I did the same thing with the 340 period 19 ( center ), and no symbol had a similar response; deleting symbol x from 1 to 63 resulted in lower scores in every case.
I suspected as much. It causes quite a disruption in the bigram landscape.
I think that it is also a good idea to explore adding or deleting a symbol at every nth position starting at position x. That would help to flush out cyclic misalignments, if there are any.
Good idea! Is that how you got the solve? I can recognize the plaintext, but you expanded to 129 symbols. My other test is still running and when it’s done I will fire up your idea, if you have any specific instructions let me know.
I have a new 340 construction hypothesis I would like to start exploring. That it exists as parts stringed together, each part having a period transposition. So smokie, I would like to ask you to make such a cipher for me (if you have the time). Anywhere between 2 to 10 parts each slightly differing in length but all with period 19.
To give you an idea, each number would be a different part but all parts have a period 19 transposition.
11111111111111111 11111111111111111 11111111111111111 11111111111111111 11222222222222222 22222222222222222 22222222222222222 22222222222222222 22223333333333333 33333333333333333 33333333333333333 33333333333333333 33333344444444444 44444444444444444 44444444444444444 44444444444444444 44555555555555555 55555555555555555 55555555555555555 55555555555555555
I think that it is also a good idea to explore adding or deleting a symbol at every nth position starting at position x. That would help to flush out cyclic misalignments, if there are any.
Good idea! Is that how you got the solve? I can recognize the plaintext, but you expanded to 129 symbols. My other test is still running and when it’s done I will fire up your idea, if you have any specific instructions let me know.
I have a new 340 construction hypothesis I would like to start exploring. That it exists as parts stringed together, each part having a period transposition. So smokie, I would like to ask you to make such a cipher for me (if you have the time). Anywhere between 2 to 10 parts each slightly differing in length but all with period 19.
To give you an idea, each number would be a different part but all parts have a period 19 transposition.
After deleting symbol 12 and un-transposing I had several rows of 17 instead of 18 symbols. I tried linking them together and solving with ZKD and got a few words. Then I was just messing around and expanded the first and last two symbols of each row to make linking them together easier. It worked well enough with your program to get a better partial solve, which was good enough. I did it because I have been thinking about making scytale messages with rearranged columns and expanding the outermost untransposed columns one by one. Something that I have been wanting to do at some point.
I was thinking that finding misalignments at equal intervals caused by skipped or added symbols should be fairly easy, and also that the message could be period 18 with a lot of adds or period 20 with a lot of skips. So I was thinking about making P = 1 to 340, and then adding and skipping a symbol at interval N, where N = 2 to 40 or whatever. The message would expand and contract in size each time. But with each N, count and score the period 18, 19 and 20 repeats. Graph and look for a spike. The graph would slope upward from left to right for skips because of the contracted message size, and downward from left to right for adds because of the expanded message size. But if there is a big spike, like with your symbol 12, then we should hopefully recognize it.
I was also thinking about checking combinations of two or three symbols to see if they are distributed evenly throughout, regardless of whether they cycle. To see if there could be more than one null added at regular intervals.
Either way, the idea is to try to figure out if he skipped or added symbols at a regular interval, such as with a skipped row or a gibberish row.
Yes, I have time. Do you just want a transposition cipher with only one key, but cut up into several pieces and rearranged? Or do you want several parts, each with a unique key? Either would be very easy to do.
I did it because I have been thinking about making scytale messages with rearranged columns and expanding the outermost untransposed columns one by one.
What do you mean?
Yes, I have time. Do you just want a transposition cipher with only one key, but cut up into several pieces and rearranged? Or do you want several parts, each with a unique key? Either would be very easy to do.
One key, several parts (with uneven lengths) having individual period 19 transpositions. Thanks.
What do you mean?
I mean try to figure out a way to solve scytale transposition ciphers that have rearranged columns. You have already done that with 19 fragments, but it was a project that I wanted to do for fun at some point.
Make a scytale message with 19 columns so that the message lines up vertically and can be easily read. Then rearrange the columns. Basically, columnar transposition. If the message was then re-drafted into 17 columns, there would still be a lot of period 19 repeats that would look just like the 340, right?
To solve, re-draft into 19 columns and un-transpose into 19 rows. They wouldn’t be in order, making it difficult to solve because words are cut in pieces and spliced together at the beginning and end of the rows. Expand column 1 and column 19 and try to solve. Then expand column 2 and column 18 and try to solve. Etc.
You probably already did that.
I will make a message with period 19, then cut it into several pieces of unequal length and re-arrange them.
EDIT: Here is smokie23.
1 4 6 17 47 34 36 2 36 49 25 37 33 18 9 12 59
42 48 45 20 20 35 13 43 52 48 34 48 32 33 12 3 7
23 36 12 24 20 48 13 41 42 20 10 53 47 20 1 56 46
30 20 44 50 25 37 14 2 45 16 23 3 35 20 31 24 46
36 44 14 1 34 47 32 33 28 17 6 37 34 54 33 13 2
32 51 50 45 43 29 35 20 3 6 47 23 36 37 42 14 48
46 22 26 41 27 50 20 28 23 24 44 35 13 42 51 9 12
29 19 2 14 25 55 9 40 14 13 11 34 52 50 41 14 27
2 3 52 14 32 43 33 25 21 38 42 34 30 22 40 28 53
46 23 52 54 36 29 20 51 52 16 59 40 31 24 3 44 32
28 32 3 47 15 1 50 8 37 17 40 14 24 39 39 46 48
23 56 24 35 33 60 36 28 5 59 51 35 24 60 47 46 48
21 43 23 44 41 9 20 52 45 50 35 43 42 49 36 12 2
21 1 12 50 29 32 13 48 36 28 23 24 51 30 2 60 52
31 33 50 22 25 3 8 37 56 20 15 9 43 29 28 1 18
15 10 44 8 43 12 32 47 12 47 2 13 9 17 51 10 23
43 14 3 35 60 30 9 35 19 24 40 1 41 42 37 2 3
57 10 32 29 58 47 8 20 4 38 13 23 14 18 52 60 55
9 36 25 59 10 27 52 48 53 12 43 1 49 38 23 13 33
36 24 60 51 47 35 39 25 41 2 3 31 20 28 52 1 46
I made a scytale cipher with 19 columns. Then I rearranged the columns and re-drafted it 17 columns. Then I un-transposed into 19 rows x 18 columns. Then I tried to solve and Jarlve, your program solved it. Then I expanded columns 1 and 18 and tried it. Then I expanded columns 1, 2, 17, and 18 and tried it. Then I expanded columns 1, 2, 3, 6, 17 and 18 and tried it.
As I expanded, the score got higher, but the solution got worse. The solution before expanding any columns was by far the best, almost perfect and I was amazed. And it had the lowest score.
1 19 4 50 7 11 10 20 13 14 16 23
18 29 20 3 23 30 26 50 29 19 32 51
20 63 36 34 38 18 26 51 13 21 1 3
42 15 43 7 44 16 45 28 19 50 31 19
46 20 19 30 49 47 50 4 19 39 10 19
5 36 7 40 23 21 17 7 51 37 25 16
22 36 31 13 58 56 36 51 8 20 13 13
21 19 41 56 22 36 46 42 40 54 13 5
17 51 19 26 39 48 51 27 34 30 53 36
55 37 38 18 7 42 23 23 29 11 34 19
38 3 54 6 2 9 27 12 61 15 31
16 19 6 22 11 25 14 28 28 31 20
40 35 42 19 19 15 50 33 36 40 44
6 5 18 6 50 8 61 23 8 3 19
34 37 12 48 53 11 2 9 38 53 55
11 3 45 6 31 30 5 10 32 16 15
3 19 19 24 16 38 19 15 26 40 33
19 16 26 49 26 19 23 27 14 60 33
11 37 17 19 43 16 46 36 20 63 56
4 6 1 19 10 19 39 43 62 20 58
2 3 5 48 8 25 11 5 14 37 17
5 36 21 41 24 50 27 37 30 52 33
34 47 37 33 39 11 21 20 22 58 41
5 51 7 32 30 50 5 36 19 48 16
47 59 40 35 17 56 51 8 52 50 54
44 28 51 20 55 23 56 28 4 57 21
50 14 57 50 16 29 59 6 28 11 11
15 33 32 40 23 9 18 1 19 21 47
29 10 61 29 3 20 20 23 62 55 31
This is similar to an early 19 column shift but I have removed the last line and I over wrapped it. Can someone chuck it In a solver please ..a week away still from firing up the new Lenovo computer.. Sorry to be a pain ..Just getting back to work now.. Darn 
@Mr lowe, no solve.
Thanks for your cipher smokie. I was thinking that the Zodiac used skytale but since it may be unpractical to wrap the entire plaintext around a cylinder he may have done it in parts. Which may or not varied a bit in length. So that means that each part (lets say 68 characters per) would have an individual period transposition. I think what you did was do a period transposition for the whole plaintext and then cut it up in parts and reorder them. That’s actually a hypothesis I find worthy of exploring too.
As I expanded, the score got higher, but the solution got worse. The solution before expanding any columns was by far the best, almost perfect and I was amazed. And it had the lowest score.
As you increase the multiplicity the score generally always goes up, especially when there is no strong solution. Higher order n-grams help with that.
I’ve been looking at the encoding again with my m_2s_cycles measurement (which was inspired smokie’s cycle measurement). I considered the following hypothesis, the 340 is split up in a number of parts, somewhat equal in length, each with an individual but equal period transposition after enconding. For instance the 340 is split up in 10 parts each having a length of 34 characters with a period 19 transposition. This hypothesis fills in two things we have been looking for, period transposition and encoding randomization.
So I ran through all part sizes from 2 to 339 with all possible period transpositions and have found nothing out of the ordinary. Remember that this is only for the encoding. I tested the routine with a cipher of 5 parts of 68 long with a period 19 transposition and as you will see in the following graphs (up to 170 part size) it returned positive. The x-axis is the size of the parts, for many part sizes there will be an unequal part at the bottom of the cipher and that was transposed also if possible.
340:
Custom cipher as mentioned:
I still need to run through the other orientations, reversed, mirrored, flipped and will report back if anything abnormal. Also plan to do the same thing for the bigrams.
I may have misunderstood. Do you want a message with a different period for each part?
Same period, but each part transposed individually. Anyway I’m happy with smokie23 for now, it is an interesting hypothesis to explore! 
I have just ran the period parts bigram test for the 340. To reiterate, the test considers the cipher to split up in parts with each part having an individual but equal period transposition. The highest scoring period is saved and for each part size and displayed in the graphs. The x-axis from left to right is the part size and the y-axis is the highest score of all periods for that part size (although it doesn’t print the particular period).
Here is a test cipher with a full period transposition. You will probably not be able to read the numbers but it peaks around 340 which is the actual part size.
And here is the 340, it peaks at around 150. That is a possible suggestion that the cipher is split up in parts with a length of about 150. I’ve also tested the other orientations and the results are similar. So either it is what it is, or a fluke, or it just another sign that something is wrong with the bigram landscape that we do not yet fully understand.
O.k. I have to ponder this. I need to leave the house for a few minutes. Then find something in the last 6 months’ spreadsheets to show you.
Instead of one big transposition, like one big grid, you are saying two or more smaller grids. Smaller units to work with, such as where cipher books show simple examples of transposition schemes and there are fewer rows and columns. No?
Instead of one big transposition, like one big grid, you are saying two or more smaller grids. Smaller units to work with, such as where cipher books show simple examples of transposition schemes and there are fewer rows and columns. No?
Well it’s a possibility. I added another variable so to say, part size. If we consider the 340 in its whole then the part size is obviously 340. Then we probably should also find a peak there if the cipher is sufficiently measurable with our measurements.
That depends on the multiplicity of the cipher and the quality of our measurements. We measure the plaintext with bigrams counts over different periods. But since we are dealing with homophonic substitution the plaintext is diffused, and as the multiplicity goes up the plaintext becomes so diffused that our regular bigram measurements stop working or become increasingly prone to outliers. We don’t have that problem with the cycles as they are not diffused (assuming), just saying.
What I’m getting at is that we could use a stronger plaintext measurement (bigrams, repeating fragments). Smokie, I would like to test out your bigram measurement. Could you explain your method again so that I can add it to my program? Thanks.
To further illustrate what I’m getting at. Here is the p1 plaintext without homophonic substitution with a 10 parts 34 characters each period 19 transposition. You can clearly see the peak, telephone pole at part 34. There is no measurement problem here, everything else is just flat.
I think that we may be having multiple conversations with each other at the same time. You want me to explain my idea about searching for misalignments at regular intervals? Do a quote.
The quote is added. I am also discussing "the measurement problem" on top of things.
