well that’s the question… how does rubik’s cube work??? is there a special system or just by random you can set all the colours right?
Well, no, it would take quite a long time to solve it turning faces at random… there are about 43 billion billion different positions. However, by learning some mildly complicated techniques it isn’t too hard to solve one systematically – the pattern I remember following was to first get the four edge pieces along one face right, then the corners of that face, etcetera. (It gets tougher as you go… generally the last four pieces are the toughest, since you need to swap them and orient them without permanently disrupting everything else you’ve done.)
It isn’t a question of one simple system - there are probably lots of solution techniques that work, each of which has variations.
And here I thought this question was going to be about the inner workings of the cube puzzle, how it held everything together and let you turn each face the way you desired, so long as the others were all straight. Which is a kinduv cool subject.
chrisk is right. When I was good I could do one from fully scrambled to perfect in IIRC 2 minutes 30 seconds. Nothing about what I did was random. There were “speed-cubing” competitions and my time was far, far slower than those of the true experts.
A little Googling will probably turn up a web site showing a technique. When the cubes were hot there were any number of slim paperbacks selling the “secret”. You might find one at a library.
Google helps. 
When the cube was hot, I got one of these solutions (on paper then, of course), refined the procedures, applied some lubrication to my cube and could consistently solve it in around 30 seconds. Sometimes I got to just over 20 seconds. I have just retried and it took me 1 minute and 15 seconds. I haven’t touched the cube for years, and now I’m going to waste the whole weekend. 
Yeah, I’m a nerd, why do you ask?
Back when the rubic’s cube first came out, I managed to solve it without anyone showing me how to do it. It’s fairly easy to get one face all the same color. Then you can figure out a fairly simple set of moves that will put all the adjacent edges into place, one by one. Solving the last face gets really difficult. I fumbled around with that for a long time, and ended up solving it mostly by luck.
Once I had it solved, someone showed me a set of instructions on two different solving methods. I had already figured out the moves to put the side edges in place, so all I had to do was memorize the moves to get the last face pieces into place. Once you have a handful of moves memorized, you can solve the cube in under a minute very easily.
Here’s a site with one method:
http://www.geocities.com/jasmine_ellen/RubiksCubeSolution.html
This site has more info:
http://www.speedcubing.com/chris/
I just bought a cube a few weeks ago. It came with a little book showing one solution (probably not the fastest solution, but the easiest one). The solution I was taught is different from that book.
The book had some interesting facts. It says that with a computer’s help, most starting positions can be solved in 17 moves, and in theory, no cube needs more than 20 moves. It also said that most human experts need about 45 moves.
I’m just proud that I can solve it now, even though it can take me 5 minutes.
Like chrisk said, it is really interesting to see how it works mechanically. Take one apart some time.
That actually happens to be the way I solved it the first time. 
My first “solving” was taking the stickers off.
The second “solving” was taking it apart and putting it back together.
Finally, a friend of mine out-geeked me by knowing how to solve it, so I had to learn in order to recover some pride.
When I was a kid, I figured out how to solve the cube on my own. I started by restoring the corners to the correct position and orientation and then working on the edge pieces. basically what I did was to work out a series of transforms that would, for example, swap two corners without disturbing anything else. (I’m sure that my process was really inefficient, but it was one I’d figured out on my own.)
Yeah, to expand on this. . .
We learned how to solve the cube in Abstract Algebra class in college.
If you repeat the same move over and over, what happens is that different sections/pieces go through “orbits”. If you repeat sequences of moves over and over, you will return to your original state. Most basic is if you just turn one face 4 times. . .you get back where you started. But if you do something like turn one face, rotate the cube, turn a face, rotate back, turn the original face, etc. . .what you’ll find is something like :
Every 6 turns, the corners are back to their original position, but the middle edge pieces have switched. (this is just an example)
OR
One corner piece has rotated but all the rest of the face is back to its original position.
Simple rules like that will show you how to rotate corners, switch two corners, switch two middle-edge pieces, etc. With just a few of these rules, you can solve the cube with nothing random about it.
I don’t care much about solving the puzzle, but I would like to know the mechanical answer of how it works.
The mechanics are actually pretty simple although the application is near genius. Inside a cube there’s a device that looks like a six-armed cross. At the end of each arm is a swivel. The center piece of each of the six sides is permanently attached to one of these swivels. On the inside of all of the other pieces are grooves and ridges. These grooves and ridges line up to form circular paths when the cube is assembled. So when you twist a side, the center piece pivots on its swivel and the surrounding pieces follow it around on the circular path.
I recently purchased a 5x5x5 cube. Anyone mastered that one without help?
The corners are the same on all cubes. After that no I haven’t solved mine yet. I only just got a 4x4 cube a month or so ago, but the baby hasn’t allowed me to play with it much.
Watch out on that 5x5, if it’s the same as a lot of the other ones the orange stickers tend to come off for some reason.
“Mastered”? No. “Could (relatively simply) master (given a week or so break from teaching and research obligations)”? I don’t doubt it at all. Really, once you’ve seen the abstract algebra sitting behind such puzzles it’s not really that big a mystery.