Is the Rubik's cube solveable in any configuration?

To clarify, you mean that each step starts with something like “Look at the cube, and see which position this cubelet is in. If it’s in position A, then do this. If it’s in position B, do something else. If it’s in position C, then do this first to put it in position A”, right?

And I think that what septimus is referring to is not teaching a technique, but rather a site where you somehow enter your cube’s current state, and then it says “Turn the top face clockwise 90 degrees, and then turn the front face 180 degrees, and then…”. Which do already exist, but they’re not something you can put on YouTube.

Set theorists might tend to disagree. Perfectly good numbers, just rather big. :smiley:

Well, you can certainly go to YouTube for tutorials on how to easily solve the cube from any scrambled position. The specific scramble doesn’t really matter for any practical purpose. So I don’t see why you would want or need a site where you enter the specific state of your cube. That just sounds like overcomplicating things massively for no reason.

Too late for edit: Unless you mean that **septimus **is referring to a cube where someone has moved the stickers around. In which case, as demonstrated, it won’t be readily solvable.

Simple way to find out if someone has sabotaged your cube: Pick a proven algorithm, and use it to solve the cube. It the cube is still messed up at the end, someone has shuffled your stickers (or, actually, you’ll be able to tell part way through the solve). In which case you need to need reassemble the cube or put the stickers back where they belong.

I know what you mean. I had one at Army AIT. We couldn’t have books, so I had one that I messed up and solved, over and over, just to have something to do on breaks, which we had a lot of. (We could have newspapers, but how long does it take to read the Baltimore Sun? I worked on crosswords a little; mostly I did the cube, over and over.) It got to where my eyes and brain were always ahead of my hands on the solving, and I knew what the sides I wasn’t looking at looked like from the configuration of the sides I could see. I probably could have learned to solve it blindfolded. You know, where you study it in the unsolved state, memorize it, and then are blindfolded, and solve from memory.

This. ETA Ref Martian Bigfoot’s post #20.

Once you’ve learned an algorithm, restoring a scrambled cube is intellectually like emptying the dishwasher. You know where everything goes, just pick and place one by one until it’s done. There’s no decision-making involved beyond the level of recognizing “*this *is a fork and *that *is a cup.”

There were many, many how-to books & pamphlets published back in the day. Each taught some variation of the known solving algorithms.

FWIW, I found the “layer by layer” method the easiest to learn, but it’s not the fastest. Just pick a side (say red) to start. Then use a few very simple moves to get all the red corner and edge pieces into place, without worrying about the other sides. Then there are two mirror-image moves you can use to place the four edge pieces of the middle layer without disturbing the top layer.

The final layer is the hardest to learn, because there are a lot of moves to memorize and they’re a little more complicated. The sequence I use goes: move the corner pieces, orient the corner pieces where needed, move the edge pieces, orient the edge pieces where needed. Then you’re done and your friends are impressed.

All my math teachers beat it into our heads that infinity isn’t a number!

Here’s an example of a finite ornamental usage. (This is in Encino, a heavily Jewish neighborhood of Los Angeles.)

(ETA: Okay, looking more closely, this is in Chicago, not Encino. But there’s one just like it in Encino.)

Some people lack the aptitude or inclination to learn and carry out an algorithm themself, but can still follow exact directions one step at a time. A website like septimus describes might be of value to such a person.

It was silly of me not to Google and see that such sites existed. I’ve never learned to solve the Cube, but sometimes restore our own cube with a solver on my laptop and I wondered if it would be an interesting exercise to construct a solving webpage. I clicked one of Reply’s links, clicked Scramble and Solve and was shown a 71-step solution. (I guess it’s much easier to find a long solution than a short solution. :wink: )

Checking for misplaced stickers may be easy, I’ll guess. Of 40 possible 3-ordered-color corner cubes, only 8 exist in a valid cube. Only 12 of the 15 possible 2-color edge cubes exist. After these constraints, what portion of cubes is solvable? Are there straightforward parity tests? I tried half-heartedly to Google for an answer but clicked only this link. :smiley:

Try the seventh post in this very thread: If you randomly assemble a cube from valid cubelets, there is a 1 in 12 chance that the result will be solvable.

One place I worked at had a Rubik’s cube out where people could play with it. On the last day I worked there, I surreptitiously did exactly this, then scrambled the cube so it wouldn’t be obvious.

You’ll also want to check that the corner cubes have their stickers in the right place, there’s two different orders for each corner. Just turn each of the eight into place individually and see if they match the centers. After that, you can always solve the cube to where there is at most 1 edge and/or 1 corner piece out of place. There’s 2 orientations for an edge and 3 for a corner. That’s a 1 in 6 chance of a cube being solvable if you have verified that the stickers are on correctly. There’s probably a relatively simple parity test for the edges. The corners are likely more complicated. I’m guessing that any parity test is going to take longer than the minute or two it takes to try and solve the cube.:slight_smile:

No difference. For easy algorithms, at least, following exact directions one step at a time is already all there is to it. When I said that there’s no thinking involved, I did mean no thinking. If you don’t want to learn the algorithm, just write it down. (I didn’t get into advanced stuff, I suppose there has to be more to it for solving a cube in five seconds).

Any algorithm that works will involve a fair number of branchings: If it’s like this, do this, but if it’s like that, do that. This may not seem any more difficult than a straight step-by-step to you, but it is to some people.

ETA: Ninja’d in reply by Chronos

I agree with **Chronos **here. In fact I almost made the same post way back earlier in the thread.

There are plenty of people that want “turn the left face clockwise, turn the right face counterclockwise” … for 200 utterly mindless steps versus

“Decide if the desired edge cube is on the bottom or not. if not, put it on the bottom.
Now twist the bottom until the desired edge cube is on the face next counterclockwise from the desired face. Then twist that adjacent face rotate it into position.” … etc for 75 steps, each more convoluted than the previous.

There’s a big difference between steps a motivated monkey could follow and steps a spatially reasoning intelligent motivated human could follow. Lots of humans prefer to only use the monkey part of their intelligence.

As an example, I can recall several nasty threads about Sudoku wherein folks argued that any puzzle which required exploring any branch was too hard and “required guesswork” to complete. They only wanted puzzles where at every stage there was at least one open cell which was fully determined by already solved cells.

For anyone who’s a real puzzle hound, those toy Sudokus aren’t even worthy of the name “puzzle”. They’re no more intellectually challenging than my earlier example of unloading a dishwasher.

Diff’rent Strokes and all that.

For what it’s worth, no 3x3x3 cube ever requires more than 20 moves to solve. Sometimes a 20-move solution might be difficult to find, but a practical computer-run algorithm should have no difficulty in finding solutions longer than 30 moves or so.

Of course, your mileage may vary for the algorithms used by actual humans. With my (known to be inefficient) method, for instance, I often (about half the time) reach a state near the end where two corner cubelets are both rotated, and everything else is solved. When I reach that stage, I consider it “almost solved”, but it still takes me 18 moves to get from there to the solution (it can be done in 16, but I don’t have that sequence memorized).

I can do those last corner rotations with my eyes closed. :slight_smile:

As can I, or behind my back. Which never fails to amaze my nieces and nephew, who have inflated it to “Uncle Chronos can solve the Rubik’s Cube behind his back!”.

I used to have an app on my iPad that let me enter each side of the cube as it was, and once all 6 sides were input, it gave me the steps (algoritm) to solve it.

Rubik’s Cube math is all over the web. 20 is referred to as "God’s number. What that means is, no cube needs more than 20 moves no matter what state it is in.

There are amazing kids/teens out there that can solve the cube in less than 10 seconds. The world record (last time ai checked) was 5.55 seconds. And yes, it is on YouTube.

There is a kid named Felix Zemdigs (I think) who owns or owned many records. For those of you that have been away from cubing since the 1980’s, you would be amazed at the puzzles that are out there. The biggest one I have (and solved) is an 8x8x8 cube, but they mass produce 11x11x11 and I am pretty sure they have a 13x13x13 now.

The world record is a 17x17x17 monstrosity built by Oskar Van Deventer (I am sure I spelled that name wrong, but google Oskar and “twisty puzzles” and you will find him.)

With the invention of the 3-D printers, twisty puzzles have exploded in complexity.

I got a cube when they first came out, in 1980. I still have it. But I had not seen it or picked it up in 30 years, until one day I saw one in a toy store.

I set about learning to solve the cube on my own, which I did. It was a great feeling, and I can still do it.

As to methods, I won’t bore you all with names, but the speed cubers all solve the first 2 layers together (called F2L) and this is called the Friedrich method. It is not that hard to learn once you wrap your head around what you are trying to accomplish, but it takes a ton of practice to master. I have not mastered it. In fact, I stink at it, but I understand the concept. It just takes an ability to recognize where the cubies are, and where they need to be (duh!), and that is something that takes more time than I have to devote to it. I also think cubie recognition is something that some people can just innately do better than others.

The teens that can solve the cube in 10 seconds or less use a combination of F2L and last layer permutation moves. It is truly amazing to watch a speed solver, especially when you can solve the cube yourself, and can’t get any faster than 1 minute.

My method is my own, and it is not optimal. However, it works. I used group theory to solve it the second time (as an adult), which is a branch of mathematics that works very nicely with the Rubik’s cube.

One final note. Rubik’s cubes are no longer under patent protection, and there are a number of companies that make excellent “speed cubes” designed specifically for speed solving. Rubik’s cubes are the originals, but they mostly suck for speed cubing. The ones people are solving in seconds are all different brands.