Put me in the weirdo corner but my program is still running on the Z340.

QT

Unless the cosmos is calling Jim Dunbar, out of the primordial abyss, otherwise the result is just overwhelming regardless of methodology.

Let’s face the facts. These gentlemen Dave Oranchak, Sam Blake and Jarl Van Eycke solved the z340 prematurely! We have not covered all potentially exciting topics yet, and with their spoiling of an overwhelming solution they have caused enormous damage to the general motivation and momentum, and spoofed our nice little party here. Shame on them!

Yes, I had so many future experiments and tests planned for the 340. We should have thought about those before so carelessly solving the damned thing!

What I see missing, and how we could move on:
1. Explanation on how to rate the level of arbitrariness that a transposition introduces into a solution, and how much of it is tolerable. To close the discussion on rejecting anagrams, while accepting transpositions (including spelling some words backwards, and change transposition for the last two lines).

I often wonder about that. We can use unicity distance to establish how long the cipher needs to be to guarantee unique solutions. But with thousands of possibilities that the 340 could have been, the unicity distance doesn’t really exist. Especially if you include things like long key lengths. For example, a Vigenere with a key length of 170, or similar running key with a long key. And worst case, the one time pad. You can only determine the unicity distance once you decide what the cipher type is.

Put me in the weirdo corner but my program is still running on the Z340.

QT

Is it because you doubt the current solution? If so, why?

Or are you looking for another layer of messages on the existing one?

Put me in the weirdo corner but my program is still running on the Z340.

QT

Is it because you doubt the current solution? If so, why?

Or are you looking for another layer of messages on the existing one?

Besides the still unresponded message I had sent to you a few weeks ago, I wouldn’t go as far as saying you haven’t found any solution. You are saying (so does the FBI, true?) to having found such solution; I’ll take that. Thus, me being very trustful, it may be correct then; at least according to your regular statements – right? So I will not doubt your idea (fact?) at all.

The only thing I’d claim is that your proposed solution is not directly derived from the Z340 cipher the way Z had actually sent it plenty of years ago. It doesn’t even fit a transposed version and nor does it when you slice the cipher into three parts. It has not even used a continuing en-/decryption method. Instead, the readable text shows irregular frequencies (e.g. letter A +42% etc.) but shows rather no related word combos (e.g. "taxi driver"); despite about 10 to be expected.

Also, I remember mentioning someday that IF the cipher is modified in combination with a possibility to enter any text on such modified cipher, there will be a strong risk of finding an ‘alternative’ solution (don’t remember which thread it was, something with the new software you are so good at [truly you are]); which would not represent the original cleartext.

Shortly, finding the K2 in the Himalaya is not having found the highest mountain over there. Simple like that. Therefore, I can’t say your solution is wrong (because I don’t know; at least not for sure) or that I would only doubt it. But I do follow my previous concerns and certainly would’nt say it is the Z original cleartext, either. To me, now this is my personal opinion, too: The message, if it is one, lacks of content; word relations; continuous method of decryption; normality regarding letter frequencies etc.

But as I said, that is just my opinion (and I will keep it for myself from time to time). Of course, there is no big need to discuss this part further as you have already explained your opinion by publishing your findings. Also, I am absolutely fine if your solution is (or is not) the absolute correct one. So far, btw, Z did not confirm – which, of course, is not the reason for my opinion (interestingly nobody had mentioned that). In any case, we should get along easily with that one.

BUT:
I think the risk for having found any text based on this different configuration of homophones (has been put in a completely different order not by Z but you guys) is very high. We talk about 26^63 different possibilities to not only find one solution or plenty of non-solutions, but also the risk of (wrong) alternative solutions, too. This risk, I am sure, increases tremendously when starting to modify the cipher itself.

So far, I am not aware about any other cipher having used so many multiple methods of encryption in only one cipher (homophones; transposition; slicing the cipher; adding prepared cleartext; shifting lines; anagramming etc.).

To me, it seems not very much plausible that Z had acted (encrypted) like this. It shall not be a secret that I see it to be a higher chance that some text had actually fit the re-structured cipher after its modifications (instead one step of encryption with e.g. 5 dyslectic errors..). To some extent, Z – my opinion again – had preferred ‘order’ in some way and last but not least, there already exist (!) multiple partial solutions with all of them being wrong, with e.g. ten or even twelve (longer) words. All I say is, to no surprise it was possible to create your result.

This result created by Z? Anybody may think about that one by him/herself. I don’t know – and I am not sure about my own opinion regarding your finding – either. Thus, yes, my prog is still running.

QT

(Pretty sure you play chess, do you? You are invited to play one day – me Elo is way above 200 – I think the whole world heard your ‘Check!’ so it was a great step anyway in the history of solving the Z case!).

while i respect the attempt of the recent news, i’ve debunked that solution as far as i can see. please see http://www.zodiac340.com

while i respect the attempt of the recent news, i’ve debunked that solution as far as i can see. please see http://www.zodiac340.com

Try rearranging the ciphertext first using "one down, two across". You are comparing the decoded message with the original 340 layout. Here is the rearranged format.

QT –

I’m working on the next video now, which goes over some further verifications and demonstrates how we think the cipher might have been put together. Maybe that will help in your analysis.

The only thing I’d claim is that your proposed solution is not directly derived from the Z340 cipher the way Z had actually sent it plenty of years ago.

I’m not sure what this means. Our solution is derived directly from the Z340 cipher, in the sense that the cipher’s key also includes the transposition steps. He sent a transposition cipher, and we decrypted a transposition cipher.

It doesn’t even fit a transposed version and nor does it when you slice the cipher into three parts. It has not even used a continuing en-/decryption method. Instead, the readable text shows irregular frequencies (e.g. letter A +42% etc.) but shows rather no related word combos (e.g. "taxi driver"); despite about 10 to be expected.

In what way does it fail to fit a transposed version? Are you referring to the errors in the 2nd part?
Irregular frequencies are to be expected when taking 340-character samples from any English text.
Certainly the 408’s plaintext doesn’t match average letter frequencies perfectly.

How do you determine "related word combos"? Why is "taxi driver" a related word combo but things like "gas chamber" are not?

I really don’t get QT’s point about letter frequency either. When you study a huge text corpus, the letter frequency will generally tend to approach "ETAOIN…" order, but in any short message, can vary quite wildly. One could look at the letter frequencies of various random posts here on this board to see that.

And Zodiac’s message contains multiple (3x) uses of the word PARADICE and AFRAID, so of course the letter A would appear more often.

Average letter frequency is a rule of thumb, not some kind of a set-in-stone law.

Anyone want to try to brute force a Z13 solution using the name Robert?

You have 1 repeat already. You have 4 of the five most common letters.

That’s a good start.

I’m working on the next video now, which goes over some further verifications and demonstrates how we think the cipher might have been put together. Maybe that will help in your analysis.

Thanks but meanwhile I have understood how your result is set-up.

I’m not sure what this means. Our solution is derived directly from the Z340 cipher, in the sense that the cipher’s key also includes the transposition steps. He sent a transposition cipher, and we decrypted a transposition cipher.

Please let me clarify: Your result is respectable but it is not a direct transformation of the Z340 at all. You claim it had been from a transposed version but this is also only partially correct, with all respect. Even if you slice the cipher, it does not work out. There still is the "Life is" section – which is the only way how you get to the combo ‘gas chamber’, btw – and even then it did not work out but had one line to be shifted and anagramming to be used to get the result over the complete length of the cipher.

Although a bit complicated, it indeed is possible to follow your thoughts from the original cipher to the result. But I think you go too far if you mention Z had sent a transposition cipher (unless you believe in your own result, which is the reason why I suggested to not discuss this too much).

Maybe you want to ask yourself:

a) Are you 100% sure that Z had used all of the steps above to create this result (cipher)?
b) Are you 100% sure that no other solution, like the Z408, exists without the steps of transposing/slicing/textinsertion/lineshifting/anagramming?

In any way, you have described the way to get to your result (and it is a result). Never denied that one. And, again, so far I do not know if this result was written by Z (or if he had sent a transposition homophone substitution with line shifting, text insertion and anagramming at all).

In what way does it fail to fit a transposed version? Are you referring to the errors in the 2nd part?

Not referring to your result in detail (because I was able follow your thoughts). I only think there could (!) exist a solution without all of those ‘steps’ (see above).

Irregular frequencies are to be expected when taking 340-character samples from any English text.

Some irregular frequencies, ok. Even up to 10%, 20% or 30%. But 42%?

Certainly the 408’s plaintext doesn’t match average letter frequencies perfectly.

Correct. But it made me check the word combos in your result. And then I realized that there is a lack of such, too. ‘Gas chamber’ would only occur because of one of your additional steps. ‘Tv show’, at least to me, is not enough to say it is a full word combo. There should be about ten. Average length of words is is shorter than expected, too. Then, what does remain? "Enough slaves" is not necessarily a combo, the words do not necessarily refer to each other. And even that term consists of 75% frequent letters..it raised one question:

Was the result ‘created’ instead of being some original Z cleartext? This question is in my head and so far it did not get out of it (sorry, couldn’t get it out).

How do you determine "related word combos"? Why is "taxi driver" a related word combo but things like "gas chamber" are not?

Referred to that one already; I’d determine that one word is related to the other not only be being followed or preceeded but also its inner meaning (e.g. taxi drives / driver uses taxi).

Overall, I’ll take it as a challenge. If it will happen to crack the cipher in the traditional way, I am fine. And so am I if your result is the last one ever to be found on Z340. So far, yours is certainly (one of the) best ones.

@Glurk

Took the challenge and did so – with your own (last) post:

Despite the use of ‘paradice’ and ‘against’ in it, it had a 7.97% letter ‘A’ frequency (compared to an expected ~8.04%).

Thus, your post is absolutely compliant to the English language. Your letter ‘A’ frequency deviated less than one (!) percent (total: 33 out of 414 letters). This, compared to a 42% deviation, is a normal frequency instead of a very irregular one.

The irregularity, imo, cannot be denied due to existing data (one may count the letters..). I am convinced it being extremely seldom that the letter ‘A’ occured with an overall 10% or 11% frequency. In one sentence? Ok, might be. But not in a text of length 340, consisting of approximately 70 words (nothing to do with the ETAOIN order).

Also, I do not want to be over-critical in this rather sensitive issue; which is why I fully accept all of your statements.

QT

http://scottbryce.com/cryptograms/stats.html

QT-

What I said (quoting myself) was:

One could look at the letter frequencies of various random posts here on this board to see that.

That’s pretty clear that I meant an assortment, not just my single post.

But I still don’t get what you are saying. Here is a chart of the letter frequencies in the 340 solution vs. expected frequencies. A few more "A"’s than expected, and a few less "T"s and "L"s. Looks totally normal to me.

One could even speculate that Zodiac INTENTIONALLY skewed the letter frequencies to increase difficulty. But IMO they look totally normal. Keep in mind the solution contains the 12-gram "IAMNOTAFRAID" twice. That’s a bunch of "A"s right there, in a perfectly readable solution.

There’s always going to be few contrarians who won’t acknowledge the solution no matter what you say.

It’s obvious really what’s happened here, ANTIFA have… gone back in time… and planted… bare with me here, I’m still figuring it out… and planted a reference to the Jim Dunbar show in the cipher… it obvious if you know this one weird trick.

What people don’t realize is that Doranchak’s computer must have privately held the preconceived notion that the cipher message contained a reference to the Jim Dunbar show. The computer acted all impartial and objective, but in reality the computer was secretly a die hard Jim Dunbar fan, and knew in it’s little clockwork heart that it’s hero would get a mention. So day and night the computer kept trying to force "Jim Dunbar" into the solution and finally out of frustration the computer just straight up anagrammed "Jim Dunbar" in, but then it’s little ROM drives got all sweaty and nervous, so before Doranchak had time to check, the computer quickly switched the result to a more vague reference to a "TV Show", crossing it’s little circuits that Dave would see it… see it and know the truth, the truth that Jim Dunbar was a god.

That white noise left over from the Big Bang? it’s actually the whispered sound of Jim Dunbar’s voice. Fractals? Actually made up of thousands of tiny Jim Dunbars within Jim Dunbars. Time travel? Jim Dunbar. Göbekli Tepe? Jim Dunbar build it. In fact when you think about it, Jim Dunbar’s pretty ho hum, I bet you’ll find him in just about anything if you stare at it hard enough.

You know what, I shouldn’t ridicule, that’s not an argument. But no ones got time to listen to every brain fart. So I’ll make a deal, if anyone can articulate a refutation that is more coherent than the found solution, I’ll listen to it.

Alright, I’ll stop. Look, a second message? maaaaaaaybe, but let’s not poke worthless pin holes in a solution that is as believable as, well I’m sure a great many things you take for granted everyday. You don’t want to go giving yourself an existential malaise.

There still is the "Life is" section – which is the only way how you get to the combo ‘gas chamber’, btw

That’s actually not true. "Gas chamber" appears in the first section, which can be strictly untransposed without any further modifications such as error corrections and ignoring text such as "Life is".

– and even then it did not work out but had one line to be shifted and anagramming to be used to get the result over the complete length of the cipher.

In my mind, the last 2 lines could have been complete filler and/or nonsense, and I’d STILL be convinced that the first 2 sections are correct decryptions.
Further, we only had to do word reversals, rather than very permissive "full" anagramming, which WOULD yield more possible plaintexts as you seem to be implying.

Although a bit complicated, it indeed is possible to follow your thoughts from the original cipher to the result. But I think you go too far if you mention Z had sent a transposition cipher (unless you believe in your own result, which is the reason why I suggested to not discuss this too much).

The way I would put it is: The solution we found is extremely strong evidence that Zodiac used a transposition scheme to create the 340.
And also: The limited length of the key (which includes substitution, transposition details, and error corrections) is still sufficiently small when compared to the entire message, such that production of the plaintext we found is astronomically unlikely to be by chance. The details of that calculation should make for a fascinating research paper some day.

As an example of key length vs cipher length: They key length of the 408 is only 54 characters. The cipher length is 408 (and readable plaintext is length 390). So by making 54 choices, you get a plaintext that is 720% longer than the information you put in. For the 340, that ratio is smaller, but still substantially larger than 1-to-1. Computing the exact ratio will require some effort.

a) Are you 100% sure that Z had used all of the steps above to create this result (cipher)?

No. The decoding steps are different than the encoding steps. We have suspicions about his encoding steps, but we can’t be 100% sure about it, because different steps can produce the same results we see in the cipher solution.

b) Are you 100% sure that no other solution, like the Z408, exists without the steps of transposing/slicing/textinsertion/lineshifting/anagramming?

No, but I would put my certainty level at around 99.9%.
Anything’s possible! But I think the math would show that an equally coherent solution with fewer of those steps (or none of them) is extremely unlikely to exist.

Not referring to your result in detail (because I was able follow your thoughts). I only think there could (!) exist a solution without all of those ‘steps’ (see above).

It doesn’t hurt to keep looking if you feel that way. If you or someone did find one, that’d be an amazing development!

Some irregular frequencies, ok. Even up to 10%, 20% or 30%. But 42%?

Yes. For example, any sample from Gadsby is gonna have skewed stats.
https://en.wikipedia.org/wiki/Gadsby_(novel)

I think the question you should be asking is: If you take a million samples of text from random English-language sources, how often do you find the distribution of A’s to be anomalous?

Correct. But it made me check the word combos in your result. And then I realized that there is a lack of such, too. ‘Gas chamber’ would only occur because of one of your additional steps. ‘Tv show’, at least to me, is not enough to say it is a full word combo. There should be about ten. Average length of words is is shorter than expected, too. Then, what does remain? "Enough slaves" is not necessarily a combo, the words do not necessarily refer to each other.

I’m having a hard time understanding how you decide which word combos are valid. Seems very subjective.

I can kind of understand what you mean if I look at other Z340 solves that have come my way, because they usually have very "forced" phrasing, and word combinations that don’t make much sense (for example, see the Graysmith solution). I don’t think our Z340 solve suffers from that problem at all.

Was the result ‘created’ instead of being some original Z cleartext? This question is in my head and so far it did not get out of it (sorry, couldn’t get it out).

I don’t believe we created or forced the plaintext we found. It’s hard to demonstrate exactly why, though, because it requires a lot of technical analysis and math to show why the key size (and the choices that go into it) is still sufficiently small compared to the cipher length. Even the line shift + skipping "life is" does not add much information to the key, given how much more text gets clearly resolved by making those choices. On the other hand, if the cipher was only, say, 100 characters long, then maybe there would have been a lot more ways to generate false solutions using the same key operations. That would be an interesting experiment to try!