The New Method / Substep / Concept Idea Thread

[Thermex]

Howdy all,
There's only 58 algorithms for each set of Ribbon Orientation (ROLL+ & ROLL-) when you account for symmetry. Add in the 72 from TTLL, and there are 188 new algorithms for someone who already knows full CFOP.

I should be able to make the algorithms soon, starting with the + cases, so expect those to come out in the next few days! It's my first time generating algs, so if anyone would like to help or can give any tips, that would be great!

One thing I forgot to mention that's very important: the ROLL/TOLL (which should we name it?) cases where the edges are already oriented are just TSLEs, so there's no need generating those. 188 cases is not a rediculous amount of algs to learn for someone looking to get faster, especially compared to ZBLL or even Neuro and Aerma's proposed method, which I believe is around 215 additonal algs (correct me if I'm wrong).
Also someone who knows both full CFOP and full CT has to learn EXACTLY 100 additonal algs, which is pretty cool. I'm not too worried about the recog, as long as you know what way the DRF corner is twisted after you solve the ribbon the U-face recognition shouldn't be too bad. It's pretty much the same as OLL recog, but there's more cases and of course they look different.

My main tips for cube explorer:
1. Make sure you look through as many algs as possible, don't just stop when you a find a decent 12 move alg towards the top of the screen. There's often some alg down there that's more ergonomic than the first one you go with.
2. When you're really stumped and can't find ANY good algs, put an x or z rotation on the cube and generate some more. You'll sometimes find something really good there.
3. When you set up a case for the FIRST time and gen algs for it, set it so it's only R U and F moves. If you don't find anything there, add in L moves. Then maybe go for a z or x rotation if that's not working.
4. Don't just go for the lowest movecount alg, go for the one that is the most ergonomic (but still pretty short). Always try to avoid string of moves like (R, F) (F, U) that require regrips.
5. As often as possible try to include multiple algs for a case (I'd say I do this for a about a fifth of the algs I've generated for Twisty EG). If you've found two algs and can't tell which is better, time yourself using them.

Okay, so I could probably figure this out myself, but does anybody know how me and JTay would set up these cases? How do you make Cube explorer ignore the permutation of the DRF corner but solve the orientation of all the pieces?

 

[TDM]

Okay, so I could probably figure this out myself, but does anybody know how me and JTay would set up these cases? How do make Cube explorer ignore the permutatiin of the DRF corner but solve the orientation of all the pieces?

 

[shadowslice e]

stuff

Call it TOLL as ROLL already refers to something else.

 

[Thermex]

@TDM thanks for that. That helped.

@JTay so the way we're gonna gen these algs is by pressing Control+Left click, then blanking out every color that's not yellow.

 

[Shiv3r]

@TDM thanks for that. That helped.

@JTay so the way we're gonna gen these algs is by pressing Control+Left click, then blanking out every color that's not yellow.

Hey, @JTay it's Justin Taylor! didnt know you had an SS account
Anyway @Thermex I need more help with better NLL algs, if anyone wants to help it would be appreciated.

 

[JTay]

Update on Ribbon:
I've decided to name Ribbon Orientation as Twisted Orientation of Last Slot. I have finished generating algorithms for all of the TOLS+ set, along with corresponding case images. The cases are sorted by corner orientation for columns and by edge orientation for the rows. Note that any of the corners could also be replaced by their opposite face's color, as the method does not care which corner is in FDR. The majority of the cases are very nice, with most being quite fingertrick friendly. I believe that average movecount for speed optimal algs is just under 11 moves. There are a few that I could not find a pleasant and efficient algorithm for, so if anyone could help find a replacement for them, that would be awesome.

https://docs.google.com/spreadsheets/d/17FsGpkiqPVtN3zUUhbIfG2AkHbUyNhmQIepc95iBWbQ/edit?usp=sharing

My next project is adding the corresponding TOLS- algorithms and generating their images as well. I'm not sure if I'll generate OLL algorithms with no regard for FDR, as I'm sure only a handful of cases would benefit. Let me know what y'all think of it!

 

[Neuro]

40 move solve using BFTT (2x2x3 Block, F2L-1 corner, Twisty Orientation, TTLL) name suggestions? Scramble from 3x3 example solve thread

Scramble: B2 L2 D' F2 D R2 U2 L2 U' F2 D' B' R2 D L' B D R' B R2 B

x2 L M2 B' M2 F//FB (5/5)
M U r' U' r2//2x2x3 (5/10)
F R' F2 U' F U' R2//F2L (accidentally solved corner) (7/17)
U R U R' U' B' R' F R F' B//Orientation (11/28)
R U R' y' R2 u' R U' R' U R' u R2//Permutation (12/40)

 

[Shiv3r]

Hey guys, we should talk about 4x4 and 5x5 and maybe even bigger methods. Here are some things I came up with recently:

4x4/5x5 method: Use Meyer or Meyer5, then after edgepairing put the middle centers a u2 away from solved, and keyhole in BL and BR, then you have a solved 2x2x3 in the back. you can then put in the final cross edge and do CFOp with a double xcross. it's Pootris4, but with the solved block on D.(and if ya hate on Pootris, one of Felik's solves in his former WR average was pootris)

6x6 I came up with a way to do L4E on 5x5 pretty quickly, which I made a video on here.

5x5: Now I also came up with a way of doing Yau5 centers really quickly(on 5x5-- on 6x6+ doing bars is better). you blockbuild one 2x3 around one center, then hold the 2x3 on the left, and then do either an r or r'(if it was a 3x3) to bring up another center, blockbuild another 2x3, then do the same move as you did the first time, blockbuild another 2x3, then do the same move(r or r') as you did before, then finish the centers. I usually finish by doing the D center first then the 2 side centers, but you can do the last 3 bars in any order.

questions/thoughts?

 

[Teoidus]

40 move solve using BFTT (2x2x3 Block, F2L-1 corner, Twisty Orientation, TTLL) name suggestions? Scramble from 3x3 example solve thread

Scramble: B2 L2 D' F2 D R2 U2 L2 U' F2 D' B' R2 D L' B D R' B R2 B

x2 L M2 B' M2 F//FB (5/5)
M U r' U' r2//2x2x3 (5/10)
F R' F2 U' F U' R2//F2L (accidentally solved corner) (7/17)
U R U R' U' B' R' F R F' B//Orientation (11/28)
R U R' y' R2 u' R U' R' U R' u R2//Permutation (12/40)

It'd be nice to see a solve where TTLL is used. For now, this just looks like a freefop solve

 

[Neuro]

It'd be nice to see a solve where TTLL is used. For now, this just looks like a freefop solve

Yeah I know, it was just coincidence, it just happened to be the F2L solution I saw but the corner ended up solved. I'll be doing more solves that should prove more useful than this one

EDIT: New solve, 50 moves but literally 2/5ths of the moves came from TTLL

SCRAMBLE: U F2 U D' R' F' D' L' U F' R U2 D2 B2 R' D2 B2 R' D2 R L

y2 D F' L U' M2 D2 F'//FB (7/7)
U2 r2 U' r' U r//2x2x3 (6/13)
F' U' F R' U2 R' U' R'//F2L-1 Corner (8/21)
y F U2 L F2 L' U2 F2 U2 F//TOLS- (9/30)
U' R U2 R' U R U' R U R' U D' R U R' D R U R2 U'//TTLL (20/50)

 

[JTay]

@Neuro Try out this alg for that TOLS-: (U) R U' R2 F R2 U R' U' R' F' R. Should end up with a U perm to finish. Much nicer!

 

[Sue Doenim]

Taking the means (taken from published algsets) and case frequency (for intuitive steps) into account, here is a good approximation of the relative movecounts of a few ZZ/Petrus LS/LL variants.

Method         Step 1        Step 2        Step 3         AUFS        Total        Alg Count
---------------------------------------------------------------------------------------------------------
ZBLL           7.04          13.5          ---            1.5         22.04        493
[color=green][b]CR             5.33          9.07          9              1.5         24.90       144 (72 w/ M)[/b][/color]
ZZ-b           10.07         13.76         ---            1.5         25.33        169
ZZ-CT          10.37         15.21         ---            2.25        27.83        197
COLL/EPLL      7.04          ~11.5         7.5            2.25        28.29        46
WV/SV          4.67          8.74          ~15            1.5         29.91        48
OLL/PLL        7.04          7.93          ~15            2.25        32.22        28

Here is a plot of movecount versus algcount for each variant and a least-squares linear regression.

What is CR? I couldn't find any mention of it anywhere. Is it a fake method for statistical reasons? Like an average or something?

 

[obelisk477]

What is CR? I couldn't find any mention of it anywhere. Is it a fake method for statistical reasons? Like an average or something?

Cardan Reduction

 

[Neuro]

Neuro and Aerma's proposed method... I believe is around 215 additonal algs (correct me if I'm wrong).

It's 187 algs. 169 ZZLL's (without mirrors), and if you keep the F2L edge in it's slot disoriented there are 2 distinct EO cases. From there 3 phasing cases (now up to 6) and the F2L corner ca be twisted 3 ways. 6*3=18 algs. If you count using mirrors/inverses, you only have 98 algs. Setting up LS that way can be a little hard to get used to though but it should give a comparable (or lower) movecount than TOLS+TTLL. I would say it's personal preference.

EDIT: If you allow the edge to be either oriented or disoriented you have ~220 algs (haven't fully accounted for symmetry yet)

 

[Neuro]

39 move solve I got using BFTT (FB to 2x2x3, F2L-1C, TOLS, TTLL)

B U2 B2 D2 F' L2 F' D2 L2 F D' B' F2 R' B' L' D' L' U2 R2

U M' B' r2 U F'//FB (6/6)
U r' U' r' U' r'//2x2x3 (6/12)
U2 R D' F' U F D//F2L-1 corner (7/19)
R2 U2 F R F' U2 R F R F'//TOLS- (10/29)
M2 U' M2 U' M' U2 M2 U2 M' U2//PLL (10/39)

The corner solved itself during TOLS but still a good solve

 

[Space Cat]

I'm not sure if joke sub-methods count, but it's still an idea.

I want to propose JLL, which isn't an acronym for anything. JLL forces a PLL skip by doing a bunch of PLL algs before OLL, and after the correct state is found, you do the OLL and it's solved.

 

[sqAree]

I'm not sure if joke sub-methods count, but it's still an idea.

I want to propose JLL, which isn't an acronym for anything. JLL forces a PLL skip by doing a bunch of PLL algs before OLL, and after the correct state is found, you do the OLL and it's solved.

I understand it's a joke method, still, you don't skip PLL by doing PLL possibly more than once.

 

[Space Cat]

I understand it's a joke method, still, you don't skip PLL by doing PLL possibly more than once.

It's technically not PLL because you're using algorithms for cases after OLL.

 

[Teoidus]

I understand it's a joke method, still, you don't skip PLL by doing PLL possibly more than once.

i think it's because the PLLs do weird things to orientation as well. He wants to use normal OLL, not pure OLL, so he has to set up a specific permutation, but 1 PLL might not cut it.

 

[Space Cat]

i think it's because the PLLs do weird things to orientation as well. He wants to use normal OLL, not pure OLL, so he has to set up a specific permutation, but 1 PLL might not cut it.

Which is exactly why it's a joke method. It's slower than 2LLL but it looks cool to non-cubers.