SudoCue Users

Posted: **Mon May 28, 2007 10:50 pm**

In ancient Japan, a Samurai who has lost his master was called Ronin. These uncontrollable outlaws often caused uproar in the country.

So much for today's history lesson...

I decided to call this Samurai-X Ronin because it is impossible to master. It has only 70 clues. To give you an impression: The full suite of techniques for the regular Samurai puzzles (with Diagonal techniques added) only manage to solve 2 cells.

If you can avoid bifurcation (a euphemism for 'guessing'), this will be a great exercise to discover multi-grid AIC's, BUG-Lites and coloring patterns. Do not forget the diagonals. They will be needed.

Code: Select all

010608000090030000000000020005000000000006000408307000000000000000000000900804000
000308090000010030040000000000000700000400000000702801000000000000000000000203005
000070000000000000000000000061000720000000000028000350000000000000000000000090000
900504000000000000000000000501802000000007000007000000000000060050060000070309000
000106007000000000000000000000902703000300000000000400010000000000090020000605030

Max AIC competition

A special forum rank awaits the player who can find the longest AIC in the solving path. It can only contain strong & weak links, no AUR's, ALS and other fancy patterns. And it must cause one or more eliminations.

This competition ends Friday June 15. Good luck!

Ruud

Posted: **Sat Jun 09, 2007 5:20 pm**

I wonder how everyone is getting on with this beauty, it did everything that Ruud promised.

I have solved it by bifurcation, it was a mammoth effort, and there really is only one solution. Now to find a simpler route.

Posted: **Tue Jun 12, 2007 12:52 am**

[quote="Glyn"]
I have solved it by bifurcation, it was a mammoth effort, and there really is only one solution. [/quote]

Glyn - Congratulations! How long did it take you, and how many (approximately) 'logical' guesses did you make to solve this monster?

I honestly don't know where to start, as I have no idea regarding chains/coloring/and the bug technique....

Which reminds me, Ruud, will the solving guide be updated soon with some of the techniques needed to solve this monster? I'd love to learn some more advanced techniques, but can't figure them out on my own.

Posted: **Tue Jun 12, 2007 5:11 pm**

lynn893, an update of the solving guide is imminent. I am currently writing material for more advanced techniques, and I will also reorganize the guide into separate pages, so it does not take so long to load.

Glyn, you are the first player who solved this Ronin puzzle. Too bad you could not solve it without guessing. I was actually hoping that some of the advanced players would have a go at it and discover some moves that would only be possible in overlapping Sudokus.

Ruud

Posted: **Wed Jun 13, 2007 8:22 am**

Ruud
It was incredibly hard, I think it could be left as a long term project. I think that either some players may have missed it as it wasn't posted on the variants page, and others may have tried solving entirely without bifurcation in which case I think they will need more time. (probably a lot more)

Did you find it was solvable using only multigrid AICs?

Perhaps now there have been some posts some more players may be curious enough to try. And once the solution is posted then everyone can direct their attention at finding the clever moves to crack it. I'll be continuing to look at it anyway.

A great puzzle much harder than Catch-22 as nothing wants to yield at all. I wonder how you created it.

Lynn

It took days, I have a lot of time on my hands due to my caring responsibilities, but not always quality time for deep reflection. So after I realised it was going to be harder I just worked on one grid branching as much as necessary within the overlapping nonet to find solutions for that grid which were unique. I think 30 or so variants of the overlapping nonet were possible. Then I tried out those into the centre grid to weed out the ones that didn't work. Even then that wasn't enough, which confirms Ruuds observation that some of the chains may need to bridge 3 or even more grids if you have a chain which goes into a grid and back out again. I also would guess that the AICs will have to be grouped as there are an awful lot of candidates left in the centre grid.