Hello all,

I'd like to offer a method of solving the crazy 4x4x4 II (hereafter the crazy) using only 1 algorithm of only 8 moves. I haven't seen anything like this on the forums or elsewhere, so I hope it's useful or at least interesting. I'm indebted to forum member robertpauljr for this.

In some recent messages with robertpauljr about the crazy, he discovered that if you apply the same sequence of moves, but do it with different layers, different parts are moved. I think his thought processes to even try that in the first place are pretty cool. He didn't add too much to that; I went ahead and experimented with what he'd claimed, and discovered you definitely could use the algorithm to solve the whole cube.

Disclaimers:

- When I say "1 algorithm", I mean that algorithm and it's symmetrical twin.
- Also, while I have solved it using this method, I didn't encounter any parity problems; I don't know if that was just lucky, or if there will be parity problems which it won't be able to deal with.
- As with other methods, there are setup moves to be done from time to time.
- And finally, it's definitely *not* a speedsolving method.

OK. Basically, it relies only on the algorithm

**Code:**

U R U' L' U R' U' L

which is simply the Corner Piece Series from the Ultimate Solution method. In that method, if you turn the upper face clockwise to begin with, you get the steps above. If you turn the upper face anticlockwise to being with, you get its mirror. Using the steps above will cycle 3 corners and leave everything else unchanged. On a standard rubik's cube, it cycles the corners UBR -> LBU -> LUF and leaves corner UFR (and everything else) untouched. I think pretty much everybody knows it.

The method is

- Solve 8 center squares using only inner slices for the turns. (I know there are other ways, but this is about using the same sequence for the whole cube.)

So, when I turn U, it's actually u instead. What this does is cycle the 2x2x2 corners exactly the same as the corners on the rubik's cube.

- Solve 8 corners using only outer slices for the turns, just the same as the rubik's cube.

- Solve the 24 outer edges using 2 outer slices and 1 inner slice.

Using **Code:**

U R U' L' U R' U' L

and outer layers up and right, and inner layer left, will cycle the 3 outer edge pieces BUl -> UFl -> RFu [not sure of the notation here].

So, I think of this as "coming from back to front on the left side".

If I did the mirror **Code:**

U' L' U R U' L U R'

then it would be "coming from back to front on the right side".

Near the end of these, there will probably be some setup moves required.

- Solve the 16 (?) inner/circle edge pairs using 1 outer slice and 2 inner slices.

Using **Code:**

U R U' L' U R' U' L

and inner layers up and right, and outer layer left, will cycle the 3 circle edge pieces BLu/ULb -> ULf/FLu -> UFr/URf

Near the end of these, there will probably be some setup moves required.

Cube complete!

As I said, I've solved it this way once earlier today, so I know that it's definitely doable. What I particularly like about it is the fact that I only have to remember 1 easy algorithm. I am quite interested in solving things easily with as little algorithm memorisation as possible, and this fits that bill.

It is, as you'll guess, a fairly slow method. Having said that, it's still a method!

I'd encourage anyone who's interested in this to try the algorithm with the different slices on a solved crazy to see its simplicity.

I hope my first "contributing" post was worthwhile. I welcome any comments and also thoughts on parity.