Okay smokie take your time.
Just added a Length operation to the manipulation solver:
Length(a,b,c)
a: 1=add, 2=remove
b: 1=at the start of the cipher, 2=at the end of the cipher
c: length of the piece
I tried it on the 408 adding/removing anywhere from 1 to 40 characters with "Multiplicity weight" setting at 1.3 and the following solve had the highest score. This operation can be used if it is suspected that there is filler at the end or start of the cipher, or if a part of the cipher was actually removed to complicate solving transposition. It could also be used if it is suspected that a contiguous cipher has been buried between filler.
AZdecrypt 0.992 (Practical Cryptography 5-grams) Manipulation, Index of coincidence Manipulation operation(s): Length(2,2,19) Score: 25115.19 Ioc: 0.06363793 Entropy: 4.17308 Chi-square: 54.1645 Characters: 389 Letters: 23 ilikekillingpeopl ebecauseitissomuc hfunitiamorefunth ankillingwildgame intheforrestbecau semanisthemoatdan gertueanamalofall tokillsomethinggi vesmethemoatthril lingexperenceitis evenbetterthanget tingyourrocksoffw ithagirlthebestpa rtofitiathaewheni dieiwillbereborni nparadiceandallth eihavekilledwillb ecomemyslavesiwil lnotgiveyoumyname becauseyouwilltry tosloidownoratopm ycollectingofslav esformyafterlif Multiplicity: 0.1388175 Characters: 389 Symbols: 54 Sequential: 2925.50 Bigrams: 53 9%P/Z/UB%kOR=pX=B WV+eGYF69HP@K!qYe MJY^UIk7qTtNQYD5) S(/9#BPORAU%fRlqE k^LMZJdrpFHVWe8Y @+qGD9KI)6qX85zS( RNtIYElO8qGBTQS#B Ld/P#B@XqEHMU^RRk cZKqpI)Wq!85LMr9# BPDR+j=6N(eEUHkF ZcpOVWI5+tL)l^R6H I9DR_TYrde/@XJQA P5M8RUt%L)NVEKH=G rI!Jk598LMlNA)Z(P zUpkA9#BVW+VTtOP ^=SrlfUe67DzG%%IM Nk)ScE/9%%ZfAP#BV peXqWq_F#8c+@9A9B %OT5RUc+_dYq_^SqW VZeGYKE_TYA9%#Lt_ H!FBX9zXADd7L!=q _ed##6e5PORXQF%Gc Z@JTtq_8JI+rBPQ
I concluded the polyliteral period test, well, the computer acted weird and all 4 instances of AZdecrypt crashed. It nearly made my intended target of 2 million restarts but I don’t have much statistics. All possible unit sizes and periods were used, this is certainly not a definitive answer if such a thing is going in the 340.
The results are strange: https://drive.google.com/open?id=0B5r0r … mdncEt5czA
Number of instances that scored higher than 20500:
340 normal: 648 <— no typo
340 reversed: 2
340 mirrored: 15
340 flipped: 65
Though I remember that average scores and bigram counts were higher for mirrored and flipped. Why would the normal orientation have so many more results over 20500 while average scores were higher for both mirrored and flipped? Since all unit sizes (polyliteral) were considered, the bulk of these transpositions move information around in pieces that are rather large and it can be noticed that most of these appear to be rearrangements of the same stuff that can be found in the 20351 scoring normal 340.
To give an example here is the normal 340 maxed out with AZdecrypt side to side with such a rearrangement:
340: transposition: ----------------- ----------------- gshiscentopernatm gmhiscentopursatm estelsilarldeiavo ostelsinalldeiava edvantocarlorsegi edcantocallarsegi nsspectingsitthat nsspectingsitthat sunterimporttheof sinterimporttheof sthercisimporital sthercrimentatcht sotablytoacanderp readymeaseconteit lativisitseletoru isperedsothformpa mentatchtreadysea lesasnaiteachipio seconteitispereds sendtheremindstop othforspalesannai orealsthcaroniath teachipionendther astrummetrclepart esundsteporealyth talmelatingsflasp careevoteadanertt romlepreslieismai deceiveupsinocost nagosonecitypasti padgeealisedvnbat simmoritalsotains hantrespetrcrepor toabanderplativis ttorperatingnflos itsolethevoteadan promrepreslieispa erttdebuiceimmino inagesonecitypayt costpadgeealisedv AZdecrypt 0.992 (Practical Cryptography 5-grams) Manipulation, Index of coincidence Manipulation operation(s): Period(2,91,134) Score: 20701.08 Ioc: 0.07376366 Entropy: 3.919348 Chi-square: 60.33494 Characters: 340 Letters: 20 gmhiscentopursatm ostelsinalldeiava edcantocallarsegi nsspectingsitthat sinterimporttheof sthercrimentatcht readymeaseconteit isperedsothformpa lesasnaiteachipio sendtheremindstop orealsthcaroniath astrummetrclepart talmelatingsflasp romlepreslieismai nagosonecitypasti simmoritalsotains toabanderplativis itsolethevoteadan erttdebuiceimmino costpadgeealisedv Multiplicity: 0.1852941 Characters: 340 Symbols: 63 Sequential: 1710.11 Bigrams: 26 HER>pl^VPk|1LTG2d Np+B(#O%DWY.<*Kf) By:cM+UZGW()L#zHJ Spp7^l8*V3pO++RK2 _9M+ztjd|5FP+&4k/ p8R^FlL9d<M+b+ZR2 FBcyA64K-zlUV+^J+ Op7<FBy-U+R/5tE|D YBpbTMKO2<clRJ|*5 T4M.+&BFz69Sy#+N| 5FBc(;8RlGFNMqG2R cT+L16C<+FlWB|)L+ +)WCzWcPOSHT/()p| FkdW<7tB_YOB*-Cc> MDHNpkSzZO8A|K;+O -*dCkF>2D(#5+Kq%; 2UcXGV.zL|(G2Jfj# O+_NYz+@^f524b.cV 4t++yBX1*:49CE>VU Z5-+|c.3zBK(Op^.f
You could have inscribed the message vertically x=22 and y=16, and then transcribed horizontally. Or perhaps you inscribed the message horizontally x=16 and y=22 and then transcribed vertically. Either way, there are 12 blank spaces. It is the "I like killing" message. I untransposed it imperfectly and recognized a lot of the words, the longest of which was "SOMETHING". Parts of the message are scrambled all over the place.
EDIT: Here is my solution. I put your solver on progressive mode, and most of the solutions scored the same. There was not much variance after about a dozen attempts.
Score: 22618 Ioc: 636 M: 151 C: 329 S: 50 ecauseyouwilltryo tgiveyoumynamebme myslavesiwillnvek illedwillbecodice andalltheihallber eborninparasthatw henidieiwithebest partofitiocksoffw ithagirlthangetti ngyourreitisevenb etterrillingexper encivesmethemostt hokillsomethinggo usanimalofalltist hemostdangerrrest becausemanwildgam einthefoefunthank illingomuchfuniti smorplebecauseiti ssilik
I thought that it could have been period 12 with the message written backwards, but that didn’t have nearly as many repeats. I don’t know why and have to look closer at that.
I am working on my system, and have a lot to do. It resembles some of the work shown in the NSA and Army manuals because it is so basic. I am not ready for diagonal yet. Please give me another horizontal – vertical with another message that I do not recognize. Keep any dimension 68 or under.
EDIT: For step 1, I made a heatmap of period x matches. It works in offsets. So you have to add or subtract the numbers on the left columns and top rows to get the period. 17 on the leftmost column + 5 on the top row = 22, where there are 66 period 22 matches. That is how I figured out 22. And period 44 at row 2, column 10 has 41 matches, in a diagonal line from the upper left cell through row 1, column 5. So that was an easy one, but like I said, I thought it could also be 17 – 5 = 12. I will post the 340 tomorrow morning.
Just an idea:
What about columnar transposition using a keyword. Im thinking of a 5 letter keyword, 5 columns, 68 rows. The first 4 rows of the 340 will fall under one column, one right after the other. Another column will be the next 4 rows in same fashion. Same for columns 3,4, and 5. The columns are arranged per the keyword (52341) beginning with the first set of four.
Soze
I am working on my system, and have a lot to do. It resembles some of the work shown in the NSA and Army manuals because it is so basic. I am not ready for diagonal yet. Please give me another horizontal – vertical with another message that I do not recognize. Keep any dimension 68 or under.
I’m interested in what you will come up with. The heatmap looks really cool, also literally.
Should I tell yet what I did with my first message?
Here’s another.
pr1_jarlve2.txt 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 15 17 18 4 2 19 20 21 22 23 24 2 25 26 27 28 2 29 30 6 31 32 33 34 35 28 32 20 36 7 31 31 11 37 4 14 38 39 40 41 1 2 9 42 43 27 44 3 13 38 7 18 32 21 45 46 23 40 11 16 47 25 25 9 46 40 26 11 27 32 22 35 48 49 34 46 41 30 1 39 12 24 19 6 10 15 42 36 45 4 3 13 34 24 16 38 2 19 4 40 14 20 50 18 9 12 44 22 40 39 5 21 46 40 17 45 41 33 37 1 37 51 7 51 36 29 10 26 6 14 27 15 25 3 33 28 31 7 34 51 52 26 49 48 32 42 31 35 18 19 13 11 21 38 36 22 26 25 53 49 23 6 5 39 43 42 20 37 41 45 47 14 22 43 12 8 33 1 52 18 18 54 7 25 47 9 46 24 51 51 1 20 10 15 27 16 33 28 17 37 34 39 29 48 47 36 49 7 3 13 38 30 21 45 22 25 10 12 9 19 25 41 42 32 26 22 46 51 40 34 19 43 11 33 51 33 24 10 52 16 54 25 31 29 4 17 15 34 23 49 6 37 14 47 27 1 48 15 3 53 13 35 28 18 32 21 38 26 20 49 12 5 30 40 6 31 14 22 11 20 4 35 8 2 36 55 27 7 42 9 54 23 6 18 28 24 45 5 52 41 21 45 29 9 46 47 26 8 32 1 31 27 16 6 29 40 48 11 17 24 51 5 8 5
Just an idea:
What about columnar transposition using a keyword. Im thinking of a 5 letter keyword, 5 columns, 68 rows. The first 4 rows of the 340 will fall under one column, one right after the other. Another column will be the next 4 rows in same fashion. Same for columns 3,4, and 5. The columns are arranged per the keyword (52341) beginning with the first set of four.
Soze
That shouldn’t keep us from solving the 340. If you want you could construct a test cipher for us.
Here is the 340 on the left and the 340 mirrored on the right. The stats aren’t anything new, but they are arranged for pattern identification. The numbers in the cells are just the period x match count. I am also going to work on highest probability score.
On the left you can see the period 19 and period 38 boxed yellow. I didn’t mark the period 39, which scores higher. Period 5 and period 10 are interesting boxed pink. A lot of cryptography books show examples of 5 x 5 transpositions. How could that cause period 19 and period 39 repeats in the final product? Perhaps adding, deleting or moving an entire row or column at some point?
Then on the right is the mirrored, with period 15, 29 and 110 marked.
Thanks for the message.
That looks very nice.
How could that cause period 19 and period 39 repeats in the final product?
340 character plaintexts typically have increased bigram counts at period 1, 2 and 3 etc. When transposed into period 19 or similar. Period 1 becomes period 19. Period 2 becomes period 38. Period 3 becomes period 57.
My theory is that period 39 is actually period 38 that shifted away from its position due to an unknown misalignment problem. The same can be observed with the mirrored 340, where it is period 15 and 29 (should have been 15 and 30). It is you who showed that period 2 repeats transposed to period 38 (with misalignment) have a much higher chance of being disturbed.
Period 5 and period 10 are interesting boxed pink.
Period 5 is very interesting. If you add a 18th column to the 340 period 5 almost matches period 19 in repeats!
Here is an image visualizing period 5 in 18 by 20, you can see the pattern resembles the period 19 pattern.
Another way to look at it, is that it has to something with that pattern, and not necessarly the period.
For the 340,
Period 5 is the first period that creates the pattern, after that comes period 19 and 38. And then the reason why slapping one extra column onto the 340 increases period 5 repeats so much becomes apparant (because of the direction change of the pattern). Let me show you.
Above is the period 19 pattern, so let me reiterate, perhaps it has something to do with this pattern.
Above is the period 5 pattern but notice that the shape of the pattern is mirrored.
Above is the period 5 pattern in 18 by 20, notice that is not mirrored anymore because we added another row, and that makes it adhere more to the period 19 pattern and therefore repeats are increased.
Somehow the information seems spread out between the periods of the pattern…
Ya folla.
The knight’s tour transposition was discussed by Barry S. and doranchak above. Glurk may also have mentioned it. The Army Manual explains it under 107. Complex Route Transposition on the actual book pages 121 and 122.
https://www.nsa.gov/public_info/_files/ … 078809.pdf
Would you like me to make a knight’s tour message to see if it can be detected or what it would look like? I would think that it should have a lot of period 15, 19, 32 and 35 repeats. And comparing those periods should yield two symbol sets that have matching symbols. Otherwise I will continue working on my simple rectangular route detection methods.
I think we should continue with our current hypothesis but I’d like to re-mention some really interesting period 19 zynchronities.
There is a magic square in the Zodiac FBI files which has the same pattern, something like that could have been used as a transposition matrix. Or a more classic magic square 18 by 18 transcribed to 17 by 20.
In Gareth’s Penn book TIMES 17, there is a section on how to construct a magic square. Supposedly he received some one ring phone calls over the span of 2 years which he wrote down on his calender, true or not, he somehow managed to interpret a "Knight’s move" out of it. Check the following image I put together from his book.
Another thing, a chess grand master and cryptanalyst by the name of Stan Vaughan has claimed a 340 solution: http://www.zodiackillerciphers.com/?p=605
I am attempting to contact Mr. Vaughan to see if he will give more information about his solution.
Any news on that attempt?
Another chess connection is the alleged 7/19/78 "I am in control" letter mentioned in Graysmith’s "Zodiac Unmasked" but has not been independently confirmed to exist:
But maybe you play chess with me. I have several cheap sets in closets all over
http://zodiackillersite.com/viewtopic.php?f=69&t=726
As far as I know, there is no direct reference to chess from the authenticated Zodiac correspondences and other evidence; they are only coming out of these secondary sources. Still, I agree that it’s worth exploring the possibility of knights’ tour transposition. But unraveling such a transposition sounds very daunting, since there is an intractable number of possible tours for a "board" as large as the 340’s grid. An 8×8 board alone has over 26 trillion tours.
I am attempting to contact Mr. Vaughan to see if he will give more information about his solution.
Any news on that attempt?
No. I don’t place any value on his claims. I had a brief correspondence with him to try to get more info. He claims his decipherment implicates Arthur Leigh Allen, but of course won’t offer any details or evidence of any kind.
Dave essentially the zodiac killer always uesed presential stamps with one exception which was the letter to Detective Toschi whereby he used an Apollo 8 stamp which had the message In the Beginning… (17 characters the message was a 17×20 340 character message and that held the key.
He also in one letter to Los Angeles paper mentioned he was playing chess with police and had hundreds of chess sets. This led me to cross check USCF membership lists off against suspects and find a match. It was indeed one of the main suspects who had died in interimMy conclusion from deciphering and other info was that Arthur Leigh Allen was indeed not only top suspect but indeed the admitted killer
Changing topics: How far into the period 15/19 scheme have you explored with regard to column ordering resulting from columnar transposition induced by a keyword? Maybe you are close but just haven’t found the right permutation of columns. Also, as you guys are exploring different transposition schemes, which of them would be compatible with the resulting pair of pivots we see in the 340? It seems to me that whatever scheme is in play would tend to produce those pivots, because they seem so improbable. What kind of schemes would create period 15/19 ngram peaks AND those pivots?
I will put knight’s tour somewhere on my to-do list. Although trying to solve the 340 by figuring out a knight’s tour transposition would be very difficult ( as if we don’t have enough problems already ), it would not be difficult to try to make a lot of practice messages with the same knight’s tour scheme but many different plaintexts to see if 340 stats can be emulated.
The repeats show up primarily as diagonal rows going in one direction, and I suspect a knight’s tour scheme would make it difficult to make so many period 19 repeats because the repeats would be diffused in other directions. Maybe at some point, if other schemes don’t work out for us, I can try something similar to knight’s tour so that it can be eliminated as a possibility. Maybe constructing such a cipher will teach us something.
It looks like a 16 x 21 grid, inscribed vertically and then transcribed right to left horizontally. Or a 21 x 16 grid inscribed horizontally and transposed vertically. Somewhere inscription or transcription is backwards. I am going to keel working on my spreadsheets, which are pretty crude. But even with a far from perfect untransposition, I was able to get a decent score and find a lot of words.
Score: 22537 Ioc: 660 M: 167 C: 329 S: 55 etomhallcompileds aroomhdrobertbobb yprincehatewaater thescoreforboysos pooomandroosiipri ncctshissneverani nhousecromtheserf ortheprovechimtoe sbeingisolatedfew ithoutofficeallow ededevelopmploret hescorcausedhisti nputfromthinspira tionentteasanroos biblethedofindmuc hcumentwrittenfro mthedallwhathelpe rtesigndotwiththe soundinbytoshwast hedoom
etomhallcompileds
aroomhdrobertbobb
yprincehatewaater
thescoreforboysos
pooomandroosiipri
ncctshissneverani
nhousecromtheserf
ortheprovechimtoe
sbeingisolatedfew
ithoutofficeallow
ededevelopmploret
hescorcausedhisti
nputfromthinspira
tionentteasanroos
biblethedofindmuc
hcumentwrittenfro
mthedallwhathelpe
rtesigndotwiththe
soundinbytoshwast
hedoom
It looks like a 16 x 21 grid, inscribed vertically and then transcribed right to left horizontally. Or a 21 x 16 grid inscribed horizontally and transposed vertically. Somewhere inscription or transcription is backwards. I am going to keel working on my spreadsheets, which are pretty crude. But even with a far from perfect untransposition, I was able to get a decent score and find a lot of words.
Score: 22537 Ioc: 660 M: 167 C: 329 S: 55 etomhallcompileds aroomhdrobertbobb yprincehatewaater thescoreforboysos pooomandroosiipri ncctshissneverani nhousecromtheserf ortheprovechimtoe sbeingisolatedfew ithoutofficeallow ededevelopmploret hescorcausedhisti nputfromthinspira tionentteasanroos biblethedofindmuc hcumentwrittenfro mthedallwhathelpe rtesigndotwiththe soundinbytoshwast hedoometomhallcompileds
aroomhdrobertbobb
yprincehatewaater
thescoreforboysos
pooomandroosiipri
ncctshissneverani
nhousecromtheserf
ortheprovechimtoe
sbeingisolatedfew
ithoutofficeallow
ededevelopmploret
hescorcausedhisti
nputfromthinspira
tionentteasanroos
biblethedofindmuc
hcumentwrittenfro
mthedallwhathelpe
rtesigndotwiththe
soundinbytoshwast
hedoom
I see the name Robert "Bobby" Prince in the word scramble. They show a listing for a older man by that name living in Sacramento, and he previously resided in Lower Presidio Heights, in San Francisco.