lichess.org
Donate

Some Puzzling Analysis

I've encountered some puzzles where you have an option between two mate in 1s and it accepts 1 and rejects the other
Haha my defensive moves rating is 2048 while my sacrifice and kingside attack ratings are both above 2400. Cant say i didnt expect this but I didnt imagine the difference would be this big
Out of curiosity: are the evaluation of your strengths and weaknesses in the "puzzle dashboard" based on only how you perform in the "healthy mix" tactics, or do they also take into account the puzzles you solve in particular theme categories?
What i encountered in new puzzles when solving in healthy mix that rly makes them frustrating sometimes is when a solution leads to a position that is +4-5 but it is unclear af and when u let engine play it out it only achieves material advantage in like 20-30 moves at 25 depth
One time solution ended with position like Queen + Rook + King + pawns vs Queen + Rook + King + same amount of pawns with one king slightly more exposed and it took computer 30+ moves to win queen for a rook, and it's not even the best example, just the one I remember.
So for me it is rly frustrating when I can see the line, but I dismiss it cause position in the end is really unclear and I am used to puzzles leading to a clear advantage. I don't think I encountered that with old puzzles that often.
It would be nice, if in the eval process one could not only have thumbs up, down but also Number of moves that should be added to the solution: See for example lichess.org/training/R2H7h, the solution given does not make any sense at all, unless you see the beautiful 24.Qf1 Re1!! 25. Qe1 Qg2# in advance.
Really nice feature. A commentary section would be amazing !
This is amazing! Finally lichess has done something really nice and kept up with expectations one could have. I am proud. I hope lichess will also work on mobile study creation and studies in general to improve them just as they did with puzzles.

This topic has been archived and can no longer be replied to.