Lest you miss it, Grant Sanderson (3blue1brown) has entered the Wordle fray now:
https://twitter.com/3blue1brown/status/1490351572215283712
[Less empirically, I wrote a little about Wordle less than a week ago HERE.]
One thing I'm not completely clear on, BTW, is when many of these recent pieces analyzing Wordle talk about "letter frequency" is whether they are simply employing well-published figures for letter frequency in English, or are they in fact using (as they ought be) letter frequency specifically in 5-letter words, as easily computable from the Wordle corpus of use? I would expect those two figures to be similar but not the same, and hopefully they are using the latter.
Also, there's tremendous emphasis placed on first word choice in the game, and more and more second word choice seems key to me for winning in 3 to 4 guesses. At least I'm finding increasingly that I can generally get the target word in 4 tries almost regardless of my first word choice so long as I spend enough analysis time on my 2nd and 3rd choices. Certain first word options simply incrementally increase the chance of 3rd guess winners.
It would be interesting to actually run a competition between some real word-maven aficionados (who have a "feel" for the game) versus some strictly (brute force) algorithmic programs and see who would achieve best scores (no time limit), and by what amount.
No comments:
Post a Comment