Lines 1,2,3,11,12 and 13 of the 340 (suggested by Olson) contain 47 individual characters.
Of these none of them repeat more than 3 times, other than +=5 and a straight up and down line =4.
Of the 47 characters, 25 of them contain their mirror image in lines 1,2,3,11,12 and 13.
That’s 0.53. I wonder if this occurs at the same percentage in all possible 6 line combinations of the 340.
I suppose it’s possible that with sequential homophones, the next homophobe after say character x is its mirror image.
For example, in the sequence:
[backward P] [backward L] [upside down V] [V]
[backward K]
The mirror image of a character might be the next homophone in the substitution sequence.
For the above, the regular and upside down V’s might be the same letter, like T.
And the regular and backward P might be L.
So it could be a word like "little."
Please forgive me I’m not sure I’m articulating this correctly or using the appropriate language.
Please check my math, I’m scribbling this on the fly.
Just some observations.
Echoes
Correction. Homophone, not homophobe. Autocorrect. This was unintentional.
What is the mirror image ratio for all the symbols in the 340? And in the 408? Could you check that Echoes?
I calculated things a little differently.
The ratio of mirror image unigrams to total unigrams in lines 1,2,3,11,12 and 13 is 0.22.
This ratio for the entire 340 is 0.30.
I don’t know if this is statistically significant.
I guess the conclusion that can be drawn from this is that the composer of the cipher used mirror image unigrams to a lesser degree in the 6 lines above compared with the full 20 line cipher. I don’t know if you can assume any 6 line combination is an adequate representation of the entire cipher.
I did not do this for the 408.
And the answer to your original question…it actually looks like it is 0.54 for 1,2,3,11,12 and 13 and 0.49 for the 340. Looks like the P approaches 0.05
And the answer to your original question…it actually looks like it is 0.54 for 1,2,3,11,12 and 13 and 0.49 for the 340. Looks like the P approaches 0.05
Then, since these rows have a ratio equal to about that of the full cipher it appears that there is nothing out of the ordinary.
Now, what is the ratio for the 408? Perhaps there is a notable difference between the 340 and the 408.
Jar—
Looks like 0.41 for the 408 vs 0.54 for the 340. So it looks like he used more mirror image unigrams in the 340 than in the 408, but for practical purposes I don’t know if this matters or is significant.
Echoes
Jar—
Looks like 0.41 for the 408 vs 0.54 for the 340. So it looks like he used more mirror image unigrams in the 340 than in the 408, but for practical purposes I don’t know if this matters or is significant.
Echoes
That is interesting. How are you counting?
I counted by hand. My calculations may need to be checked, I’m doing this on the fly.
I counted total unigrams used that also had a mirror used in the body of the cipher/total unigrams in the cipher.
I thought it would also be interesting to know if a unigram and its mirror were used in succession before the next mirrored unigram was used but at first look right from line 1 it seems this isn’t the case. What do you make of the repeated sequences RJI and BCV at right angles? These are mirrors too.
Echoes
Hey Echoes,
interesting analysis! You could also try to make a map that shows the distances and the distribution of mirrored and normal letters.
The pattern you address is called "pivots". Unfortunately, we do not know what exactly is the reason for this, but there have already been some ideas and theories.
Do you know the following link yet? All observations made so far are listed there:
http://zodiackillerciphers.com/wiki/ind … servations
Translated with www.DeepL.com/Translator
Thanks for the info Largo. A little too advanced for me but the statistical analysis of these ciphers is very interesting. The question becomes how do these analyses translate into a partial way to solve for the professional and unprofessional alike.
Meant "practical" way…
When looking at the unigrams without taking into account their frequencies.
Only counting horizontally mirrored unigrams:
408: 10 out of 54 = 18.51%
340: 14 out of 63 = 22.22%
Also including rotation and vertically mirrored:
408: 14 out of 54 = 25.92%
340: 20 out of 63 = 31.74%
– The extra for the 340 come from the symbols that the 408 does not have. The half filled circles and a few more.
This may help (some of my earliest work):
408:
340:
So how would you interpret those results?
