3blue1brown made two videos on a program that he built to automatically beat Wordle.

However, there are multiple goals which can be optimized for.

In his videos, he optimized for the lowest average. This makes sense as a basic goal.

However, giving yourself the best possible chance of solving the puzzle in 3 or less words is an equally good goal. Or the best chance of getting the correct answer within the 6 allowed guesses.

If there was ever a professional Wordle league, (and maybe there is), then it might make sense to have a progressive scoring schema to reward extremely low guesses. Such as something like

Number of Guesses	Points Awarded
1	10
2	6
3	3
4	1
5	-1
6	-5
fail	-10

Optimizing for this scoring schema would be a completely different challenge.

If chess scoring was used when playing against a friend, where you get 1 point for a win and 0 points for a tie, then you’d look amongst strategies that are similarly likely to let you get it in 4, for the one that gives the best chance of getting it in 3.

---

Mathematics can help us reach our goals, but defining goals is left up to us. A topic that is adjacently explored here.