[Help Thread] Roux Discussion and Help

[GuRoux]

cancel [M,U] solution is fastest i think.

 

[PenguinsDontFly]

Or skipping it completely by doing ULUR differently or solving UFUB instead. @TDM style

 

[shadowslice e]

My splits. I'm probably slightly faster at the CMLL that it looks because it's hard to press the spacebar and execute quickly. Any tips? I guess it's going to be the SB.

 

[PenguinsDontFly]

View attachment 6769
My splits. I'm probably slightly faster at the CMLL that it looks because it's hard to press the spacebar and execute quickly. Any tips? I guess it's going to be the SB.

yup, SB.

 

[wir3sandfir3s]

View attachment 6769
My splits. I'm probably slightly faster at the CMLL that it looks because it's hard to press the spacebar and execute quickly. Any tips? I guess it's going to be the SB.

That's not Mac, and definitely not Windows...

 

[shadowslice e]

="3s, post: 1630, member: 35540"]That's not Mac, and definitely not Windows...

No it's Firefox on Ubuntu cause it runs better on the old laptop that I use because I got given it for free so I tend to use it for stuff. The only thing it doesn't do very well is gaming or some things like word or similar but I have another computer which does that well (though technically it's not mine) plus I don't really care if it gets stolen because it's probably worth very much

 

[GuRoux]

View attachment 6769
My splits. I'm probably slightly faster at the CMLL that it looks because it's hard to press the spacebar and execute quickly. Any tips? I guess it's going to be the SB.

second block a lot.

 

[Daniel Lin]

How exactly do you influence EO during CMLL?

If all edges are oriented you preserve the EO. And if you get an arrow, then you preserve the EO. But what should you do if you get a 4flip? Leave it alone, or do an OLLCP to flip em all?
And what about one on top one on bottom? should you do M2 OLLCP? Or force an arrow???? what do you do in each specific case?

 

[shadowslice e]

How exactly do you influence EO during CMLL?

If all edges are oriented you preserve the EO. And if you get an arrow, then you preserve the EO. But what should you do if you get a 4flip? Leave it alone, or do an OLLCP to flip em all?
And what about one on top one on bottom? should you do M2 OLLCP? Or force an arrow???? what do you do in each specific case?

The way most people do it (if they do it)is to have 2 different algs which have different effects on EO. There's no real hard and fast rule but the key word here is influence as opposed to control as in partial edge control (like in VHLS or ZBLS). What people tend to do is their fastest alg unless it leads to a 6-flip or destroys an arrow or EO skip in which case they do the other alg (unless they both destroy in which cases tough). Many second algs from what I've seen are the COLLs for the cases which is nice as they preserve EO. Personally, most of my alternate algs are only very slight variations such as f R U R' U' f' as opposed to F R U R' U' F' etc so I don't go to the trouble of learning entirely different algs usually.

 

[Teoidus]

4c recognition w/ 1 look, tracking BU recommended but not absolutely necessary

Don't AUF ULUR over to their proper places. After placing them with an M move, look at:
Non-U sticker of the edge that will end up at BU after solving ULUR (going to refer to this sticker color as a "filter" color C)
F center (will just call this F)
FD sticker (= FD)
To disambiguate some other cases, we'll also look at the U stickers that will be on the M slice after AUF'ing to solve ULUR. Let:
~UB = sticker about to be at UB
U = U center
~UF = sticker about to be at UF

Now:
if C == F == FD, cancel the AUF into U2 M' U2 M. Example: alg.cubing.net;
else if C == FD, the first moves are M2 U2 (the rest is pretty easy to see; it'll end up being something like M2 U2 M' U2 M' or M2 U2 M' U2 M). Example: alg.cubing.net;
else if C == F (=> C == sticker that will be at FU), look at U stickers:

if ~UB == U, M' U2 M U2. Example: alg.cubing.net;
else if ~UF == U center, M U2 M U2. Example: alg.cubing.net;​

else if !(C == F || C == FD), look at U stickers that will be on the M slice after AUF

if ~UB == U, M' U2 M' U2. Example: alg.cubing.net;
if ~UF == U, M U2 M' U2. Example: alg.cubing.net;​

Skips, bars, columns, dots should be obvious. Not solving ULUR straightaway also gives you time to easily cancel into all that stuff.

 

[Daniel Lin]

What people tend to do is their fastest alg unless it leads to a 6-flip or destroys an arrow or EO skip in which case they do the other alg (unless they both destroy in which cases tough).

what if it's an M2 away from an arrow? Should you preserve EO and then do M2? or ignore EO?

Personally, most of my alternate algs are only very slight variations such as f R U R' U' f' as opposed to F R U R' U' F' etc so I don't go to the trouble of learning entirely different algs usually.

Yeah I know you don't have learn lots of entirely different algs, but there must be an optimal way of doing things, right?

 

[Teoidus]

What I've seen most people do is:
- have two sets of CMLLs, so you can get rid of 6flip cases
- know exactly what these CMLLs do to EO, so you don't have to pause before LSE

So ~84 algs makes you immune to 6flips and eliminates any pauses after F2B--not bad at all really.

If you happen to have UL and UR close to oriented + placed on DFDB, you can do Pinkie Pie (OLLCP, max 9 move LSE)

 

[Umm Roux?]

What I've seen most people do is:
- have two sets of CMLLs, so you can get rid of 6flip cases
- know exactly what these CMLLs do to EO, so you don't have to pause before LSE

So ~84 algs makes you immune to 6flips and eliminates any pauses after F2B--not bad at all really.

If you happen to have UL and UR close to oriented + placed on DFDB, you can do Pinkie Pie (OLLCP, max 9 move LSE)

If you adjust the m-slice prior to CMLL, you would not need to learn 84 algs, just a bit of quick thinking

 

[Teoidus]

That doesn't always work; some CMLLs don't affect EO or flip non-m-slice edges

But yeah it's true that many "alternate CMLLs" are really just CMLLs with wide moves instead of regular moves (i.e. what you were saying with conjugation with M/M')

 

[Daniel Lin]

ok I see. I guess I'll just learn one more CMLL set

About what percent of CMLLs solve a different case when conjugated with an M? What about M'?

 

[Teoidus]

I don't know for sure, but it's a lot. If you look at PDF's CMLL doc he lists an alternate alg for every case, and many of them are just the previous alg with wide moves inserted (quick example: F R' F' R U R U' R' -> F R' F' r U R U' r' -> tada, can preserve EO in caseyou see an easy arrow case or something, wide sunes do all sorts of flips, etc)

you can also probably use most of the ZBLLs that you've learned that are nice and short since you know for sure that they preserve EO

 

[Jlvs2run]

How would you solve this, and your comments please.
U' M U' M' U' M' U M U M U M' ~ scramble
I'm following the guides, but 32 turns for LSE seems like way too many.

I just now tried this again, and it's 13 moves for me now, nice to see the difference.
M' U M' U' M U' M' // EO
U2 M' U2 M // LR
U' M2 // L4E

 

[Cryoo]

Here are my splits. I'm surely slightly faster at CMLL/CPEC that it looks because it's hard to press the spacebar and execute fastly. Do you have any tips? I think my worst step is FB

 

[GuRoux]

work mostly on first block and second block, maybe lse.
fb: try to plan the whole step in inspection.
sb: really just need to do a lot of solves to improve lookahead.

 

[Cryoo]

work mostly on first block and second block, maybe lse.
fb: try to plan the whole step in inspection.
sb: really just need to do a lot of solves to improve lookahead.

What do you actually mean by " maybe LSE " ? In case you didn't know, i'm Thibaud Ou from Facebook