Homophonic Substitution Cipher

How it is calculated

The text is converted to upper case and only the letters A to Z are encoded; spaces, digits, punctuation and accented letters are skipped and not counted. For each letter, take its index i (A = 0 … Z = 25) and its position p among the letters (first letter = 0). The code is (3 × i + p) mod 99, written with two digits and separated by spaces.

Because the position is part of the formula, the same letter gets a different code each time it appears, which is the many-to-one idea behind homophonic ciphers. Unlike a real homophonic cipher with a fixed table of alternatives, the code here is a simple rule, and it is reproducible: the same message always gives the same output. Codes wrap around after 98 (the next one is 00). This is a teaching demo only, with no key and no decoder, and it must not be used to protect anything.

Common mistakes

• Expecting a decoder or a secret key. The tool only encodes with a fixed rule; anyone who knows the rule can reverse it, so it offers no security.
• Counting spaces and punctuation as letters. Only A to Z are encoded, so 'Hello, World' gives 10 codes, not 12.
• Typing accented or non-Latin letters such as ş, ğ or ü. They are ignored; a text with no A to Z letters gives no result.
• Assuming the same letter always gets the same code, as in a plain substitution cipher. Here it depends on the letter's position in the message.
• Thinking this is how historical homophonic ciphers worked. Real ones used a table with several symbols per common letter; this is a simplified position-based demo.