lichess.org
Donate

I found a puzzle with 2 solutions

@mendlero said in #10:
> @DJdatadriven I agree that there are puzzles with multipule ways to win, but usually the alternative also counts as correct answer in this puzzle g2 counts as wrong
Have you reported this to lichess
g2 is wrong. Even according to stockfish ( the evaluation drops to -0.7 from -7.3) and with perfect play, White can hold.
@HollyMu said in #9:
> @Schtaeve : yes you are right. but there are thounds of puzzels, which train you "loose allway the rook to win the queen".
> This one is out of this box ....

Before responding with things like this, please can you play it out on Stockfish. There's another pawn on the c file. You get to keep your rook and it's impossible for white to prevent the promotions of both the g and c pawns without losing their rook.

One outcome is absolutely crushing (rook vs two pawns and white king nowhere near their pawns to protect it). The other outcome is a queen against a rook and two pawns, of which white can hold their own with perfect play and it's practically a drawn position.
<Comment deleted by user>
@Schtaeve
White cant hold with perfect play if black plays perfect too after ...g2 kxd4 if you open your opening explorer you will see that black is wining or you can check my study about the puzzle where I show the variation after g2 if both sides play perfectly
(I put the study in the first post)
@DJdatadriven :
Ok, lets take your look.
The king do not beat the rook, but kills the g-pawn.
Same thing.
After the way (after g2) the black side will win with the other pawn.
The king will move false what ever offered he takes.
Schtaeve told me not to "fakenews", but I really do not understand lichess (in this point).
But lets say, I learnt some chess while thinking about this situation and so we all won .....
Pushing g2 wins. There is no much doubt about it, however, you can still make a queen in either side by playing Rd3+ and follow the actual solution. You will get the rook for both pawns. So you end up in a K+R vs K+2p.

But if you go for the g2 push and allow the capture of the rook, you will end in a K+Q vs K+R+2p. Its doable, but objectively harder if not impossible if you get is this kind of position. You need to be very precise. So going for Rd3+ is the safe way.

lichess.org/editor/8/5k2/4R3/5P2/4P3/2K5/8/6q1_w_-_-_0_1
This is unfortunate, to be punished for finding a good move. Thankfully these puzzles are relatively rare and disappear into the crowd when you continue doing lots more puzzles.

Had this puzzle been from a physical book, then the intention of such an ambiguity would be clearly pedagogical: you are supposed to ponder all the possibilities in a position and not quit calculating as soon as you find the first line that doesn't lose.

Obviously the ambiguous puzzle would have to be assigned a much higher rating than a similar position without this ambiguity. But a Lichess puzzle was found automatically (by running a tactics-search algorithm on the games played on Lichess) and they get their rating not by the judgment of some master player but by how well or badly Lichess players do and what their ratings were. The puzzle's rating is not objective and static but changing and relative to the history of the community's attempts at it.

The problem here is that many lower rated players (lower than the objective rating of the puzzle) who should lose against the puzzle (losing rating points and increasing the puzzle's actual rating) will solve the puzzle the easy way and despite having quit calculating early. So luck is very much an issue here and ideally shouldn't be.

So again, in a printed book you'll find such puzzles and learn from them to take your time, consider all the candidate moves. They prepare you better for real play that way. In puzzles we tend to slip into an automatism where we stop looking for alternatives as soon as something looks ok. In real play what seems ok might turn out to be losing and it is crucial to look for more than one ok line, hold them firmly in the mind, and compare and contrast their individual ok-ness.

Cheers!

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