Random Cubing Discussion

[RyanReese09]

Sounds good! I'll race you to sub-30 if you're down for it Right now I can't even pull off consistent sub-60. I'm also going full color neutral with Roux. I figure it's time I did it anyway, and learning a new method is a great excuse to try.

I'm going with white/yellow on my top/bottom with any sort of horizontal holding position (aka YG, YO, WB, WG)

I'm ~ 40. Times vary so much depending on how fast I can recognize a good way to blockbuild.

Let the race commence >.

 

[DavidWoner]

Sounds good! I'll race you to sub-30 if you're down for it Right now I can't even pull off consistent sub-60. I'm also going full color neutral with Roux. I figure it's time I did it anyway, and learning a new method is a great excuse to try.

Kir would tell you to keep white/yellow fixed on L/R. He says it makes finding/placing L/R edges during LSE much easier.

 

[oll+phase+sync]

Kir would tell you to keep white/yellow fixed on L/R. He says it makes finding/placing L/R edges during LSE much easier.

I also prefer w/y at L/R and think it's much more beginenr friendly, but as far as I know Waffle uses w/y at U/D.

personally when I tried to do Roux with with yellow on top, I preferd to first place UL or UR and do second edge plus EO watermanish.

Is there a big green opinion on this?

 

[oll+phase+sync]

I tried to learn some fancy stuff for the non matching blocks Roux case ( first block D-color is differnt from second block d-color)

Doing Roux this way is great because you have so much more freedom for your blocks

Currently instead of fancy I go for simple by using TOTAL DISORIENTATION:

Wich means I use one out of seven standard CO algs (Sune, Bruno, FSexyF', ...) to make my L/R colors face up. (this is total disoriented now)

But there only 4 cases:

F (RUR'U')x3 F'
R2URUR'U'RUR'UR (or what ever DoubleSune variation you like)
RU'L'UR'ULUL'UL (Pi case with bars on top)
R'F2 y'U R2U'R2U'R2 yR (PI case with X on top)

Recog in the same order (a same color pair on the side must have same U-color, too):
2 same color pairs
no same color pairs
1 same color pair
X case

Edit: the only thing wich makes this methode not absolut beginner friendly, is that EO gets a little more difficult with non matching blocks.

 

[NeedReality]

I also prefer w/y at L/R and think it's much more beginenr friendly, but as far as I know Waffle uses w/y at U/D.

personally when I tried to do Roux with with yellow on top, I preferd to first place UL or UR and do second edge plus EO watermanish.

Is there a big green opinion on this?

This is actually kind of odd, because I messaged BigGreen asking him about color orientation earlier today.

I personally use Y/W for U/D, though I'm wondering if switching to fixed side colors would be better (or if fixed top/sides together is too restricting). I found it a bit easier to transition from Fridrich with Y/W on U/D because that's how I did F2L and such, so it was easier to see how pieces matched together for blocks.

 

[Cubenovice]

If even Chris is getting into Roux I have no excuse not to try it.
It is the only “main” method I have not looked into yet…

Regarding the non-matching blocks:
Inspired by the usage of pseudo blocks in FMC I have been using these more and more in CFOP, ZZ and Petrus solves over the last couple of weeks.
I find it makes solving more challenging and thus much more fun.

Double face turn pseudo blocks work just fine for EO and LL recognition but the quarter turn pseudo blocks take some more effort.

 

[riffz]

Regarding the non-matching blocks:
Inspired by the usage of pseudo blocks in FMC I have been using these more and more in CFOP, ZZ and Petrus solves over the last couple of weeks.
I find it makes solving more challenging and thus much more fun.

:\

 

[Athefre]

Kir would tell you to keep white/yellow fixed on L/R. He says it makes finding/placing L/R edges during LSE much easier.

You mean Gilles originally would tell someone to use white/yellow on L/R.

 

[Athefre]

Edit: the only thing wich makes this methode not absolut beginner friendly, is that EO gets a little more difficult with non matching blocks.

The only solutions I've found are:

1.

a. At the end of the second block, since it usually ends in R/R' always adjust the centers to match that block (or the opposite block if you choose).
b. During Step 3, or even during the second block, watch the edge opposite of the the block you adjusted the centers to. So if you adjusted centers to match the right block, watch the UL edge. If you adjusted the centers according to the left block, watch for the UR edge. This edge you are watching is the one that makes non-matching orientation more difficult to recognize. If you know where it is, then you've solved the problem.

2. Cubacca's second LSE idea. And various other LSE methods that don't focus on orientation.

 

[oll+phase+sync]

1.

a. At the end of the second block, since it usually ends in R/R' always adjust the centers to match that block (or the opposite block if you choose).
b. During Step 3, or even during the second block, watch the edge opposite of the the block you adjusted the centers to. So if you adjusted centers to match the right block, watch the UL edge. If you adjusted the centers according to the left block, watch for the UR edge. This edge you are watching is the one that makes non-matching orientation more difficult to recognize. If you know where it is, then you've solved the problem.

Great! and simple !! I can't beleave I never thought of it. I always tracked both UR and UL edge instead of just one.

 

[miniGOINGS]

Just got a 16.58 while warming up (6th solve of the day, I average low 20s).

F2 B2 D' B R L2 F' B' L' F2 U2 L' D2 L' U' L2 B2 F L F2 R L' D' B2 F
x' y2
U R' U2 M F u' M2 u (8)
r' U' R2 U' M U r U' r' U' M' U R' U2 R (15/23)
R U R' U R U r' F R' F' R (11/34)
U M' U M' U2 M' U M' U2 M' U2 M' U M' U2 M2 U2 M' (18/52)

Sad thing is, my blocks were pretty bad, and I just found a 10 move CMLL alg for that case. >.<

Some how the L6E was a lot of moves, but had no U' or M, so it was faster than normal I guess.

 

[JackJ]

Neat-o solution... blocks came out of nowhere. O_o

U2 R' U B' U L B' D2 U' F R2 L2 D F D U B D' B F2 L2 R2 F' B L

x2 F U2 R' F R U
y z' x' F' R U R' L' U' L
F U' y2 x' R U' R' U' R U R' U2 R U' R'
F' r U R' U' r' F R
U R U' R U R U R U' R' U' R2 U' F2

Time: 18.86. Pretty good.

 

[cmhardw]

Can anyone confirm this from experience? Does anyone else do this?

I've noticed that fairly often in my Roux solves that at the very end I come down to a commutator 3 cycle during LSE. Most of the time this happens after orienting edges, but my current pb single solve (31.91) it happened immediately after finishing corners.

Looking at it, there are two scenarios where you can end with a commutator 3 cycle, rather than 1) orient edges 2) place L and R edges 3) permute the rest.

Situation 1) After having done F2B and U layer corners, you have a commutator 3 cycle (Probability: 1/72)
Situation 2) After having done F2B, U layer corners, and orienting the last six edges you have a commutator 3 cycle (Probability: 1/9)

Therefore the probability of ending with a commutator 3 cycle is:
\( \left(\frac{1}{72}\right)+\left(\frac{71}{72}\times\frac{1}{9}\right)=\frac{10}{81} \approx 12.34\% \)

So it's likely that in an average of 12 solves, that at least one solve will end in a commutator 3 cycle rather than the steps I listed above.

Does anyone else do this during Roux? Is this worth pursuing further?

 

[Athefre]

Can you give some examples?

 

[DavidWoner]

Does anyone else do this during Roux? Is this worth pursuing further?

http://www.youtube.com/watch?v=0NKjizrcGUs

The difficulty for most people is recognizing them quickly enough, especially if they are not oriented.

 

[oll+phase+sync]

Sometimes I spot a 3 cycle after orientation, but most most of the time I ignore them. (I don't use direct commutator solving for these cases, though)

If UR/UL are placed befor orientation 3-cycles apear quite often and it is esay to spot.
But I can only regcognize, the correct 3-cycle fast, if centers are already aligned.

 

[cmhardw]

@Athefre: In my solve what happened was that after solving F2B, and then corners, before beginning LSE, I noticed that a lot of stuff was solved. I aligned the centers with an M layer turn, and noticed that I was down to a 3 cycle of edges. Since I am still very much a beginner with Roux (and dare I say it an expert with commutators) I solved the cube with: x' z' E R U' R' E' R U R' and done.

I figured that I could orient edges, place L/R edges, then permute the M layer edges. But, at this stage at least, my recognition and alg knowledge is so minimal for Roux that it would be much faster to just solve the cube directly in 8 turns

I do understand that for those at a much higher level in Roux solving, that the recognition could be much slower and also with any necessary cube rotations and such that this method could actually end up being slower. For now I will use commutators when I see them, but if I end up sticking with Roux long enough to get sub-30 at it, I may try to move away from this for a more "proper" Roux style ending.

@Woner, @oll+phase+sync: Thanks for the responses. I'm gathering that using comms at the end of a Roux solve is ok only if it doesn't kill your recognition completely, or slow you down over the "proper" Roux ending. At last for now, ending with a comm is definitely preferable for me over ending with the proper Roux steps, as I am still so new to Roux solving in general.

 

[StachuK1992]

Is there any reason we don't have a notation for A B A? I suggest [A| B] for that, and for A B A B (I've yet to see a case for this) I guess [A|| B] could work.
Example:
[R U' R'| U2 F' U' F U] =
R U' R' U2 F' U' F U R U' R'

not much shorter, but it helps me memo.

 

[cmhardw]

Is there any reason we don't have a notation for A B A? I suggest [A| B] for that, and for A B A B (I've yet to see a case for this) I guess [A|| B] could work.
Example:
[R U' R'| U2 F' U' F U] =
R U' R' U2 F' U' F U R U' R'

not much shorter, but it helps me memo.

Per had a post from a while back talking about this.

There's also a great post made by someone, I think Lucas or Michael, about how to extend this to double conjugations. I can't find it at the moment, but I'm still looking.

 

[StachuK1992]

Per had a post from a while back talking about this.

There's also a great post made by someone, I think Lucas or Michael, about how to extend this to double conjugations. I can't find it at the moment, but I'm still looking.

I think a notation for A B A would be pretty useful. I'm going to just use the notation I've been using for now.

Thanks for the link.