I’m getting increasingly frustrated by the puzzle quality. Puzzle difficulty is determined algorithmically using an Elo system (where essentially puzzles play against players). That sounds great in theory. The problem is that there’s no penalty for guessing, in fact, the UI forces you to make a guess. This means the rating partly measures guessability instead of solvability. At my low-ish rating, I get a mixture of puzzles I can realistically solve, and extremely hard puzzles where the obvious dumb guess happens to also be the correct solution. I’m not sure if anybody else is annoyed by that as much as I am.
The biggest benefit, in my opinion, of doing puzzles is building pattern recognition. If that obvious dumb guess is consistent for the structure presented in the difficult-but-guessable puzzle, then I think there's still value to doing it.
