Puzzles that are not instructive

I like the lichess puzzles and puzzle streak, but sometimes you come accross puzzles that are not instructive. This is an example:

lichess.org/training/W50J2

You don't win any pieces in this puzzle. The material is equal at the end of the captures (though the puzzle doesn't go to the end of all the captures). How are we supposed to learn from a puzzle like this? I guess white is significantly better because white's pawns are all the way up in black's grill? But I don't see much learning value, especially for a puzzle that's rated 1400-1500.

I wish lichess had some human curation of these puzzles.

I see your point. This happens to me also when there is more than one way to win material.
Its because the computer is picking these scenarios and not a person. I agree, its very irritating.
My record for puzzle streak is 39. Whats yours?

Agreed. The computer sees an advantage of +3 at the end, but you won't win material neither checkmate for another 10 movers if both sides play correctly. Of course, the black king is so restricted that he is easy to attack, but still - I would expect a more obvious outcome from a puzzle.

@greysensei said in #1:
> How are we supposed to learn from a puzzle like this?

If you can't see the difference between the winning line and the alternatives, then the puzzle is precisely putting the spotlight on a flaw in your current chess understanding. How is that not helping your learning?

What is the learning value of a puzzle with an obvious solution? How are you supposed to learn from something you already know?

@pepellou Repetition. Thats how we learn. It helps spot tactics during a game. Not all tactics require you to be a GM to solve them.. sometimes just simple pattern recognition can be enough.

@pepellou said in #4:
> If you can't see the difference between the winning line and the alternatives, then the puzzle is precisely putting the spotlight on a flaw in your current chess understanding. How is that not helping your learning?
>
> What is the learning value of a puzzle with an obvious solution? How are you supposed to learn from something you already know?

I agree with your logic if the puzzle was higher rated. You don't win any material, but you have to evaluate the resulting position in the line as winning for white. However, at the level where this puzzle is, it does not seem appropriate. I solved it correctly, but only because I was thinking... Hmm what would a 1400 do here? Exchange sac followed by check. You can tell that this puzzle was algorithmically selected. I don't think you would find such a puzzle in a book of chess puzzles, like the woodpecker method or something similar.

@greysensei
You win 1 pawn in this puzzle.

Without that tactic, black had a completely winning position, extra pawn, rooks are doubled, some threats of back rank mate,

After this tactic, black advantage is completely gone, and white have equal or better position.

So think of it as a tactic to save white from losing.

@greysensei said in #6:
> I don't think you would find such a puzzle in a book of chess puzzles, like the woodpecker method or something similar.

Why not?

I think it's an instructive puzzle. The very fact that it forces you to look beyond the material difference makes me think it should be instructive for all of you complaining about it.

> I solved it correctly, but only because I was thinking... Hmm what would a 1400 do here?

Well, I can tell you that's not how I solved it :-)

But that's the thing with puzzles and their rating, isn't it? Not everyone will solve them using the same thought process. A complicated tactic sometimes gets a low rating because everybody solves it when the solution is the most natural move, even when weak players miss checking important lines that are not the main line. Some other times an easy puzzle gets a high rating because everybody is failing it. I've even seen some horizon effect in puzzles, where some puzzles are solved more by weak players than strong ones, because the strong ones see a problem that the weak players don't, and then they don't find a solution for the problem so they try a different line...

But the point is that the rating is a result, not something given. Same as with players (where rating measures performance and not strength, and there are so many factors causing it other than strength).

I can tell this is the kind of puzzle many weak players will solve because "that's the only tactic I see so I'll do it". When after solving it you raise the question "why is that the solution and not any other move?" and you don't have an immediate answer, I think you're exactly asking the right question and you're on your path to learning something. If then you ask yourself "why is this rated 1400?" then maybe you're onto something or maybe not (I honestly think that reflecting about it could lead you to understand ratings more than chess), but if your conclusion is that the puzzle is not instructive based on the rating then you have a problem.

Thinking that the puzzle should be obvious to you because it's 1400 is like thinking you're never going to lose to someone who's 1400. And saying it's not instructive because it's not obvious is such an oxymoron ...

@greysensei said in #1:
> I wish lichess had some human curation of these puzzles.

I wish that too. We should create a study to collect these instructive puzzles :-)

By the way, there is a human curation of puzzles. Even though there's no direct button to do it when you're training, some puzzles have been reported to be wrong in the forums and have been removed (those were different cases though, as they were objectively wrong which can happen sometimes depending on the engine's analysis depth when creating them).

So if you see a puzzle that's wrong or should be removed, I think you're doing the right thing by reporting it in the forums. I just disagree with you specifically about this particular case :-)

lichess isn't a big money company. They have an automated algorithm that creates fantastic puzzles 95% of the time with 3% of the resources available compared to chess.com.

Besides that I also think the puzzle is very educational although a bit questionable when you consider 1400 rated players are confronted by it.

Edit: percentages completely made up, but I hope they transfer the meaning.