@smokie, I added a numerical version. Is your spreadsheet okay with 50 symbols?

Yes, 50 symbols is fine. I used some of the formulas from my heat map spreadsheet so that I can analyse one row with a single click. Will get back soon.

I’ve done some small sample tests with the Dan Olson scheme (row period 10) and it’s another interesting thing. When applied to the 340 it makes bigrams spike @ period 36 for 41 repeats (unmirrored). It can be added that such a transposition would be just about right for the randomization of the cycles observed in the 340 versus the 408. And it’s also consistent with the magic square transposition matrix scheme once more. Some things to consider here…

My 2-symbol cycle measurement (capped 340):

408: 270
340: 180
408 after doing row period 10: 169 <—

Here are the results for your +/- 50 symbol message.

The leftmost column in bold is the row number. I ran three trials of 30 shuffles for each row. The "br" stands for original row order rank by count of period 19 repeats in the entire message after 30 shuffles. The "scr" stands for original row order rank by total score for period 19 repeats. I averaged the three trial values in the two rightmost columns.

For example, see Row 1 trial 1 in the blue box at top. I shuffled row 1 thirty times and the original message ranked #1 by period 19 repeat count and score as compared to the 30 shuffles. The same for trials 2 and 3. In other words, in a total of 90 random shuffles of row 1, there was not any shuffle that resulted in a higher repeat count or score as did the original symbol order. EDIT: Rows 3, 5 and 14 also had such results.

Row 11 looks like the best candidate for gibberish, but with 90 shuffles of row 12, there were also many that resulted in higher period 19 repeat counts and scores as compared to the original symbol order.

With rows 4 and 16, an average of 7 out of 30 shuffles resulted in higher period 19 repeat counts as compared to the original symbol order. And an average of 8 out of 30 shuffles resulted in higher scores as compared to the original symbol order.

Crude as it may be, I say row 11 or row 12, or perhaps both of them. There are also high numbers at row 13, so I wonder if one row as gibberish could cause a disturbance with adjacent rows.

You said that there were a lot of period 19 bigram repeats, but the 50 symbols makes me suspicious. Is this a transposition scheme? If it is, then I would like to compare with a message with similar symbol and period 19 repeat count, but which is not a transposition scheme.

I haven’t done the 340 yet, but will later this afternoon.

EDIT: It must have been row 12. I took row 11 out, moved rows 12-20 up, and put row 11 at the bottom. The number of period 19 repeats decreased slightly. However, I did the same thing with row 12, and there were 8 more period 19 repeats and the overall score increased substantially.

.

.
Now for the 340.

With rows 1, 7, 9, 13, 14 and 17 I could not get more period 19 repeats or higher scores with even 90 shuffles of those rows. With several other rows, 90 shuffles only resulted in a few more period 19 repeats or higher scores. With rows 4, 6, 10, 15, 19 and 20, I could get several shuffles resulting in more period 19 repeats or higher scores.

However, it was easy to shuffle row 3 and get more period 19 repeats and higher scores. The red box shows the not shuffled count and score.

For a stark comparison, here is row 8, where 30 random shuffles of that row resulted in much fewer period 19 repeats and score:

.

.
I took row 3 out of the 340, shifted rows 4-20 up one row, and put row 3 at the bottom. The number of period 19 repeats went up to 73 and the score went up from 528 to 547. There are several diagonal rows of period 19 repeat symbols that pass through the area without disturbance.

EDIT: Here is one un-transposed version of the 340 with row 3 moved to the bottom, showing the corresponding diagonal rows of period 19 repeats as period 1 repeats. I un-transposed by picking up the symbols left to right top to bottom, and writing them down top to bottom left to right. I don’t know if this is the correct un-transposition, of course, but any un-transposition should link the diagonal period 19 repeats into period 1 repeats somewhere in the un-transposed message.

.

.
O.k., so here is a comparison with smokie12. Below is the transposition tracing map for smokie12, where I skipped a plaintext during transposition at rows 6 and 14. The numbers show symbols positions, not symbols.

Here are the results for the row shuffling experiment, showing that if any row is gibberish, it would be row 6. Interestingly, with 90 shuffles of row 14, there were few increases of overall period 19 repeats or score.

.

.
Here are the smokie12 period 19 repeats. There are a few short diagonal rows of period 19 repeats passing through the distortion area.

I moved row 6 down to the bottom as with the 340 to see what happens with the diagonal rows. Unlike the 340, no substantial change.

I would like to continue experimenting with the transposition scheme. Make test messages, with a lot of period 19 bigram repeats caused by transposition and not caused by transposition, if possible, and find ways to compare with the 340.
.

smokie ..do you have or can you get a z 340 colored chart of all known period 19 bigrams and post it up please

smokie ..do you have or can you get a z 340 colored chart of all known period 19 bigrams and post it up please

I think that you must want it to be easier to visually match up the period repeats. Good suggestion; I will work on that soon.

Yes smokie.. The visual of the Period 19 repeats are real interesting. Also wondering this. An old Boy Scout code trick rap the code around a tube. The tube had to be the exact same radius to read it again, Could this produce period 19 bigrams

Its amazing that it takes very few " tetras" moves to get this together. when you bring all the 19 repeats back to one space apart using column shift. Only one does not play well. Not sure why they don’t clash with each other more and make it difficult to structure. Probably a simple explanation. But I can’t see why yet.. The reason I’m looking at this is just to try and find how why the 19 period structure came about.
EDIT realignment fix .. move column 1 through to 5 down two. segments.

Here’s something I whipped up to visualize the bigrams:

http://zodiackillerciphers.com/period-19-bigrams/

Put your mouse cursor over each button to see the corresponding bigrams in the cipher text.

Let me know if it doesn’t work for you.

Mr. Lowe, I have been trying to figure out how the period 19 repeats came about for a few months now. The boy scout trick seems as plausible as any, and personally work toward the simple because it suits my skill set.

Doranchak, that bigram program is really cool. Maybe Jarlve could put a link to it on page 1 of the thread.

Doranchak, that bigram program is really cool. Maybe Jarlve could put a link to it on page 1 of the thread.

Thanks. I will soon add the period 15 scheme to it, too.

Thanks doranchak.. That’s real handy.. I did not realize there were so many other 19 bigrams. Strange thing again I just mind checked and flipped back and forth and nearly all of the ones that I missed lined up except one. Which I think if I column shift line 5 down and 6 up it will repair itself with no adverse effect on the rest. It is now very very interesting. I would expect there to be many misses where they would not line up.. I am not near my "tetras" set up at the moment.. Also I look forward to the 15s It will be interesting if the 15s slot in and work without disrupting the period 19s and even more interesting if they are already aligned.

And maybe I’ll have a few that won’t fit because of a random chance 19 bigram. Mmmmmmmmm