• Welcome to the Speedsolving.com, home of the web's largest puzzle community!
    You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

    Registration is fast, simple and absolutely free so please, join our community of 40,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!

BH Resource Center

byu

Member
Joined
Dec 18, 2008
Messages
2,021
Location
California
WCA
2009YUBR01
YouTube
Visit Channel
I think this might be useful for the several people who are interested in starting the BH method (including myself)

I compiled up as much information as I could find at the moment and put it all together in one place. These pages are very good references that I used:

Speedcubing BH Corners - http://speedcubing.com/chris/bhcorners.html
BH Method Thread - http://www.speedsolving.com/forum/showthread.php?t=11909
BH Method Thread - http://www.speedsolving.com/forum/showthread.php?t=10756
BH Corners Text - http://dbeyer.110mb.com/BHcorners.txt
Viewpoint Shifting - http://www.speedsolving.com/forum/showpost.php?p=117456&postcount=45
Speedcubing BH Edges - http://speedcubing.com/chris/bhedges.html

For a person who is just beginning the BH method, you'll want to figure out how each of the following types of cases actually work.

I've explained them in my own words here, dbeyer has also explained them.

There are 378 algorithms in the BH method for corners, but you shouldn't learn them all. They should be intuitive. Here's the types of cases you can encounter (for BH corners):

GENERAL BH INFORMATION
TERMINOLOGY
Cubie -- A piece, as a whole, with no specific permutation or orientation there of. Just to make a reference to a location on the cube. Standard U/D first.
URB or DRF for example.

Interchangability -- There are different ranges and degrees of interchangability. Looking at the cubie locations as a whole. A piece is interchangable with every other piece in one turn (htm), except it's polar opposite.
Interchangability on the Layer: This is the broadest form of interchangability. Looking at the cubies as a whole.
The six cubies interchangeable with the URB;
ULF, DRF, DLB, UFR, UBL, DBR
NOT the DFL

Interchangable on the Plane-orbit:
The U plane is the URB, UFR, ULF, UBL.
The F plane is the FRU, FUL, FLD, FDR

Interchangable on the Slice-orbit:
The URB's R-slice-orbit;
URB, BRD, DRF, FRU.
The FUL's U slice-orbit;
FUL, LUB, BUR, RUF

Adjacent Cubies -- To be next interchanged by 1 quarter turn
URB and FLU for example.
FDR and RDB
Opposite Cubies -- To be interchanged by 1 half turn.
URB and DLB for example.
FRU and FLD
Polar Opposites -- The two cubies with are not interchangable in anyway shape or form with just 1 turn (HTM);
URB and DFL for example.

Oppsosites have interesting characteristics.
Either the pieces are interchangable on the plane or slice.
URB and ULF for example on the U plane.
URB and DRF for example on the R slice.

Or the opposites are not interchangable. That means they are twisted opposites.
URB and LFU for example.
URB and FUL perhaps.

Lets move the LFU one quarter turn.
There are 6 quarter turns possible, L, L', F, F', U, U'
U and U' have no effect on the cube state really, both cubies are being moved. Because both are in the U layer.
The other 4 quarter turns possible bring the LFU to another location in the U layer, or to the location of the URB's polar opposite.
Both turns that keep the LFU in the U layer (the URB and the LFU's common layer) make the URB and the LFU interchangable on either the U plane or R slice.
This of course also applied to the URB. You can make 6 moves, U, U', R, R', B, or B'
2 which have no effect,
2 bring the URB into the LFU's plane-orbit or slice-orbit.
2 bring the URB into the LFU's polar opposite.

So determine if two pieces are interchangable opposites, or twisted opposites.
Here is a nice characteristic for setting up commutators.
Interchangable opposites, are good to have, because well you've found your part B to the commutator in an 8 move case.
Twisted opposites are good, because you if you have twisted oppsites, and the third piece is interchangeable (adjacent or opposite) with one of the first two, you've found your insertion.
URB and ULF are in a cycle. You could interchange the two with U2.
URB and LFU are in a cycle. You could insert with FR'F' or R'FR

Thanks to dbeyer for submitting the terminology.

KINDS OF BH CORNER COMMUTATORS
Pure 8 moves
This is just a standard 8-move commutator. In the form ABA'B', A is three moves, and B is one move. It is always ABA'B' or BAB'A' to solve the cube from a pure position.

Example of Pure Commutator:

Do F R B R' F' R B' R' on a solved cube

The case you are looking at right now is UBR->UFR->LFU

This is a standard 8-move commutator. In the format ABA'B' this is what each should do in an 8 move commutator.

A = Put one corner in place (3 moves)
B = Interchange another corner for the corner solved in A to the position that corner A was in (1 move)
A' = Solve the corner brought over by move B (3 moves)
B' = Undo the interchanging move (1 move)

In this case, we want A to put one corner in place. To do this, we do R B R', which solves UFR. So we have

A = R B R'

Next comes B, which moves a new corner (in this case, UFL) into the position where UFR is right now. To do this, we do F. SO we have.

A = R B R'
B = F

After that we have A', which is the inverse of A, which will solve what was UFL before. Now we have:

A = R B R'
B = F
A' = R B' R'

Lastly, and if you're following along this should be extremely obvious, we need to undo B with B', which is just undoing the interchanging move, so F'. We now have.

A = R B R'
B = F
A' = R B' R'
B' = F'

The entire commutator written out is

R B R' F R B' R' F'

That is exactly what you will see if you go to the BH website on speedcubing.com, the case looks like this:

(URB UFR LFU) R B R' F R B' R' F' (8 HTM) Direct Insert

A direct insert is a kind of pure 8-move commutator. This is how you would intuitively find an 8-move commutator without memorizing the algorithm.

A9 9 moves
A9's are very similar to Pure commutators, except they have a 1-move setup and undo-setup. The thing that makes this so special is that the 1 move cancels out with the first part of A. In the for SABA'B'S', S and A have a cancellation.

Here is an example of an A9 commutator.

Do B2 D B' U2 B D' B' U2 B' on a solved cube.

The commutator we are looking at is URB->UBL->RUF, which is an A9 commutator. What we are going to look for is 1 move that does two things.

1) Sets up an 8-move commutator.
2) Cancels into the A of the 8-move commutator.

So, in this case, if we do the move B2, it sets up for a pretty simple 8-move commutator. And the A part of the commutator, as we can see, is B' U2 B, which will cancel into the B2. So, if we use the format:

SABA'B'S'

We now have

S = B2
A = B' U2 B

Now we need to find the B. B is pretty simple, it's D, to bring in the next corner. So now we have.

S = B2
A = B' U2 B
B = D

Now, we simply do the inverse of A to put the next corner in place.

S = B2
A = B' U2 B
B = D
A' = B' U2 B

Next, we reverse B.

S = B2
A = B' U2 B
B = D
A' = B' U2 B
B' = D'

And finally we undo the 1 move setup S that we did to start the A9 commutator. This won't cancel like the first time we applied the commutator. So now we have this:

S = B2
A = B' U2 B
B = D
A' = B' U2 B
B' = D'
S' = B2

The entire commutator, without cancellations looks like this:

B2 B' U2 B D B' U2 B D' B2

But of course, B2 B' simplifies to just B, so now we can write it like this:

B U2 B D B' U2 B D' B2

And this is exactly what you will find in the BH corner website on speedcubing.com. The case looks like this:

(URB UBL RUF) B U2 B D B' U2 B D' B2 (9 HTM) A9

Columns 11 moves
dbeyer explains column commutators as person preference from cyclic shifts or A9's. I am still researching how cyclic shifts work, but as for columns, I know that if you do one setup move, you'll get an A9 with no cancellations except for the one within the A9, therefore giving you 11 moves.

Here is an example of a Column commutator.

Do R U L2 U R' U' L2 U R U2 R' on a solved cube.

The case you are looking at right now is URB->ULF->DFL.

So, the basics of what we are looking at is a new setup move to setup to an A9 case. This setup move should be 1 move, and we'll call it M. The pattern we are looking at for a Column case is:

MSABA'B'S'M'

The setup move that we're going to use is R, which will set us up for a 9 move commutator. So we have:

M = R

Now, we're going to look for the S, which was part of the A9 case (see above). The mve that we're going to use for this is U. So now we have.

M = R
S = U

The next step is to find the normal 8-move commutator from this position. The A that we're going to use will be U R' U' so now we have.

M = R
S = U
A = U R' U'

Next, we're going to do B, which will bring in a new corner piece into position. This move is L2. So now we have.

M = R
S = U
A = U R' U'
B = L2

Next, of course, we are going to undo all of the moves we did in the order of the pattern shown above (MSABA'B'S'M')

So now we have.

M = R
S = U
A = U R' U'
B = L2
A' = U R U'
B = L2
S = U'
M = R'

The commutator without cancellations looks like this:

R U U R' U' L2 U R U' L2 U' R'

But, the two Us in a row become U2, so the commutator looks like this:

R U2 R' U' L2 U R U' L2 U' R'

Which is exactly what the case looks like in the website:

(URB ULF DFL) R U2 R' U' L2 U R U' L2 U' R' (11 HTM) Columns

Orthogonals 10 moves
An orthogonal commutator is basically very similar to an A9, except for the fact that it doesn't have any cancelations like the A9 does.

Cyclic Shifts 10 moves
The cyclic shifts is a commutator with a 2 move setup, a 4 move A, and a 1 move B but because the setup and the end of A' can cancel each other out, they finish the commutators while doing the first 2 moves of A' at the end.

Example of a cyclic shift: Do R' F U2 F' R F R' U2 R F' on a solved cube.

The case you are looking at is URB FUL BLU.

S= F R'
B= U2
A= R F' R' F
B'=U2
A'= F' R F R'
S'=R F'

That's all the parts of it written out completely. Written out, the commutator looks like this:

F R' U2 R F' R' F U2 F' R F R' R F'

But of course, the last two moves cancel each other out, so now we have this commutator:

F R' U2 R F' R' F U2 F' R (10 HTM)

Thanks to Rubixcubematt for giving this information on cyclic shifts.

Per Specials 12 moves
A per special works like this. This is the commutator:

A = U F2 U' F2 U'
B = R2

Let's look at the A part of the commutator. The two interchangeable corners are URB and DRF, interchangeable by an R2 turn. The lone corner is ULF. We are going to insert ULF into DRF without having any other net affect on the R slice. We will do this in the following way:

View the URB, UR, and UFR block as one block that will not be destroyed at any point of our solving. View the DBR, RB, DR block as a 2x2x1 block that will also never be destroyed, nor moved, during the A part of the commutator. This leaves the FR and DRF block. This block will be destroyed during the A part.

The first part of the A of the commutator is U F2. This brings the 3x1x1 block starting at UR down to DF. It also throws the FR 2x1x1 block up to the FL and ULF. Lastly, it moves the lone corner over to UBL. The next move is U'. This brings the lone corner from UBL over to ULF "decapitating" the FL and UFL block's corner and replacing it with the lone corner. Lastly, we will put the 3x1x1 block which is now in DF back to UR with F2 U'. Notice this has replaced the R layer back together, and has placed the lone corner at DFR. This replacement happened with the "decapitating" U' turn. Now interchange with the B part of the commutator by doing R2. Next we will indo the A part. Bring the UR 3x1x1 block to DF with U F2. This moves the 2x1x1 block that was at FR and DRF to FL and ULF. Now decapitate this block in the other direction with the turn U. Now place the 3x1x1 block, in DF now, back to UR with F2 U'. Lastly, do the second interchange move, R2, to complete the commutator.

Thanks to Chris Hardwick for giving information on Per Specials.

VISUALIZING COMMUTATORS
(URB RFD ULF) is very easy and is solved with : L D' L' U2 L D L' U2

My thought process would go like this:
Image in my mental journey for this cycle would be KC or K.C. the intials of a good friend of mine from high school who was an excellent drummer. Seeing him playing drums tells me KC is my cycle.

I can immediately see that ULF is interchangeable with the buffer via a U2, so I mentally label RFD as the "lone corner". I see that I can insert the lone corner into the U layer with L D' L' and not affect anything else in the U layer. Remember that I call the ULF spot the "action spot" (see my tutorial). This is the spot in the interchangeable layer in which I insert the lone piece. Now I see if the lone corner, RFD, really is supposed to go to the action spot ULF, and it does. Ok, so I do the insertion or A part of the commutator and I execute L D' L'. In my mind I picture the 3 corners as dark gray blobs on a light gray cube. I don't care at all about the colors, I only want to see the locations. Now that I have done the insertion move, the A part of the commutator, I do the interchange move. I actually execute U2 to interchange the buffer at URB to ULF.

This last part is very important. I now clear my mind of mentally visualizing *anything* at all. I'll describe what I am actually doing at this step in a minute, but bear with me for one second.

Literally, do not picture anything in your mind. Just know that you have just done the A part of the commutator followed by the B part. Now just blindly and without visualizing it undo the A part with the A' or inverse of the insertion. After that blindly, and without visualizing anything, undo the interchange move with U2.

For every single commutator I execute I am only mentally visualizing the first half of the commutator (either the A then the B, or the B then the A depending on if the commutator is ABA'B' or BAB'A').

Now I mentioned I would tell you what I am actually doing in just a second. What I am actually doing after the first half of the commutator is going back to my mental journey and recalling my next image. If I have a memory delay, then this is advantageous because it gives me a couple seconds head start to recall the piece (remember that I am physically executing the inverse of the two commutator parts at this very second). This works to shorten my delays if I have memory lapses, and also to make for smooth no pause solving when I don't have memory lapses.

Thanks to Chris Hardwick for submitting this information.

USING POLES TO DETERMINE COMMUTATORS
Look check this out:
There are 30 cases that are 10 moves.
There are two case types:
18 cyclic shifts
12 orthogonals.

Orthogonal vectors are at a right angle to the plane. You have 3 stickers on a plane.
Here is your perspective of determining if a case is orthogonal. That is to create a right angle. Pretty much your x,y,z axes of a 3-d graph.

Place the cube in one of the "rubik's cube stands" like that comes with a Rubik's Cube fresh out of the box. Pretty, much the tip of one corner is on the table, and the other is pointed upright. These simulate the north and south poles.

Anway, from a bird's eye view overhead, you see 3 faces. Call each of these planes. There are three adjacent corners to the north pole.
Okay, for example: Let the north pole be the UFR.
Now the 3 adjacent cubies are the ULF, URB, and DRF, cubies being cycled. Only 2 combinations of sticker patterns are possible to create a Orthogonal case. The direction of the cycle doesn't matter. It could be URB -> ULF -> DRF or it could be URB -> DRF -> ULF. (These are not orthogonal cases, just cubie cycles).

Since the URB is my buffer. I start with the U sticker of the URB. Now I know the 3 cubies that must be cycled. Now a sticker from each plane must be chosen. One from the U, the F, and the R planes (or their parallel planes).
So one sticker from the U has been chosen, the buffer is in the U plane, URB.
Now, if I was to cycle to the RFD, I must pick the FUL. If I did the FDR, I must choose the LFU.


If the URB is the Buffer:
URB -> (L)FU -> (F)DR; URB -> FDR -> LFU
URB -> (F)UL -> (R)FD; URB -> RFD -> FUL
URB -> (L)FU -> (B)DL; URB -> BDL -> LFU
URB -> (F)UL -> (L)BD; URB -> LBD -> FUL
URB -> (F)DR -> (L)BD; URB -> LBD -> FDR
URB -> (R)FD -> (B)DL; URB -> BDL -> RFD

So if they are all on different planes: Well then you have an orthogonal case!
Its awesome. A simple quarter turn setup, will give you interchangability. Then you can do an 8 move commutator. It's very veratile for the setups. Remember, its always just a quarter move. And afterwards its always just an 8 mover.

Thanks to dbeyer for submitting the information

ORBITS
Lets look at an Orbit.

The Controls--
Two adjacent pieces
Turning the F layer
The UFR cubie

Let this piece be AnI to the Buffer.
URB and RUF for example.
Turn F, you have slice-orbit opposites,
Turn F2, you have twisted polar opposites.
Turn F', you have plane-orbit opposites
Turn F4, you have Adjacent non-Interchangable (AnI)

Let this piece be interchangeable on the slice-orbit
URB and FRU
Turn F, you have twisted opposites
Turn F2, you have twisted polar opposites
Turn F', you have twisted opposites
Turn F4, you have Adjacent Slice-Orbit (AsO)

Two adjacent pieces
Let this piece be interchangable on the plane-oribit
URB and UFR
turn F, you have twisted opposites
turn F2, you have parallel polar opposites
turn F', you have twisted opposites.
turn F4, you have Adjacent Plane-orbit (ApO)


The controls--
Two opposite pieces
Turning the F layer
The ULF cubie

Interchangable Opposites
URB and ULF
turn F, you have AnI
turn F2, you have interchangable opposites
turn F', you have twisted polar opposites
turn F4, you have interhchangable opposites

Twisted Opposites - On the Plane (F plane)
URB and FUL
turn F, you have AsO
turn F2, you have twisted opposites
turn F', you have twisted polar opposites
turn F4, you have twisted opposites

Twisted Opposites -- On the Slice (F slice)
URB and LFU
turn F, you have ApO
turn F2, you have twisted opposites
turn F', you have twisted polar opposites
turn F4, you have twisted opposites

Of course if you were to turn the L layer, and use the ULF cubie, you would, wind up with inverted and swapped results of FUL and LFU.
Just look at these trends. Notice the characteristics of the oribits. This will allow you to see the correct setups for orthogonal cases. This will allow you to see the correct insertion for cyclic shifts.
This will allow you to find the correct cancelations on A9s, and Columns cases.

It's a very powerful method. Its not a list of algorithms to memorize. Its a freestyle method, that gives you a vast understanding of the cube's properties. This method goes into muscle memory. And using the techniques described here, you can actually figure out any cycle, and how to solve it by recognizing relationships.

You first recognize two pieces and their locational relationship to one another. What really sets it apart though is the 3rd piece. Just like you could recognize a lot of relationships between corner and edge pairs, but you need to reference the c/e pair to where the solved pair belongs.

Thanks to dbeyer for submitting this information

I'm hoping to expand this resource center with all the information necessary for learning the BH method, since there's a lot of interest in it now. I'm still working on learning it, and I'm going to use this resource center as a reference. Chris, Mike, Daniel, and everyone else who knows a lot about BH, post information below that I haven't already covered in the BH Resource center and I'll add it to this original post. That way people won't have to go digging through lots of threads to find the information they're looking for (like what I'm doing now).
 
Last edited:

Rubixcubematt

Member
Joined
Jul 23, 2008
Messages
793
Location
in a house
YouTube
Visit Channel
I think that the cyclic shifts is a commutator with a 2 move setup, a 4 move A, but because the setup and the end of A' can cancel each other out, they finish the commutators while doing the first 2 moves of A' at the end. E.G. do R' F U2 F' R F R' U2 R F' so this is URB FUL BLU.
so S= F R' B= U2 A= R F' R' F B'=U2 A'= F' R F R' S'=R F'
But since the last 2 moves of A and the S' can cancel each other out, the commutator becomes F R' U2 R F' R' F U2 F' R (10 HTM)

Hope this helps.
 
Last edited:

dbeyer

Member
Joined
Sep 12, 2006
Messages
396
Location
Harrington, Delaware
He is correct, and it is also can be looked at as taking the commutator itself, and starting at a different point in the alg.

ABA'B'
A: RF'.R'F
B: U2

By starting at the dot (.), proceeding and wrapping around doing all 10 moves, you have actually done a legitimate algorith. Just like with a sine wave, if you shift to the right or the left, the sine wave curve is still a sine wave, just shifted. As such if you take a reference to the state of the cube at the same point in time, as if there wasn't a horizontal shift, you would get a different result (cycle, or value). The properties of the commutator are still there, yet there are different results with respect to time.
 

deadalnix

Member
Joined
Apr 6, 2007
Messages
560
WCA
2008SECH01
An explaination of the name convention (that I don't get at all) for simple 8 moves case could help.

Anyway, I have no problem to figure out how to solves these cases, but I can't say if I'm doing a drop and catch, direct intert, toss up or whatever.
 

byu

Member
Joined
Dec 18, 2008
Messages
2,021
Location
California
WCA
2009YUBR01
YouTube
Visit Channel
I updated the Resource Center to display the explanation of cyclic shifts, copied pretty much from Rubixcubematt's description of them. I'm still working out Per Specials.
 

byu

Member
Joined
Dec 18, 2008
Messages
2,021
Location
California
WCA
2009YUBR01
YouTube
Visit Channel
Hehe... I'm just as bad as you are, if not worse. It's basically recognition. If you can see a three move A, then you know it's an 8 move. If you see a setup to a three move A, you have a 9-move (A9) or a 10-move (Orthogonal), depending on cancellations.
 

Rubixcubematt

Member
Joined
Jul 23, 2008
Messages
793
Location
in a house
YouTube
Visit Channel
I think i found out the per special case as well, well at least for this case.

Do L2 U B2 U' B2 U' L2 U B2 U B2 U' this makes it the URB DLB ULF case.

A= U B2 U' B2 U' B= L2 A' = U B2 U B2 U' B'= L2

so the commutator in full is
U B2 U' B2 U' L2 U B2 U B2 U' L2

BTW, no problem with helping out with the cyclic shifts, its a good thing that your doing creating this thread.
 
Last edited:

cmhardw

Premium Member
Joined
Apr 5, 2006
Messages
4,115
Location
Orlando, Florida
WCA
2003HARD01
YouTube
Visit Channel
I'll explain the Per Specials, because I find them by far the most beautiful corner case - no question.

There are 2 main algorithms in the BH method for Per Special cases. One of them is 12 move commutator, and the other quite honestly is an alg Daniel and I found using ACube. I can't speak for Daniel, but I am still entirely baffled as to how this algorithm works.

I'll explain the commutator:

A = U F2 U' F2 U'
B = R2

Let's look at the A part of the commutator. The two interchangeable corners are URB and DRF, interchangeable by an R2 turn. The lone corner is ULF. We are going to insert ULF into DRF without having any other net effect on the R slice. We will do this in the following way:

View the URB, UR, and UFR block as one block that will not be destroyed at any point of our solving. View the DBR, RB, DR block as a 2x2x1 block that will also never be destroyed, nor moved, during the A part of the commutator. This leaves the FR and DRF block. This block will be destroyed during the A part.

The first part of the A of the commutator is U F2. This brings the 3x1x1 block starting at UR down to DF. It also throws the FR 2x1x1 block up to the FL and ULF. Lastly, it moves the lone corner over to UBL. The next move is U'. This brings the lone corner from UBL over to ULF "decapitating" the FL and UFL block's corner and replacing it with the lone corner. Lastly, we will put the 3x1x1 block which is now in DF back to UR with F2 U'. Notice this has replaced the R layer back together, and has placed the lone corner at DFR. This replacement happened with the "decapitating" U' turn. Now interchange with the B part of the commutator by doing R2. Next we will indo the A part. Bring the UR 3x1x1 block to DF with U F2. This moves the 2x1x1 block that was at FR and DRF to FL and ULF. Now decapitate this block in the other direction with the turn U. Now place the 3x1x1 block, in DF now, back to UR with F2 U'. Lastly, do the second interchange move, R2, to complete the commutator.

Hope this helps to see how the Per Special commutator alg works. My description of the "decapitating" turn is more violent than I really picture it during a solve, but I think it makes it clear what's going on in that part of the step.

Chris
 
Last edited:

byu

Member
Joined
Dec 18, 2008
Messages
2,021
Location
California
WCA
2009YUBR01
YouTube
Visit Channel
Thanks Chris, I'll add that to the resource center. I still am having trouble executing commutators blindfolded. I used Ryan Heise's tutorial for how to do commutators, and that makes it REALLY hard to do them blindfolded. Chris, can you walk me through your thought process for a really easy 8-move commutator? Like, if you only know the 2 pieces you shoot to from UBR, what will you think in your head?

EDIT - Oh, and Chris, are you done with the BH edges page yet?
 

cmhardw

Premium Member
Joined
Apr 5, 2006
Messages
4,115
Location
Orlando, Florida
WCA
2003HARD01
YouTube
Visit Channel
Thanks Chris, I'll add that to the resource center. I still am having trouble executing commutators blindfolded. I used Ryan Heise's tutorial for how to do commutators, and that makes it REALLY hard to do them blindfolded. Chris, can you walk me through your thought process for a really easy 8-move commutator? Like, if you only know the 2 pieces you shoot to from UBR, what will you think in your head?

(URB RFD ULF) is very easy and is solved with : L D' L' U2 L D L' U2

My thought process would go like this:
Image in my mental journey for this cycle would be KC or K.C. the intials of a good friend of mine from high school who was an excellent drummer. Seeing him playing drums tells me KC is my cycle.

I can immediately see that ULF is interchangeable with the buffer via a U2, so I mentally label RFD as the "lone corner". I see that I can insert the lone corner into the U layer with L D' L' and not affect anything else in the U layer. Remember that I call the ULF spot the "action spot" (see my tutorial). This is the spot in the interchangeable layer in which I insert the lone piece. Now I see if the lone corner, RFD, really is supposed to go to the action spot ULF, and it does. Ok, so I do the insertion or A part of the commutator and I execute L D' L'. In my mind I picture the 3 corners as dark gray blobs on a light gray cube. I don't care at all about the colors, I only want to see the locations. Now that I have done the insertion move, the A part of the commutator, I do the interchange move. I actually execute U2 to interchange the buffer at URB to ULF.

This last part is very important. I now clear my mind of mentally visualizing *anything* at all. I'll describe what I am actually doing at this step in a minute, but bear with me for one second.

Literally, do not picture anything in your mind. Just know that you have just done the A part of the commutator followed by the B part. Now just blindly and without visualizing it undo the A part with the A' or inverse of the insertion. After that blindly, and without visualizing anything, undo the interchange move with U2.

For every single commutator I execute I am only mentally visualizing the first half of the commutator (either the A then the B, or the B then the A depending on if the commutator is ABA'B' or BAB'A').

Now I mentioned I would tell you what I am actually doing in just a second. What I am actually doing after the first half of the commutator is going back to my mental journey and recalling my next image. If I have a memory delay, then this is advantageous because it gives me a couple seconds head start to recall the piece (remember that I am physically executing the inverse of the two commutator parts at this very second). This works to shorten my delays if I have memory lapses, and also to make for smooth no pause solving when I don't have memory lapses.

Hope this helps,
Chris
 

byu

Member
Joined
Dec 18, 2008
Messages
2,021
Location
California
WCA
2009YUBR01
YouTube
Visit Channel
Wow! I actually understand how to see 8-move commutators in my mind. Thanks Chris! Now a new problem has risen... I can't seem to figure out A9s in my mind or Orthogonals. It's hard to find a one move that sets up two interchangeable pieces.
 

dbeyer

Member
Joined
Sep 12, 2006
Messages
396
Location
Harrington, Delaware
Look check this out:
There are 30 cases that are 10 moves.
There are two case types:
18 cyclic shifts
12 orthogonals.

Orthogonal vectors are at a right angle to the plane. You have 3 stickers on a plane.
Here is your perspective of determining if a case is orthogonal. That is to create a right angle. Pretty much your x,y,z axes of a 3-d graph.

Place the cube in one of the "rubik's cube stands" like that comes with a Rubik's Cube fresh out of the box. Pretty, much the tip of one corner is on the table, and the other is pointed upright. These simulate the north and south poles.

Anway, from a bird's eye view overhead, you see 3 faces. Call each of these planes. There are three adjacent corners to the north pole.
Okay, for example: Let the north pole be the UFR.
Now the 3 adjacent cubies are the ULF, URB, and DRF, cubies being cycled. Only 2 combinations of sticker patterns are possible to create a Orthogonal case. The direction of the cycle doesn't matter. It could be URB -> ULF -> DRF or it could be URB -> DRF -> ULF. (These are not orthogonal cases, just cubie cycles).

Since the URB is my buffer. I start with the U sticker of the URB. Now I know the 3 cubies that must be cycled. Now a sticker from each plane must be chosen. One from the U, the F, and the R planes (or their parallel planes).
So one sticker from the U has been chosen, the buffer is in the U plane, URB.
Now, if I was to cycle to the RFD, I must pick the FUL. If I did the FDR, I must choose the LFU.


If the URB is the Buffer:
URB -> (L)FU -> (F)DR; URB -> FDR -> LFU
URB -> (F)UL -> (R)FD; URB -> RFD -> FUL
URB -> (L)FU -> (B)DL; URB -> BDL -> LFU
URB -> (F)UL -> (L)BD; URB -> LBD -> FUL
URB -> (F)DR -> (L)BD; URB -> LBD -> FDR
URB -> (R)FD -> (B)DL; URB -> BDL -> RFD

So if they are all on different planes: Well then you have an orthogonal case!
Its awesome. A simple quarter turn setup, will give you interchangability. Then you can do an 8 move commutator. It's very veratile for the setups. Remember, its always just a quarter move. And afterwards its always just an 8 mover.
 

dbeyer

Member
Joined
Sep 12, 2006
Messages
396
Location
Harrington, Delaware
Orthogonals are a very special correlation of three pieces being cycled. All three cubies are adjacent to the same cubie, and hence none are adjacent to another cubie in the cycle. That is to say, the cubies themselves (not the stickers) Are all mutually interchangable by a double turn.
The ULF is interchangable by U2 with the URB.
The ULF is interchangable by L2 with the DLB.
The ULF is interchangable by F2 with the DRF.

(U, L, F)2 <-- See??
Likewise the pattern applied to the URB, finding it's 3 opposite cubies. doing U2, R2, or B2. You will see that they are all mutually opposite and all interchangable.

URB, ULF, DRF, and DLB are opposites. (R/L follow U/D)
DFL, DBR, UFR, and UBL are opposites. (F/B follow U/D)
By choosing URB as the buffer, our focus is the URB subset of opposites.
If anybody was to use another buffer, you should determine which set your buffer is in.
Is it an R/L plane cubie, or a F/B plane cubie does [R/L or F/B] follow U/D?

Now Orthogonals, using this concept of opposite in the R/L plane. You choose 3 cubies. One is of course the buffer. Then you choose one sticker from each plane. And if you've meet all these requirements. You have an orthogonal case.

Lets use the same subset of cubies for our cycles, but lets use a different limiting factor for the sticker selection. The selection of sticker permutation is limited to one set of parallel planes. Since we start with the URB. That's limiting the cycles to the U/D stickers of this subset. That is to say that in this effect, all three cubies are mutually opposite and interchangable.

URB -> ULF -> DRF is a 12 move case. They are all mutually opposite, and all mutually interchangable.

URB -> ULF -> DRF; URB -> DRF -> ULF
URB -> ULF -> DLB; URB -> DLB -> ULF
URB -> DRF -> DLB; URB -> DLB -> DRF
 

dbeyer

Member
Joined
Sep 12, 2006
Messages
396
Location
Harrington, Delaware
Here are the other 18 cases that are 10 moves.
These are cyclic shifts.
Similar to Orthogonals in several respects:
Rather than the other two pieces being mutually opposite to the buffer, the two other pieces in the cycle are mutually adjacent to the buffer. So this cycle is contained to a single layer.
Either the U, R, or B layer, if your buffer is the URB.
As such, these cycles, do not involve the Polar opposite (the DFL)
A sticker from each set of planes must be in the cycle (the U/D, F/B, and L/R).
Cyclic Shifts: One piece, and its two adjacent cubies are being cycled.
Orthogonals: One piece, and its two opposite cubies are being cycled.
--
Cyclic Shifts: One sticker from each set of planes is being cycled
Orthogonals: One sticker from each set of planes is being cycled

When you are determining if a case is a cyclic shift. Determine the middle piece, that is the piece adjacent to both.
URB -> ULF -> UFR;
UFR is the middle piece.
URB -> UFR -> UBL;
URB is the middle piece.
Now sticker permutation, obviously the opposite cubies are both adjacent to the middle piece.

There are three types of interchangability.
Interchangability on the Plane
Interchangability on the Slice
Interchangability on the Layer
URB and ULF are interchangable on the U plane.
Likewise: URB and UFR are interchangable on the U plane.
URB and DRF are interchangable on the R slice.
Likewise: URB and FRU are interchangeable on the R slice.

The third concept of interchangability is on a layer. You are looking at the piece as a whole. You do not look at sticker permutation. You just make a reference to the cubies, rather than the sticker permutation.

the URB is on 1 plane, and 2 slices.
Each plane and slice-orbit has 4 pieces on it. Hence, the URB is interchangable with 3 pieces on the U plane, 3 pieces on the R slice, and 3 pieces on the B slice.

Now any cubie is interchangable with respect to the layer with 6 of the 7 other cubies with one turn (HTM). Every cubie but it's polar opposite.
Adjacent cubies are of course interchangable on that layer.
Adjacent cubies can have three characteristics. The first two were described already, interchangability on the plane, or on the slice.
If neither exist, then the cubies are adjacent and non-interchangable.
Here we coined the term AnI (like the anny in Danny).
 

dbeyer

Member
Joined
Sep 12, 2006
Messages
396
Location
Harrington, Delaware
So cyclic shifts are contained to three cubies on one layer.
The opposite cubies on the layer are both AnI with respect to the middle cubie. With both cubies being AnI with respect to the middle cubie, you are then creating a cycle with a piece from each plane.

For clarification: When I say each plane, let it be understood -- a plane or its parallel, (speedcubing correlation, a face or its opposite side)
 

dbeyer

Member
Joined
Sep 12, 2006
Messages
396
Location
Harrington, Delaware
Here is a quick glossary list:

Cubie -- A piece, as a whole, with no specific permutation or orientation there of. Just to make a reference to a location on the cube. Standard U/D first.
URB or DRF for example.

Interchangability -- There are different ranges and degrees of interchangability. Looking at the cubie locations as a whole. A piece is interchangable with every other piece in one turn (htm), except it's polar opposite.
Interchangability on the Layer: This is the broadest form of interchangability. Looking at the cubies as a whole.
The six cubies interchangeable with the URB;
ULF, DRF, DLB, UFR, UBL, DBR
NOT the DFL

Interchangable on the Plane-orbit:
The U plane is the URB, UFR, ULF, UBL.
The F plane is the FRU, FUL, FLD, FDR

Interchangable on the Slice-orbit:
The URB's R-slice-orbit;
URB, BRD, DRF, FRU.
The FUL's U slice-orbit;
FUL, LUB, BUR, RUF

Adjacent Cubies -- To be next interchanged by 1 quarter turn
URB and FLU for example.
FDR and RDB
Opposite Cubies -- To be interchanged by 1 half turn.
URB and DLB for example.
FRU and FLD
Polar Opposites -- The two cubies with are not interchangable in anyway shape or form with just 1 turn (HTM);
URB and DFL for example.

Oppsosites have interesting characteristics.
Either the pieces are interchangable on the plane or slice.
URB and ULF for example on the U plane.
URB and DRF for example on the R slice.

Or the opposites are not interchangable. That means they are twisted opposites.
URB and LFU for example.
URB and FUL perhaps.

Lets move the LFU one quarter turn.
There are 6 quarter turns possible, L, L', F, F', U, U'
U and U' have no effect on the cube state really, both cubies are being moved. Because both are in the U layer.
The other 4 quarter turns possible bring the LFU to another location in the U layer, or to the location of the URB's polar opposite.
Both turns that keep the LFU in the U layer (the URB and the LFU's common layer) make the URB and the LFU interchangable on either the U plane or R slice.
This of course also applied to the URB. You can make 6 moves, U, U', R, R', B, or B'
2 which have no effect,
2 bring the URB into the LFU's plane-orbit or slice-orbit.
2 bring the URB into the LFU's polar opposite.

So determine if two pieces are interchangable opposites, or twisted opposites.
Here is a nice characteristic for setting up commutators.
Interchangable opposites, are good to have, because well you've found your part B to the commutator in an 8 move case.
Twisted opposites are good, because you if you have twisted oppsites, and the third piece is interchangeable (adjacent or opposite) with one of the first two, you've found your insertion.
URB and ULF are in a cycle. You could interchange the two with U2.
URB and LFU are in a cycle. You could insert with FR'F' or R'FR
 
Last edited:

DavidWoner

The Punchmaster
Joined
May 7, 2008
Messages
3,936
Location
Kansas City, MO, USA
WCA
2008WONE01
YouTube
Visit Channel
Edit: whoops just saw that chris already explained Per specials, but thankfully it looks like I understood them properly, even if I conveyed the idea poorly. And I agree, once you understand these, its apparent how beautiful they actually are.

lets look at this case:
(URB ULF DRF) U F2 U' F2 U' R2 U F2 U F2 U' R2 (12 HTM) Per Special

(U F2 U' F2 U', R2)

although I think it will make more sense if we look at this way:

[(U F2: U'), R2]

Note that ULF and DRF and interchangeable by F2 (we will call this lowercase a), and DRF and URB are interchangeable by R2 (lowercase b).

Now what we try to do is swap the non-buffer pieces (swap a) without disturbing the R face of the cube. What U F2 does is move the buffer out of the way, and changes swap a from F2 to U'. So you perform swap a, then undo the conjugation. We have now swapped ULF and DRF without disturbing the R face, so we can just exchange pieces with using R2 without any problems. The rest is a simple matter of how commutators work, what with undoing the first half.

I hope this made a bit of sense. If not I will try to clarify a bit more later.
 
Last edited:
Top