Roux Method?

[waffle=ijm]

and CLL permutes and orients at the same time so probably 7 Fridrich OLLs are vaild

 

[pjk]

Everyone always talks about how Roux using no algs, and Fridrich has tons to memorize. For Roux CMLL, you're looking at 40 or so cases. For full Fridrich, there is 57 OLLs and 21 PLLs. If you learn an alg a day, it really doesn't take long to learn algs. Amount of algs shouldn't scare you. If you're into it, 80, 100, 500 algs is not too many, because you do it for fun.

 

[rubiksfriend]

Uh-oh, betrayed by noob-post-count! XD I've heard being color-neutral in Petrus is a must, but with CFOP, I assume most people have optimized F2L for only certain pieces so being color neutral would be very difficult. Although, if Petrus users (the top ones) can so easily expand to a 2x2x3 block, then finish of the 2 layers and recognize cases, it must be possible with CFOP. If a Petrus user learnt all ZBLL, I think it could be the fastest method out there (does Johannes Laire know some?) As for holdouts, one never knows what a method can accomplish with enough work behind it. Also, brunson, which method are you using?

Waterman face building should go like this(usually): build 1x2x3 block, pair up remaining face corners and edge. Avg. time- 5 sec? (3 for first block, 2 for second?)

 

[AvGalen]

The biggest study (initiated by LarsvdB) showed that Cross-neutrality saved 1 move on average. Some very fast people (like Erik) are already opposite color neutral and would save even less moves

For Petrus being color neutral was more important in saving moves.

And Johannes knows a lot of algs, including many ZBLL

 

[deadalnix]

begining with a 2*2*2 bloc save 1 move on average compared with cross. With no color neutrality, you can have goos results too.

I think the colors must ne the same on OE step (opposite color neutral can be easy here), so you can start with any 2*2* bloc but you have only one choice for extention of the bloc.

 

[brunson]

Everyone always talks about how Roux using no algs, and Fridrich has tons to memorize. For Roux CMLL, you're looking at 40 or so cases. For full Fridrich, there is 57 OLLs and 21 PLLs. If you learn an alg a day, it really doesn't take long to learn algs. Amount of algs shouldn't scare you. If you're into it, 80, 100, 500 algs is not too many, because you do it for fun.

I made it a mission to learn all my OLL cases and it got to be a grind at the end.

Now I've been taking the attitude of, "Learn new algs when I feel like it, but better to learn how the alg works rather than just memo the moves". I'm easing into ZBF2L using that philosophy. I have cases 1, 10 and 13 printed and posted on my wall and I play with them when I'm not goofing off posting here.

Of course, it's causing my OH to drag on perpetually as I'm refusing to learn OH algs for PLL.

 

[waffle=ijm]

OLL is such a drag for me. That's why when I still did Fridrich (Before switching to Roux) I tried using ELL and CLL instead of OLL and PLL.

 

[fanwuq]

My main problem is edge orientation. If FRUR'U'F' didn't take only 1.5 second to recognize and execute, I would have switched to Roux or Petrus. I really don't want to learn the other 20 or so OLLs. I end up wasting 3-10 seconds on that step (often ending up doing the BLD pure flip) when I try these 2 methods. The EO recognition is too slow for me. Block building isn't a problem for me. I often do it in my Fridrich F2L.
For Roux, How many optimal algs do I have to learn for Placing top 2 edges and orient all edges?

 

[deadalnix]

For Roux, How many optimal algs do I have to learn for Placing top 2 edges and orient all edges?

0, you just need to understand what you do

 

[fanwuq]

I do understand. I just end up either M2 BLD or 3cycle BLD to solve them on cases when you can't resolve orientation with one MU'M' or rUR'U'MURU'R' and inverse.
So How to flip one edge on D and one edge on U without messing up corners?

 

[deadalnix]

If you ask this question, you don't understand orientation on roux method.

In simplified version, if you put f or b center on top, you will flip the 4 edges when you do a U or U'.

If you want to orient an U and an D edge, put them on opposites position on the cube using U layer. Then put the F or B center on top. You'll have 1 misoriented edge and 3 well oriented edges. Flip the 4 edges (so you'll have 3 misoriented and 1 well oeiented edge). You are now in a MUM like case

 

[waffle=ijm]

when using roux the orientation is a little bit weird at first. here are the rules to orienting for a roux solve


An edge is correctly oriented if you can solve it with a sequence in the group generated by <U, D, L, R, F2, B2>, i.e.:
- For UL-edge and UR-edge, the orientation is correct if the L/R-facelet is on F/B/L/R-side.
- One of the four other edges is correctly oriented if a facelet matches the adjacent M-center's color or the opposite M-center's color.

Orienting correct can be done using only M/U moves

 

[rubiksfriend]

Recognizing bad edges in Roux is soooo easy! Learn some easy "sequences" on his site, then optimize them and find your own.

 

[waffle=ijm]

I'm so used to orienting. that i can do L and R pieces in while I orient. I don't know how I do it though. It just come naturally.

 

[deadalnix]

I'm so used to orienting. that i can do L and R pieces in while I orient. I don't know how I do it though. It just come naturally.

Yeah, that's the idea. But no alg is needed.

Orientation is easy to recognise in a speed context : all the edges with no U/D color on top are misoriented. For FD, if an U/D color is in F, he is misoriented. You can deduce the orientation of the DB edge by parity.

You must have an U/D center on top for this. In a FMC context, a more complex definition can lead to a better solution, but I think it souldn't used in speed.

 

[waffle=ijm]

i got a sub 15 once (lucky CLL) because i was able to fix the L/R pieces while orienting

2 Blocks - 10.xx seconds
CLL - lucky
M slice orient (with L/R) - 2.xx seconds
M slice perm - 2.xx seconds

 

[Dirk BerGuRK]

Yeah, that's the idea. But no alg is needed.

Orientation is easy to recognise in a speed context : all the edges with no U/D color on top are misoriented. For FD, if an U/D color is in F, he is misoriented. You can deduce the orientation of the DB edge by parity.

You must have an U/D center on top for this. In a FMC context, a more complex definition can lead to a better solution, but I think it souldn't used in speed.

Hey deadalnix,
I would agree with you to a certain extent. Its important to recognize some lucky cases in step 4 because they are worth looking for. Check out Gilles optimization section under step 4 if you haven't already. Cases like 1a are my favorite and important since all the orientation cases have to go through that stage. As with all the steps, a little imagination goes a long way.