It’s nice to have the observations laid out. I’ve often pondered the apparent symmetry, but I hadn’t noticed that it’s exactly half the alphabet. The symmetry is very striking though, it just looks mirrored.

The way I see it is that there are only a few possibilities:

1. He wanted it to be solved and understood that it would need a verifiable solution method
2. He wanted it to be solved, but failed to understand that he had constructed an unverifiable solution method
3. He didn’t want it to be solved, and either wrote gibberish, or intentionally ensured it would be unverifiable

The recent solution to the Z340 has tipped the scales towards the first two possibilities.

Since the only worthwhile assumption for us to labor under is the first possibility, we should labor under it.

So, assuming option 1, what methods are available to someone who wants to construct a valid, solvable and verifiable 13 character cipher? There must be methods.

One such method the Zodiac could have used is the previous cipher keys. By reusing them in the Z13 he would have been able to construct a verifiable cipher. Another way would be to include strong clues in the letters. Notice I say "strong", because it would need to be a pretty big (obvious) hint, a hint that verifies something cryptic, could not itself be overly cryptic. So trying to line pumpkins up with mount Diablo will see us all in the grave long before a valid solution is uncovered.

This line of enquiry also recommends that we set aside methods that would not be verifiable. So looking for generic words like "kill" won’t work. Just look at how many people generated random Zodiac words in the Z340. This approach won’t work. The Z340 wasn’t solved by finding a word like "paradice", but by finding a sentence that made a peculiar and singular reference to a then recent event in the Zodiacs career. However neither the Z340 method, nor the Z408 will work to solve such a short cipher.

Another possibility is some kind of self verifying cipher. I’m not exactly sure how this could be done, but in this scenario the cipher might have multiple parts that verify each other. One example was fayz’s recent attempt at solving the Z32, I don’t think it’s the right solution, but I noticed something about it. He tried to find coordinates in the Z32, and in his first attempt he found a set of coordinates and the remaining letters could be anagrammed to spell the name of a road near those coordinates. Unfortunately the road was a fair way from the coordinates. But it got me thinking about something else, the strongest part of the solve was that it used a method that was self verifying. Normally an anagram would be remote, but if you had coordinates that pointed precisely to the anagram solution the coordinates would in effect corroborate the anagram and vice versa. It occurred to me that there aught to be many such ways of doing similar things, not necessarily with coordinates, but just by having something about each part of the solution, or the cipher itself in some way validated the result. A sort of cipher internally verifiable, or mutually verifiable cipher.

With that in mind, we could approach the Z13 by trying to think of all the ways a verifiable 13 character cipher could be constructed and excluding all the methods that would be unverifiable.

Does anyone know, or has anyone read, of a method of constructing a reliably verifiable cipher that is only 13 letters long? As in one where the solver does not have the key to begin with.

It’s nice to have the observations laid out. I’ve often pondered the apparent symmetry, but I hadn’t noticed that it’s exactly half the alphabet. The symmetry is very striking though, it just looks mirrored.

The way I see it is that there are only a few possibilities:

1. He wanted it to be solved and understood that it would need a verifiable solution method
2. He wanted it to be solved, but failed to understand that he had constructed an unverifiable solution method
3. He didn’t want it to be solved, and either wrote gibberish, or intentionally ensured it would be unverifiable

The recent solution to the Z340 has tipped the scales towards the first two possibilities.

Since the only worthwhile assumption for us to labor under is the first possibility, we should labor under it.

So, assuming option 1, what methods are available to someone who wants to construct a valid, solvable and verifiable 13 character cipher? There must be methods.

One such method the Zodiac could have used is the previous cipher keys. By reusing them in the Z13 he would have been able to construct a verifiable cipher. Another way would be to include strong clues in the letters. Notice I say "strong", because it would need to be a pretty big (obvious) hint, a hint that verifies something cryptic, could not itself be overly cryptic. So trying to line pumpkins up with mount Diablo will see us all in the grave long before a valid solution is uncovered.

This line of enquiry also recommends that we set aside methods that would not be verifiable. So looking for generic words like "kill" won’t work. Just look at how many people generated random Zodiac words in the Z340. This approach won’t work. The Z340 wasn’t solved by finding a word like "paradice", but by finding a sentence that made a peculiar and singular reference to a then recent event in the Zodiacs career. However neither the Z340 method, nor the Z408 will work to solve such a short cipher.

Another possibility is some kind of self verifying cipher. I’m not exactly sure how this could be done, but in this scenario the cipher might have multiple parts that verify each other. One example was fayz’s recent attempt at solving the Z32, I don’t think it’s the right solution, but I noticed something about it. He tried to find coordinates in the Z32, and in his first attempt he found a set of coordinates and the remaining letters could be anagrammed to spell the name of a road near those coordinates. Unfortunately the road was a fair way from the coordinates. But it got me thinking about something else, the strongest part of the solve was that it used a method that was self verifying. Normally an anagram would be remote, but if you had coordinates that pointed precisely to the anagram solution the coordinates would in effect corroborate the anagram and vice versa. It occurred to me that there aught to be many such ways of doing similar things, not necessarily with coordinates, but just by having something about each part of the solution, or the cipher itself in some way validated the result. A sort of cipher internally verifiable, or mutually verifiable cipher.

With that in mind, we could approach the Z13 by trying to think of all the ways a verifiable 13 character cipher could be constructed and excluding all the methods that would be unverifiable.

Does anyone know, or has anyone read, of a method of constructing a reliably verifiable cipher that is only 13 letters long? As in one where the solver does not have the key to begin with.

One way to solve an "unsolvable" cipher is if you know what cipher method was used, and have a fairly limited number of potential keys to use to solve it. For example the Vigenère cipher requires a keyword and an alphabet sequence, and then a ciphered message. The number of potential keywords could be very low, such as "MyNameIs" or "MyNameIsZodiac" or "Zodiac", something he would figure the reader to think of easily. But the alphabet sequence is more difficult to figure out, especially since he used symbols instead of just the alphabet. Figuring out the alphabet sequence to use could reveal what letters to use instead of the symbols, and then type the full ciphered message with the resulting letters instead of letters and symbols.

Once you have that, you could likely decipher it by trying a few keywords.

The most striking features are probably the chance that it represents the first half of the alphabet, and the 3-2-1-1-1-2-3 sequence of alphabet/symbol sequence.

But yeah, it’s quite a lot of potential ground to cover, so I think this is a case of the more eyes the better, otherwise we’ll waste our time

I like the idea of a common encryption method, something "down to earth". This seems like one such, are there any others?

As I am fond of mentioning about the Z13:

ROBHTHEHIPPIE and ROBETHEHIPPIE fit the symbol distribution pattern in its most simple form.

ROBH fits with the rh on the Desktop Poem.

ROBE fits with Robert Emet (Emmet) the Hipie that you can unscramble from the leftover letters in the 408.

Z did scramble some of the words in the 340 solution, correct? If so, we know anagrams are one of his strategies. It would only take one substitution and one unscramble to get to one of those solutions.

Thanks for the observations, egg!

I hope you don’t mind that I incorporated them into my list at http://zodiackillerciphers.com/wiki/ind … ations#Z13

Formidable ideas Mr Egg
No doubt , something that can be checked is the way to , try to at least decode both , z32 and z13.

You can work with (yours)keys of your choice in various encoding forms on this site http://rumkin.com/tools/cipher/

It would also be interesting to read the army manual well and check something, I mean, encoding and decoding modes, around that is not available online.

Marcelo

4- It is possible to linearly sequence part of the alphabet mirrored around the central ciphered character: nmlklmn

Egg ,
Observing your thoughts I noticed that besides what you observed, we can do the same, considering that the third Taurus symbol is the letter L (considering the mirroring that you observed) Note that the L of Taurus ends in N (of the alphabetical symbols) passing over the symbol of "Aries" – what gives L-M-N
See the Image

This gives a theoretical key value for the non-alphabetic symbols

M L M
Regarding the alphabetic symbols, I can imagine a few things
1-may be worth the symbol itself
2-Keys of 408/340
3-Other type of encoding

In another topic I posted about the symmetry of the coding too , but I didn’t notice this great find with mirroring. : http://www.zodiackillersite.com/viewtopic.php?f=25&t=5137&p=83628#p83628
See the image

I liked your post very much, I speculate, without any certainty, that it is an "auto key".

Marcelo

This is the identity letter
By the way have you read how
the last cipher Is written
and made?

I think it is a misconception that several common 13 letter names share the symbol distribution of this cipher and could fit the suspiciously symmetrical structure. I believe the unlikely pattern implies non-linearity. The three circled 8s in the middle are the most curious to me. Unless his name is Steve Nelson I think we have to imagine a transposed solution. Perhaps the answer is split and multidirectional with each half containing a name, perhaps formed through every second symbol. This proposed solution of mine contains extra letters which makes it about as likely to be correct as Zodiac supplying his actual name but I think there could be value in this transposition method.
/Users/andrewjamesgray/Desktop/Z13 MVEILNLBLIEMB Melvin Belli MB.png

I checked about 30,000 real names with that cipher, none matched. Odds are nice if you find a real match..
QT

I checked about 30,000 real names with that cipher, none matched. Odds are nice if you find a real match..
QT

Eddie Resendes fits (if you allow homophonic assignments). Real person, but born in 1989.
https://www.imdb.com/name/nm6334390/

Ernie Zerenner is another one. Found a real person with that name, born in 1941.

I checked about 30,000 real names with that cipher, none matched. Odds are nice if you find a real match..
QT

Eddie Resendes fits (if you allow homophonic assignments). Real person, but born in 1989.
https://www.imdb.com/name/nm6334390/

Ernie Zerenner is another one. Found a real person with that name, born in 1941.

I’ve mentioned this before but William Lawson works. The same guy who admitted to hitting Isobel Watson with his car then stabbing her. Also, if you use the Z408 key all the letters work from n-z, half the alphabet. (I think it’s the E/N that’s reversed if I remember correctly). Also the eight ball symbol, the one never used before or since, it’s the letter L. Why is that significant? L solved the z408 (kiLL). The z340 also has some symbols that work from a-m. All true. Check me for correctness.

I’ve mentioned this before but William Lawson works.

William Lawson only works if you allow rearrangements, which allows very many more names to work, too.

I’ve mentioned this before but William Lawson works.

William Lawson only works if you allow rearrangements, which allows very many more names to work, too.

Yes, it does. But can anyone find another name that fits that has some connection, some relevance, some reason to look at him? I don’t think anyone thinks that the z13 cypher is in the correct order. Just like I don’t think anyone thought the z340 was in order.
After being on this board for a couple of years now, it’s so common for responses to be negative. And I’m okay with it, but what if this is the correct name? Cyphers are your strong point, what are the odds that letters in this name correspond with symbols from the two other keys? Out of 8 letters I think it’s 6. (Im going off memory). And that there is the resemblance of a method, the alphabet being cut in half. I think it’s also relevant that a Vallejo detective thought the Watson crime to be better than 50/50 zodiac.
The early disregard for the name maybe just that, a little early.