Seeks optimal guessing strategy using letter frequency analysis on all remaining viable words as incoming guess/score combinations shrinks corpus. Initial solutions corpus ripped out of the wordle website code.
yarn install
yarn global add ts-node # To use the TypeScript REPLIn a ts-node REPL:
import {WorldeSolver} from '.'; var s = new WordleSolver();
// Guesses and their scoring can be chained together statefully to shrink down the corpus as scores are appended.
// Scores are 5-letter strings of `x` for an exact match, `i` for a letter to be included, and `e` for letters to be excluded.
s.score('irate', 'eeeee').score('lousy', 'ixiee').getBestGuess()
// Returns 'mogul'. Or, just omit `.getBestGuess()` to see the reduced `corpus` over time.- Best starting word is along the lines of:
irate,later, or possiblyadieu. Full analysis would require comparing different solve strategies with different solving words across the corpus of target words. This is left as an exercise to the reader. - Find common consonant clusters, front and back. This makes it easier to figure out what to do if the midpoint of the word is found to be a vowel.