Live data from Hacker News

Adversarial Wordle

qntm.org

71–72 of 72 posts

Re: Adversarial Wordle

#71
post #51

Earlier quoted context omitted.

I've written a Wordle solver in F#. It's interesting because your first word contains the five most probable letters. My solver starts off by selecting, at random, one of the possible words that have these five letters. Could you say more how you've determined your solver to be deterministic in 6 steps? Also, I'm not sure I understand what this means: > It selects a word to minimize the number of words assuming the w…

So, my algorithm is deterministic because there isn't any random element in it. And then it appears the adversarial doesn't either because the game plays out the same way every time. So, to generate a single guess, the algorithm will basically try every single word in the dictionary against every possible remaining word. For any given guess, it tracks the count of various possible scorings. The scorings are the posit…

Thanks for the elaboration. I see what you mean by deterministic now. Is it also deterministic in the sense of always getting an answer within 6 guesses (maybe convergent is a better word)? (Maybe it's not possible to guarantee this. That's something I've been wondering.)

My solver is a naive one at the moment, where I've only done a little analysis to get a good starting word. From there, my solver just tracks the results from each guess and randomly selects a word that meets the constraints for the next guess. That actually works surprisingly well, and it hasn't failed Wordle yet.

I've been meaning to go back and do a more thorough analysis and create a more targeted strategy.

Re: Adversarial Wordle

#72
post #21

This is great! A few days ago, I wrote a Wordle solver in C. It selects a word to minimize the number of words assuming the worst-case green/yellow. So, it sounds like these two are in direct conflict. My solver is deterministic, and it looks like the adversarial is too, so they always play the same game: S E R A I (0 green, 0 yellow, 429 words remain) M O L D Y (1 green, 0 yellow, 35 words remain) C E N T U (0 greem…

I'd love to see your code! My strategy[0] (not-yet-automated) doesn't aim to minimize in the worst case, but rather to minimize the expected size of the set of possible words. [0] https://blog.scubbo.org/posts/cheating-at-word-games/

Have you done any thinking about hard mode? Getting trapped in a 'deep' pattern (?ight: e, f, l, m, n, r, s, t, w) too soon seems like a problem. But this is way outside my bailiwick.
Post reply on HN