# I have a puzzle and a solution... it's the method I'm after

The cipher message is:

J B A Z R X K Z N

T W K O Y E Z

T O N G L A N D

B R I D G E

How did I (or more accurately my father) arrive at the answer?

• First thing I noticed is ROT13(2 zber yrggref va pvcuregrkg guna va cynvagrkg. Guvf zrnaf ab punapr vg'f Pnrfne fuvsg be vgf znal inevnagf (yvxr Ivtrarer), fvapr nyy gubfr pvcuref qba'g punatr grkg yratgu). Otherwise, I'm drawing a blank. Commented Nov 5, 2018 at 1:32
• Is the mathematics tag an independent clue?
– WAF
Commented Nov 5, 2018 at 9:49
• I'd like to add either reverse-puzzling or puzzle-identification to this. @Nixy Do you know the answer to your own question? Commented Nov 5, 2018 at 14:34
• The last clue was never solved. Apparently over the years various people contacted my father for further help and my dad even send a further clue out to someone in the same code to see if that would help, but it didn't. Apparently, even the army was asked about it! My dad just gave me the answer that was long sought after and asked me if I could work out how it was done. I couldn't and so joined stack exchange after googling about codes for a couple of hours. I tagged it the way I did simply because I didn't know better, sorry! I also don't know the answer, but if you can explain it to me...
– Nixy
Commented Nov 5, 2018 at 18:44
• Treasure hunt for a case of whisky organised in the 1980s by a squash club in Kirkcudbright -- is this related to this treasure hunt here? Commented Nov 5, 2018 at 19:13

Method:

The method is a sliding-alphabet cipher similar to Vigenere, or more precisely, Quagmire 2. As @Evargalo notes, the first letter of each word is likely to be a key. Write the two substitutions below a standard alphabet as if solving a polyalphabetic system, with the key letters outside the table to the left:

  tongland    bridge
J BAZRXKZN  T WKOYEZ

abcdefghijklmnopqrstuvwxyz
J K  N  R    X ZA    B
T  W YZ E O        K


There are a few possibilities for the type of system. A plain Vigenere with standard alphabets is excluded immediately. A Quagmire 3 (an unknown alphabet sliding against itself) is excluded by finding the proportion E:g::K:r between the T line and plain line which clashes with a:g::K:R in the J line. There are too few letters to be completely confident in any of the others. The most convincing (to me, at least) is a Q2 system, with a standard plain alphabet and a cipher alphabet beginning with the keyword AEIOU:

  abcdefghijklmnopqrstuvwxyz
J K  N  R    XYZA    B
KLMNPQRSTVWXYZAEIOUBCDFGHJ
VWXYZAEIOUBCDFGHJKLMNPQRST
T  W YZ E O        K

The keying is unusual - the key letter is aligned with plain z instead of plain a. A Q2 would normally key these as K and V instead of J and T but the system otherwise fits.

• I'm so impressed by how many old cipher questions you've been able to come in and solve or sweep up after previous posters. Do your knowledge and solution strategies stem from professional work or are you a cipher hobbyist? :)
– Stiv
Commented Jun 21, 2023 at 12:14
• That's a funny question to answer. I do this for fun, but the way I got started was by failing to solve a nasty puzzle hidden in Noita and then having the thought "Weren't there, like, professional codebreakers at one time? Did they have technical literature, training manuals, that kind of thing?". And then I read everything William F Friedman had ever written. So I did learn from the pros, in a way. Commented Jun 21, 2023 at 12:16
• But not a professional yourself, then - that's impressive. The ciphers used in the puzzles you've been solving are so wide and varied I had wondered if this was part of the day job for you. Well, I'm glad you're here to tidy up these unanswered/incomplete questions - cipher questions don't always get a lot of love around these parts! (Expertise is often required...)
– Stiv
Commented Jun 21, 2023 at 12:22
• Well, thank you! Feel free to ping me if you see something old I might like. :) Commented Jun 21, 2023 at 12:30

I have not solved it, but I have noticed a pattern that may help future solvers. Or maybe it is just a dead end ?

Consider that the first letter of each line is a key, and then you solve letter by letter.

Then you have

(key=J) BAZRXKZN -> TONGLAND
(key=T) WKOYEZ -> BRIDGE

If we use a

Cesar's code, moving each letter to another one according to their place in the alphabet.

Then

The distances from the cipher to the solution are quite regular, around +14 in the first word and around +5 in the second word:

Indeed:

B->T : +18
A->O : +14
Z->N : +14
R->G : +15
X->L : +14
K->A:+16
Z->N : +14
N->D:+16

And:

W->B : +5
K->R : +7
O->I : +20 (?!)
Y->D : +5
E->G : +2
Z->E:+5

The irregularities I cannot explain.

• B->T should be +18 rather than +12 should it not? Commented Nov 5, 2018 at 12:07
• The oscillation can be explained if it is not Cesar's but a keyword cipher. Commented Nov 5, 2018 at 13:13