lichess.org
Donate

Can these puzzles be removed from lichess please?

I've been combing through the lichess pawn endgame puzzles over the last couple of weeks and I've occassionally run into a few bad puzzles:

#GfCey

#fzqXY

Both of these puzzles have the same flaw, and I recieved them back-to-back! Both puzzles require the solver to notice that an otherwise correct candidate move is incorrect due to 3 move repitition. The players who played the games which these puzzles were extracted from are 500 rating points below the rating of the puzzles and they decided to repeat moves. Anyone who would be expected to find the winning move (i.e. meeting the puzzle rating requirement) would not have repeated moves twice already. I'm aware you could argue its an interesting way to ask for alternative ways to win, however I'll get on to that now:

The first puzzle is most offensive in that there are only 2 candidate moves, both of which would be winning under normal circumstances. The second move of the puzzle is extremely easy to find by process of elimination. Anyone who notices the 3 move repitition would be able to "solve" the "puzzle".

The second puzzle is again 2 different move orders (that both work) for the same idea.

lichess.org/study/dq0kwyrE
This is just a study I wrote about another position (the one in chapter 1) that keeps occurring as an incorrect solution. It's a tricky win and the correct solutions to those puzzles are much simpler to win than it would be to win the position in the study. In this case I understand that a winning move can be incorrect due to the comparative difficulty and I think it's fine that puzzles like this exist.
#2
For the first one Ke5 is a repetition because black can just play Ke3. Threefold is moves 55, 58, 60. It would be winning for black if there was no repetition. In a puzzle like that I would insta play Ke5 so I kind of agree with OP here.
Same thing goes for the second one, instead of a6, Ka7, white could play Ka7, a6. It's the same thing, same line and same result if we don't consider the repetition of moves 50, 52, 54 with last black's move being Kc7.

I know of puzzles with the same "problem" that got removed after a bunch of people disliked them, so go ahead and do so.
@Deadban for your first answer i kind of understand but for the second answer, white can't play Ka7, a6 because it would result in a draw if black moves to Kc7, which white doesn't want. and on move 50, 52, and 54, even though black repeated the position 3 times, white didn't, since white moved a6 when black moved Kc7 for the third time, and according to Wikipedia, a threefold repetition can only happen if the same position occurs three times in a game, not an opponent's move three times.

I think for your first answer it's just about how the opponent played a bad move and u took advantage of it. I see those types of puzzles quite often and i dislike them too, as bad moves are unlikely to happen in real games. I agree that they should be removed.
and i don't see a repetition on moves 55, 58, or 60
#4 The repetition in the first puzzle is at moves 56-58-60 (my bad). In the second puzzle if white plays Ka7 instead of a6 (which is what OP mentioned) and then black responds with Kc7, that's a repetition. In my previous comment I just said that both orders would work if not that Ka7 first falls into a threefold.
#6 You keep missing the point dude. You keep referring to the puzzle solution and not looking at what OP and I are saying. There is no pawn recapture in the line because instead of d4 black plays Ke5, to which white responds Ke3. Then there is the d4 push.

Also stop pinging me please.
Guys in second puzzle Ka7 definitely does work, if black plays Kc7 then u go a6. The only reason it does not work is because of previous moves which make it threefold. But the point of a puzzle is to find a solution from the given position. The end user should not have to check for previous moves played in the irrelevant reference game to reach that position.
There are some really odd puzzles.

Doing a puzzle streak and came across this sub 1600 rated puzzle. Whereby you give up 2 rooks for a queen and pawn. Even stockfish finds it tough to turn this into an advantage - for some 40 moves.

At the end of the puzzle - taking basic piece value into account - you are actually down a pawn. Your opponent has two rooks for a queen. It doesn't feel solved at all.

lichess.org/training/95hxL

Is there an easy way to report sub par puzzles - give it the down arrow does not feel enough - especially when it has been played 4000+ times.
In terms of system u have a point but in terms of puzzle no. The solution is to win a queen,pawn and a bishop for two rooks leaving you with material advantage of queen and bishop vs two uncordinated rooks. The problem is puzzle ends before entire solution. After Kxb7 Qb5+. If Ka8 Qc6+. as soon as King touches a dark square white goes Bf4+ and will atleast win the black bishop outright if it blocks.

This topic has been archived and can no longer be replied to.