Pascal has forestalled me. I was implying the guess should be a random choice from the o/s words. I recalled now that my algorithm chose the o/s word which maximized the expected information content of the feedback.
On Mon, 24 Jan 2022 at 15:09, 'Pascal Jasmin' via Programming < [email protected]> wrote: > I would start with a 5 word dictionary sorted by english language > inclusion frequency of each letter. > > Each guess results in a pruning of this dictionary, (3 prunings for: just > words with letter in right position, no letter, and letter in wrong > position) but previous guesses can be discarded. The strategically optimal > next guess is always the word at the top of the pruned dictionary. > > > > > > > On Monday, January 24, 2022, 09:09:17 a.m. EST, bill lam < > [email protected]> wrote: > > > > > > The game wordle in the website. > > https://www.powerlanguage.co.uk/wordle/ > > Any idea of how to generate a list of candidates words after new > information from each row? > > Eg, > > T H I E F # all wrong > R A D O N # O and N are there but misplaced > > what is the list of words for the 3rd row? > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
