"Better" response in multi-move puzzles

[ddugovic]

Puzzles are generated by playing the best moves for the opponent which produces a good puzzle experience most of the time. However some of these computer lines don't pose a challenge to the user and can be played without much thinking. In these cases there is often a nicer or more challenging line "hidden" in the variants the computer didn't play.

There are multiple reasons to rather use such more challenging variants:

• The puzzle is of higher quality
• The puzzle receives a more appropriate rating as some users might have played the initial move without considering the more challenging response hidden in the variants.

The linked puzzle in #1 is just an example. The reason 1... Rb3 is a good move is because the white queen can no longer protect f3 which allows the black queen to use f3 in the mating variants. However this knowledge is never tested in the actual puzzle. After 1... Rb3 the computer plays 2.Kg1 and even an absolute chess beginner can then find the trivial 2... RxQ. A more challenging response to 1... Rb3 would be 2. Qd3 or 2. Qd2 where then black has to prove whether they really understood the motives behind 1... Rb3.

The reason I brought up Chess Tempo is that it is a well established and feature rich chess puzzle site, no matter what one might think about their UI. Lichess only offers a small subset of Chess Tempo's features which were polished in over a decade of development time. Naturally Chess Tempo already had to solve a lot of the problems which occur with generated puzzle sets, so it makes sense to look at their approaches and check whether they also make sense for Lichess.

As Lichess now offers the same voting system as Chess Tempo and experiences the same problem with subpar computer continuations, it would only make sense to offer the same solution: vote-based alternative responses.

I offered this suggestion for a simple reason: It's a good approach to improve the puzzle experience on Lichess. Obviously Lichess doesn't need to go that way and could just let these puzzles get down voted instead. This has the benefit of avoiding to introduce additional complexity to the UI. In the end I would be interested in the opinion of the community and Thibault's vision for Lichess' puzzles.

https://lichess.org/forum/lichess-feedback/suggestion-better-response-in-multi-move-puzzles?page=2#12

[ornicar]

Much simpler: downvote these puzzles. Then only the best will remain.