Zodiackillersite.com

Homophonic substitution:

Consider a 5-symbol cycle:

Code: Select all

ABCDE
ABCDE
ABCDE
ABCDE
ABCDE


Now meet the anti-cycle, its opposite.

Code: Select all

AAAAA
BBBBB
CCCCC
DDDDD
EEEEE


Anti-cycling can be detected by finding the period with the lowest cycle score (it cycles negatively versus randomizations), or symbols that uniquely appear in certain sections only. And it forms fake cycles at higher periods! The impossible achieved, transposition after encoding, very far fetched though but certainly interesting.

Example cipher (low ioc):

Code: Select all

.#.X"X.##AW%5"M58
-?-Y3PE-A7AE(M,PY
S49WA7A(,I*-49_/S
3_XA88D_%>D#!06<-
D_/@-4I*TF(/?FY6P
(F<6_.(/@F<IZ/!6H
0FTI9Z6HD<G8I4G88
'QXDRRZQO&'@DH00=
)&ZO&'@&OQ\':@*=R
#=H0&^5;T;HY;=:=\
;);H?;7:+T:NG10+B
B=10UQ9CC[YX\[44>
2BNG02*RBN+?+\$5G
C$[42$.\$NG$>N+12
!2V2>KRR?VCV?[C1K
15LCL!KYVL1!LJJ$N
VKSL)]XKJ#]!>KJJ?
]Y[O]OU\JL)]\K>KJ
J1[$%K)]U[9OUWL,]
?]Y39E"UMP>.J#7*U


I wonder if it could act as a pivot and/or box marker mechanic with certain transpositions because of many palindromic patterns in the ciphertext (see next codebox):

Though I think it may actually be the other way around. In terms of probability, if anti cycles create these palindromic patterns at its encoding direction and cycles at a higher periods, then normal cycles would create cycles at its encoding direction and palindromic patterns at higher periods.

Code: Select all

Input versus reverse: 17.64%
------------------------------
.#.X"X.#         
-?-     A7A     
    A7A         
                 
                 
                 
                 
       QO&       
       &OQ       
       ;T;   =:=
;);             
                 
          +?+   
                2
!2V2    ?VCV?  1K
1 LCL           
                 
   O]O       K>K
                ]
?]               


The "#.X"X.#" fragment comes from "LIKEKIL" (ilikekilling), and because of the nature of the anti-cycle, "AAAAABBBBBCCCCCDDDDDEEEEE", short fragments of the ciphertext act as 1:1 substitution and make these kind of repeats much more probable.

I have this new encoder spreadsheet where I can transpose and encode 96 messages at a time, and a whole book in just a few minutes.

I am still fine tuning it, but am considering adding a cycle pattern library. A list of different types of cycle patterns that I can think up, like moonrock. Then select from the list before encoding. I have to figure out how to do it.

At some point consider using the library to score the cycles, and detect which of the patterns have been used for encoding. See what level of accuracy the detection is, and if pretty accurate, then apply to the 340.

doranchak: The TV show was excellent. Thanks.

Thanks smokie. Glad you liked the episode.

Your cycle pattern library idea sounds awesome!

Jarlve wrote:Now meet the anti-cycle, its opposite.

Code: Select all

AAAAA
BBBBB
CCCCC
DDDDD
EEEEE


Interesting idea. If anti-cycling is used, we should see a lot of symbol imbalances (all A's occurring before all B's, all B's occurring before all C's, etc). A symbol balance test should reveal this.

Another test could be to look for the anomalous pairs of symbols where all occurrences of one symbol are seen before/after all occurrences of a second symbol.