Advanced Roux Techniques

[PenguinsDontFly]

This is a thread in which we discuss existing advanced techniques for the roux method, question their usefulness, expand on ideas that already exist, and create new ones.

This video briefly explains 6 advanced techniques that already exist:

 

[shadowslice e]

There's also one special arrow case which is useful where either UL/UR or UF/UB are on the "point" of the arrow and D where you can do M then AUF so it's in the right place. It's sort of like misoriented centres with extra skips. I'll make a video tomorrow (hopefully) if this was a bit unclear.

Also, experiment with UL/UR inserts to get nicer 4c cases.

 

[TDM]

I was thinking of making a video of every LSE trick I could find in the summer, to try to get everything into one place - from basic ULUR insertion tricks to things like misoriented centres and BU prediction. Not started it yet because of exams, but do you want me to post it here once I'm done?

 

[PenguinsDontFly]

I was thinking of making a video of every LSE trick I could find in the summer, to try to get everything into one place - from basic ULUR insertion tricks to things like misoriented centres and BU prediction. Not started it yet because of exams, but do you want me to post it here once I'm done?

Yes, definitely! That would be awesome!

 

[TDM]

Yes, definitely! That would be awesome!

Do you think one big video would be better or a series? Also, what do you think I should include? Here's my list so far:

Spoiler

ULUR with M' U2 M' or M U2 M
Influencing ULUR during EO
OLLCPs/multiple algs per CMLL
BU prediction
Easy 1-look edge permutation - L5E (EP5)/EPLLs/Parity EPLLs
Misoriented centres
(Skipping dots? Don't know if I should include this or not, it's not useful for 2H)

It's a bit short, is there anything else you can think of?
Also, the way I do EO+ULUR I never get both ULUR edges in the U layer so I won't ever find your UFUB method useful without an EO skip. I don't know how to put that part in the video(s) without contradicting myself.

 

[PenguinsDontFly]

Do you think one big video would be better or a series? Also, what do you think I should include? Here's my list so far:

Spoiler

ULUR with M' U2 M' or M U2 M
Influencing ULUR during EO
OLLCPs/multiple algs per CMLL
BU prediction
Easy 1-look edge permutation - L5E (EP5)/EPLLs/Parity EPLLs
Misoriented centres
(Skipping dots? Don't know if I should include this or not, it's not useful for 2H)

It's a bit short, is there anything else you can think of?
Also, the way I do EO+ULUR I never get both ULUR edges in the U layer so I won't ever find your UFUB method useful without an EO skip. I don't know how to put that part in the video(s) without contradicting myself.

I think a bunch of videos and just discussion on this thread would work.

Thats a good list. I can't seem to think of anything else.

 

[TDM]

I think a bunch of videos and just discussion on this thread would work.

Thats a good list. I can't seem to think of anything else.

Cool, I think more shorter videos would be less work since I can plan each one separately.

Alright, I'll start some time next week then. Can anyone else reading this think of something to include?

 

[IuriG]

Do you think one big video would be better or a series? Also, what do you think I should include? Here's my list so far:

Spoiler

ULUR with M' U2 M' or M U2 M
Influencing ULUR during EO
OLLCPs/multiple algs per CMLL
BU prediction
Easy 1-look edge permutation - L5E (EP5)/EPLLs/Parity EPLLs
Misoriented centres
(Skipping dots? Don't know if I should include this or not, it's not useful for 2H)

It's a bit short, is there anything else you can think of?
Also, the way I do EO+ULUR I never get both ULUR edges in the U layer so I won't ever find your UFUB method useful without an EO skip. I don't know how to put that part in the video(s) without contradicting myself.

I think skipping dots is useful for 2H, I use it quite a lot.
And I aways get at least one of the L R edges on the last layer, but I still find UF/UB useful on OH, since I have time on OH to look for both LR/FB edges and influence the best, sometimes I can chose both being in the botton over only one, or even a skip.

 

[IuriG]

And it's useful on 2H, but because I'm actually neutral betwen both, so I just solve what I see first. So it's a look ahead advantage, I guess.

 

[obelisk477]

So was 6-flip CMLL a bad idea?

 

[PenguinsDontFly]

So was 6-flip CMLL a bad idea?

No, I just forgot about it. Sorry. Well if you think about it, 6 flip CMLL falls under the "influencing EO during CMLL" category.

 

[obelisk477]

No, I just forgot about it. Sorry. Well if you think about it, 6 flip CMLL falls under the "influencing EO during CMLL" category.

I didn't mean there should be a video for it lol. I'm just not convinced the algs are worth it. Maybe they are, but I feel like you can't predict positions of UL/UR as well

 

[ottozing]

Not sure why you did u' M u for that F2B minus centers example instead of u M' u', but whatever

Also, what are your thoughts on the thing where you solve F2B with UL/UR (or UF/UB) oriented in the D layer and then do an OLLCP into trivial LSE? I remember seeing Alex say it was a good direction for Roux to go in, and since you're one of the people picking up where Alex left off in terms of pushing Roux I was curious as to what you thought

Cool video too btw

 

[PenguinsDontFly]

Not sure why you did u' M u for that F2B minus centers example instead of u M' u', but whatever

Also, what are your thoughts on the thing where you solve F2B with UL/UR (or UF/UB) oriented in the D layer and then do an OLLCP into trivial LSE? I remember seeing Alex say it was a good direction for Roux to go in, and since you're one of the people picking up where Alex left off in terms of pushing Roux I was curious as to what you thought

Cool video too btw

Woah that is so simple...why do you always find these?

I talked to Alex about it and he says he thinks it is good, but I would have to disagree. I tried learning OLLCP, but gave up part way through once I found out it wasn't worth it. My main argument for this is that the moves you waste orienting and placing the ULUR edges on D are the same moves you would do if you solve them after OLLCP. Now we just have 1) orient 2 edges on D and 2) OLLCP. My problem with this is that you are still doing extra moves before corners, and the OLLCP algs are worse than CMLL. Thus, you are wasting time in second block and CMLL just to orient edges, which is arguably one of the easiest parts of LSE. I think the best CMLL advanced thingies we have at the moment are CLS, CMLLEO, and the "learn 2 algs for each case" method.

 

[TDM]

Also, what are your thoughts on the thing where you solve F2B with UL/UR (or UF/UB) oriented in the D layer and then do an OLLCP into trivial LSE?

I'm not PDF, but I don't think this is a good idea. If you're wanting to solve specific pieces into the D layer like that, why not do CFOP but use ULUR for two of the cross pieces? I think that would be faster.

That would then be the same speed as CFOP, but after OLLCP you would be left with something slower than EPLL. So surely this method would be slower than CFOP with OLLCP?

 

[Teoidus]

Since most of the techniques mentioned thus far center around LSE: is there any benefit to simultaneous blocks?

 

[PenguinsDontFly]

Since most of the techniques mentioned thus far center around LSE: is there any benefit to simultaneous blocks?

It can be more efficient in some cases. For example, it is more efficient to solve/preserve a pre-made second block pair with the edge to make preserveinstead of to break it while doing first block.

 

[Daniel Lin]

Good video.
to part 3. Non-matching blocks
CMLL Recognition with Non-matching blocks is not difficult if the blocks at the bottom of the opposite color.
I want to add an explanation. Sure, it's already written, but I have not found.

Spoiler: How to Use Non-Matching blocks are not trained NMCMLL

When the CMLL Recognition add validation sameness stickers belonging to the Down blocks.
Special attention to the stickers Front and Back.

I will call the colors, as in the video: White Down on the FB blue and Yellow on SB green.
Features:
- White and yellow assume same color (for the definition of OLL);
- Color Red and Orange Front and Back - the same or the opposite, depending on the accessory to the left block or right block
= Have the same stickers of the Down unit - the ordinary meaning as in CMLL;
!!!= Different stickers Down (white or yellow sticker). At various stickers (one has a white sticker and the other yellow sticker) - the value labels are reversed. If two green - believe their "opposite", and if the opposite color stickers (one green and the other blue) consider them the same.
Example of Video https://www.youtube.com/watch?v=L7pIPmSo24c&feature=youtu.be&t=211

:

Red with white sticker and orange and yellow sticker - the same color (this Fronts different blocks);
Blue and Green - the opposite (Left, Right - do not depend on white, yellow)

need to do

R’ F’ R U2 R’ F2 R U’ F U F’

Perhaps it will help to use "Non-matching blocks"
Sorry for my English

If the blocks are off by an R2 I can just do it intuitively. But for the hard to recognize ones I use athefre's method

 

[Shiv3r]

the 6-flip CMLL would be great, I think that it could be learned intuitively however with the use of commutators before and after algs, which I do sometimes to force a 4-flip