Any more roux?

[5BLD]

Ok, I am praying that I am not asking something silly, or that has been asked before...
Are there any extra little bits to add to the awesome but simple method that is roux? Like in fridrich, you can 'add on' MGLS, WV, VH, ZB etc. Little variations of the method. Can you somehow influence the second block during the first block? Or even influence both before even starting the blocks? Forcing CMLL skips during blocks? Combining steps? I am by no means an expert at all, but I have been experimenting and I find that making some premoves before the blocks is somewhat useful for reducing move count although it's not particularly practical(long inspection).

 

[Kirjava]

non matching blocks, last six edges hacks, CLL+EO

 

[5BLD]

Lol what you just said inspired/reminded me to have a go at non matching blocks- thanks! What do you mean by LSE hacks? And would CLL+EO be done intuitively with double layer moves and M slices inserted (edit: just realised that it is very hard to do), or would it be algorithm based? I like the idea- it would allow looking ahead for L/R edges and stuff.

 

[waffle=ijm]

roux optimized CLS into KCLL?

 

[5BLD]

But how do you go about doing that waffo? I need to eat more waffles- they help me work out acronyms.
Oh i see what you mean. I bet no ones done this before, have they?
Edit: Ive spotted YOUR little article haha. I'm gonna read it now. Looks interesting

 

[Kirjava]

And would CLL+EO be done intuitively with double layer moves and M slices inserted (edit: just realised that it is very hard to do), or would it be algorithm based?

Alg based and based on 'tweaking'. starting an alg with r instead of R to flip some edges, etc.

As for LSE Hax, there are a lot of random ones dotted around the forums. Mostly for forcing skips.

 

[5BLD]

Ahh I see what you mean! So like doing an M' before a sune would flip these (points to imaginary cube) edges because of the way the algorithm permutes the edges!It is an interesting concept. This all sounds familiar... Did I read it somewhere here before i joined the forums?
I am going to have a little go now, although I realize there's going to be a large number of cases- probably would learn them by discovering them rather than learning them from an alg sheet.

 

[waffle=ijm]

dotted around the forums.

lolpun?

Optimizing LSE is also a way to go. You don't really need more algs but it is an easy way to get fast.

 

[5BLD]

Yeah that's exactly what I was looking for waffo
I am very lazy...
I noticed that 4a and 4b can be fused together quite easily...

 

[FatBoyXPC]

That "M' before Sune then M because of the way it permutes edges" thing you mentioned is referred to as Fat Sune. You actually get 4 OLL cases out of that I know you don't need to know OLL for roux, but I was just saying.

 

[Cool Frog]

Yeah that's exactly what I was looking for waffo
I am very lazy...
I noticed that 4a and 4b can be fused together quite easily...

As in one fluent motion by looking ahead, or simultaneous (like a skip)?

 

[Kirjava]

U'M'U2M2U2M'U
UM'U2MU2M2U'
UM'U2M2U2M'U'
U'M2U2MU2MU
UM'U2MU
RUR'UM'U2MURU'R'
M2U'M'U2M'U'M2
M2UM'U2M2U2MU'M2
RUR'UM2U2M2URU'R'
RU'r'U'M'UrUr'

Some I just thought up off the top of my head. Learn to recognise them when they appear.

 

[thierce]

I think learning to solve LSE cases more effectively is pushing your speed pretty much

Does anyone else know this phenomenon:
When starting LSE, I sometimes do AUF before turning the M-Slice without being able to give a reason!
Sometimes U, sometimes U', sometimes no AUF at all... But it still occures in some of my solves and I cant come up with a reason or where that habit comes from

Edit:
@Kirjava: I shat bricks when I tried some of the algos! Thanks!

 

[5BLD]

Kirjava: wow, I guess that's what I call 1-look step 4. Did you find these by experience?

Cool frog:I mean both. It is quite common to get the easy 4b 'skip' when one l/r edge is opposite its correct spot, while the other is in the D layer. Or yes, in a smooth motion, too, unlike in fridrich where it is almost impossible to forsee the PLL. But you probably know more than me!

Thierce: yes, I used to do that! Old habits from fridrich. What i do now is instinctively orient the centers- when it might be better not to. Or it might be better to have this instinct. I'm not sure, really.

 

[Cool Frog]

Kirjava: wow, I guess that's what I call 1-look step 4. Did you find these by experience?

Cool frog:I mean both. It is quite common to get the easy 4b 'skip' when one l/r edge is opposite its correct spot, while the other is in the D layer. Or yes, in a smooth motion, too, unlike in fridrich where it is almost impossible to forsee the PLL. But you probably know more than me!

Thierce: yes, I used to do that! Old habits from fridrich. What i do now is instinctively orient the centers- when it might be better not to. Or it might be better to have this instinct. I'm not sure, really.

Those seem like cases I already do, I don't mind misoriented centers because I can usually force 4 edges flipped and have easy cases.
Yeah, my opposite layers seems very smooth. Just tracking the pieces during EO

 

[JonnyWhoopes]

I guess this would be a good place to ask this:

I've recently started to actually practice Roux again. I wanna give that last push and come out sub-15 (the sixth person to do that with Roux?). How does one go about recognizing EO with misaligned centers? Because, every time I fix centers, I feel that I'm missing out on some optimizations that I just don't see.

 

[waffle=ijm]

use the same and opposite rule with the pieces and centers you are given with. exclude the UL and UR pieces and align them so your U color is on top with those pieces.

 

[Athefre]

It's usually only useful in a speedsolve if you are tracking the L/R edges before getting to that step.

It's fun, but difficult.

 

[5BLD]

Yeah, my aim is sub-15, and I'm not far now... But it's as if the sub-20 barrier really is a barrier. As in, some days I go past it into 16-18, some days i get stuck at 20-24.
Is it better to try to go fast with tiny pauses or go slow and smooth? I always have a 1-2 second pause before cmll and that tiny split second before 4c.
For blocks I go slow, but always forget to try to see the cmll case.

 

[Rpotts]

hey a good site for CLL recognition -> david woner and rowan escher's new site

click on the case and it gives multiple recognition angles. Unfortunately it's not finished yet so no L, Pi or H cases yet (though Pi and H are easy to recognize comparatively.)