Zodiac Discussion Forum

Recap: three interc…
 
Notifications
Clear all

Recap: three interconnected Zodiac puzzles solved

3 Posts
2 Users
0 Reactions
928 Views
 urik
(@urik)
Posts: 61
Trusted Member
Topic starter
 

My recent posts have addressed separately three Zodiac puzzles (Z13, ZF symbol, and "sorry no cipher" note), which turn out to be solvable and related to each other as pieces of a larger structure involving also Z340 and Z32, with the three ciphers being encapsulated by the ZF symbol. It is worthwhile to consider these elements combined, as this has interesting implications, and the way the pieces fit together boosts the confidence in each individual solution.

The ZF symbol elegantly summarizes the transposition rules of the three ciphers Z340, Z13, and Z32, when interpreted as follows:

           Z340 = knight
         O ➡ O    move
Z32?   ⬈       ⬊
     O ⬅  ⬅  ⬅ O
      Z13 = 3 left

The symbol was too ambiguous to make sense before the transposition rules of two ciphers became known, but at least now we can use it for Z32. The dot arrangement is uncharacteristically precise, supporting its interpretation as conveying the transpositions. Furthermore, while the symbols on the Halloween envelope and card differ somewhat (observe the angles between lines and the position of the z), the dots are kept precisely the same, further supporting the above interpretation. This dot arrangement is rather clever, and Zodiac uncharacteristically wrote the symbol also on the envelope, perhaps because he was proud of himself. So it now seems more likely that the initials ZF (or perhaps ZVF, or similar) really had some special significance for Zodiac, even if the symbol was inspired in part by other sources (like comics or cattle brands).

While we needed the Z340 and Z13 transpositions to understand the otherwise ambiguous ZF symbol, its elegance (if you agree with my interpretation above) now reciprocatively supports the bottom edge of the dot pattern (j=-3 transposition), and hence the "slaves slayer" message that solved Z13 and fixed j. A redundant "e" padding found in my first Z13 analysis was anticipated but not mid-sentence, thus penalizing the p-value and introducing some tension, which was relieved subsequently once the "sorry no cipher" note was understood as explicitly clarifying the padding/separator and its central position (see image below). We didn’t understand the note before solving Z13, so again, the note now reciprocatively supports "slaves slayer" and thus j=-3. Direct and strong support for "slaves slayer" stems from the focus on "paradice slaves" (sic) in the otherwise mostly understandable/unoriginal Halloween card, and on "slay" in the otherwise bizarre SLA letter. These pieces now all seem to fit nicely.

Focusing on Z13: solving it by utilizing the Zodiac vocabulary (not an original idea in cryptanalysis) produced a highly statistically significant and robust answer (even before independently corroborating the padding, and despite an unsupported but unsurprising letter swap). The solution is supported by additional evidence, in particular an unexpectedly frequent reuse of the Z408 key, mainly with (and with Greek, as suggested previously on this board) but also without key inversions; see key and key comparison below. Nevertheless, and although the padding was already explained in the note, I entirely relaxed the vocabulary constraint, and ended up again with the same "slaves’ slayer" solution when focusing on constrained messages in depth (also a standard cryptanalysis technique) with key inversions (I including Greek, per suggestions on the board, but this can be relaxed). This makes me cautiously optimistic about Z32 being solvable too, for example in the strategy outlined here: one can always decipher a message (of arbitrary length and homophones) if it is sufficiently in depth. Here, we indeed have evidence for depth, as well as the diagonal transpositions inferred from the ZF symbol, and several additional clues on the content of the message.

Zooming out, we see that besides taunting the police, the Halloween card was largely an attempt to hint at the three cipher transpositions or brag about their clever arrangement. The "sorry no cipher" and SLA messages seem specifically targeted at facilitating a solution to Z13, indicating that Zodiac was eager to advertise himself as the (deranged, humoristic, or copycat) "slaves’ slayer". Zodiac may have repeated this fairly rare phrase elsewhere, thus associating himself with other putative communications. Most other elements in the Halloween card (13 eyes, number 14, four-teen, "you’re doomed") seem like taunts unrelated to the ciphers, but the pumpkin and its positioning may perhaps pertain to Z32 or the supposed bomb: as in other Zodiac puzzles, we may realize what a clue means only after solving Z32. Seeing a coherent solvable structure in the above puzzles, we can be cautiously optimistic that the "Sierra club" card or other riddles attributed to Zodiac can be solved, if given sufficient efforts and all relevant materials.

Zooming further out, Zodiac presents as an amateur with zero cryptographic (and certainly no cryptanalytic) training, did not show much skill, intelligence, or organization when encrypting his messages, and generally appears sloppy and perhaps (according to letter analyses) dyslexic. But admittedly, coming up with the interconnected puzzles above does require higher than average intelligence and creativity. Now switching to a more speculative territory. Zodiac was probably self-taught and confident in additional topics. Elements in the Halloween card (by fire, by gun, etc.) still seem irrelevant and likely inspired by 1952 comics. If Zodiac saw these comics at an impressionable age (rather than as an adult tool or hobby), then his most likely DOB would typically be estimated around 1940-1946, and direct contact with big bad Zode around 1964 would suggest a young staff. But this contradicts evidence for an older Zodiac, and unfortunately, a conversation between Cheney and Allen suggests that the impact of the Zode story reached quite far in some circles (thus drawing the heat to unlucky Allen). The most probable explanation in my opinion is a sophisticated Zodiac, role-playing the character of Zode (with the comics reference, perhaps the Belli fascination, the logically inconsistent "slaves slayer", etc.), and smart enough to cover his tracks and live to an old age in plain sight.

What do you guys think?


[Edited for clarity]

 
Posted : May 16, 2021 4:16 pm
(@veroh)
Posts: 21
Eminent Member
 

:)
You’re thinking out of the box and that is certainly a good thing.

I am interested in how you found "j=3" in this scheme?

Roger

 
Posted : May 22, 2021 10:43 pm
 urik
(@urik)
Posts: 61
Trusted Member
Topic starter
 

🙂
You’re thinking out of the box and that is certainly a good thing.

I am interested in how you found "j=3" in this scheme?

Roger

It seems to me we have no choice here but to think like a Zodiac: looking for conspirative patterns everywhere is a slippery slope, but is justified in the special case of bizarre Zodiac letters like the Halloween card, riddled with taunts and clues.

The j=-3 (minus necessary) was my starting point here, in an (admittedly non-original) attempt to decipher Z13 by constraining it to the Zodiac dictionary. This is a very old technique: given multiple and repetitive communications from a source of limited vocabulary (think of WW1 military status reports), a new cipher can be more easily solved if one is able to reproduce the message as a combination of words drawn exclusively from a dictionary compiled from previous communications. This method can be generalized, for example by further constraining to, or prioritizing, the most likely words, enforcing fixed text templates (think of any status report…), etc.

So my code scanned all possible values of j with all possible combinations of prioritized words in a Zodiac dictionary (compiled by a Wolfram Mathematica code I can share, based on Zodiac correspondences here), allowing for padding as in Z408. First, I got no viable solutions. But when I allowed for one letter swap (also based on previous ciphers), I got one unique and sensible (if ridiculous: "slaves slayer", with a separator in the 7th symbol) solution for j=-3, and no solution for any other j. Hence, accepting this solution also fixes j=-3 (I didn’t notice the ZF dot pattern till later). Note that unlike my interpretations of the ZF symbol and "sorry no cipher", which cannot be rigorously proven and can only be judged by the number of questions they resolve, this Z13 solution can actually be proven statistically. Namely, if the p-value threshold imposed on the cryptanalyst (folding in a letter-swap penalty, if needed) is in the range I consider reasonable, then this solution is statistically proven correct and unique. (Such statistical proofs are formally always necessary, although omitted in practice for sufficiently long and self-evident cipher solutions; I don’t really see a point in proposing a cipher solution that cannot be thus tested.)

 
Posted : May 25, 2021 6:06 am
Share: