The New Method / Substep / Concept Idea Thread

[Hazel]

Here's another roux-ish varient of mine:
1) CPBlock
2) 2x2x1 in RB
3) EODB
4) 2-gen Winter Variation
5) L5E

EDIT: Here's another method idea:
1) First 2 blocks - pretty much a 2x2x3 in DB minus the DB edge itself
2) EODB - solve DB while solving EO
3) finish F2L however you like
4) ZBLL

 

[Neuro]

Here's another roux-ish varient of mine:
1) CPBlock
2) 2x2x1 in RB
3) EODB
4) 2-gen Winter Variation
5) L5E

EDIT: Here's another method idea:
1) First 2 blocks - pretty much a 2x2x3 in DB minus the DB edge itself
2) EODB - solve DB while solving EO
3) finish F2L however you like
4) ZBLL

1st method, I don't really see it. Roux seems better with a similar alg count. CP Block is difficult, just allowing the 2x2x1 in RB isn't very smart, and 2GWV usually isn't very efficient.

2nd idea is decent; although I'd suggest allowing either DB or DF to be variable as it would probably make blocks easier. Allows for a lot of F2L/LL variants (like in all EO methods) so there's flexibility there also

 

[Hazel]

2nd idea is decent; although I'd suggest allowing either DB or DF to be variable as it would probably make blocks easier. Allows for a lot of F2L/LL variants (like in all EO methods) so there's flexibility there also

Very true: you could use CT, many ZZ varients, Winter Variation/PLL, WVCP/EPLL, COLL/EPLL, ZBLL, CPLS/2GLL, WVCP/L5E... the options are endless, and if you know a lot of algs you could just use whichever is best for each solve.

 

[Hazel]

Sorry for the double post, but this is mostly unrelated to my last one.
I created a spreadsheet of Last 5 Edges algorithms (when they're all already oriented, the U edges + DF) if anybody wants to use them for a method. There's only 12 of them. I allowed commenting in case anybody finds better algs then the ones I have.
https://docs.google.com/spreadsheets/d/1EEvR0dRzO3NqgG_GBfQMa9w3vEuFdfH7GCd63xJNeao/edit?usp=sharing

 

[Neuro]

Sorry for the double post, but this is mostly unrelated to my last one.
I created a spreadsheet of Last 5 Edges algorithms (when they're all already oriented, the U edges + DF) if anybody wants to use them for a method. There's only 12 of them. I allowed commenting in case anybody finds better algs then the ones I have.
https://docs.google.com/spreadsheets/d/1EEvR0dRzO3NqgG_GBfQMa9w3vEuFdfH7GCd63xJNeao/edit?usp=sharing

I'd think recog would be easier if you held the DF edge in UB so you can see 3/4 stickers at once

EDIT: Most of these cases are optimal using Roux 4b/4c so if you are good at LSE these would be very easy to learn

 

[Thom S.]

So I have this idea for translating Roux to 6x6(possibly 5x5 too) and 7x7 while not using huge M moves. I would wish that Neuro or Shiv3r could see this.

1. Opposite Centres on L and R(possibly on U and D if you pair on the E slice)
2. First Block(better do your blocks as long as you have all the freedom)
3. Second block
4. Solve the two layers Adjacent to the outer ones in kind of an ObLbL approach
5. CMLL
6. Solve the Wings Adjacent to the Corners on the U layer(I estimate a big Algorithm count but not a huge one)

now you have two two-layer deep Roux blocks and four 1x2x2 blocks acting as corners

7. reduction of the last four Centres and last six Edges
8. 3x3 Stage with LSE

 

[shadowslice e]

So I have this idea for translating Roux to 6x6(possibly 5x5 too) and 7x7 while not using huge M moves. I would wish that Neuro or Shiv3r could see this.

1. Opposite Centres on L and R(possibly on U and D if you pair on the E slice)
2. First Block(better do your blocks as long as you have all the freedom)
3. Second block
4. Solve the two layers Adjacent to the outer ones in kind of an ObLbL approach
5. CMLL
6. Solve the Wings Adjacent to the Corners on the U layer(I estimate a big Algorithm count but not a huge one)

now you have two two-layer deep Roux blocks and four 1x2x2 blocks acting as corners

7. reduction of the last four Centres and last six Edges
8. 3x3 Stage with LSE

So adapted stadler? Could be interesting as the algs for the wings exist. I'd be interested to see what the movecount is like. Looks like it has quite a lot of pros but I'll mention my main issue with it: you'll have a hard time doing centres as you will frequently have to break blocks in order to solve the edges which could add inefficiencies. If you can find a good way to so this I'll gladly be on board with it (not that I do bug cubes much).

 

[Neuro]

So I have this idea for translating Roux to 6x6(possibly 5x5 too) and 7x7 while not using huge M moves. I would wish that Neuro or Shiv3r could see this.

1. Opposite Centres on L and R(possibly on U and D if you pair on the E slice)
2. First Block(better do your blocks as long as you have all the freedom)
3. Second block
4. Solve the two layers Adjacent to the outer ones in kind of an ObLbL approach
5. CMLL
6. Solve the Wings Adjacent to the Corners on the U layer(I estimate a big Algorithm count but not a huge one)

now you have two two-layer deep Roux blocks and four 1x2x2 blocks acting as corners

7. reduction of the last four Centres and last six Edges
8. 3x3 Stage with LSE

Can you give an example solve? I'm having a hard time figuring it out

I don't really do anything bigger than 4x4, but if I did I'd imagine I'd use something like this

LR Centers
FB
Rotate as in Meyer and do edge pairing
Rotate back and solve SB with DF/DB edge and any other edge oriented.
OLLCP
Rotate to solve with <RU>

 

[Abram Lookadoo]

new method (blocky ribbon)

1) 2x2x2 block
build a 2x2x2 block in back left
2) F2L - 1 corner
solve two 1x2x2 blocks in DFL and DBR, and while placing the two blocks place the edge FR oriented
3) tols
orientate all pieces with 1/173 algorithms
4) ttll(?)
permute all remaining pieces with 1/93 algorithms

 

[Thom S.]

So adapted stadler? Could be interesting as the algs for the wings exist. I'd be interested to see what the movecount is like. Looks like it has quite a lot of pros but I'll mention my main issue with it: you'll have a hard time doing centres as you will frequently have to break blocks in order to solve the edges which could add inefficiencies. If you can find a good way to so this I'll gladly be on board with it (not that I do bug cubes much).

Yes, it's quite like Stadler.
Could you give me a link to the Wing algorithms, as I'm very unexperienced with big cubes and don't know where to look for?
Yeah, there are still problems but my first thought for 6x6 is that in step 7 you kind of have the centre situation as in Yau when you only have one face to rotate(and only the inner r slice instead of the whole right half). For Edges I have to look more into Edge pairing methods, but(as a nub at big cubes) my first guess is that since everyone seems to be ok with 3-2-3 we could do a commutator to have one dedge pair solved and then use 2-3(as the UL and UR spots could easily used as buffers)

Can you give an example solve? I'm having a hard time figuring it out

There will be some in the near future.

 

[alwin5b]

Can anyone explain to me how CPFB is done (corner permutation first block), or point me to a post where CPFB is explained? I can't seem to find anything about this.

 

[Abram Lookadoo]

i have not seen any instructions about cpfb, so i created my own way of doing it
this is to help you understand cp tracking (from teoidus) http://teoidus.github.io/dump/2GR/CPLine/CPLine.html

0) locate and track cp
1) build a 1x2x2 block in LBD
2) build the FL DFL pair
3) place the pair while manipulating cp to the “solved state"

 

[Neuro]

Can anyone explain to me how CPFB is done (corner permutation first block), or point me to a post where CPFB is explained? I can't seem to find anything about this.

Personallly I use this method:

Inspection: rotate until WBO corner is in DBL with white on D
1: solve the edge belonging in DL in telation to the DBL corner using <RUFrufMES>
2: solve CP with DFL corner (use Teoidus's 2GR CP tracking/solving) <RUFru>
3: solve BL FL edges <RUruME>

There are a lot of different ways to solve it, but I find this easiest

 

[Teoidus]

I think @Arc 's method is best:

1. Track CP through square (without DBL fixed, instead keeping track of where the faces that will become <R,U> are--he calls it "floating CP" which is pretty apt)
2. Solve last pair + CP. I don't know how he does this one, but e.g. force last edge somewhere + algorithm would be ~6 CP * 3 CO * 7 DLF = 116

 

[alwin5b]

another method:

1. CPFB
2. COLine (orient all corners and place the E layer edges oriented in the E layer)
3. solve the DR corners relative to each other (so the R sticker of both DR corners is on the same face)
4. L7E similar to WaterRoux (~ 70 algs)

some details:

2. COLine: place a pseudo-pair (see my previous posts) and orient the other E layer edge. Then apply one of 104 algorithms, which are similar to TSLE algs, but better, because only one piece needs to be preserved (the E layer edge belonging to the pseudo-pair).

3. I don't know yet where to put the DR corners, it seems to me keeping them in the U layer would be best, but the L7E algs require the DR corners to be in DR.

 

[Neuro]

another method:

1. CPFB
2. COLine (orient all corners and place the E layer edges oriented in the E layer)
3. solve the DR corners relative to each other (so the R sticker of both DR corners is on the same face)
4. L7E similar to WaterRoux (~ 70 algs)

some details:

2. COLine: place a pseudo-pair (see my previous posts) and orient the other E layer edge. Then apply one of 104 algorithms, which are similar to TSLE algs, but better, because only one piece needs to be preserved (the E layer edge belonging to the pseudo-pair).

3. I don't know yet where to put the DR corners, it seems to me keeping them in the U layer would be best, but the L7E algs require the DR corners to be in DR.

in regards to L7E, it seems to be agreed that the best strategy ATM is to insert one of the final 2 edges for R in while doing EO and do L6EP like in Roux 4B/4C. This can be done with algs or intuitively. If you really want to do this method, you may want to put the corners in D oriented (swapped or solved) and use algs to solve all the corners at once. Personally, I think that the method wouldn't be worth it. It seems to me that CPFB would be quite difficult to plan in inspection; and solving the corners (in my or your way) with regard to E slice edges would be difficult to do quickly. I could be wrong though, good luck!

 

[Hazel]

Here's a 2Gr method -
Step 1 - EOLine
Step 2 - CPLine - solve DL, DLF, and DLB while solving cp.
note: as of now I'm pretty sure steps 1 and 2 could be done in either order.
Step 3 - Finish 2x2x3
Step 4 - Solve the rest of the cube 2-gen!

 

[mDiPalma]

sounds good but i dont understand your corner-naming convention.

 

[Teoidus]

looks good but i don't understand your method making convention

 

[Hazel]

sounds good but i dont understand your corner-naming convention.

So the tree bottom-left 3 pieces are solves as well as CP