http://en.lichess.org/training/17409
24...Nh3+ 25.Kh1 Nf2+ is completely winning as well due to 26.Kg1 Qe3! and should not be rejected.
24...Nh3+ 25.Kh1 Nf2+ is completely winning as well due to 26.Kg1 Qe3! and should not be rejected.
While both 25. ...Nf2+ and Rf2+ are winning, Rf2+ wins the queen more quickly and is more to the point. From an engine perspective 25. ...Rf2+ also gets a decidedly better evaluation (a difference of about 2.5 points according to stockfish). Lichess puzzles, whether you like it or not, require you to be as precise as possible, not just winning.
my point is not that Nf2+ should be accepted as the solution, but it quite obviously is worth a "good move, but not the best" - we are talking about a valuation of +5 vs +7 here.
There are quite a few posts already about this and I understand that for some reason the lichess makers opted for this solution.
A completely and utterly winning alternative should never be rejected as a fail though - at least IMHO.
Anyway, lichess is a fine site and I can certainly live with those tactics trainer flaws. Kudos to all the people who have contributed.
There are quite a few posts already about this and I understand that for some reason the lichess makers opted for this solution.
A completely and utterly winning alternative should never be rejected as a fail though - at least IMHO.
Anyway, lichess is a fine site and I can certainly live with those tactics trainer flaws. Kudos to all the people who have contributed.
You do have a point. The only times I have noticed getting the "good move, but not the best" message is in mating sequences when you play a move that requires one extra move to mate than the minimum. It's worth noting that the puzzles are automatically generated, which may limit how well one can automate when the "good move, but not the best" is appropriate.
The lichess puzzles have been good for me for calculation and making sure to notice all details of the position since the lichess puzzles usually require the absolute best move to get the puzzle correct. While sometimes winning moves are considered wrong, I don't mind because there are many times where finding the most accurate move is necessary, and the lichess puzzles test whether you can find that best move even though it may not be "necessary" in the given position. Obviously lichess isn't the best site for puzzles, but it does have value, and the automatic generation will give us an endless source of puzzles. I also do puzzles at chesstempo.com.
Anyway, I'm not really meaning to argue against your point. Just pointing out a few reasons why I don't mind the way grades puzzles.
The lichess puzzles have been good for me for calculation and making sure to notice all details of the position since the lichess puzzles usually require the absolute best move to get the puzzle correct. While sometimes winning moves are considered wrong, I don't mind because there are many times where finding the most accurate move is necessary, and the lichess puzzles test whether you can find that best move even though it may not be "necessary" in the given position. Obviously lichess isn't the best site for puzzles, but it does have value, and the automatic generation will give us an endless source of puzzles. I also do puzzles at chesstempo.com.
Anyway, I'm not really meaning to argue against your point. Just pointing out a few reasons why I don't mind the way grades puzzles.
another IMHO awful example of a "fail" that should be a "good move, keep looking":
http://en.lichess.org/training/18100
...Qf2+ Kh1 Rxf1 ("correct" is Qxh2#)
http://en.lichess.org/training/18100
...Qf2+ Kh1 Rxf1 ("correct" is Qxh2#)
next one:
http://en.lichess.org/training/15498
...Kg7 f3 Bc5+ fails??! (Rh8 only move accepted as "correct")
http://en.lichess.org/training/15498
...Kg7 f3 Bc5+ fails??! (Rh8 only move accepted as "correct")
http://en.lichess.org/training/33463
Qxf7+ is +7 but gets rejected (should be "good move"), the "correct" Rxf7+ is +8.
Qxf7+ is +7 but gets rejected (should be "good move"), the "correct" Rxf7+ is +8.
This topic has been archived and can no longer be replied to.