I’ve been looking at the subtle statistical quirks in a classic substitution cipher lately—some patterns that seem to leak key information. Have you come across anything similar in your analyses?