Your network blocks the Lichess assets!

lichess.org
Donate

Open sourced positional puzzles

Kudos to the effort.Would like to see more of these.

Kudos to the effort.Would like to see more of these.

Very interesting @neilgd72 I will have a look at it, and if you don't mind, I might build a front end for it on my website.

Very interesting @neilgd72 I will have a look at it, and if you don't mind, I might build a front end for it on my website.

Very nice! I hope Lichess adds them to their list

Very nice! I hope Lichess adds them to their list

Muito bom!!! Sinto falta de ver exercícios competentes de estratégia. Vocês estão no caminho certo, e a comunidade agradece!!!

Muito bom!!! Sinto falta de ver exercícios competentes de estratégia. Vocês estão no caminho certo, e a comunidade agradece!!!

@HollowLeaf said ^

Very interesting @neilgd72 I will have a look at it, and if you don't mind, I might build a front end for it on my website.

Go for it!

@HollowLeaf said [^](/forum/redirect/post/442CsEoV) > Very interesting @neilgd72 I will have a look at it, and if you don't mind, I might build a front end for it on my website. Go for it!

This is honestly pretty similar to the work of "Woodpecker method 2" ... except before the humans went through everything and found the puzzles to make the book actually useful.

I went through about 30 of these positions and about 30% of them are actually tactics after all (trapping a piece, or saving your piece to the only square it can go to). About 30% of them seemed somewhat interesting since you had to change the pawn structure in a way that was good for you. But even then, most of the time your move is forced because there is no other way to avoid losing a pawn. About 30% of them seemed genuinely useful. Either it was an example of prophylaxis or an example of avoiding/attempting a trade of a good piece for a bad piece.

As always with AI or any other statistical analysis, it may be helpful to generate a starting point. But to be actually helpful, you kinda need to use your own brain. We haven't automated the art of providing truly useful information just yet!

This is honestly pretty similar to the work of "Woodpecker method 2" ... except before the humans went through everything and found the puzzles to make the book actually useful. I went through about 30 of these positions and about 30% of them are actually tactics after all (trapping a piece, or saving your piece to the only square it can go to). About 30% of them seemed somewhat interesting since you had to change the pawn structure in a way that was good for you. But even then, most of the time your move is forced because there is no other way to avoid losing a pawn. About 30% of them seemed genuinely useful. Either it was an example of prophylaxis or an example of avoiding/attempting a trade of a good piece for a bad piece. As always with AI or any other statistical analysis, it may be helpful to generate a starting point. But to be actually helpful, you kinda need to use your own brain. We haven't automated the art of providing truly useful information just yet!

OK, I'll ask: how did people select these puzzles?

OK, I'll ask: how did people select these puzzles?