# Mastering the Rubik's Cube

1. Dec 11, 2012

### MarneMath

My friend has been working on mastering the Rubik's cube lately. He can solve it now and make new designs. One day he asked me for a pattern so, I told him I wanted the red center piece enclosed by the green cubes, and the green center piece enclosed by the red cubes. I think this is impossible, but he beleive it is. Does anyone know if this can be done?

The Cube looks like this:

Placing the cube flat on a cube, the color of the face facing me is red. Moving right the colors are blue, orange and green and then back to red. The face at the bottom is yellow and the face on top is white.

2. Dec 11, 2012

### justsomeguy

Re: The Rubik's Cube

Not possible, as red and green are adjacent. It can be done with red and orange, blue and green, yellow and white, etc.

3. Dec 11, 2012

### wuliheron

Re: The Rubik's Cube

The center squares in a standard 3x3x3 cube don't move relative to each other. You can move all the other pieces around them, but the centers themselves form a jack or spindle shape that never changes.

4. Dec 11, 2012

### Containment

Re: The Rubik's Cube

You can do it but it involves taking the cube apart and putting it back together wrong :)

5. Dec 11, 2012

### Staff: Mentor

Re: The Rubik's Cube

If the cube has color stickers, just move them around and pretend like you did it.

6. Dec 11, 2012

### Staff: Mentor

Re: The Rubik's Cube

Evo likes how Jim thinks.

7. Dec 11, 2012

### Jimmy Snyder

Re: The Rubik's Cube

Start with the cube in the solved state. Then film yourself wildly twisting and turning it into some horrible mess that looks unsolvable. Then show the film backwards.

8. Dec 11, 2012

### Staff: Mentor

Re: The Rubik's Cube

Lol!!

9. Dec 11, 2012

### wuliheron

Re: The Rubik's Cube

One of my favorite patterns to make on the cube is hybrid between 90 degree rotated crosses and ordinary checkerboards. Combined when people try to solve for the ordinary checkerboard pattern which is easy it instead flip-flops to another checkerboard pattern as if they had simply made a mistake. You can keep some people busy forever that one.

10. Jul 6, 2013

### Staff: Mentor

Borrow a lego Rubik cube machine:

Last edited by a moderator: Sep 25, 2014
11. Jul 6, 2013

### wuliheron

That's a odd video. The cube is usually about 13 to 17 at most moves from being solved and to make that video they must have programmed a more complicated solution. World record holders sometimes just happen to recognize the pattern the cube is scrambled in and solve it in maybe 17 moves. The cube is the shape everybody loves though and nothing else feels as good in your hands or brain.

12. Jul 6, 2013

### Staff: Mentor

Yes, it solves in way too much moves, which most likely means it uses some relatively simple and naive algorithm. I remember one such algorithm was to solve upper layer first, then repeat some sequence of moves that rotates corners of the middle layer in some particular order, till they happen to be in the right places, then repeat some other combination of moves that rotates some of the pieces of the bottom layer in some particular order till they fit, then use exactly the same approach for the remaining pieces - and you are ready. Key to success is a selection of particular sequence of moves for each stage - it needs to shift pieces you want to move without moving all other pieces.

Kind of a brute force approach that you can teach to a monkey, and it typically requires much more moves then the optimal solution, but it works.

13. Jul 6, 2013

### wuliheron

I prefer to think of the cube as a closed system myself and, it is my toy after all. Once you understand the power of context over content the cartoons become obvious and it explains why children under five can have a better chance of success solving the cube.

14. Jul 6, 2013

### BruceW

lulwut?

15. Jul 6, 2013

### rcgldr

Rubiks Cube, 3x3x3, and Professor's Cube / Rubik's x5, 5x5x5, have centers mounted to a 3 axis "star" and can only be rotated. One variation of Rubik's Cube had 4 of the centers marked with arrows to provide a bit of an extra task to solve the cube. Rubik's Revenge, which was 4x4x4 had movable centers, and adjacent sets of 4 center pieces could be swapped.

The way I initially solved Rubiks Cube was top layer, middle layer, and bottom layer. The top was straight forward, the middle layer used a pattern of moves that could be repeated to solve the bottom layer. The bottom layer had to be solved by getting corners into correct position, then rotated so colors match centers. Another sequence could be used to rotate any 3 bottom edge pieces, and a final set of moves to flip them so colors matched centers.

The standard method is to do all 8 corners, then top, bottom, middle edge pieces. First all corners are moved into place then rotated so colors match. Then top and bottom edge pieces are moved and rotated into place. The middle edge pieces were moved into place, then a pattern called "Rubiks maneuver" was used to flip two middle pieces to get colors to match centers if needed.

For Rubik's Revenge, 4x4x4, the top and bottom sets of 4 center pieces could be fixed first. Then the middle sets of 4 center pieces could be fixed, but solving the corners first elminated having to remember the color orientation. Then the cube could be solved similar to a 3x3x3 cube, except two new patterns not possible with 3x3x3 cubes could occur since edge pieces are in sets of two. It is possible to have two pairs of edge pieces swapped (on different sides) and 1 pair of edge pieces rotated (colors on wrong side). The 4 center pieces have skinny "legs" to connect them into slots in an inner sphere and are easily broken, so you have to be careful with the 4x4x4 cubes.

The 5x5x5 cubes don't introduce any new moves. The only "new" pieces are the center "edge" pieces, but these solved in the same manner as the center "corner" pieces (like the 4x4x4). The Professor's Cube uses the one time standard of white and blue on opposite sides. The Rubik's x5 has white and blue adjacent.

Somewhere in my archives, I have a windows 3.1 version (it will run on any version of Windows) of a virtual Rubik's cube that lets you set the number of pieces.

Last edited: Jul 6, 2013
16. Jul 6, 2013

### Staff: Mentor

That's what I wrote about. It is guaranteed to work, but it is far from optimal.

17. Jul 6, 2013

### rcgldr

It was slower, but it required fewer patterns and you didn't have to know about "Rubik's maneuver" to rotate (to get colors to match) a pair of edge pieces. I figured out how to disassemble and reassemble the cube, then wrote down what patterns of moves did, then use that to solve the puzzle. I'm also slower in that I keep the top on the top, rather than flip the puzzle over once the top is done, to solve what was the bottom.

Last edited: Jul 6, 2013
18. Jul 8, 2013

### 2sin54

The cube is 20 moves at most from being solved. The world record for the fewest moves made to solve a scrambled cube was 20 moves.

19. Jul 8, 2013

### wuliheron

According to world record holder Tony Snyder, the cubes used in championships are scrambled using 13 moves and simply reversing them would solve the cube. His record in a competition is 19, but 13 to 17 is what a computer can do and his goal is to replicate the feat.

20. Jul 8, 2013

### Staff: Mentor

Wikipedia entry on optimal solutions for Rubik's Cube lists a "computer-assisted proof" that at most 20 moves are required.

21. Jul 8, 2013

### wuliheron

I believe that is correct, but 17 is more like the average and 13 the minimum.

22. Jul 8, 2013

### BruceW

well, 13 is only minimum because in the championships, they only scramble them with 13 moves, right?

23. Jul 8, 2013

### wuliheron

They use a formula that produces the most scrambled cube possible, so 13 is the minimum for a completely scrambled cube.

24. Jul 8, 2013

### collinsmark

25. Jul 8, 2013

### BruceW

nice links. 35 years of computer time. phew! So when the cube is in an initial configuration which takes the longest to solve, it takes exactly 20 moves to solve it? (if I understood right)

edit: and to put it in other words, even in the worst-case scenario, it is possible to solve in 20 moves?