[Help Thread] Roux Discussion and Help

[RubiksJake12]

Congrats!

Was that an unintentional CMLL skip, our you forced it?


This was massively helpful. I kept seeing the A2 A3, etc. notation and had no idea what was going on. I'll try to teach it to myself now.


A2, A3, ..., H6 are the names of the CMLL (CLL, COLL) cases. The letter stands for the orientation, for example, A is where all the corners oriented. Then numbers indicate which pieces are needed to swap. 1 is for nothing, 2,3,4,5 indicate different adjacent piece swaps, 6 is diagonal swap. Gilles Roux uses this kind of notation, see his CMLL table: http://grrroux.free.fr/method/Step_3.html
I don't know if he invented the notation, but it seems the general case notation in Roux CMLL.

For a two look CMLL, you need to know:
- Bx, Cx, Dx, Ex, Fx, Gx, Hx (where x can be any number) for the orientation, and
- A2 (adjacent swap) and A6 (diagonal swap) for permutation.

Note that A2 = A3 = A4 = A5 because of AUF, and also H2 = H4 and H3 = H5

There is a different type of case notation, see this table: http://www.speedsolving.com/wiki/index.php/CxLL

I personally recommend that you should start to learn full CMLL once you are about 30-40 seconds on average. Otherwise you can improve your times much easier by focusing on F2B (and CMLL) and lookahead.

For determining what you need to focus on, see the guides of http://rouxtorial.webs.com/miscellaneous.htm and http://wafflelikescubes.webs.com/sub15guidetoroux.htm

Thank you very much for this response, it was tremendously helpful. I was so confused because I kept seeing this notation everywhere I Went. I still don't understand it fully, but I'll keep reading your post until I fully understand Thanks for your time and effort!



Decided to ask this question here instead of the One Question Answer Thread. Hope that's ok.

For the Roux method, do you have to learn the algorithms for the last step? (orienting and permutating the last edges). Because I can solve it intuitively and it always leads me to 1 of 2 PLLs I already know and can execute quickly.

Basically, are you supposed to learn algorithms for this step of Roux, or can I do it my own way? Is learning the algorithms faster once you commit them to muscle memory?

 

[GuRoux]

Thank you very much for this response, it was tremendously helpful. I was so confused because I kept seeing this notation everywhere I Went. I still don't understand it fully, but I'll keep reading your post until I fully understand Thanks for your time and effort!



Decided to ask this question here instead of the One Question Answer Thread. Hope that's ok.

For the Roux method, do you have to learn the algorithms for the last step? (orienting and permutating the last edges). Because I can solve it intuitively and it always leads me to 1 of 2 PLLs I already know and can execute quickly.

Basically, are you supposed to learn algorithms for this step of Roux, or can I do it my own way? Is learning the algorithms faster once you commit them to muscle memory?

it doesn't matter what way you learn it, as long as you can do it well in the future. you definitely should not be using pll of oll for the LSE. make sure you are doing it like most other roux users.

 

[Renslay]

Decided to ask this question here instead of the One Question Answer Thread. Hope that's ok.

For the Roux method, do you have to learn the algorithms for the last step? (orienting and permutating the last edges). Because I can solve it intuitively and it always leads me to 1 of 2 PLLs I already know and can execute quickly.

Basically, are you supposed to learn algorithms for this step of Roux, or can I do it my own way? Is learning the algorithms faster once you commit them to muscle memory?

Step 4a) (edge orientation): I learnt them as algorithms, but they are exceptionally easy; practically, all of them are "how to bring EO case into the E31 case" (where 3 is unoriented on the top, 1 is on the bottom, which can be solved with M* U* M*). But you can (and should) experiment with different solutions, not just one alg for one case. For example, if there are two wrong edges on the bottom, and two adjacent on the top, you can do AUF + M2 + E31, or you can also do M2 first, then AUF, then E31.

Another example, where wrong edges are:
1) UB + UR = M' U' M' (U2 M' U* M*)
2) UB + UL = M' U M' (U2 M' U* M*)
3) UF + UR = M U M (U2 M U* M*) OR AUF to 1) or 2) because M' is better then M
4) UF + UL = M U' M (U2 M U* M*) OR AUF to 1) or 2) because M' is better then M

And so on, and so on, there are many other solutions to the rests, where you can mirror, inverse or transform the algorithm into a different one. (U* means you can choose between U and U', similar for M* - you can influence your Step 4b) with it!).
Also note that when all the six edges are misoriented, an easy, direct solution can be R U' r' U' M' U r U r'. Futhermore, an interesting thing, that if you have 6 misoriented edges, and you have your UL/UR edges in UF and UR places, then after the algorithm above, your UL/UR edges will be in DF and DB - an easy case for Step 4b!

Step 4b) (UL/UR edges): this should be a piece of a cake. Full intuitive. For weird cases, M' U2 M is your friend.
Step 4a) and Step 4b) can be usually merged, because you can solve the E31 case in many ways.

Step 4c) (M-perm): algorithmic, although all of them are easy, short and few. Practice recognition and lookahead!

With practice, the whole Step 4 can be done with one, fluent solving. Keep in mind that it is not as wired as for example an OLL + PLL solve; you can affect the sub-steps easily, even merge them.

 

[RubiksJake12]

it doesn't matter what way you learn it, as long as you can do it well in the future. you definitely should not be using pll of oll for the LSE. make sure you are doing it like most other roux users.

But I AM doing PLLs >.< I don't know how else to do it. I guess I'll watch more videos and try to catch on. Oh well.

Step 4a) (edge orientation): I learnt them as algorithms, but they are exceptionally easy; practically, all of them are "how to bring EO case into the E31 case" (where 3 is unoriented on the top, 1 is on the bottom, which can be solved with M* U* M*). But you can (and should) experiment with different solutions, not just one alg for one case. For example, if there are two wrong edges on the bottom, and two adjacent on the top, you can do AUF + M2 + E31, or you can also do M2 first, then AUF, then E31.

Another example, where wrong edges are:
1) UB + UR = M' U' M' (U2 M' U* M*)
2) UB + UL = M' U M' (U2 M' U* M*)
3) UF + UR = M U M (U2 M U* M*) OR AUF to 1) or 2) because M' is better then M
4) UF + UL = M U' M (U2 M U* M*) OR AUF to 1) or 2) because M' is better then M

And so on, and so on, there are many other solutions to the rests, where you can mirror, inverse or transform the algorithm into a different one. (U* means you can choose between U and U', similar for M* - you can influence your Step 4b) with it!).
Also note that when all the six edges are misoriented, an easy, direct solution can be R U' r' U' M' U r U r'. Futhermore, an interesting thing, that if you have 6 misoriented edges, and you have your UL/UR edges in UF and UR places, then after the algorithm above, your UL/UR edges will be in DF and DB - an easy case for Step 4b!

Step 4b) (UL/UR edges): this should be a piece of a cake. Full intuitive. For weird cases, M' U2 M is your friend.
Step 4a) and Step 4b) can be usually merged, because you can solve the E31 case in many ways.

Step 4c) (M-perm): algorithmic, although all of them are easy, short and few. Practice recognition and lookahead!

With practice, the whole Step 4 can be done with one, fluent solving. Keep in mind that it is not as wired as for example an OLL + PLL solve; you can affect the sub-steps easily, even merge them.

I don't know half of what you said, I'm sorry. I'm extremely new to Roux. I'm an avid Fridrich user with an avg of just under 20 seconds, but I picked up Roux like two days ago. I'm still learning some notations, ie. I don't know what "how to bring EO case into the E31 case" means. EO=edge orientation, but the numbers are giberish to me.

By "edge orientation" , does that simply mean putting the top edges on top, and the bottom edges on the bottom? Or does it mean putting the top edges in their correct place (even if not flipped properly) and the same for the bottom?

And I think the last thing I was confused about was when you said things like "(U2 M' U* M*) What do asterisks mean? D:

Oh, and out of curiousity, is the quickest way to do M' moves with one's thumb?

 

[TDM]

And I think the last thing I was confused about was when you said things like "(U2 M' U* M*) What do asterisks mean? D:

I believe they mean that those moves can be done as either U/M or U'/M', depending on your preference/whichever affects the next step in a more useful way.

 

[pijok]

I don't know half of what you said, I'm sorry. I'm extremely new to Roux. I'm an avid Fridrich user with an avg of just under 20 seconds, but I picked up Roux like two days ago. I'm still learning some notations, ie. I don't know what "how to bring EO case into the E31 case" means. EO=edge orientation, but the numbers are giberish to me.

Ignore what he said and watch/read some tutorials first. For example: http://www.youtube.com/watch?v=bltwWaHOhJs

By "edge orientation" , does that simply mean putting the top edges on top, and the bottom edges on the bottom? Or does it mean putting the top edges in their correct place (even if not flipped properly) and the same for the bottom?

neither nor

Oh, and out of curiousity, is the quickest way to do M' moves with one's thumb?

I think even with nose is better than with thumb
Here you go: http://www.speedsolving.com/forum/showthread.php?40427-How-I-perform-M-moves

LSE is usually done like this:
4a) Orient all edges
4b) Solve UL+UR
4c) Solve M Layer

 

[Renslay]

But I AM doing PLLs >.< I don't know how else to do it. I guess I'll watch more videos and try to catch on. Oh well.



I don't know half of what you said, I'm sorry. I'm extremely new to Roux. I'm an avid Fridrich user with an avg of just under 20 seconds, but I picked up Roux like two days ago. I'm still learning some notations, ie. I don't know what "how to bring EO case into the E31 case" means. EO=edge orientation, but the numbers are giberish to me.

By "edge orientation" , does that simply mean putting the top edges on top, and the bottom edges on the bottom? Or does it mean putting the top edges in their correct place (even if not flipped properly) and the same for the bottom?

And I think the last thing I was confused about was when you said things like "(U2 M' U* M*) What do asterisks mean? D:

Oh, and out of curiousity, is the quickest way to do M' moves with one's thumb?

It's edge orientation, not permutation. Just like in OLL, you it doesn't matter where the pieces are going. All you care is how is a piece in the given place, not where.

Doring Step 4a), you only "twist" the 6 edges (if needed), and they can move anywhere.

If you solve the blocks with white on bottom, then your goal during Step 4a) is to make sure that your top and bottom faces are all white and yellow (could be mixed), including the centers. Note that white and yellow can be mixed, even the yellow center can be on the bottom.

For example, take this LSE scramble (hold your cube in a way that you would hold your cube after the solved blocks):
F L2 U2 R2 B' R2 U2 L U' L R' F R F2

Correct edges: UL, UB, UR, DB. These edges has white OR yellow on top/bottom side.
Wrong edges: UF, DF.

The solution is U2 M' U M + U M' U M'

Note that the first part (U2 M' U M) bring us to the case where we have 3 wrong edges on the top and one in the bottom. This is the easiest case, can be solved with M U* M* or M' U* M* (yes, U* means U or U'). So, as the second part, we either can do U M' U M' or U M' U' M' or U M' U M or U M' U M'. For a beginner, it doesn't matter.

Now, Step 4b) is putting your UL/UR edges (those that originally belongs to UL and UR) to ther correct places, together with the corners. Now they are indeed at UL and UR, but not with the corners. Solution: U M2 + U' M2 U'. First part brings them down to DF and DB (U M2), rotate corners (U'), put them in place (M2 U'). Also an alternative solution: U' M2 U' M2 U.

Step 4c) is where you permute the last 4 edges. Here it is simply M' U2 M U2 M2.

Full solution:
U2 M' U M U M' U M' U M2 U' M2 U' M' U2 M U2 M2

Advanced solution (with so called misoriented centers - just show you that later you can merge the steps):
U2 M' U M U M U' M2 U M' U2 M' U2 M'

If you want, I can give you all my "algorithms" for Step 4a) (the edge orientation step).

 

[Bestsimple]

I've been trying some roux OH recently but i've been really struggling with my M moves while using table abusing. The M's are fine it's just the M move that makes the cube slip and slow the solve down a lot.

 

[cubingboss]

lean the top layer against the table when you do Ms and use your ring finger to push it.

 

[Bestsimple]

So for the M move is the technique you're referring to is pushing the BD piece to BU using your ring finger?

 

[RubiksJake12]

I believe they mean that those moves can be done as either U/M or U'/M', depending on your preference/whichever affects the next step in a more useful way.

OH! I've never seen this before, thank you!

Ignore what he said and watch/read some tutorials first. For example: http://www.youtube.com/watch?v=bltwWaHOhJs


neither nor


I think even with nose is better than with thumb
Here you go: http://www.speedsolving.com/forum/showthread.php?40427-How-I-perform-M-moves

LSE is usually done like this:
4a) Orient all edges
4b) Solve UL+UR
4c) Solve M Layer

Hahaha, thank you for the links. I've been watching Waffle's videos lately and I've been watching random solves using Roux method simply to get an idea of what it looks like/moving styles/fingertricks etc. Thank you for the concise answers.

Edit:

I think even with nose is better than with thumb
Here you go: http://www.speedsolving.com/forum/sh...erform-M-moves

Ok, so I meant M moves. I'm sorry. I used this website http://badmephisto.com/pll.php to learn my Olls. The ONLY M' move i've ever had to make in solving a rubik's cube (because i use fridrich) is in the Z perm that I'm aware of anyway. I never tought myself cube notation because I thought I understood it properly. Today, after a year of speedcubing, I realized that I've had my M and M' prime turns backwards because of the Z perm I think haha. I've been performing the Z perm by doing M2 U M2 U M' U2 M2 U2 M' U2 unknowingly because you can do M OR M' and it still will solve the case. Since I used an M and it solved it, I assumed that I correctly executed an M' move O_O I'm literally blown away that I've gone this long without knowing I've been doing moves wrong. It explains how I haven't been able to properly do some Roux algorithms I've tried recently as well. I'm completely stupid.

Bottom line, I'm an idiot. My real question is: How do I properly/efficiently do an M' move? And your link has helped me. I thought I'd just share this bit of mindboggling stupidity on my end to explain to you why I use my thumb to do "M'" moves....I mean I use my thumb to do M moves

If you solve the blocks with white on bottom, then your goal during Step 4a) is to make sure that your top and bottom faces are all white and yellow (could be mixed), including the centers. Note that white and yellow can be mixed, even the yellow center can be on the bottom.

Oh, duh. My apologies, I was thinking stupidly.

If you want, I can give you all my "algorithms" for Step 4a) (the edge orientation step)

This would be great. I'm watching tutorials, but having algs to reference if need be would be fantastic. A PM would be much appreciated if you have the time.
I very much appreciate the time you've put in to explaining this. I'm getting a better understanding of this step now from posts on here, tutorials, and watching random solves using Roux. It's a very interesting and intuitive method.

 

[GuRoux]

So for the M move is the technique you're referring to is pushing the BD piece to BU using your ring finger?

or middle if after an M2 for me.

 

[Bestsimple]

I don't understand you u do M2 M?

 

[Renslay]

My Step 4a) edge orientation algorithms

First you need to understand that the most basic EO case is when you have 3 unoriented edges on the top and one on the bottom layer (I will always list the unoriented edges):

1a) UL,UF,UR,DF = (M' U M) or (M' U M') or (M' U' M) or (M' U' M'). You can use any of these "algorithms", they all solve the case. I simply refer them as M' U* M*.
Also you must learn the mirrored case:
1b) UL,UB,UR,DB = M U* M*

In other EO cases, most of the time the solution this: bring the actual case to 1a) or 1b). Sometimes there is a short direct solution too.
Note that before the EO alg, you probably need an AUF.

So, here is a bunch of algorithms. The first one which I usually use (most comfortable for my fingertricks), but I listed many alternatives.

2a) UB,UR = M' U' M' + U2 1a) = M' U' M' U2 M' U* M*
2b) UB,UL = M' U M' + U2 1a) = M' U M' U2 M' U* M*
Or you can AUF it to 2a). You can also construct algoritms to UF,UR and UF,UL, but those would contains many Ms instead of M's, which is slower. In that cases, I recommend you to AUF to 2a) or 2b).
alternative (direct solution, OLL):
UF,UR = r U R' U' M U R U' R'
I don't like this, becasue in case of 2a) or 2b), you ends up with U* M* - you can choose between 4 endings, with you can manipulate the next step.

3) UF,UB = M' U' M + U 1a) = M' U' M U M' U* M*
alternatives (mirrored cases):
M' U M + U' 1a)
M U M' + U' 1b)
M U' M' + U 1b)
alternative (direct solution, OLL):
R U R' U' M' U R U' r'
Same note here as in 2) alternative.

4) DF,DB = M2 UF,UB = M' U M' + U' 1b) = M' U M' U' M U* M*
alternatives: mirrored cases, similar to 3). Basically, 4) = M2 3).

5a) DF,UB = M' U' M + U' 1a) = M' U' M U' M' U* M*
alternatives: mirrored cases.
5b) DB,UF = M' U M' + U 1b) = M' U M' U M U* M*
alternatives: mirrored cases. Basically, 5b) = y2 5a) y2

6) UF,UL,UB,UR = M' U2 M' + U2 1a) = M' U2 M' U2 M' U* M*

7) DF,DB,UL,UR = M2 6) = M U2 M' + U2 1a) = M U2 M' U2 M' U* M*

8) DF,DB + two adjacent on the top = M2 + AUF + 1a) or 1b)
8a) DF,DB,UL,UB = M2 U' M' U* M*
8b) DF,DB,UB,UR = M2 U M' U* M*
8c) DF,DB,UR,UF = M2 U' M U* M*
8d) DF,DB,UF,UL = M2 U M U* M*
You can also use AUF of course if one of them is much faster for you. For example, 8c) = U' 8b).

9) DF,DB,UL,UB,UR,UF = R U' r' U' M' U r U r' (direct solution)
alternatives: do something like M' U* M* which brings the case into one of the previous ones. For example:
M' U M' + U2 5a) = M' U M' U2 M' U' M U' M' U* M*
But I highly recommend you the direct solution.

 

[Renslay]

I don't understand you u do M2 M?

I think he refers to:
M' = push BD sticker to DF with ring finger,
M2 = do the same as M' and push the new BD sticker to DF with middle finger. Similar to the U2 combo with index finger + middle finger.

 

[GuRoux]

I don't understand you u do M2 M?

If I do an M2 then maybe a U or U' then an M, I do the M with my middle since it's already there. If the previous M move was M' or M, i usually use my ring.

 

[Bestsimple]

Wait, are we both talking about OH?

 

[RubiksJake12]

I've got the method down sort of finally, and the hardest part is definitely the L6E. Any tips?

 

[GuRoux]

I've got the method down sort of finally, and the hardest part is definitely the L6E. Any tips?

keep practicing it, learn to do step 4a with one look, step 4b with one look, and step 4c with one look. And after that, blend them together for no looks.

 

[RubiksJake12]

keep practicing it, learn to do step 4a with one look, step 4b with one look, and step 4c with one look. And after that, blend them together for no looks.

Oh god Alright. I like having a goal because it's easier for me to focus and improve. Otherwise I jump around on different little things. anyway, are the L6E usually the most difficult for people when learning Roux? That and block building I assume. Which do you think is more "important"?