I have made several compositions which could rate as the hardest ever and the next one is probably not quite in that category. But it is pretty hard and a nice change from the standard batch unleashed every time the hardest question comes up again - about once per month.
.
One of my older studies. Not sure how hard it is. The annotation is a bit old, so there's probably way too many exclamation marks.
Another extremely hard puzzle! White to move and checkmate in 12! Found it on https://www.youtube.com/watch?app=desktop&v=FN8hVAV20hQ!
One of my older studies. Not sure how hard it is. The annotation is a bit old, so there's probably way too many exclamation marks.
That's a pretty cool study with careful tempo play. Note however that 2. Nf5!! is a very strong move but not really zugzwang. The latter would assume by definition that white had no winning plan if black could "pass" but he does! Of course, there are real zugzwangs later in the solution.
The question of difficulty is a complicated one with engines in the game. Like, your study (and mine too) is very easy for engines while it is a massive task for humans. Some of the hardest puzzles are quite simple for men but pose a big challenge to engines. And some are both though for different reasons. There is also paradox in the human approach. As we have learned that composed puzzles feature weird but interesting moves we tend to prioritize those in the search order and quite rightly so. Does that mean they are not that hard? The common game engines use no meta-arguments from the human environment. For them much of the trouble is in the first move because that is the one furthest removed from the ultimate justifications. They only ratify the beginning after they found the end (see e.g. post#128) while we often see the sign posts along the way telling us we are on the right track - or not! By engine standards our studies are `easy`, they solve them in seconds, for humans they are `hard` especially yours which consists of subtle maneuvering and zz situations. No really attractive `big` surprises to guide us through the maze. Which is OK for your study type but hard for us solvers!
When it comes to "best" matters are clearer to me than on "hardest". Compositions are primarily about artistic content and much less about difficulty which rates them above most "engine-hard" puzzles even when these feature an amazing "first move". The sum total of "interesting choices" in our studies is much higher than in engine type studies. So far, for who knows what the future holds in store for us!
Final note: Though we tend to forget about it on chess.com, solving a puzzle ought to include "playing both sides". In many compositions the black choices are as important - sometimes even more important - than the white choices. For proper judgement, give everyone a piece of paper and a pen and let them figure out the whole thing! Btw, in solving tournaments you even get a complete chess set and you can try all you like. Only the written text on the paper counts!
Beautiful queen domination in the last study! The material point scores are even but you would expect the black queen to easily hold in this position.
I wonder, are the domination scenarios found by engine or some kind of AI-software? Or did the software only analyze them?
As a composer I noticed that black could win if his a-file pawn is one square more advanced in the later stages. That however would require black to sac his queen on g1 and not promote when he can, but play ..Kxg3 instead. The extra twist is what composers are looking for to get the max out of their ideas.
.... and then white wins with his material advantage ...
Your text shows you have no clue what you are talking about. Do you live on planet earth?
Here is a nice mate in 12 by Dutch composer Johan Christoffel Van Gool that borrows the idea of Konrad Bayer's "Immortal Problem" which is sacrificing all of your officers and having a lone pawn deliver the check mate:
White to move and mate in 12:
And here is the "Immortal Problem" by Konrad Bayer:
White to move and mate in 9
Here is a nice mate in 12 by Dutch composer Johan Christoffel Van Gool that borrows the idea of Konrad Bayer's "Immortal Problem" which is sacrificing all of your officers and having a lone pawn deliver the check mate:
White to move and mate in 12:
And here is the "Immortal Problem" by Konrad Bayer:
White to move and mate in 9
Composers commonly do not duplicate the framework of earlier compositions literally. There is an essential improvement over Konrad's original - the first extension any composer would think of. Van Gool sacrifices all white pieces in the game array, while Konrad stops at 5 out of 7!
Here is a nice mate in 12 by Dutch composer Johan Christoffel Van Gool that borrows the idea of Konrad Bayer's "Immortal Problem" which is sacrificing all of your officers and having a lone pawn deliver the check mate:
White to move and mate in 12:
And here is the "Immortal Problem" by Konrad Bayer:
White to move and mate in 9
Composers commonly do not duplicate the framework of earlier compositions literally. There is an essential improvement over Konrad's original - the first extension any composer would think of. Van Gool sacrifices all white pieces in the game array, while Konrad stops at 5 out of 7!
Indeed, I should have specified that all officers that were on the board were sacrificed for Bayer's, and that all officers that are in the game's array for Van Gool's.
Outside of the field of chess problems, there are even a few real-life games in which a player managed to sacrifice all or nearly all of his pieces as the most efficient path to the win.
One of our chess.com members (a GM from Europe) has just such a game.
EDIT:
Found it! GM Serper vs Nikolaidas.
Sacrifice all the pieces!! - Chess.com
Found it! GM Serper vs Nikolaidas.
After playing such a game you will have to feel you're gonna be the next world champion 
Thx for the best puzzles
Yes, it's a good one. It is so good that it is always reposted when someone asks for the hardest puzzle. In this thread it is already in post #53 and I wouldn't be surprised to find it several more times here! (Checked; just posted twice!)