The New Method / Substep / Concept Idea Thread

[ProStar]

Ok, back with another idea that probably already exists. It's a combination of CFOP + Roux.

F2B - Solve First 2 Blocks like you normally would do for Roux
CMLL - Same as Roux
EO - Orient Yellow/White edges like normal for Roux, then use M moves to place the remaining cross edges
EPLL - Same as CFOP

Example Solve(excuse my laughably horrific F2B):

z2 U R2 L' U2 L U2 L' U L U2 L U L' R U R' U' M U' M' y R U R' U2 R' U R U' y R U' R' // F2B ( I don't mean any disrespect to Roux users, I'm just awful lol )

R U R' U' R' F R F' // CMLL

U M' U' M U2 M' U' M // EO pt. 1

M' U2 M // EO pt. 2 (I know I could've done U M' U' M U2 M' U M for EO, this is just so it's easier to understand)

U' M2 U M U2 M' U M2 U' // EPLL(+AUF)

 

[Cubinwitdapizza]

Ok, back with another idea that probably already exists. It's a combination of CFOP + Roux.

F2B - Solve First 2 Blocks like you normally would do for Roux
CMLL - Same as Roux
EO - Orient Yellow/White edges like normal for Roux, then use M moves to place the remaining cross edges
EPLL - Same as CFOP

Example Solve(excuse my laughably horrific F2B):

z2 U R2 L' U2 L U2 L' U L U2 L U L' R U R' U' M U' M' y R U R' U2 R' U R U' y R U' R' // F2B ( I don't mean any disrespect to Roux users, I'm just awful lol )
R U R' U' R' F R F' // CMLL
U M' U' M U2 M' U' M // EO pt. 1
M' U2 M // EO pt. 2 (I know I could've done U M' U' M U2 M' U M for EO, this is just so it's easier to understand)
U' M2 U M U2 M' U M2 U' // EPLL(+AUF)

This has been proposed by a lot of people, but no EO and before the EO. This is usually not very efficient, but this version might be better than placing L and R and then L4E. I can do some comparison solves for refrence.

 

[Deleted member 36445]

Ok, back with another idea that probably already exists. It's a combination of CFOP + Roux.

F2B - Solve First 2 Blocks like you normally would do for Roux
CMLL - Same as Roux
EO - Orient Yellow/White edges like normal for Roux, then use M moves to place the remaining cross edges
EPLL - Same as CFOP

Example Solve(excuse my laughably horrific F2B):

z2 U R2 L' U2 L U2 L' U L U2 L U L' R U R' U' M U' M' y R U R' U2 R' U R U' y R U' R' // F2B ( I don't mean any disrespect to Roux users, I'm just awful lol )
R U R' U' R' F R F' // CMLL
U M' U' M U2 M' U' M // EO pt. 1
M' U2 M // EO pt. 2 (I know I could've done U M' U' M U2 M' U M for EO, this is just so it's easier to understand)
U' M2 U M U2 M' U M2 U' // EPLL(+AUF)

thats how i solve roux

 

[Pyjam]

It's inefficient. Have you noticed that the first move of U-Perm, H-Perm, and Z-Perm is M2 ? and then deconstruct the DF-DB line.

 

[ProStar]

It's inefficient. Have you noticed that the first move of U-Perm, H-Perm, and Z-Perm is M2 ? and then deconstruct the DF-DB line.

I'm afraid I don't understand what you're saying.

 

[Pyjam]

If you look how this PLL U works : M2 U' M' U2 M U' M2, you see that it deconstructs DF-DB to solve UL-UR instead.
The PLL H and Z do the same thing.

 

[PapaSmurf]

For the ESO method: just start off by building a 2x2x2. It averages 6 moves if you get good.
Instead of solving the whole E slice, just solve the FL and BR edges. That will only be another 4 ish moves.
EO+cross is an interesting step. I'd guess about 8-10 moves?
2 keyhole F2L pairs which would be maybe 10 moves.
An F2L pair, around 8 moves.
ZBLL, 16 moves (inc. AUF).
All in all, an interesting method, rotationless and good ergonomics, althougn mixed. Probably a low-ish movecount of <50 moves. Would be cool to see if it's good, but would need ZBLL (493 algs) to unlock its proper potential.

Spoiler: Example solve

alg.cubing.net

alg.cubing.net

Overall, a pretty lucky solve, but it feels like on of those methods that are lucky just because they are. Could have potential.

 

[ProStar]

(I asked this another time but don't remember where so I don't know what the answer was)

How many algs would it take to to the last slot while orienting and permuting corners? Obviously the set would be enormous, but it could be like OLS; you learn only certain subsets(like WV or SV for OLS)

 

[Sue Doenim]

The case count would be somewhere around 6000.

 

[Etotheipi]

The case count would be somewhere around 6000.

thats a large number.

 

[ProStar]

thats a large number.

OLS has over 17,000 total. No one learns OLS, they learn certain subsets. Recognition would be harder then OLS though.

 

[Deleted member 51471]

I have this idea. It's probably very bad, but here it is :
1. Solve the entire cube ignoring the orientation of every piece.
2. Orient the E-Slice (1 algorithm)
3. Apply pure OLL on one side. Try orienting the side as much as you can, if finish with one twisted corner or edge, don't worry. If you have both, make sure that two bad pieces are next to each other after pure OLL.
4. Do the same thing on the other side.
5. If you have twisted corners, edges, or both, align them so they are on the same layer, then rotate the cube and apply another pure OLL so they are both solved simultaniously.

 

[Cubinwitdapizza]

I have this idea. It's probably very bad, but here it is :
1. Solve the entire cube ignoring the orientation of every piece.
2. Orient the E-Slice (1 algorithm)
3. Apply pure OLL on one side. Try orienting the side as much as you can, if finish with one twisted corner or edge, don't worry. If you have both, make sure that two bad pieces are next to each other after pure OLL.
4. Do the same thing on the other side.
5. If you have twisted corners, edges, or both, align them so they are on the same layer, then rotate the cube and apply another pure OLL so they are both solved simultaniously.

Seems pretty bad Lol. No offense.

 

[ProStar]

So, it's like Petrus, except instead of EO you do CO. Then, when F2L is finished, you have all the corners oriented. Then you can do one alg and solve the rest of the cube. The LL set would have less algs than ZBLL, cause corners instead of edges would be oriented. I can't figure out how to do CO properly though.

 

[Etotheipi]

So, it's like Petrus, except instead of EO you do CO. Then, when F2L is finished, you have all the corners oriented. Then you can do one alg and solve the rest of the cube. The LL set would have less algs than ZBLL, cause corners instead of edges would be oriented. I can't figure out how to do CO properly though.

CO is a pain XD. ask WoowyBaby about how to solve it, he's an FMCer who uses domino reduction sometimes, he'd know.

 

[fortissim2]

So I was looking at the Zipper Method, and I kinda thought of a variant for it. It goes like this:
1. Cross
2. F2L - 1 edge
3. CLL
4. L5E
What I like about the second step is that if you get a bad F2L case, you can just insert the corner and continue with the other corner-edge pairs, which can be faster than doing normal F2L.
Instead of doing OLLCP, you could do CLL (smaller alg set) and move on to the first step of L5E.
The L5E part can be split into 2 parts: solve 2 top edges to where they belong in the cube with an algorithm, and L3E, using 1 algorithm to solve the rest of the cube.

There might be more efficient ways on doing the last 5 edges, but for now I think it's good.

Edit: I also included an example solve to show you guys how it works.

https://alg.cubing.net/?setup=F2_D-...2_top_edges U2-_R-_U-_R-_U-_R-_U_R_U_R_//_L3E

 

[PapaSmurf]

Justin (the inventor of Zipper) has already come up with that and it's called Zipper-b, but it's almost certainly better than normal Zipper. The algs are found here.
You one look L5E for the best.

 

[ProStar]

So I was looking at the Zipper Method, and I kinda thought of a variant for it. It goes like this:
1. Cross
2. F2L - 1 edge
3. CLL
4. L5E
What I like about the second step is that if you get a bad F2L case, you can just insert the corner and continue with the other corner-edge pairs, which can be faster than doing normal F2L.
Instead of doing OLLCP, you could do CLL (smaller alg set) and move on to the first step of L5E.
The L5E part can be split into 2 parts: put 2 yellow edges to where they belong in the cube with an algorithm, and L3E, using 1 algorithm to solve the rest of the cube.

There might be more efficient ways on doing the last 5 edges, but for now I think it's good.

It wouldn't be normal L5E. Normal L5E does the 4 U layer edges and a D layer edge.

 

[PapaSmurf]

L5E just means last 5 edges, so it doesn't really matter. Just as CLL in Waterman is different to CLL in CFCE, L5E in Zipper is different to other L5E.

 

[WoowyBaby]

So, it's like Petrus, except instead of EO you do CO. Then, when F2L is finished, you have all the corners oriented. Then you can do one alg and solve the rest of the cube. The LL set would have less algs than ZBLL, cause corners instead of edges would be oriented. I can't figure out how to do CO properly though.

CO is a pain XD. ask WoowyBaby about how to solve it, he's an FMCer who uses domino reduction sometimes, he'd know.

Petrus with CO would make the right block F2L blockbuilding a whole lot worse, because with EO you can do it iwth just R and U moves, while without it you require F and / or B moves, and, also, it is basically impossible to building your right block without breaking CO.
CO isn't hard to do. In fact, it takes the same number of moves to do as EO, but the benifits of CO aren't much. It doesn't make blockbuilding any easier like EO does and it barely makes the corners any easier to solve, and it heavily restricts what you can do, for example, if you have CO it's impossible to solve the cube without breaking CO, unlike EO, unless........ along with CO you also have EO on two axis as well; so then you have domino reduction. Then it's actually really dang good