Classifying Rubik's Cube Grips

blade740

Mack Daddy
Joined
May 29, 2006
Messages
851
WCA ID
2007NELS01
YouTube
Visit Channel
Basic Rubik's Cube grips can be defined by the location of each thumb on the cube. We will use an ordered pair of integers to describe this, where 0 translates to a thumb on the D face. This usually means the DL or DR face, but really your thumb can float a bit. 1 means a thumb on F, 2 means a thumb on U. A thumb on the B face can be represented by either -1 or 3, depending on which grip you're coming from. And so on, and so forth. The two most common grips are (1,0), which means the left thumb is on F, and the right thumb is on D, and (1,1), where both thumbs are on the F face. After that, grips that I commonly use are (1,2), which is really just a mirrored verson of (1,0), and (1,-1), which I think of as square-1 grip (along with (1,1)).

Generally your wrists have a range of motion to permit 5 total unique grips, from a -2 grip like (1,-1), to a +2 grip, like (1,3). You can extend further than this in theory, but it puts strain on the wrists (and I've never used an alg that required me to do so). Of course there are also unique grips that aren't as easy to classify with this system, but these are generally specific to individual algorithms. Performing R and L moves changes your grip, as do grip changes (where you pick up your thumb and move it to a different face).

Once you figure out where the thumb is, it's very easy to see which moves can be easily performed with each hand. With your right hand in a 1 position, you can perform U and D' pull moves. With your left hand in 1, it's U' and D. With your right hand at 0, it's F and B', at -1 (or 3) it's D and U', and I think you can figure it out from there. You can also perform "push" moves in the opposite direction (for example, U' and D with your right hand in the 1 position) You can always perform R and L moves with your wrists, as well as M and M', with either hand, in any grip. Wide turns follow these as well, though they generally require you to shift your thumb a bit. I suppose you could also add E and S, which go along with U and F. There are also moves that can be performed with the thumb (the inverse of the move that can be performed with the index finger in each grip). These wind up with the thumb on the L/R face, which will require a regrip to get back to a standard grip.

In a (1,0) grip, you can perform:
Pulls:
U'
U2'
F
F2
D
D2
B'
B2'

Pushes:
U
U2
F'
F2'
D'
D2'
B
B2

Other:
R
R'
R2
L
L'
L2
M
M'
M2
M2'

In a (1,1) grip:
Pulls/Pushes:
U
U'
U2
U2'
D
D'
D2
D2'

Others:
R
R'
R2
L
L'
L2
M
M'
M2
M2'

In a (1,2) grip:
Pulls:
U'
U2'
B
B2
D
D2
F'
F2'

Pushes:
U
U2
B'
B2'
D'
D2'
F
F2

Others:
R
R'
R2
L
L'
L2
M
M'
M2
M2'

In a (1,-1) or (1,3) grip:
Pulls:
U'
U2'
D
D2

Pushes:
U
U2
D'
D2'

Others:
R
R'
R2
L
L'
L2
M
M'
M2
M2'

Anyway, thoughts? I should really take some pictures, but I can't be bothered at the moment. I'd also really like to see a solving program that found algs requiring minimal grip changes.
 
Last edited:
About how many grip changes are made during a solve? How big are the grip differences between Roux and CFOP? Things like this are interesting. Looks like you put a lot of time into this.
 
Interesting idea. I can see this being useful for grouping algorithms on the wiki or finding algorithms that are predominantly a certain grip.
Of course there are also unique grips that aren't as easy to classify with this system, but these are generally specific to individual algorithms.
I don't know about that. I frequently use the left index finger push to do U when in the (1,0) and (1,2) grip, and I'm sure I'm not the only person.

EDIT: THAT'S what I was thinking of, the Gripper program. I remember hearing about it somewhere.
 
Last edited:
Anyway, thoughts? I should really take some pictures, but I can't be bothered at the moment. I'd also really like to see a solving program that found algs requiring minimal grip changes.

This sounds amazing. But I would need to see it, I can't picture it. I'm fairly new to speedcubing, and I don't know that many algorithms. There are so many agls available, and I really want to pick the ones that suit me best. This thread could help! I'll try reading it a couple of times again, but like I said, I could really use pictures (or a video?).
 
I use this notation:
{f:f} for your (1,1)
{f:d} for your (1,0)
{f:u} for your (1,2)
{f:b} for your (1,-1)
{f:b'} for your (1,3)
 
Back
Top