A different shift for every position, driven by a repeating keyword. That defeats simple frequency analysis — until you find the key length. This tool encrypts, decrypts, and then breaks the cipher using Kasiski spacing and the index of coincidence.
Paste ciphertext into the input above, then run the two classical steps: find the key length, then solve each column as a separate Caesar cipher. Works best on 200+ letters.
English sits near 0.067; random text near 0.038. The lengths scoring highest are the likely ones. Click a bar to try it.
Each column is a Caesar cipher; the best shift for each is found by chi-squared against English.
A repeated trigram usually means the same plaintext met the same part of the key. The key length tends to divide those spacings.
| Sequence | Positions | Spacings | Divisors |
|---|
E(x_i) = (x_i + k_(i mod m)) mod 26
Write the keyword repeatedly under the message. Each plaintext letter is shifted by the letter above it, so the same plaintext letter encrypts differently depending on where it sits.
That is the whole idea: a single letter frequency table of the ciphertext looks nearly flat, because the tall E bar has been smeared across m different positions. For three centuries this was called le chiffre indéchiffrable.
It is really m independent Caesar ciphers interleaved — and that is exactly how it falls.
Two nineteenth- and twentieth-century ideas, both in the panel above:
The tool shows the mechanism — the slides show why it is built that way.
Finding a key length from nothing but ciphertext is the moment cryptography becomes detective work. One-on-one tutoring builds that instinct step by step.