What was the Zodiac killer thinking anyway? "Oh, they cracked my cipher. I’ll make my next cipher unbreakable by constructing an off-road plaintext with out of range english letter frequencies." It doesn’t seem likely, and then, what was he thinking anyway?

i don’t think it’s a leap to assume he was surprised how quickly the 408 was solved, especially since it basically didn’t say anything – he claimed "in this cipher is my identity" but in reality the cipher appears to have been an excuse to drum up hype. so i’m guessing it was a big hit to the ol’ serial killer ego to have some mom and pop from nowheresville solve it in a week. all of his communications were of the "haha you’re wasting all this time and you’ll never catch me "type, so again i don’t think it’s a stretch that the 340 was intended to never be solved. based on the way he sought thrills (imo more through written communication than the actual murders) i don’t think he’d make it gibberish. i think he would get his kicks from the fact that it does say SOMETHING but that thousands of people will spend decades trying to figure out exactly what it says.

all that to say i doubt it’s the same type of cipher as the 408 for two reasons – 1) it was shown that a homophonic substitution cipher could be cracked easily, 2) he wanted to prove to himself that he was the smartest guy in the room. some interesting ideas that have been suggested are that other writings have clues (i think this is possible but would focus more on either the dripping pen card or the sf chronicle letter he sent within a day or two of the cipher since that’s where his head was at the time) or that the solution, however convoluted, is a lot simpler than the attempts being made to solve the cipher (e.g. – shift every fifth letter nine times, or he omitted all of the letter "e", etc).

side note – i think it’s weird that the nov 9 letter to the chronicle says "up to the end of october i’ve killed seven people" as though he wrote the letter earlier than he sent it. in other words, why not just say "so far i’ve killed seven people" unless you’re not planning to send the letter for a while. i mention this only because it’s not necessarily the case that the nov 9 bomb letter and the 340 cipher were created in the same span of time. in fact, the dripping pen letter with the 340 says "…before you hear the bad news. you won’t get the news for a while yet…" which seems to me like he’s talking about the bomb letter (the bad news = he’s making a bomb) but they get that letter the next day. again, this all fits in with my theory that he didn’t live in the area and sent the letters when he was passing through, but for the purposes of the thread i simply mean to state that it’s possible we can extract more information from the nov 8 and nov 9 letters together, although there’s some evidence he may not have intended them to be sent so close together.

Thanks for your reply masootz,

I agree with your thinking, he really got a thrill out of these ciphers.

or that the solution, however convoluted, is a lot simpler than the attempts being made to solve the cipher (e.g. – shift every fifth letter nine times, or he omitted all of the letter "e", etc).

Funny that you mention that because I have been looking at letter shifts the last couple of days. It could explain why the "+ symbol" is so dominant in the 340.

Letter omittance (not in the form of using words without that letter) is a good idea and I’m going to look into that straight away. I wrote a program a la ZKDecrypto which enables me to test things like that out quickly.

I can only hope there are some clues.

I took another look at the prime phobia observation (by Dan Johnson) of certain symbols in the 340 mentioned here: http://zodiackillerciphers.com/wiki/ind … servations

The most frequently occurring symbol, +, occurs 24 times. Only once does it fall on a prime-numbered position in the cipher text (counting from 1 to 340), against expectations.

340 numerical: https://www.dropbox.com/s/agb38v5qzase4 … s.txt?dl=0
408 numerical: https://www.dropbox.com/s/7zb1ul620uxd1 … s.txt?dl=0

I compiled 2 texts of the numerical version of the 340 and 408. Which for each symbol show the position they fall on (if you treat the ciphers as being on a line). With prime and evil number counts and modulo’s. If there is something to it possible patterns might be observed here.

For example, symbol 19 (the "+ symbol") in the 340 looks like this in the file:

symbol 19: 20 40 64 65 72 81 105 128 133 140 142 159 162 172 201 211 237 238 255 276 282 290 291 340
count: 24 prime: 1 evil: 15
mod 2: 0 0 0 1 0 1 1 0 1 0 0 1 0 0 1 1 1 0 1 0 0 0 1 0
mod 3: 2 1 1 2 0 0 0 2 1 2 1 0 0 1 0 1 0 1 0 0 0 2 0 1
mod 4: 0 0 0 1 0 1 1 0 1 0 2 3 2 0 1 3 1 2 3 0 2 2 3 0
mod 5: 0 0 4 0 2 1 0 3 3 0 2 4 2 2 1 1 2 3 0 1 2 0 1 0
mod 6: 2 4 4 5 0 3 3 2 1 2 4 3 0 4 3 1 3 4 3 0 0 2 3 4
mod 7: 6 5 1 2 2 4 0 2 0 0 2 5 1 4 5 1 6 0 3 3 2 3 4 4
mod 8: 4 0 0 1 0 1 1 0 5 4 6 7 2 4 1 3 5 6 7 4 2 2 3 4
mod 9: 2 4 1 2 0 0 6 2 7 5 7 6 0 1 3 4 3 4 3 6 3 2 3 7
mod 10: 0 0 4 5 2 1 5 8 3 0 2 9 2 2 1 1 7 8 5 6 2 0 1 0
mod 11: 9 7 9 10 6 4 6 7 1 8 10 5 8 7 3 2 6 7 2 1 7 4 5 10
mod 12: 8 4 4 5 0 9 9 8 1 8 10 3 6 4 9 7 9 10 3 0 6 2 3 4
mod 13: 7 1 12 0 7 3 1 11 3 10 12 3 6 3 6 3 3 4 8 3 9 4 5 2
mod 14: 6 12 8 9 2 11 7 2 7 0 2 5 8 4 5 1 13 0 3 10 2 10 11 4
mod 15: 5 10 4 5 12 6 0 8 13 5 7 9 12 7 6 1 12 13 0 6 12 5 6 10
mod 16: 4 8 0 1 8 1 9 0 5 12 14 15 2 12 9 3 13 14 15 4 10 2 3 4
mod 17: 3 6 13 14 4 13 3 9 14 4 6 6 9 2 14 7 16 0 0 4 10 1 2 0
mod 18: 2 4 10 11 0 9 15 2 7 14 16 15 0 10 3 13 3 4 3 6 12 2 3 16

Modulo is a simple mathematical function (binary input) which tries to fit the "modulo" as many times as possible into a number without exceeding it and then outputs the rest, for example: 17 mod 2 = 1. Because 2 fits in 17 eight times (16) and then the rest is 1 (17-16). So if the output is 0 it means the number is divisible by.

340: https://www.dropbox.com/s/levhwlmppmrln … t.txt?dl=0
408: https://www.dropbox.com/s/yyuch6nurwewz … t.txt?dl=0

I personally think there is not much out of the ordinary here and I just want to share some data. Both the 340 and the 408 exhibit prime phobia for certain symbols. Which I believe is due to the characteristics of the plaintext and inflation of this by the homophonic distribution. But I like to keep an open mind.

Update,

I started working on columnar transposition in my program. I fully transpositioned the 408 this way as a test cipher, so far it doesn’t seem to catch on so I have to figure out what the problem is. The searchspace is the permutations of 17 elements (17! or 355687428096000 possible combinations) which should be very doable for a heuristic search but I believe that the hill to climb here is a very narrow and steep one for the ngrams. And it might get some elements right in a row but at the wrong offset.

I may need to use a random key layer to get better starting points, or an entirely different approach. Best matching couples, tripples and permutate from there… Getting it to work may be big, as then I can use it as a base for other possible transpositions.

A question for the cipher guys, do you think it actually possible/likely the 340 is columnary transposed?

https://www.dropbox.com/s/r0t0zat1ng23y … 3.jpg?dl=0

Letters are colored brighter by corpus frequency.

For example, symbol 19 (the "+ symbol") in the 340 looks like this in the file:

symbol 19: 20 40 64 65 72 81 105 128 133 140 142 159 162 172 201 211 237 238 255 276 282 290 291 340

i started doing the same this weekend but was interrupted by life. do you have data on the space frequency between letter repetitions? if you do, i’d rather not reinvent the wheel. basically for example, symbol 19 would be 20, 24, 1, 7, 9, 24, 13, 5, 7, etc – it’s obvious he was irregular with his character usage (some repeating characters are next to each other while others only show up at opposite ends of the cipher) but i want to see if there’s a discernable pattern in the repetition and compare that to the frequency we’d expect characters to be repeated.

FWIW, my poi knew computer programming and used Cobal and Fortran languages. He explained the basics to me and I remember him talking about binary and something about making a mistake halfway into something and not wanting to start over.

The dropped letter idea is interesting as in speed writing as one would do when taking notes in school. Has anyone thought of "pig latin?" Or Roman Numerals?

masootz is this okay? If you want anything added let me know I should be able to program it easily.

symbol 1: 49 251 26
symbol 2: 175 71
symbol 3: 63 22 56 29 19 29 52
symbol 4: 95 150 75
symbol 5: 14 34 1 8 24 78 18 83 41 23
symbol 6: 51 34 65 35 31 62
symbol 7: 49 33 71 66 40
symbol 8: 52 55 43 76 16
symbol 9: 71 218
symbol 10: 74 13 212 21
symbol 11: 66 42 59 16 19 43 31 20 30
symbol 12: 230 36
symbol 13: 34 71 17 142 12
symbol 14: 170 13 78 27
symbol 15: 28 71 7 102 48
symbol 16: 52 32 10 12 23 43 41 43
symbol 17: 59 19 42 173
symbol 18: 112 82 13 103
symbol 19: 20 24 1 7 9 24 23 5 7 2 17 3 10 29 10 26 1 17 21 6 8 1 49
symbol 20: 14 112 21 13 22 13 24 21 25 29 4
symbol 21: 23 57 18 98 45 41
symbol 22: 25 55 23 84
symbol 23: 39 29 35 36 24 77 35 19 16
symbol 24: 83
symbol 25: 75 78 147
symbol 26: 17 241 8 3 15
symbol 27: 103 49 137
symbol 28: 87 84 32 26 9
symbol 29: 108 28 23 92 31
symbol 30: 28 35 101 48 77
symbol 31: 35 39 47 33 76 76
symbol 32: 91 103 41
symbol 33: 12 242 4 13
symbol 34: 113 20 40 30
symbol 35: 207
symbol 36: 74 36 42 27 16 24 17 23 26
symbol 37: 32 68 46 14 70 56
symbol 38: 70 46 14 80
symbol 39: 101 109 81
symbol 40: 24 44 15 23 50 55 35 37
symbol 41: 72 38 32
symbol 42: 156 92 31
symbol 43: 110 148
symbol 44: 29 133 115
symbol 45: 198
symbol 46: 60 187
symbol 47: 66 71 39
symbol 48: 102 60 78
symbol 49: 50
symbol 50: 26 71 21 18 14 25
symbol 51: 11 8 48 21 37 11 9 59 25
symbol 52: 120
symbol 53: 69 46 32 5 10
symbol 54: 89 129
symbol 55: 61 16 84 67
symbol 56: 151 33 14 28
symbol 57: 163
symbol 58: 110 120
symbol 59: 128
symbol 60:
symbol 61: 42 48
symbol 62: 186
symbol 63: 55 73

capricorn,

I think something like pig latin would not pose a problem for a deciphering program, and what do you mean by roman numerals? I have checked for single dropped letters but nothing came out of it. And I also believe something like that would have been discovered by human efforts long ago.

I believe the Zodiac understood why the 408 was so "easily" cracked and really did his best to come up with a cipher that did not have the same problem(s).

capricorn,

I think something like pig latin would not pose a problem for a deciphering program, and what do you mean by roman numerals? I have checked for single dropped letters but nothing came out of it. And I also believe something like that would have been discovered by human efforts long ago.

I believe the Zodiac understood why the 408 was so "easily" cracked and really did his best to come up with a cipher that did not have the same problem(s).

Unless the solve for the 408 was incorrect. I know a lot of effort has been put in to solving the other ciphers, but has anyone looked as hard at the method used to solve the 408 to see if it is sound and there are no other possibilities? Just thinking outside the box, but what if the solve was incorrect and Z knew it was so he kept going after realizing it could not be solved?

IIRC, there are lots of L’s, C’s, X’s, V’s and I’s in these codes. These letters = numbers in Roman Numerals.

I am stymied by these codes but just suggesting this as a possible hint or clue that might help decipher the codes.

IMO, whoever Z was/is could not be so smart as to outwit all the many sophisticated attempts to solve this. I’m trying to think of things he could have done that have not been tried.

The Hardens solution is questionable in my mind. I could not really follow their explanation of how they solved it and then there was the question raised about Mrs. Harden and a book or something she left behind after she died.

If they were correct, then I agree with whoever said Z was upset it was solved so quickly so he changed his method.

What if he, like my guy, studied a foreign language in school, ie. German and knew just enough of it to write something in that language and guessed that everyone would assume it was written in English?

I am just trying to think of things that have not yet been tried and that would be something anyone so inclined with normal or slightly above normal intelligence could make up. Sorry if these things have already been suggested…there is just too much to remember and I don’t have time to go back and review. I really don’t remember too much about all the posts about the codes either as so many of them are way over my head and just too complicated.

capricorn,

I think something like pig latin would not pose a problem for a deciphering program, and what do you mean by roman numerals? I have checked for single dropped letters but nothing came out of it. And I also believe something like that would have been discovered by human efforts long ago.

I believe the Zodiac understood why the 408 was so "easily" cracked and really did his best to come up with a cipher that did not have the same problem(s).

Unless the solve for the 408 was incorrect. I know a lot of effort has been put in to solving the other ciphers, but has anyone looked as hard at the method used to solve the 408 to see if it is sound and there are no other possibilities? Just thinking outside the box, but what if the solve was incorrect and Z knew it was so he kept going after realizing it could not be solved?

I like the idea. There could be an additional message encoded in the 408 but then I suppose it would not run full length or it would be a subset at least. The errors might not be errors etc. Take for instance the Belli letter: http://bermuda-triangle.org/assets/imag … 0_1969.jpg The errors in here are just to lame.

IIRC, there are lots of L’s, C’s, X’s, V’s and I’s in these codes. These letters = numbers in Roman Numerals.

I am stymied by these codes but just suggesting this as a possible hint or clue that might help decipher the codes.

IMO, whoever Z was/is could not be so smart as to outwit all the many sophisticated attempts to solve this. I’m trying to think of things he could have done that have not been tried.

The Hardens solution is questionable in my mind. I could not really follow their explanation of how they solved it and then there was the question raised about Mrs. Harden and a book or something she left behind after she died.

If they were correct, then I agree with whoever said Z was upset it was solved so quickly so he changed his method.

What if he, like my guy, studied a foreign language in school, ie. German and knew just enough of it to write something in that language and guessed that everyone would assume it was written in English?

I am just trying to think of things that have not yet been tried and that would be something anyone so inclined with normal or slightly above normal intelligence could make up. Sorry if these things have already been suggested…there is just too much to remember and I don’t have time to go back and review. I really don’t remember too much about all the posts about the codes either as so many of them are way over my head and just too complicated.

I’ll keep an eye on the roman numerals. I believe Zodiac could have outwitted everyone because it is for example known that it is harder to solve a test then to make it. I find it a very good exercise trying to come up with things that have not been tried before.

Different lanuguages already have been tried to some extent, ZKDecrypto comes with Italian, German and Spanish language ngrams. I remember reading doranchak also tested out allot of languages that have been found in the Guttenberg project. But it could still be I guess. Or even multi-languagal?

The Hardens solution to the 408 is very sound. It is a homophonic substitution cipher where every specific symbol maps to a specific letter in the alphabet, nothing more. The entropy is such that only one good solution can come out of it, as shown by various deciphering program ranging from amateurs to university studies. There could be something else there but then it seems likely that it is hidden somehow in the words/text themselves, or the last 18 characters, or the errors, or a combination of those.

masootz is this okay? If you want anything added let me know I should be able to program it easily.

symbol 1: 49 251 26
symbol 2: 175 71
symbol 3: 63 22 56 29 19 29 52
symbol 4: 95 150 75
symbol 5: 14 34 1 8 24 78 18 83 41 23
symbol 6: 51 34 65 35 31 62
symbol 7: 49 33 71 66 40
symbol 8: 52 55 43 76 16
symbol 9: 71 218
symbol 10: 74 13 212 21
symbol 11: 66 42 59 16 19 43 31 20 30
symbol 12: 230 36
symbol 13: 34 71 17 142 12
symbol 14: 170 13 78 27
symbol 15: 28 71 7 102 48
symbol 16: 52 32 10 12 23 43 41 43
symbol 17: 59 19 42 173
symbol 18: 112 82 13 103
symbol 19: 20 24 1 7 9 24 23 5 7 2 17 3 10 29 10 26 1 17 21 6 8 1 49
symbol 20: 14 112 21 13 22 13 24 21 25 29 4
symbol 21: 23 57 18 98 45 41
symbol 22: 25 55 23 84
symbol 23: 39 29 35 36 24 77 35 19 16
symbol 24: 83
symbol 25: 75 78 147
symbol 26: 17 241 8 3 15
symbol 27: 103 49 137
symbol 28: 87 84 32 26 9
symbol 29: 108 28 23 92 31
symbol 30: 28 35 101 48 77
symbol 31: 35 39 47 33 76 76
symbol 32: 91 103 41
symbol 33: 12 242 4 13
symbol 34: 113 20 40 30
symbol 35: 207
symbol 36: 74 36 42 27 16 24 17 23 26
symbol 37: 32 68 46 14 70 56
symbol 38: 70 46 14 80
symbol 39: 101 109 81
symbol 40: 24 44 15 23 50 55 35 37
symbol 41: 72 38 32
symbol 42: 156 92 31
symbol 43: 110 148
symbol 44: 29 133 115
symbol 45: 198
symbol 46: 60 187
symbol 47: 66 71 39
symbol 48: 102 60 78
symbol 49: 50
symbol 50: 26 71 21 18 14 25
symbol 51: 11 8 48 21 37 11 9 59 25
symbol 52: 120
symbol 53: 69 46 32 5 10
symbol 54: 89 129
symbol 55: 61 16 84 67
symbol 56: 151 33 14 28
symbol 57: 163
symbol 58: 110 120
symbol 59: 128
symbol 60:
symbol 61: 42 48
symbol 62: 186
symbol 63: 55 73

you just saved me hours of work! thanks a bunch!