Last few moves were something like:
Since white moved last, black has the move and mates in 1 with Nxc3#
Mate in 1 [retro]
Sort:
BigDoggProblem wrote:
Last few moves were something like:
Since white moved last, black has the move and mates in 1 with Nxc3#
That’s correct. Good work!
I suppose this one wasn’t as hard as the other ones, since it’s clear (at least to experienced solvers) that Ng1 is promoted, being free on its own. However, the pawn moves are precise, and only a black bishop can be captured on b7: a knight is one move too slow, whereas a rook/queen would have checkmated on f1. The bishop got there via c8, visiting the home square of the original light-squared bishop.
daStrwbrry wrote:
BigDoggProblem wrote:
Last few moves were something like:
Since white moved last, black has the move and mates in 1 with Nxc3#
That’s correct. Good work!
I suppose this one wasn’t as hard as the other ones, since it’s clear (at least to experienced solvers) that Ng1 is promoted, being free on its own. However, the pawn moves are precise, and only a black bishop can be captured on b7: a knight is one move too slow, whereas a rook/queen would have checkmated on f1. The bishop got there via c8, visiting the home square of the original light-squared bishop.
I should have held off on posting the solution. Would have been interesting to see if anyone else would have tried to solve it.
BigDoggProblem wrote:
I should have held off on posting the solution. Would have been interesting to see if anyone else would have tried to solve it.
Yeah, I suppose there might have been some others who could possibly solve it, but just needed more time. I also would have expected some "Nxc4#" replies from those who know nothing about these types of problems, but fortunately that hasn't happened (yet).
Maybe send the solutions via PM next time? At least I can acknowledge that you solved it while others are still thinking about it. Alternatively, you could just wait for a few more hours until you think they have had enough time: you can decide what's best. Still, you solved it correctly 
daStrwbrry wrote:
I suppose this one wasn’t as hard as the other ones, since it’s clear (at least to experienced solvers) that Ng1 is promoted, being free on its own.
Do we get any explanation? Why would the g1 knight be promoted?
magipi wrote:
daStrwbrry wrote:
I suppose this one wasn’t as hard as the other ones, since it’s clear (at least to experienced solvers) that Ng1 is promoted, being free on its own.
Do we get any explanation? Why would the g1 knight be promoted?
Well, that’s the retro part of the puzzle. With these types of puzzles, we are only concerned about what’s possible and what’s impossible; we have to determine how this position can be reached via legal moves. Here it is possible to work out that the knight is promoted (as well as other things), just from looking at the position shown on the board. To do this we need to analyse the position in terms of legality: this is retro-analysis.
To start the retro-analysis, look at the captures from each sides. Pawn structures are usually the first thing to look at: for this position, both sides have doubled pawns, so both sides have made at least 1 capture with pawns. White is missing only 1 unit, and black is missing 2. White’s missing unit, which seems to be the a-pawn, must have been captured on the d-file, but it seems impossible: it needs 3 captures to the d-file, but black is only missing 2 pieces. However, there is a trick…
Another problem you will need to deal with is how to reach the pieces in the bottom left corner - the “cage”. Specifically these pieces:
This is just a brief part of the analysis. Once you work out the captures from both sides, as well as how to get to the “cage”, you should be able to work out why the knight must be promoted, or why black has the move.
Usually with these puzzles (but not always), the free piece(s) [i.e. the piece(s) not stuck in the cage] unpromote (when retracting, or unplaying moves) to release the cage, or to provide tempo moves for the other side. That was what my little comment with Ng1 being promoted was about.
The a-pawn could have promoted, sure, or could have been captured. The g1 knight might be a promoted piece or not. The latter options might be impossible, but I have no idea why. There's a zero percent chance that I'm willing to put in some weeks to solve and understand this, that's why I asked for an explanation. But whatever.
@BigDoggProblem already gives the last moves at post 2 which are required to reach the position, which I treat as a solution. But if you need further convincing that the knight on g1 is promoted, try creating a game that reaches the position with that knight not being promoted. The game doesn't have to be realistic - both sides could make horrendous moves - as long as it reaches the position. You should find out that it is impossible. The only way to reach the position is via the moves at post 2, where the knight is promoted.
I will give a full explanation in a few days if no one else wants to explain the retro-analysis, just to convince you that the knight must be promoted. I have given some hints at post 7 on where to start analysing, and the moves at post 2 might help understand what must have happened in the position. I do understand though that this puzzle isn't like the normal mate in 1 puzzles which you typically see on this site, since it requires a different approach to solve it: you have to think about what must have happened in the past, rather than trying to checkmate the enemy king.
daStrwbrry wrote:
@BigDoggProblem already gives the last moves at post 2 which are required to reach the position, which I treat as a solution. But if you need further convincing that the knight on g1 is promoted, try creating a game that reaches the position with that knight not being promoted. The game doesn't have to be realistic - both sides could make horrendous moves - as long as it reaches the position. You should find out that it is impossible. The only way to reach the position is via the moves at post 2, where the knight is promoted.
I will give a full explanation in a few days if no one else wants to explain the retro-analysis, just to convince you that the knight must be promoted. I have given some hints at post 7 on where to start analysing, and the moves at post 2 might help understand what must have happened in the position. I do understand though that this puzzle isn't like the normal mate in 1 puzzles which you typically see on this site, since it requires a different approach to solve it: you have to think about what must have happened in the past, rather than trying to checkmate the enemy king.
The sequence that BDP gave is not "the solution". The solution is to prove that it's the only possible way and all others are impossible.
The problem with this is that there are more ways to reach this position than atoms in the universe. Maybe most of them are impossible and only one remains. As I said, I don't feel I have the required days / weeks / years / centuries to solve this on my own, that's why I asked for an explanation.
Also, there is absolutely no need for those condescending snide remarks about "this is not a mate-in-on puzzle". I know what retro-analysis is.
Whatever. This conversation has became sour for no reason. I'm out. Bye.
magipi wrote:
daStrwbrry wrote:
@BigDoggProblem already gives the last moves at post 2 which are required to reach the position, which I treat as a solution. But if you need further convincing that the knight on g1 is promoted, try creating a game that reaches the position with that knight not being promoted. The game doesn't have to be realistic - both sides could make horrendous moves - as long as it reaches the position. You should find out that it is impossible. The only way to reach the position is via the moves at post 2, where the knight is promoted.
I will give a full explanation in a few days if no one else wants to explain the retro-analysis, just to convince you that the knight must be promoted. I have given some hints at post 7 on where to start analysing, and the moves at post 2 might help understand what must have happened in the position. I do understand though that this puzzle isn't like the normal mate in 1 puzzles which you typically see on this site, since it requires a different approach to solve it: you have to think about what must have happened in the past, rather than trying to checkmate the enemy king.
The sequence that BDP gave is not "the solution". The solution is to prove that it's the only possible way and all others are impossible.
The problem with this is that there are more ways to reach this position than atoms in the universe. Maybe most of them are impossible and only one remains. As I said, I don't feel I have the required days / weeks / years / centuries to solve this on my own, that's why I asked for an explanation.
Also, there is absolutely no need for those condescending snide remarks about "this is not a mate-in-on puzzle". I know what retro-analysis is.
Whatever. This conversation has became sour for no reason. I'm out. Bye.
From my point of view, you brought all the sourness in with you. There was indeed no reason; you made a choice to have that attitude.
There is no point in trying to explain complex problems to someone who will not meet halfway and make some effort to learn.
Off you go, then.
FieryChallenger wrote:
Nxc4 is mate!
...if only it were White's turn. But it's not.
magipi wrote:
The sequence that BDP gave is not "the solution". The solution is to prove that it's the only possible way and all others are impossible.
I do like to see analysis/explanation too, since it does give the sense of a more complete solution. However, if the last moves are shown, this suggests to me that the solver understands what had happened in the position. It's unlikely that someone posts correct but random last moves without understanding the puzzle. Additionally, I recognise BDP as a strong solver with retros in various other threads.
The problem with this is that there are more ways to reach this position than atoms in the universe. Maybe most of them are impossible and only one remains.
Sure, there are several billion games that reach the position, but the point is that in all of these games, Ng1 is promoted, and black has the move. There is no other way to reach the position - if you do the analysis - without Ng1 being promoted, or with white having the move. Certain illegalities prevent other ways to reach the position.
As I said, I don't feel I have the required days / weeks / years / centuries to solve this on my own, that's why I asked for an explanation.
If you think that you cannot solve the puzzle, or feel pressured that you have to solve it, then don't bother. No one is insisting that you have to solve the puzzle and work out all the details. However, you won't get the explanation if you don't try to solve it. Sure, there might be others who can and are willing to explain it. But not always.
This might seem a bit harsh, but the explanation should really be for people who at least make a good attempt to try and solve the puzzle. Giving explanations for people who don't make an effort to solve it in the first place seems a bit pointless to me.
Also, there is absolutely no need for those condescending snide remarks about "this is not a mate-in-on puzzle". I know what retro-analysis is.
Sorry if I appeared condescending, I just wanted to assist you in trying to solve the puzzle and make sure you understood what is going on, but that would be underestimating you. Nevertheless, if you do know about retro-analysis, why not make an effort and try to solve it yourself?
I also talked about obvious things for "experienced solvers". I will refrain from mentioning these things in the future since I realise that this might be confusing for other people, and it might put them off when trying to solve the puzzle. Sorry if you were confused by this; next time I will stop talking about "experienced solvers" stuff.
Whatever. This conversation has became sour for no reason. I'm out. Bye.
Fine, go. I don't like having internet arguments as much as you do. But if you do return here, just remember this:
If you don't want to solve the puzzle on your own, then you don't have to. But if you want an explanation for the puzzle, there's not much point in giving it to you if you don't at least make a decent effort to try and solve it. Besides, if I just gave away the explanation away every time, then that would just spoil the whole point of the retro: figuring out why black is the one who mates in one, not white. That's no fun if I just gave you answers.
@magipi The way @BigDogProblem posts the solution to this RA-problem is common. It starts from the critical position, defined as the point from which the forward play to the problem diagram is "more or less" unambiguous. "More or less" is also common since making RA-problems is difficult and the author is granted some latitude in the precise moves and sequences. The parts which show up in every possible solution (like the fact that Ng1 was promoted) are pluses of the composition while the minuses that are not in every proof game are considered of lesser importance. For instance, all moves before the critical position are considered irrelevant since there are so many legal move orders.
The remainder is based on the required skill level. If you can't comfortably retract a proof game from the critical position posted by BigDogProblem, you don't qualify as an RA-solver. This is the same as for endgame studies. Before you can successfully solve endgame studies. you need considerable knowledge of basic endings. Though the situation for studies is simpler due to the availability of engines and tablebases which do not yet exist for RA-problems (as far as I know).
RA-problems have another property in common with endgame studies. Solutions are not PROOFS though they look a bit like it. If they were, why do you think hundreds of studies were found cooked after the tablebases entered the stage? A proof by mathematical standards can never be undone. They were simply good looking move series rated by skilled observers!
The final factor for deciding soundness of compositions comes from allowing reversing the burden of proof. A hundred years after publication and awards any composition can still be nullified by a 12 year old kid finding a solution dual. Compositions are never safe - until complete tree searches can be accomplished by future quantum-systems.
A new retro problem I made.
mate in 1