Random Cubing Discussion

[LearningCode]

LearningCode: the averages are affected by skipped cases; should this work? You can get fairly drastic changes just by skipping one case with a fast/slow time.

EDIT: I mean your CLL trainer btw

I know =x
So it's better if I don't let skipped cases affect it?

[EDIT]
A simple fix, really xD
I changed one line of code, skipped cases no longer affect average times ^^

[EDIT=2]
Anyone think I should extend this thing for other sets of algorithms?
(Like OLL, PLL, Ortega, ELL)

 

[Kirjava]

By wich way/methode did you recognize the corner case so fast

I just use my normal recognition and if a corner is from a different layer I invert the F/B colours.

 

[oll+phase+sync]

If anyone was ever wanting to be a non-matching block user with ZZ or Petrus but didn't want to memorize the 83 NMCOLL cases then have awkard EPLL, or use OCLL then struggle with PLL recognition, here are sequences for something that makes more sense.

http://www.athefre.110mb.com/NMLL.html

As with Roux, the only advantage would come from the blocks. You better be good at blockbuilding and be able to take advantage of the lucky cases, else this isn't any better than regular blocks with OCLL+PLL.

As far as I understand I have to figure out the corner orientation case by myself first (are there any tricks as this schould be the difficult part EDIT2: Understood now not difficult at all just as normal OCLL plus option for opposit colors.
When I got the A-G case , I just compare F/B L/B ...



EDIT Now the Cube is displayed correctly (had to invert my input) EDIT2 but this cube exsample does not belong into this methode!


When I scramble like this R2ULU'RUL' U2 I know I have to do a Niklas and F/B should tell me this. I just don't see what F/B measn

 

[Athefre]

It's not about orienting U colors. Opposite orientation (R2 block) is easy to recognize. I don't know of any tricks for recognizing U colors when you have "adjacent" blocks (R block). It's difficult because maybe the U colors on the left should be yellow but the ones on the right should be oriented to be green. Which yellows and which greens do you choose to orient? There are three of each. So you have to spend your time looking for the two orange+yellow corners and the two red+green corners. Too time consuming.

On the page I posted, you forget the U colors and instead orient L/R colors to L/R. That includes edges. Then you permute the L+M+R layers.

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

EOLine - DU'F'UL'U'B'D2
F2L - R'L2UL2UL' RUR2U2R'URUR'
Separate - U' RU2R'U'RUR'U'RU'R'
Permute - U2 lzR2x'U2rUr'U2lU'l y'U2M'

B L R2 B2 R U F B2 U L' B' F2 U' R' L U L2 R B2 U' D' B' D' U R'

EOLine - UL'U'BR'U2R2D
F2L - UL2U'L2 ULU2 D'F2D R2U'R'U'R
Separate - U2 R'U2RULU2R'U'RU2L'
Permute - U' FR2U'L'UR2U'LUF' U'R

R2 B L2 R2 B2 D2 U F D' F' U' L' U L D2 F R F2 B' L R' F B' L F2

EOLine - U'F'UL'R'B'RD'
F2L - UR'U2 R2URLU L2U'L'U'LU'L'U2 R'ULU'L'
Separate - RU2R'U'RU'R2U2RUR'UR
Permute - RU'R'URU'LUL'Ux'U2RU2R2 xU2rR2

 

[Athefre]

See any two cases that match? Keeping recognition from being possible without AUF or tilts.

Spoiler

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

 

[Andrew Ricci]

Some MU U Perms. I didn't find them on the wiki, and they are quite nice, so let me know if anyone uses them.

Ua: M' U2 M U M' U2 M U M' U2 M

Ub: M' U2 M U' M' U2 M U' M' U2 M

 

[JonnyWhoopes]

Some MU U Perms. I didn't find them on the wiki, and they are quite nice, so let me know if anyone uses them.

Ua: M' U2 M U M' U2 M U M' U2 M

Ub: M' U2 M U' M' U2 M U' M' U2 M

I've been using these for about 3-4 months. They're very intuitive.

 

[Andrew Ricci]

I've been using these for about 3-4 months. They're very intuitive.

Yeah, I thought someone (or multiple someones) would be using them/know them. I found them just by messing around with M moves.

 

[Reinier Schippers]

I know them but I prefer a 2gen RU alg I also like to play around with M's i dunno why

 

[Kirjava]

Yeah, I thought someone (or multiple someones) would be using them/know them. I found them just by messing around with M moves.

All MU EPLL is just Roux.

 

[Cyrus C.]

What do you guys think of learning EG-1 before CLL? If you solve one face, the probability of having an adjacent swap is 2/3, solved is 1/6. Wouldn't it make sense to learn EG-1 first and get 2/3 of the cases out of the way?

 

[JonnyWhoopes]

All MU EPLL is just Roux.

All MU ELL is just Roux.

 

[Kirjava]

All MU ELL is just Roux.

pureflips.

 

[JonnyWhoopes]

pureflips.

Granted.

 

[Escher]

What do you guys think of learning EG-1 before CLL? If you solve one face, the probability of having an adjacent swap is 2/3, solved is 1/6. Wouldn't it make sense to learn EG-1 first and get 2/3 of the cases out of the way?

Do it.
It's not as easy because no OLLs overlap with EG-1 as they do with CLL so you have 'more' to memorise, especially if you aren't familiar with COLL recognition already. It's also probably not as fast on it's own, but don't take my word for it.

 

[qqwref]

If you're sure you want to learn EG, it's more useful to learn EG-1 first than to learn CLL first.

 

[miotatsu]

on the topic of that human kociemba idea discussed back a ways in this thread: I wonder how many moves/how fast that could be
also while I'm at it does anyone have algs for the "Advanced Human Thistlethwaite" method by pseudoprogrammer/grant slatton? I have been trying to get in contact with him but he doesn't seem to go online very often...

 

[qqwref]

on the topic of that human kociemba idea discussed back a ways in this thread: I wonder how many moves/how fast that could be

a minimum of 37 moves and 12.86 seconds, it's impossible to go faster or more efficient

 

[miotatsu]

well I am more interested in what kind of move count you could probably average with it, not the absolute minimum

 

[qqwref]

OH (or 2h?) trick to put in an unconnected pair while flipping edges:

(U2) r U' r' U' r U r'