This should NOT be a 1991 rated puzzle.

lichess.org/training/GvtT4 In this puzzle, it is a simple knight fork which even 900 rated puzzle players can solve and they gave it has 1991 rated. So, it’s a glitch!

Puzzle ratings go up and down based on how many times they are solved or failed, just like your rating goes up and down with wins and losses. If I had to guess, too many people capture the rook on g1 with check before they notice the fork, thus causing more failures than you'd think and making the puzzle's rating go up over time.

Oh man, @Rneelagiri. It is just voting, it doesn’t really matter with rating puzzles. And of course, if any player, including low and high rated ones, win any puzzle, so why would they dislike them!? They will of course like it, because they got it correct.

@RadRuss below 1200 guys would capture the rook, but I am 2022 defenieltly looking for a fork. Any above 1700 player would like for the fork.

well, to be honest in puzzles 1900 its not like super high difficult puzzle, i'm on 2100 and im pretty bad in puzzles ofc

I agree with @RadRuss

People got the puzzle wrong. That caused the rating of the puzzle to go up and the rating of the people to go down.

What else do you think we should do to punish people for their glitch of getting a puzzle wrong that you got right?

By the way those same people are getting puzzles correct that you're getting wrong.

@FuriousArjun You're probably right, but puzzle rating is based on more than just votes. I actually was under the impression the voting was unrelated to the rating at all, but I could be wrong about that part of it.

NOTE: @Rneelagiri deleted their post which was on #2 after my original one.

Yeah. Also, no need to point that out btw. You could just edit ur post :)