2x2x2 Blindfolded Tutorial
A first step to learning how to solve the original 3x3x3 Rubiks cube could be learning how to solve the 2x2x2. Or maybe you just want to see if you can do it with the smaller cube. Solving the 2x2x2 blindfolded can be done considerably easier than solving the 3x3x3. The method described here is probably the simplest method there is. Variations of it has probably been invented by many, the idea is not that unique but I don't think you can come up with anything simpler.
There are a number of different methods for remembering large amounts of random data and you could use any method that you can find on Internet pages on blindfolded solving of the Rubik's cube. You will need only the part for memorizing the corners on a 3x3x3 cube. But the amount of data to memorize for solving the 2x2x2 is typically far less than half of what it is for a 3x3x3 cube and instead of learning a memorization method I think you should first try to memorize the whole cube visually (no tricks, just memorizing by looking where each peace should go). If you can do that (I don't think you will find it that hard, I have memorized the 5x5x5 which is roughly ten times as much data visually a couple of times) you don't need to spend a lot of time on methods for memorizing and are soon about to start solving. Should you need to use a method that makes it easier to memorize I think that the one described by Stefan Pochmann in the M2/R2 Method tutorial is a suitable one. There are better methods, but they generally need more preparation work before you can start using them.
When you are learning to memorize you should always solve the cube so that the colors on the solved cube are on the same sides every time. On a cube with the most common color scheme you could for instance choose to have white on the top face and green on the side facing yourself (the orientation used when scrambling for competition for a cube with the most common color scheme). You can choose any other orientation that suites you but choose one and stick to it. A good tip is to have a solved cube on the table in that orientation to look at while you are learning.
Memorizing is now done like this:
Start by looking at the down, right, front piece. Where should it be on the solved cube? Specifically: where should the bottom face be? Where that face should be is the first thing you should memorize, no more and no less. Then you do the same thing for the face that is in the place you just memorized, that is: memorize where it should go. You continue to do this until you get to the piece that should be in the down, right, front position. If you can manage to move all the pieces the way you memorized them, all of the pieces you have moved will be correctly placed. If you are lucky all pieces on the whole cube will then be correctly placed. If there still would be some pieces that are not in the correct positions you have more than one cycle in the orientation you started. If you want you can break up that cycle by moving the correctly placed down, right, front piece to where one of the incorrectly placed pieces are and continue as before until you get back to that same position and you will get the bottom, right, front piece where it should be again. This is described in any method you can find for solving the 3x3x3 cube blindfolder that you can find on the Internet. But, to make things as simple as possible there is one thing you should note: You can choose to hold the cube in a number of different orientations when you start (24 to be precise). If you want to avoid multiple cycles you could try different start orientations. It turns out that most scrambles can be solved with only one cycle which greatly simplifies the process of solving a cube blindfolded. On the few scrambles that can't be solved with one cycle it seems that you can always find an orientation where the second cycle is just two pieces. Now, if you turn the whole cube so that one of those two pieces are at the bottom, right, front position you just have to memorize one more move and of course how you turned the cube to get it that way because that is how you will turn it during solving just before you solve the last two pieces. For some people it is as easy to see how to do this without actually turning the cube as it is by doing it, but if you find it easier to actually turn the cube you can do that. Just remember how you turned it so you can turn it back (and forth again during solving). If you are lucky, that's it. Quite often however, there will be one or more pieces that are correctly placed from the start but incorrectly turned or, because there are two cylces, one piece from each cycle can't be correctly turned. Turning a piece can be done by sending the down face of the down, right, front piece to where any of the faces of that piece is and then sending that piece back, now to the place where the first face (the one you now have at the bottom of the down, right, front position) should be. So, you memorize a chain of two moves where the first move is to any face of the incorrectly turned piece and the second is where the face that was in that position should go, exactly like it were two different pieces. When you have done this with all incorrectly turned pieces (if any) the whole cube will be solved. The last piece will always end up correctly turned because having only one piece incorrectly turned is impossible without disassembling the cube.
A simple example:
To understand the examples you need to know the standard notation for scrambling a Rubik's cube. If you think you have understood the ideas so far you don't need to spend time on this but it is pretty simple, actually. If you have a cube that does not have the standard color scheme the colors in the example will be wrong but you should be able to translate them. If you scramble a cube with the standard color scheme with the white side up and the green side facing you with the following scramble: D R D' you will have the blue, red and white piece in the down, right, front position with the blue face down. That face should be in the up, right, back position facing away from you. Memorize that position. In that position is the white, red and green piece with the white face where the first face should go. That face should be in the up, right, front position facing up. Memorize that position. In that position is the orange, green and yellow piece with the orange face up. That face should be in the down, left, front position facing left. Memorize that position. In the down, left, front position is the yellow, green and red piece which is the one that should be placed where you started. Moving DRF -> BUR -> DFL (where the letters denote the sides of the cube and the first letter indicates which face we are talking about) will solve this scramble.
The last thing is the execution. You only have to learn one algoritm that swaps two pieces. There are many different algorithms that do that but a very simple one that you can learn in 30 seconds is one that my son discovered. In standard notation it is: L' U' L' U' L' U' L' U' L' U' L' U' L' U' L' x2 z. If you don't know this notation it means: turn the left face anti clockwise a quarter turn and then the top face anti clockwise a quarter turn and repeat those two moves totally seven times, then turn the left face anti clockwise a quarter turn, then roll the whole cube half a turn forward and then turn the whole cube a quarter turn clockwise from the side facing you. This will make the bottom side of the bottom, right, front piece and the right side of the bottom, right, back piece to be swapped. What you now have to do for each of your memorized positions is finding a sequence of moves (called setup moves) that places that face at the right side of the down, right, back position, execute the algorithm above and then reverse the sequence of moves you did just before. This will, one by one, place each of the pieces correctly on the cube.
Another approach that is worth mentioning that uses an algorithm that takes a little longer to learn is to use the Old Pochmann method for corners. This is a method for the ordinary Rubik's cube but the corner part of it, which is also only one algorithm, could be used for the 2x2x2 cube. It differs in the way that it is the upper face of the upper, right, back piece and the down face of the down, left, front piece that are swapped instead of the bottom side of the bottom, right, front piece and the right side of the bottom, right, back piece as with the other algorithm. There are two advantages of learning this algorithm instead of the first one:
Learning to do it
When learning you don't need to start by memorizing and solving blindfolded. You should start with a solved cube and just executing the algorithm to see what happens. Then, for every face of the pieces, you should try to find sequences of moves (the setup moves) that place them on the right side of the down, right, back position. There never need to be more than three moves. You don't have to memorize them all if you can just hold them in your memory during execution but it will be much easier, go much faster and go wrong less often if you don't have to think them up everytime during solving. For the setup move sequences that are one or two moves there really is only one choice but the ones that require three moves can be done in different ways. For instance: L U2 B' and U' L' B2 are equivalent. Choose some that suite you and stick to them. Doing everything the same way all the time is the key to making few errors. You could now try moving pieces using your setup moves and see that they work. When you start to feel confident you could scramble the cube and solve it piece by piece while looking at it. If something goes wrong chances are that you forgot the x2 z part at the end of the algorithm, it seems that it is easy to forget that in the beginning.
If you can solve the cube this way while looking and you can memorize it you can solve it blindfolded. Take your time when memorizing until you get your first successful solve. If you spend much effort on recalling during execution, the risk of screwing up somewhere increases. A video camera can maybe be of help when trying to find out where you go wrong (if you do).
A complete example:
Scramble the cube with: F2 U' F2 U R' F2 R U F2 U2. This scramble is actually solvable with one cycle but keep it in the same orientation after the scramble because that will require a two cycle solve where two corners will end up incorrectly turned which is the worst case you will ever get.
At the down, right, front position you have the red, white and green piece with the red face down. That should go to the right side of the up, right, front position. In that position we have the orange, green and yellow piece with the yellow face to the right. That face should go to the bottom side of the down, left, front position. In that position is the red, green and yellow piece which is the one that should be in the down, right, front position where we started. So the sequence for moving the pieces in this cycle is DRF -> RUF -> DLF. You may note that the last piece will arrive incorrectly turned. But then the two upper pieces in the back layer are swapped. Turning the whole cube with x2 y' will get the red, white and blue piece at the up, left, back position to the down, right, front position with the blue face down. That face should go to the back of the up, right, back position which will now, after we turned the whole cube, be on the right side of the down, right, back position. At that position is the orange face of the orange, white and blue piece and that means that the orange face of that piece will end up at the up side of the up, left, back position. The orange side should be on the left side, so it will be incorrectly turned. So, after swapping the last two pieces you return the cube to its original position by reversing x2 y', which is y x2. The upper face of the up, left, back piece should go to the left side so now you use the swap algorithm to shoot the down face of the down, right, front piece to the upper side of the up, left, back position. After that you shoot to the left side of the same position and the cube will be solved.
So you have: DRF -> RUF -> DLF, do x2 y', DRF -> RDB, do y x2 and then DRF -> ULB -> LUB. Shooting DRF to RUF is done by placing the face on RUF at RDB, executing the algorithm (L' U' L' U' L' U' L' U' L' U' L' U' L' U' L x2 z) which we from now will call A, and moving RDB (which is now another piece) back to RUF. RUF -> RDB can be done with U2 B2. The whole sequence for doing this will be: U2 B2 A B2 U2. The next target is DLF which can be moved to RDB with L2 U B'. You get L2 U B' A B U' L2. Now, for swapping the two last pieces: x2 y' A y x2. No setup moves are needed because the second piecs is already in the right position. Finally turning the last piece (DRF -> ULB -> LUB): U B' A B U' and B2 A B2 and we are finished.
The complete solve: U2 B2 A B2 U2 L2 U B' A B U' L2 x2 y' A y x2 U B' A B U' B2 A B2.
What to do next
This is far from the fastest way to solve the 2x2x2 blindfolded. As this is not an official event at competitions there is no real point in perfecting a method for solving the 2x2x2 fast. But the 2x2x2 could be used when learning the corners of a method for the 3x3x3 though, if you want to go on and try that. The simples method for solving the 3x3x3 is probably the method called Old Pochmann. It consist of only four algorithms and simple setup moves much like the ones for the 2x2x2. It is more complex than the 2x2x2 as you will have to consider multiple cycles (there can be quite a few). And it is of course more to memorize and execution will be many more moves than for the 2x2x2. It is not a fast method either and if your goal is to compete you should learn a faster method from the start. The M2/R2 Method (by the same Pochmann) is not that hard to learn and was once considered a good method and many still use it for solving the edge pieces but nowdays there are better methods for the corners. The best one is probably the BH Method (by Beyer and Hardwick) which is three cycle optimal (look it up). Complete BH is 378 algorithms (many are symmetrical variations of each other) but subsets of it can be used to solve all the corners (in a non optimal way) if you should want to start using it before you have memorized all algorithms. Getting really fast requires learning lots of algorithms, putting a lot of time in setting up a personal memorizing system and practicing a lot.