An introduction to blindfold solving
This guide is incomplete. Actually, that's an understatement; I've barely even started. In the meantime check out Shotaro Makisumi's guide at cubefreak.net.
Leyan Lo taught me how to solve the Rubik's cube blindfolded during a six hour road trip in January 2005. Three months later I became the first woman to successfully solve a cube blindfolded in competition at Caltech Spring 2005. At the time of this writing, my best time in competition is 1:23.96, placing me at 21st in the world.
Blindfold solving involves inspecting and memorizing the starting state of the cube, then executing the solve without looking at the cube. It is not about using tactile feedback or other cues to determine differences between pieces. In competition, time starts as soon as the competitor begins memorization and stops when execution is finished.
One could theoretically memorize the position and orientations of every piece and then track all of them as he executes his normal speedsolve solution*. However, this is much more difficult than a method which solves the cube using algorithms that only affect a few pieces at a time. For example, the PLL algorithms.
(*This is in fact the basis of "speed-blindfold" solving, an unofficial event. Unlike the official blindfold solving event, the competitor is given unlimited time to memorize the cube and only the solving phase is timed.)
Try this. Take a solved cube. As you know from studying the PLLs, the algorithm to cycle the UF, UL and UR edges clockwise is R2 U R U R' U' R' U' R' U R'. Now do: L' (R2 U R U R' U' R' U' R' U R') L
You'll notice that what you just did was cycle the UR, UF and LF edges. The L' move at the beginning was a setup move which placed the LF edge in the UL position. After executing the PLL, the L at the end cancels the setup move. In a nutshell, this is how blindfold solving works. (Setup moves) (Algorithm to do something) (Inverse of setup moves)
This tutorial is an introduction to the 3-cycle orient-permute (3OP) method of blindfold solving, in which all pieces are oriented and then permuted in separate steps. This was the favored method of Shotaro Makisumi, Leyan Lo, Tyson Mao, myself, and other top blindfold solvers for a few years, during the dawn of competitive blindsolving. Recently, Stefan Pochmann's new M2 method for edges has been shown to be a very competitive and possibly superior alternative to 3OP. M2 is a considerably more advanced method. Beginners to blindfold solving are encouraged to start out with Pochmann's original method, now known as "Classic Pochmann."
This tutorial assumes you know how to solve a Rubik's cube and are familiar with the concepts and notation. You should also know the PLL algorithms well enough to be able to do them without looking at the cube. At a minimum, you should know the U-perms, the A-perms, and at least one permutation that swaps two corners and two edges, e.g. the T-perm. However, the more you know, the easier it will be for you (Exceptions: the four G-perms, which are never used in blindfold solving unless you manage to be unnecessarily clever).
I would also advise you learn how to do the U-perms from the front and the back. It makes execution more efficient.
|R U' R U R U R U' R' U' R2|
|R2 U R U R' U' R' U' R' U R'|
|R' U R' U' R' U' R' U R U R2|
|R2 U' R' U' R U R U R U' R|
Unlike sighted solving, blindfold solving is best accomplished by holding the cube in the same orientation every time. I hold my cube with yellow on top and orange on front. Of course, this is arbitrary. Pick whatever orientation you like.
Step 1: Orientation of Edges
We define an edge to be oriented correctly if it can be put in its solved position using moves in the set (U D F2 B2 R L), i.e. quarter turns are allowed on all faces except the F and B faces. To recognize whether an edge is oriented correctly or incorrectly, there are just a few simple rules:
If the edge is in the U/D layer:
- If the edge has a U/D color sticker, it is correctly oriented if the U/D sticker is on the U/D face; otherwise incorrect.
- If the edge has no U/D color sticker, it must have a R/L color sticker. The edge is correctly oriented if the R/L color sticker is on the F/B/R/L face. It is incorrectly oriented if the R/L sticker is on the U/D face.
If the edge is in the middle layer:
- If the edge has a U/D color sticker, it is correctly oriented if the U/D sticker is on the F/B face; otherwise incorrect.
- If the edge has no U/D color sticker, its R/L color sticker must be on the R/L face; otherwise incorrect.
Once you have recognized the orientation of your edges, you can solve them using these algorithms:
M U M U M U2 M' U M' U M' U2 flips UF and UB
(M U)*4 (M' U)*4 flips all four edges on the U layer
If the edges you want to flip are not in the particular positions needed, simply use setup moves. For example, to flip UF and BL, use B' as your setup move. Remember to keep track of any setup moves/cube rotations you do so that you can reverse them afterward.
Step 2: Orientation of Corners
Each corner, having three stickers, can be oriented one of three ways. Corner orientation is easier to recognize. Every corner has a U/D color sticker on it. If this sticker is on the U/D face, the corner is correctly oriented.
Just as you can't flip only one edge, you can't twist only one corner.
THIS PAGE IS VERY INCOMPLETE