained. I am looking forward to further episodes. Thank you very much!

 
Posted : April 13, 2020 11:23 am
(@beijinghouse)
Posts: 34
Eminent Member
 

Great video Dave!

I knew the gist of the Graysmith story but enjoyed learning the details of the timeline of how his solve came to exist. I knew it was amateurishly created and wrong, but I still liked seeing you calmly dismantle it with the FBI docs + simple demonstrations.

It’s sort of quaint to see someone like Graysmith from pre-internet times be so embarrassingly out of their depth. It’s perhaps even tragic when you consider he appears to have found the correct experts at the time in the FBI / ACA who could have set him straight, yet he still ended up deluding himself that they were wrong and he was right… all while likely fabricating supportive affirmations from those sources that he succeeded when he was almost certainly told the opposite in both cases. It’s easy enough to imagine how he ended up in this situation. This was Graysmith’s first book. He was probably desperate to ink a deal with a big New York publisher. I assume he simply exaggerated things in his pitch meeting … including claiming he knew who the killer was and that he had solved the z340, etc. So after he cashed his advance and banged out his book, he probably felt he simply had to deliver a solve to his editor, even if he knew in his heart (or from reading the FBI’s letter) that he was likely wrong.

Have you considered analyzing and responding to this:
http://tsjuzek.com/blog/z340.html
https://www.ep.liu.se/ecp/158/013/ecp19158013.pdf

It’s basically the most well-done analysis trying to show z340 is a pseudo-cipher / hoax with no information content.

While it makes some interesting points, I feel like the author is really bending over backwards to get the result they want and also pushing their analysis tools past their breaking point.

Distilling and explaining the subtle errors in this work is not nearly as easy as it is with Graysmith, but would also be more interesting.

I’m not sure the best way to organize a critical discussion of their work, but as a starting point, you could easily show that small changes to the 408 key could trick their metrics into declaring the 408 was even more likely to be a pseudo-cihper than the 340.

You could also touch on beale1 and show a case where you can effectively confirm that a cipher is a hoax by finding the dummy message + encoding technique.

To be fair, the author of this analysis fully accept that 340 could be a true message with some added transposition step or different encyperment method.

But they also explicitly stress that if it’s the same author, it would be hard to believe they could make such a weak cipher first and then make such a well-concealed one. But you could point to proof in later Zodiac letters that he read about how the Harden’s deciphered his 1st message so easily, and their public description of how they did it (ie, seeing bigrams) is exactly the knowledge Z would have needed to make the exact adjustment to a future key that would get this paper’s metric to classify it as a pseudo-cipher.

 
Posted : April 13, 2020 8:33 pm
doranchak
(@doranchak)
Posts: 2614
Member Admin
Topic starter
 

It’s easy enough to imagine how he ended up in this situation. This was Graysmith’s first book. He was probably desperate to ink a deal with a big New York publisher. I assume he simply exaggerated things in his pitch meeting … including claiming he knew who the killer was and that he had solved the z340, etc. So after he cashed his advance and banged out his book, he probably felt he simply had to deliver a solve to his editor, even if he knew in his heart (or from reading the FBI’s letter) that he was likely wrong.

That is a very plausible explanation. The book would not have been as exciting to read if he had just shrugged his shoulders and said, "Who knows?"
Other aspects of the book support your explanation, such as the way he invents narrative details and dialog when he tells the stories of the victims. He seems to adopt an omniscient viewpoint that can’t possibly come from direct evidence.

Distilling and explaining the subtle errors in this work is not nearly as easy as it is with Graysmith, but would also be more interesting.
I’m not sure the best way to organize a critical discussion of their work, but as a starting point, you could easily show that small changes to the 408 key could trick their metrics into declaring the 408 was even more likely to be a pseudo-cihper than the 340.

That’s a really good idea for an episode. My goal is to make the videos easy to understand so that’d be an interesting challenge with that paper. Thanks for reminding me about that paper – I had forgotten about it!

But they also explicitly stress that if it’s the same author, it would be hard to believe they could make such a weak cipher first and then make such a well-concealed one. But you could point to proof in later Zodiac letters that he read about how the Harden’s deciphered his 1st message so easily, and their public description of how they did it (ie, seeing bigrams) is exactly the knowledge Z would have needed to make the exact adjustment to a future key that would get this paper’s metric to classify it as a pseudo-cipher.

Good points. We don’t even yet have a very reliable way of distinguishing between known encipherment systems (besides a rough process of elimination and guesswork), so it’s a bit presumptuous to me to assume we can detect a non-cipher. Especially since the space of homemade encipherment systems has an unknown size.

I think his "caveat" conclusion is more correct (that the cipher is an "advanced" cipher).

http://zodiackillerciphers.com

 
Posted : April 14, 2020 12:57 am
Marclean
(@marcelo-leandro)
Posts: 764
Prominent Member
 

Nice Vídeo
Thanks 4 make this . :) :)
I only have to visualize it about 8 times until I understand, the difficulty with the language and mainly, difficulty with code.
One day, who knows, you might put subtitles in Spanish or even Portuguese for us !!! :D
Marcelo

https://zodiacode1933.blogspot.com/

 
Posted : April 14, 2020 1:07 am
(@fishermansfriend)
Posts: 132
Estimable Member
 

cool! Can’t wait to watch. Big fan of your work.

 
Posted : April 14, 2020 4:40 pm
 Khys
(@khys)
Posts: 154
Estimable Member
 

This is good stuff. Thanks.

 
Posted : April 14, 2020 8:59 pm
(@anderson110)
Posts: 55
Trusted Member
 

Really nice job. Very cool graphics and animations!

 
Posted : April 19, 2020 6:18 am
(@largo)
Posts: 454
Honorable Member
 

I have an idea for another episode:

How about an explanation why we can be sure that z408 was solved correctly. There’s always the claim that z408 was never solved. One could, for example, mention that different solvers independently come to the same result. Moreover, the context of the plaintext fits the overall picture. The key is absolutely plausible and the few errors (A == S) etc. can be easily explained, because the cipher symbols are similar.

Translated with www.DeepL.com/Translator (free version)

 
Posted : April 25, 2020 11:16 am
Zamantha
(@zamantha)
Posts: 1588
Member Moderator
 

Great Thread “Lets Crack Zodiac”.
Thank you DoranC, always appreciate your new ideas!

Zincerely, Zamantha*

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
If Zodiac ever joined a Z forum, I’m sure he would have been banned for not following forum rules. Zam’s/Quote
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
MODERATOR

 
Posted : April 25, 2020 11:35 am
doranchak
(@doranchak)
Posts: 2614
Member Admin
Topic starter
 

I have an idea for another episode:
How about an explanation why we can be sure that z408 was solved correctly. There’s always the claim that z408 was never solved. One could, for example, mention that different solvers independently come to the same result. Moreover, the context of the plaintext fits the overall picture. The key is absolutely plausible and the few errors (A == S) etc. can be easily explained, because the cipher symbols are similar.

Great idea! I thought about doing an episode about the different solvers, so that might be a natural fit for that topic. It might also fit with the question "how do we know Z340 is not just simple substitution", by showing how test ciphers can be easily cracked by all the solvers but Z340 still remains a mystery.

http://zodiackillerciphers.com

 
Posted : April 25, 2020 12:34 pm
(@f-reichmann)
Posts: 30
Eminent Member
 

I would expect the evidence that 408 was solved correctly to the only existing reasonable solution can be proven with unicity distance, as in https://en.wikipedia.org/wiki/Unicity_distance . Unless, of course, someone would doubt that it is a homophonic cipher. Which I would reject as arbitrary, everything could me a one time pad, that would be a useless discussion.

It could then continue with crypto artwork: Are there homophonic ciphers possible where both the cipher symbols, and the clear text, are meaningful? I sometimes get an impression when I look at 340, the „her“ and „zodaic“ resembling symbol sequences. No idea if these are artefacts, but they inspired me to think about how such an artwork could be constructed, independent of Zodiac.

Last but not least, it could b e demonstrated that if 340 is homophonic, then there is only one meaningful solution expected. And that all solvers failed, despite being able to crack homophones at 340 length and 63 symbols.

How about an overview of how the solvers work?

 
Posted : April 26, 2020 12:44 am
Page 1 / 2
Share: