In summer 2015, I already knew how to solve a Rubik’s cube using the beginner’s method. Not just that, I had spent quite some time teaching my classmates how to solve it. Having watched a video of someone solving it blindfolded, I decided to take it to the next level and learn the blindfolded solve.
I started my search online, and sure enough, there were several resources on blindfold solving. In this article, I’m going to discuss what I found on the internet, give you a brief idea on how exactly a blindfold solve is done, and how I devised a slightly different method myself.
So let’s start with what exactly we try to achieve in a blindfold solve. In a Rubik’s cube, there are “algorithms”, sequences of moves that achieve certain positions. For example, if you want to rotate a corner, there would be a sequence of moves to do this. The algorithm is written using alphabets that represent which layer to move.
Here is a simple algorithm: R’ D’ R D. What this algorithm tells me to do, is first rotate the right layer, then the down layer, then the right layer again, and finally the down layer. However, when I talk about the right layer, there are two possible directions of rotation. To distinguish between these directions, we use the R’ (read as right-inverted) and R. The former refers to a counter-clockwise rotation, whereas the latter refers to a clockwise one.
When we do a regular solve of a cube, we use algorithms that achieve a certain transition at a time, while disturbing the rest of the cube. Only towards the very end, we make sure to use algorithms that achieve transitions without disturbing the rest of the pieces.
In a blindfold solve, we focus on using algorithms that do not disturb the rest of the cube, from the beginning itself. This means that there is a certain piece on the cube, and we call this the shooting location, from where we move a piece to another location. You can understand this better through an example.
In a cube, we have two kinds of pieces- edge pieces and corner pieces. In the image below, A, B, C and D are the edge pieces, whereas 1, 2, 3 and 4 are the corner pieces.
In this scenario, let’s just say that I choose B as my shooting position. This means that whatever algorithm I have in memorized will move the piece which is right now at B, to another position (say D). Suppose the scrambled cube is such that the piece at B needs to be moved to position C, I would bring the piece C to the position D using some moves, then perform my algorithm, and move the piece back to C. While doing all of this, I make sure not to disturb the rest of the cube at all.
When I do the above, the piece at C now comes at B, meaning that the algorithm I know is a swapping algorithm. So, I just look at what piece is present at C, and where that needs to go in a solved cube. Let’s say it needs to go to a position O (not shown on the cube), and the piece at O needs to be at position M. Then, in my head, I have the alphabet sequence C-O-M. I have already named all my pieces such that when I say an alphabet, I know where that piece is.
Thus, I shoot from B to C, and the piece at C comes to B. Now I shoot from B to O, and the piece at O comes to B. Finally, I shoot from B to M, and the cube is solved! I follow a similar method for the corners.
In summary, what I have in my mind is just a sequence of plus and minus, which represents the rotation of corner pieces, followed by a sequence of numbers, which tells me the shooting positions for my corners, and at the end a sequence of alphabets which tells me the shooting positions for my edge pieces. Typically, the +/- sequence has 7 characters, there are around 6 or 7 numbers, and around 10 alphabets. By remembering all of these together with 5 algorithms (2 pairs of algorithms are very similar to each other, and 1 separate algorithm), I am able to solve a Rubik’s cube with my eyes closed!
The following section goes more into detail about how I devised my own algorithm by eliminating the parity condition and reducing the number of algorithms required.
When looking up online, I found an algorithm called the Y-perm to solve the corners. This algorithm takes care of the corner rotation as well as position. However, I was having some difficulty using this algorithm. Hence, I used a corner rotating algorithm, which was much simpler, and paired it with a the L-perm algorithm to solve corners. The L-perm simply swaps the corner positions 1 and 2, while at the same time swapping the edges B and A. To solve corners, I simply make sure that my corners are rotated correctly, and then use the L-perm to shoot from the position 1.
Parity is a special condition that comes up when I finish solving the corners. Since my corner swapping algorithm also swaps edges B and A, I get a slightly disturbed cube if I solve an odd number of corners. So out of 8 corners, let’s say 3 were already in their correct positions, and I just had to solve 5. In such a scenario, I end up getting the piece A at my shooting position instead of the piece B.
Traditional blindfold solving methods dictate that one must use an algorithm called the R-perm to eliminate parity. However, the R-perm was another algorithm that I found difficult to apply. Hence, I devised a simple solution. When memorizing the cube, I take note of whether I’m solving an odd number of corners or not. If I am, instead of looking at where my piece B needs to be in a solved cube, I simply look at where my piece C needs to be, and start my alphabet memorization from there. This trick eliminates the need for a parity algorithm!
Hope you were able to get a general idea of how a Rubik’s cube is solved blindfolded, and liked the method that I devised using improvisations to the traditional methods!
P.S. I successfully solved the Rubik’s cube blindfolded in front of an audience of around a hundred people in summer 2015 during an adventure camp!