Longest King Maze Challenge

A King Maze is a type of problem that follows these 3 rules:

1. Only king moves may be made by both sides

2. The white king must try to reach a certain "goal square"

3. The black king must do his best to stop white from reaching the goal square (including stalemating and checkmating himself so there are no more legal king moves).

What is the longest king maze you can design?

Here's an example where the goal square is a1:

Also, for those of you who have come across this topic before here's a link to a collection of some of the best puzzles: https://en.lichess.org/study/d8WFHDZ0 (apologies for advertising the enemy).

I made one, though it's not quite legal, as I need the black king to make moves he wouldn't make to get the solution to work (turns out it was harder to get one correct than I thought), but it shows the power of triangulation, which is worth something, I suppose.

For next time, I'll try to make something that is legitimitely correct. 

There is a standard fairy chess variant called "goal square" created by a dutch composer. It simply states that white must place a unit (any unit) on a specified goal square in a certain amount of moves. Obviously, black will do anything to prevent it, even get himself mated!

Unfortunately, I couldn't find a reference on the Internet with the full information on this variant. Also, I am not sure of the inventor's name though it was probably Hans Burbach. I thought I'd let you know anyway to show the idea of "goal squares" has been around.

Good news for you! The diagram is legal! It's a close call but it all works after black makes 2 captureless knight promotions on a1 and h1. With these hints you should be able to produce a proof game yourself if you so desire.

By "not quite legal" I think he meant not valid as a King Maze because Black didn't do his best to stop White from reaching the goal.

Ah, you may be right. And may be he didn't realize how close he'd come to the edge of the 'other' legality. Anyway, he is now warned when making further changes to his puzzle.

@Player8-CHAMPION You've been very kind to the black king giving him many squares, which does make it tricky to find a solution. If the black king shuffles between a7 and b8 then white cannot make progress. You're right in recognising the triangulation theme is useful though. It is used a lot in some of the later puzzles I linked which you may find interesting.

@Arisktotle That's cool, I can find some chess maze puzzles, but none where both sides move. It would be cool to see other examples of this type of puzzle though, thanks for letting me know.

I found one example by Hans Burbach & Jac. Haring, 1989, Probleemblad. Others should be somewhere in my archives of decades ago; if only they were ordered .. (sigh) 

The stipulation for the problem below is to reach goal square b3 with any white unit in 3 moves. 

The main line is simple but the real content is in the try lines as you can see in the solution tree.

Tricky!  I was sure I had it with Rxg4.

I found that this problem, a 4th prize winner by Johannes Jacob Burbach in 1988, was contributed to yacpdb by Tobias Verhulst on 4/24/2014.

@Arisktotle Nice puzzle, it seems so easy at first glance! Here's the longest king maze I could come up with (not quite a legal position but it is a legal king maze). I'm not sure if anyone will be able to extend it, it's quite tricky! The goal square is a1.

Actually, quite a long way from a legal position. If someone is interested in extending a solution, I suspect they will wait for a legal position to start from.

@n9531l Haha, there are 16 white pieces and 16 black pieces which is much better than some other puzzles of this type have been! For this type of puzzle starting from a legal position does not concern me, it's just a nice bonus if it does.

However, I do have a long puzzle starting from a legal position which I'll post if anyone's interested, although the one above is much nicer in my opinion and a far greater challenge to try to extend.

The question of whether starting from a legal position should be a requirement or just a nice bonus is something we can agree to disagree about.

Is your long puzzle starting from a legal position longer than #1?

@Chess_Dogg : The unofficial meta-convention on 'Fairy Escalations' allows you to break any chess rule including the legality of positions. The condition is that you must have 'something to show for it', preferably a lot. It is not worth wasting legality to extend the maze with a single move. However, I was already impressed by post #1 and I trust that #11 will deliver the goods I expect from it as well. Just need to find a hole in my agenda big enough to fit in your maximized King Maze. Currently it tells me I should have gone to bed an hour ago!

Yes. The legal puzzle is just over 3 times longer than #1 and #11 is over 4 times longer. 

I hope you enjoy solving it Arisktotle. It's nowhere near as simple as #1 so may require a good nights sleep!

That's useful information. I hadn't even recognized this as a Fairy problem, since all the pieces and moves are normal chess pieces and moves. I guess it's the constraints on what the pieces are allowed to do that makes it one.

Fairy conditions are pretty much anything outside regular stipulations. There're fairy pieces (nightrider, grasshopper etc), fairy conditions that change the fundamental rules (Circe, Alice etc) and in a strict sense, heterodoxes (S#, H#, H= etc) are also types of fairy conditions, changing the objective but not other rules.

King mazes as presented here are fairies. I was also put off by the illegality of them at first... until I tried to make one, which is definitely overkill incarnate (see link in #2; looking back, I'd replace wRh5 and wBh7 with a wQh7, and maybe also scrap the whole thing and rebuild it. wK reaches h1 in 41.)

#11 should be worth the effort - the slightly different 222-mover ten months back certainly was. I think it's fully worth the illegality. Not sure I can afford the time again though. :-)

@Remellion:  I know about s# and h#, but yacpdb also has a lot of r# problems. How do those work? I tried some online searches but must not have looked for the right thing.

Edit:  I searched some more and found reflexmate. I guess that's what these are.

Here is the current longest maze starting from a legal position. Not as hard to solve but definitely quite hard to extend and keep the legality. It'd be interesting to see peoples ideas.