Here's a variant question about Wordle strategy that's AFAIK unsolved, which I'm currently working on (my program is too slow). Namely: yes, any optimal strategy will involve adapting your second guess to the response to the first guess, but what if you'd like to play two blind guesses instead of one? (Many human players seem to do this in practice.) In that case, what would be the best two starting words (and decision tree thereon), and how much worse is it than the globally optimal strategy?

In other words: what's the cost of a second blind guess?

Maybe going even further: is it possible to still win the game with three blind guesses, or four? What is the smallest subset of "blind" guesses after which the possible solution set is "small" (say at most a word or a handful: less than some specified k).