Reply To: Daily Cipher
@Puzzling_Pelican
Thank-you, I wasn’t expecting to see your upload your code. Your round_midpoint function helped me figure out the second triangle adjustment. I don’t think I would have got there without it even though I had most of the other elements in place. I didn’t use a spiral grid creation routine, just a straight list of the 91 points which I could fill programmatically with a generated key, but I had a very similar grid referencing. With the midpoint adjustment and a few tweaks my code now drops out the correct solution.
Just for interest, I will look at brute-forcing this. My quadram scoring works with letters (obviusly), and you divide ‘illegal’ letters with 0 & 1 – easy enough to replace. However, I need to think whether I score decoded texts with lots of the other non-alphabet characters used, or just skip over them. Even just a small key change from the correct solution could mean lots of non-alapha characters, and a brute-force could fail to zero in on a solution if I skip too much. Just ignoring the non-alphabet characters can cause the scoring to go way too low. My quadram routine gives the correct text a score of -17458. I have had much lower scores when ignoring non-alpha characters.
Unlike Playfair, it is not just double letters ‘LL’, ‘OO’ etc to worry about. I had figured ‘THELIGHTHOUSE’ would be a good crib, and was (slowly) some way down a hand decryption. In the trigrams THE-LIG-HTH-OUS the combination ‘HTH’ would clearly not make a triangle, so I could see that additional plain text padding would be required.
Thanks again for an interesting final cipher puzzle.