Is there something wrong with Lichess' puzzle ratings?

before it was 1209
now 1963

Solving puzzles should be a pastime like crossword puzzles.
Do you see ratings on a cross word puzzles?
All that was needed was having the puzzles listed with the number of pawns on the chessboard or the number of pieces.

Players could then solve puzzles by number of pieces.
Example: I want to solve puzzles with no major pieces on the chess board.
Or only with mayor pieces and no pawns.
Or puzzles with all the chess pieces on the chessboard.
Or mate in 4.

A rating does not hone your chess solving skills.

okay 1-203i2h xc what

like and share

When you start a new account, the puzzles are easy which allows you to boost your rating. When you solve more, the puzzle rating goes higher.

@Toscani

chess puzzles=crossword puzzles
crossword puzzles≠rated
chess puzzles need be listed with the number of pawns on the chessboard or the number of pieces.

syllogism of the day ;)

@Toscani
Puzzles coud be presented the way you say, but did you see crossword puzzles use only words with 4 or 5 letters?

But everybody say what they would like to have for us, of course.

Puzzle ratings are a good motivation and feedback for me. By being thrown back often times I realized I need to look more deeply into the puzzles. And it works, in that it changes my puzzling habits or style in a way that I solve more constantly.

Does it transfer to my games? Yes, it does, while I just hang pieces, all the same ;)

My deduction and theory ... Epistemology: truth, belief and justification.

"Organon Lichess puzzles": The puzzles must be entertaining and have a solution so we can get a centipawn value for our wack at the puzzle. Every time we make a move, the centipawn loss is compiled to give us a skill level, based on the positional centipawn loss created by the move.

No more rating, just a skill level indicator. Peoples second best move or third best move would have value too.
A compromise on chess puzzle solutions is the best way to go.

@Toscani
"Organon Lichess puzzles", nice name! You should convince them to open a second puzzle section.

Lichess puzzles are a lot about having stockfish as a trainer. And would it teach us strategy instead of tactics, centipawn evaluation would make a lot of sense.

All the while, tactics puzzles in general, and especially those here on lichess, seem to work well with a rating system. In all tactical puzzles, you only win when you win. The reason you shan't get half a point for a drawing continuation is probably that, unlike in a game, you are told there is a solution. Sometimes (but not in lichess puzzles) you have to secure a draw, which makes it more obvious why the lesser solution is not rewarded.

Lichess pick their puzzles in a way so that they contain but only-moves. That's why centipawn evalutation would be the wrong meter. When you misstep, you may retain a playable position, but the difference between the 2nd best choice and the winner is always so, that the 2nd best is not winning, (very often even losing).

Stockfish is often playing surprising and challenging moves when we get ours right; at least as often it plays moves that leave us in wonder, because it does not play the move we had to beat in our calculations, (or feared most in our guesswork).

So the rating we get is the rating we deserve for playing stockfish from a winning position! But- Stockfish's own rating plays no role: The puzzle quality rating is a rating derived from the succesful/unsuccesful solving activity of the users.

My only question to dear @thibault and the other dear makers would be, how did a puzzle rating start, on release, are they all 1500 (like us solvers start) or 2000+ (matching the average of the ratings the players whose games are used in a puzzle had), or something else, based on some mathematical judging of their difficulty?