lichess.org
Donate

I'm puzzled

[Bug report]
http://ru.lichess.org/training/56519
Although
26.Be6 Ra7 27.Rec7 Qxh2+ 28.Kxh2 a5 29.Qe7# (mate in 4)
- the move 26.Be6 was rejected with "puzzle failed". In the same time the "right solution" is:
26.Qg7 Re8 27.Rxe8+ Kxe8 28.Bf7+ Kd8 29.Qg8+ Qf8 30.Qxf8# (mate in 5)
Thanks.
The same discussions go over at chesstempo, there however you will never fail a puzzle for finding a longer mate. 'Best' move is a relative thing, the shortest mate is not necessarily the most straightforward.
Come up with a method to systematically calculate 'progress' in a chess game, and this discussion can come to an end.
Chesstempo's 'Try again' sufficiently well. Instead, failing a move because of longer mate, or worse, because of engine evalutation 4.5 vs 4.2 or 3.6 (as I understand from http://en.lichess.org/forum/lichess-feedback/httpdelichessorgtraining38363) is counter to learning. It's a whole other can of worms concerning the chess understanding of engines which is not accurate to the centipawn - different engines don't produce the exact same output - and outright wrong in certain positions.

Isn't there an algorithm for validating a solution? In which case, shouldn't you be looking for a bug to fix rather than disabling puzzles?
[bug report]
lichess.org/training/58726
Right solution:
1.... Rh8 2.Qf3 Rh1+ 3.Kxh1 Qh2# (mate in 3)
Alternative 'N+2'- solution, that was rejected ("puzzle failed"):
1.... Qh2+ 2.Kf2 Qxg2+ 3.Kxe3 Bc5+ 4.Kf4 g5+ 5.Ke5 Bd4# (mate in 5)

This topic has been archived and can no longer be replied to.