A tool intended to decrypt letter-substitution ciphers with high accuracy using letter frequencies. Words sourced from this repository.
By calculating data from 370k English words, I believe an algorithm can somewhat accurately estimate the solution to a letter-substitution cipher.
True variables I plan to take into account:
- Probability of letter being in word (of certain size aswell)
- Proportion of letters in all words
- Proportion of letters in words that contain the specific letter
I will run a Chi-Square goodness of fit test to every possible combination, returning the top several lowest p-values and their respective combinations, with the goal of hving the correct solution in said results.