Hardest Mate in 4 of All Time

hey. i analysed the position(/diagram). it cannot be reached from normal chess starting position by legal valid moves.

if anyone can prove me wrong he will get 1 month diamond membership gift from me.

to prove you have to show the game. not by faulty logic like #24.

There are several posts which show why the puzzle diagram is legal. Note that discussions like the one in #24 are less about the legality of the position than about the logic of the solution. You may find that logic weird but that doen't affect the legality of the diagram. Watch your diamonds before someone grabs them!

you are incapable of showing the game. because there is no legal game.

you just showed an invalid diagram. what does retro analysis say about invalid diagram??

I have tried creating a sequence of moves that lead to that position in the diagram but it seems unbelievably hard or impossible.

I looked around for chess algoriths that would create a legal game to prove that a position is legal but there don't seem to be any. I mean, what is the definitive way to prove a position's legality other than producing a legal sequence of moves that lead to it? I would assume there would be forks of chess engines to deal with this problem ...

You won't lure me into the boring activity of producing proof games. Somebody will show you though. Proof games are not normal chess games. They are full of outrageous sacrifices and (under)promotions. For retro-specialists diagram #1 is very easy. Since you claim the diagram itself is illegal (apart from the solution), you do not even require to prove castling rights. That makes it very, very easy.

There are some engines to verify "shortest proof games" but I don't know about proof games in general. May be Remellion knows.

posted here. lets see if anyone can solve it 

https://www.chess.com/forum/view/fun-with-chess/is-this-position-legal2?page=76

comment #1506

I've never tried doing proof games, but this one seemed like it might be fairly easy and would be a good one to start with. My first attempt is shown below. I'm sure it could be made shorter with some work, but why bother?

White to play

Let me start by separating two issues. Your first objection in this thread was that more information about the castling rights should have been provided by the OP. I have explained about the Codex and how a solver could know what the conventions are and what should be assumed about the castling rights. Since you haven't responded to those explanations, I suppose you are satisfied with them.

The second issue is that you are not happy with the RS-convention. I can understand that and you share it with many problemists. Even the Codex committee that drew up the latest version of the retro-articles doesn't like RS a great deal and ranked it secondary to PRA. However, I am of different opinion and will explain why. It all begins with the questions you posed:

Q: Does the past of a diagram exist? A: No

Q: Has the position been set up arbitrarily? A: Yes

Those are absolute facts and any retro-composer will confirm them. Are we finished then? No! Why not? Because this state of affairs didn't make problemists happy and that probably includes you. Examples: would you like formations like "wPg2, wPh2, wPh3" in a problem? or "wBh1, wPg2"? Or allow both black and white castling in diagram #1? Probably not. Why not? Because these positions and plays are "ungamelike". What does "ungamelike" mean? It means that they couldn't occur in a real game - the same as saying "they have no legal past".

Because they are no real pasts but we still liked to have pasts, we had "to pretend". Everything in the conventions is fictional, every right and every past is created. However, there is more than one way to look at a fictional past and 2 of them are presented in the Codex as PRA- and RS-logic. The PRA-approach views the world as an analytical system; you extract knowledge of the past from the couch in your home by using your brain (a priori). The RS-approach views the world in the same way our physical world works; you must go out and observe things and do experiments to know things (a posteriori).

Your intuition told you that PRA is the right approach. Split the fictional past into cases controlled by different conditions and solve each one of them. There are however a number of issues and objections related to PRA:

1. PRA looks very scientific by taking into account all possible pasts, but factually it doesn't do that. In problems similar to #1 (but with a PRA solution) it does not take into account the case where neither white nor black has castling right from the past. To justify that, PRA imports the preferenced condition from the basic castling-convention (which is RS).

2. You come into a town where there are 2 restaurants on opposite sites of the street. You ask a local what's the best place to eat. He says: there are 2 restaurant owners in this town, mr Tran and mr Trump. Mr Tran is very friendly and he advertises on the daily puzzle. Mr Trump on the other hand is always out campaigning and he doesn't give a damn about his outlets. I don't know which restaurant belongs to whom, but I can tell you one thing "if a restaurant belongs to mr Tran, eat there and if it belongs to mr Trump, don't eat there". Do you now know where to eat? The point is that partial analysis only delivers partial knowledge and not everybody is happy with that outcome. A PRA-solution tells you that the problem can be solved but not by which moves.

3. You would expect PRA to be consistently applied to all retro-active conditions but it is not. Please check out #73. This is designed as an RS-puzzle and everybody will be very happy about that. Try to analyze it through a PRA-type logic and you will see how difficult that is. Particularly when you try to generalize the findings into rules for selecting partial cases in any diagram.

The foregoing is to show the weaknesses in the PRA-approach. Of course, RS has its own weaknesses but also has its strong points, for instance it has one solution. A critical issue is this one: how can you claim you are observing events - and derive knowledge about the past from them - while you create the events yourself? The answer is: "it is a role play and the role of the player is different from the role of the observer". Isn't that a strange thing, "roles"? No, it's not because solving chess puzzles has always been a role play! As a solver you always had to play both sides (forget about computers) and you must act for each side in their own interest, the best move for white, and the best move for black. That is a role play. Kids and newbees are sometimes confused about that. Isn't it stupid to play the best moves for the "opposite side"? The way to logically look at RS solutions is through the eyes of the observer role: "these are the variations we have seen being played; white mates in 4 in all of them and therefore the goal has been achieved". The observer doesn't know and doesn't care from which past and analysis they have been derived.

That would be an abstract from some pages in my book. I don't expect to have convinced you of the validity of RS-logic as there is lot more to it. I only hope you have recognized that there is more than one way to fill the information void left by the absent pasts of chess compositions.

There is no MATE IN 4!!!!!!!!!!!! hes a really good troll

I think, you just earned yourself a few diamonds. Congrats!

Well done on the proof game. Of course with this game Black cannot castle.

I figured a game where Black can't castle would be simpler than one where Black can castle but White can't. However, I don't plan to try to verify that.

What is your take on comments #103 and #112?