I stumbled upon a puzzle. The greatest minds consulted by the FBI could not solve it. It is not certain that two notes are not just gibberish but the FBI wants them decrypted if possible to solve a crime.
One thing I have learned in my life is how to solve problems. Since the greatest minds have failed, some other information must be needed. Since I like to prognosticate/read tea-leaves, perhaps I can help.
The situation seems to have been that a murder victim for many years scribbled notes no one could read. This could be gibberish. If it is not there must be a simple way of encrypting/decrypting. Otherwise a young person of ordinary capability would not create and carry such notes around. He must have had a quick and easy way to write them and to read them. Thus, a lot of complicated block cyphers and stuff can be ruled out. Perhaps I am wrong and the guy was a genius…
Looking at the notes is painful. The printing is poor. That it is printed suggests the guy was not particularly literate, so I will chuck the possibility of one-time keys from great works of literature etc. That is, there must be no bulky key.
One of my first observations is that the combination of letters, SE occurs a lot. Random letters would have that occurring much more rarely, perhaps once in 500+ characters of the alphabet (26×26=676) but it occurs on almost every line. In fact, it occurs multiple times on several lines.
In case the writer uses two letters to substitute for one letter, I sorted the message out in pairs of letters. There are several other marks, so this could have a high error rate. For the time being, I will ignore dashes.
I count 37 “SE”s on page 2 and 22 on page 2.
Most of the first page thus analyzed consistes of “SE”, “S something”, and “T something” with a bit of “U something”, “V something”, and of course “W or X something”. That is a clue that I am onto the right track. If I can figure out what S, T, U, V, and W stand for I will have the job almost done. I would expect “S something” to be “E”.
This has all been accomplished with vim, and sort…
I am guessing that these letter-pairs are vowels:
pogson@nb:~/Downloads$ grep -i e ../wordlist|wc
421337 421337 4648919
pogson@nb:~/Downloads$ grep -i a ../wordlist|wc
380800 380800 4174192
pogson@nb:~/Downloads$ grep -i i ../wordlist|wc
373356 373356 4230559
pogson@nb:~/Downloads$ grep -i o ../wordlist|wc
307954 307954 3507828
pogson@nb:~/Downloads$ grep -i u ../wordlist|wc
182117 182117 2032091
pogson@nb:~/Downloads$ grep -i y ../wordlist|wc
102564 102564 1159290
Perhaps he is only switching around vowels…
Nope. E, L, M, and N both start lots of pairs and the second letter of each pair is often a consonant.
To be continued …