The Novel Area of Cryptic Crossword Solving
journal.frontiersin.orgI'd love to see more of a breakdown of the makeup of cryptics: what percentage of the answers involve anagrams, reversals, homonyms and so on? And how many clues encode the answer in pieces? Seems like a lot of puzzles use a composite of clues like that, with anagrams coming in second. Homonyms and reversals seem pretty rare. The "& lit" type clue is almost never present.
But I wonder: are there regional differences? Have certain types of clues fallen out of favor over time? Can I follow puzzlemakers that shoot for more variety?
One type of clue that's fallen out of favour, according to a book I read about it a while back, are those which require knowledge of, for example, the classics. It used to be that you needed a pretty good private education, but as English speaking has become more international, and as fewer and fewer English speakers (as a percentage) have that sort of education, those clues are less likely to be useful.
Different setters take different approaches; it's why the papers almost almost name the setters, and why it's a big deal when one retires as they're sort of unique.
poking around http://www.fifteensquared.net/ is one way to do a quick (manual) survey - you can focus on specific setters or papers, and count the clue types from the annotations without needing to sit and solve them all.
Good research from French city daily reading matter!