Waffle's Roux Tutorial

[5BLD]

Ya, I solve Roux. My colours are white or yellow on the sides, and either orange or green on top.

Hm... interesting. Orange or green? Hm... well Stachu has already asked what I was gonna ask.
But what do you average? It might be fun analysing each others' videos and stuff.

 

[oll+phase+sync]

You found the one case that doesn't solve a 3-cycle as a 3-cycle. But it's also a case that really doesn't matter. It only adds one more move to that case, and that is made up for by not having recognition time.

So the solution is: M2UMU2MU2M2U2 compared to the old way of M2U'M2U2MU2M.

On my first try to solve F2 M' F2 M U M2 I even needed 2 more moves M2UMU2M'U2M2U2M2 than the old way, but this case never happend to me in real solves.

But using this in praxis feels really good because silly 'errors' like last 4b move = U first 4c move = U2 won't happen anymore.

 

[5BLD]

Thinking back to nouxb times... Ah yes. My old thread. Stuff that was fired at me and I didn't understand any of it. Well now I do...
Any of you actually tried Roux CLS? Or any of you use KCLL in practice?

 

[waffle=ijm]

Roux optimized CLS+A set of KCLL was something I tried.

 

[5BLD]

Did it work?
I tried both too but for KCLL it seems such a small advantage; and CLS was awkward.
However, the reason could be just because I'm not used to it.

Btw, I did not try them together like you did... That seems even more awkward... But as I said earlier, it might be because I'm not used to it.

Ive also been experimenting with 'setting' up the second block during the first...

 

[Kirjava]

Or any of you use KCLL in practice?

All the damn time.

 

[5BLD]

Meh. So you discovered all these cases over time? Or just used them for a few algs?

 

[Kirjava]

I don't understand you. I know multiple CLLs for each case and use a specific one to give me a good case. If there's an EO skip I'll do COLL, for example. Or sometimes I'll do M2 before CLL to skip EO. Happens a lot more in OH.

 

[Escher]

I use KCLL and I do about 100 Roux solves a year.

 

[5BLD]

I don't understand you. I know multiple CLLs for each case and use a specific one to give me a good case. If there's an EO skip I'll do COLL, for example. Or sometimes I'll do M2 before CLL to skip EO. Happens a lot more in OH.

Ah right I see what you mean now. You know many algs for each case and know how it affects LSE. Well that's what KCLL is... Right, so it's just from ages of experience then...

Which reminds me. I need to go finish off learning COLL for CFOP and Petrus.
And now, Roux... Not that it's necessary but it'll give me an advantage.

 

[Cool Frog]

Or any of you use KCLL in practice?

KCLL
Kirjavas CLL
(LL part confuses me)

Ohya, anyone notice EO is harder to identify on crazy colour schemes.
Waffo has his old avatar

 

[Kirjava]

(LL part confuses me)

The LL part /is/ weird. Blame CMLL.

 

[StachuK1992]

KCLL
Kirjavas CLL
(LL part confuses me)

Ohya, anyone notice EO is harder to identify on crazy colour schemes.
Waffo has his old avatar

Is he not solving the Corners of the Last Layer, the Kirjava way?

 

[Cool Frog]

Is he not solving the Corners of the Last Layer, the Kirjava way?

Nobody is, they are just doing something SIMILAR to the way he does it. nobody can be just like Kirjava, because he is well, Kirjava
(just how you are you, and me is me)

 

[StachuK1992]

I guess I can't do MGLS because I'm not Lucas or Shotaro. Good to know.

 

[Kirjava]

Call it KCEO if it makes you feel better. It really just needed a term so people knew what I was talking about.

 

[blakedacuber]

just out out of curiosity what does the "M" stand for in CMLL?

 

[RyanReese09]

M (slice).

 

[oll+phase+sync]

Roux optimized CLS+A set of KCLL was something I tried.

...Any of you actually tried Roux CLS?

I'm not sure I get the CLS concept right within Roux:

- Step3 get replaced by CLS (M-Slice, EO disruption allowed) -- Is there some "commutator magic" available to do this without so many algs?
- Step4a get replaced by CP + EO in one go ( A subset of KCLL if I'm right) -- Is there a List of algs somewhere ?

Knowing A KCLL would also be a simple and efficient way to handel NMCLL I believe.

 

[5BLD]

Well, for CLS I tried MGLS as I was too lazy to come up with my own algs. Yes commutators are possible though.
Personally I don't think it's particularly practical.

KCLL is vaguely defined so we all have different ways of doing it.
Either learn lots of algs for every case, or as Thom said earlier, know how it affects edges and therefore know which one to use (over time it is inevitable you know multiple algs for each case.)
Now I do the latter; I know COLL too and also know multiple algs for most of the cmlls. So I incorporate it into my solves now by acknowledging the edge orientation and doing a specific one to give me a good case.

To clarify, KCLL is CMLL but influencing EO.