Log in Sign up

What is in this category

These entries cover unsolved ciphers and coded messages whose plaintext is almost certainly English or German. We left out mysteries in unknown or invented languages (such as the Voynich Manuscript or the Rohonc Codex) and anything that has been solved.

EntryYearLanguageLengthDifficulty / notes
Kryptos K41990English97 lettersOfficial plaintext clues; method unknown
Zodiac Killer Ciphers Z13 and Z321970English13 + 32 symbolsProbably too short to prove any solution
The Dorabella Cipher1897English87 symbolsLikely substitution, but with a twist
The Beale Ciphers1885English520 + 618 numbersBook cipher; possibly a hoax
Ricky McCormick's Encrypted Notes1999English~30 linesProbably personal shorthand
The Somerton Man (Tamam Shud) Code1948English~50 lettersProbably initial letters of words
The D'Agapeyeff Cipher1939English196 digit pairsTextbook challenge; maybe encryption errors
The WWII Carrier Pigeon Message (Bletchingley)~1944English27 × 5 lettersCodebook (+ one-time pad?); needs the book
Cicada 3301: The Liber Primus2014English~55 pages of runesModern stream cipher; large ciphertext
The Paul Rubin Cryptogram1953EnglishSeveral linesProbably nulls / masking
The YOG'TZE Note (Günther Stoll, 1984)1984German6 lettersNot cryptanalysable; context puzzle

A note on German ciphers

Many well-known German "unsolved" cryptograms have been solved or explained in recent years. That is largely thanks to the readers of Klaus Schmeh's blog Cipherbrain:

  • The Rilke cryptogram (WWII) turned out to be typing practice, not a message.
  • The SS radio message to Lippert (1944) is most likely a forgery.
  • Most of the 1918 ADFGVX messages were broken by George Lasry and others.

Many small German items (encrypted postcards, diaries) are still open. Schmeh's blog is the best place to find them, and new ones appear there regularly.

General tips for would-be solvers

  1. Start from the best images. Retyped transcripts often contain errors that make a solution impossible.
  2. Know the context. Who wrote it, when, and for whom tells you which ciphers they could have used.
  3. Respect the unicity distance. Very short ciphertexts (Z13, Z32, YOG'TZE) have many "solutions". Without a key that explains how, a readable text proves nothing.
  4. Test against known facts. For Kryptos K4, any method must reproduce EAST, NORTHEAST, BERLIN and CLOCK. For Beale, it must explain the Gillogly strings.
  5. Use good tools. CrypTool 2, dCode and AZdecrypt handle substitution, transposition and homophonic ciphers.
  6. Publish your method, not just your plaintext, so others can check it.

Good background sources

Put together with AI as a quick reference I keep for my own work. Entries like this are usually not reviewed in detail by a person.