It's not optimal at all, but works for every wordle word; it's built by finding the word (guess) that minimizes the maximum size of the possible words (to guess), a bit like Knuth Algorithm for Mastermind.
For an optimal strategy, a backtracking solution with pruning / maybe a branch and bound approach would work, but I didn't manage to implement a fast enough implementation for my small 8 threads CPU.
It's not optimal at all, but works for every wordle word; it's built by finding the word (guess) that minimizes the maximum size of the possible words (to guess), a bit like Knuth Algorithm for Mastermind.
For an optimal strategy, a backtracking solution with pruning / maybe a branch and bound approach would work, but I didn't manage to implement a fast enough implementation for my small 8 threads CPU.