Void Parity Resolution/Algs

[AlphaSheep]

I have also seen where I can get my edges into a zig zag pattern where they need to move in a way such as:
UF > UR > UL > UB > UF

These are called W perms, and there are two possible cases - Wa and Wb. I never remember which is which.

One trick I like is to set up the four edges onto the M slice so you can cycle them in a single move. For example: F' L F R2 D' moves all four edges onto the M slice. You can then cycle them and fix parity in one move, then undo the setup moves. So the two cases are
F' L F R2 D' (M') D R2 F' L' F
F' L F R2 D' (M) D R2 F' L' F

An alg that's better suited to speed solving that I've never gotten round to learning is
M U' M' U2 F2 U' M2 U2 M U' F2
And then just replace all the U' moves with U moves to get the other case.

 

[Solvador Cubi]

Awesome! thanks AlphaSheep. Those are nice and helpful.

I've also seen some of those other cases already have names too, like the counter- and clockwise pair are "O-Perms".
do any of the others? e.g. the two 2-edge swaps? I'm thinking !-Perm and /-Perm thoughts?

I was thinking about updating the Void Cube page on the wiki: https://www.speedsolving.com/wiki/index.php/Void_cube
With at least a little more description and some LL algs.

Reading the current text there helped me realize that 2 corners can be swapped too!
Are there any known algs to correct that case?

 

[TDM]

These are called W perms, and there are two possible cases - Wa and Wb. I never remember which is which.

One trick I like is to set up the four edges onto the M slice so you can cycle them in a single move. For example: F' L F R2 D' moves all four edges onto the M slice. You can then cycle them and fix parity in one move, then undo the setup moves. So the two cases are
F' L F R2 D' (M') D R2 F' L' F
F' L F R2 D' (M) D R2 F' L' F

An alg that's better suited to speed solving that I've never gotten round to learning is
M U' M' U2 F2 U' M2 U2 M U' F2
And then just replace all the U' moves with U moves to get the other case.

Isn't there only one case? The second alg solves the same case from a U2.

 

[Solvador Cubi]

This is what I'm seeing:

to move edges in this direction: F > L > R > B > F
Then this is the alg: M U' M' U2 F2 U' M2 U2 M U' F2

to move edges in this direction: F > R > L > B > F
do a U' then the above alg.

Does that sound right to anyone else?

 

[Solvador Cubi]

Does anyone know of an algorithm to solve the void cube corner parity where you would need to Swap 2 Opposite (diagonal) Corners

It seems like a decent way is the combination of an E Perm + O Perm
but is there anything known that is shorter?

thanks.

 

[Teoidus]

I think that the best way for CFOP users to solve void cubes is to use CFCE. ELL + parity algorithms will be far more friendly compared to these diag swap algs

 

[Solvador Cubi]

I agree with you Teoidus about changing up one's method a bit to avoid those situations.
I'm a CFOP 4LLL solver (with EPLL last) so I wouldn't normally need that diag swap alg.

If someone finished F2L and was presented with that case (unlikely, but still possible) then knowing one alg (preferably 16 htm or less) would be nice.

I'm also just trying to compile a complete list of Void Cube LL scenarios and algs.

 

[Martin Orav]

I'm not sure if this has been posted in this thread before, but I used the alg M2 U/U' M/M' U'/U M2, that I figured out myself in like 5 minutes one time I was bored and couldn't find a good parity alg for a void cube. It preserves F2L, COLL and CPLL, and flips all the LL edges. I learned the OLL case 47 to make that part faster. The OLL 47 is also really easy to recognise so it makes the OLL a bit faster in general, but that's not the topic to talk about here.

Today, I was bored, and since I learned intermediate LSE like a week or so ago, I was thinking that what if I just did an M' or an M move and then did LSE EO and then fixed the DF and DB edges. After trying out different ways to do do it I found this alg M' U M U2 M' U2 M' U' M U' M U2 M' (U'). It's 13 moves STM so pretty good considering the fact that its a 2-swap. If you used EOLR, by looking at DF and DB edges as UL and UR edges you might get a better alg.

Since there was already a 2-swap in the first post in this thread, it doesn't help much, but it's an alternate alg.

If there was an <M, U> subgroup optimal solver it would be really good, because if you use RL and FB mirrors along with inversing pretty much every sequence of MU moves can be very fingertrickable. There might be one in cube explorer, but I don't have a computer so I can't check.

This way we could also generate all the parity EPLLs (or ELLs, although using edge control and CLL on 3x3 with parity EPLLs might be better (idk)) since they will be fingertrickable, as said above. If someone can tell me how to use cube explorer on an apple computer, I'll try it since my mother has a computer, but it's apple.

Now if you are a Roux solver, you can just use the EOLR case, that you get, when you do an M/M' move after the first EO+LR or recognise it before LR. That might be faster, but I'm not sure. Also, I'm not sure if this would work, but by not doing just M', but deciding whether to do M or M' you might be able to skip dots. Not sure if this would work though. Again, with an <M, U> subgroup optimal solver we could generate parity M slice cases, which is probably the best for Void cube speedsolving with Roux.


Edit: now I read through the thread, and I saw an idea that's probably better for Roux (solve UL & UR, after CMLL, solve M slice in one alg). No parity ever. The amount of cases is really low as well, it's 27, including the skip, (for void cube) (can someone confirm, please) and the recog doesn't seem too bad either.

 

[Solvador Cubi]

I wanted to follow up on this thread with an answer to my own question from a few months ago.

The best alg I "found" (using online solvers, actually) to swap 2 diagonal corners is:
M' F' M F2 U2 F' M U2 F U2 B' U2 B (L2 F') 2

I also put all my Void Cube notes together in this one page info-graphic:
http://solvexio.cf/app/#/Void_Page

It organizes 10 parity fixing algs for solves using CFOP or CFCE methods.


-= Solvador Cubi

 

[1001010101001]

Use roux logic.

 

[Electro Web]

Opposite
(UF <--> UB) : (M' U2)2 B U B' (M' U)4 B U' B' M' (looking for a better one, though)

I too had been looking for a better algorithm, and after looking and looking I couldn't seem to find one that could be easy to remember, so after a long time of looking I decided to take matters into my own hands and I made my own algorithm, that combines the knowledge I've gained from solving the 4x4 parity and the knowledge I've gained from solving the void cube parity and this was the algorithm I came up with:

Opposite
(UF <--> UB): (U R U' R') U' (M' U)2 (M' U')3 U' (M' U') (r U R' U')

 

[Solvador Cubi]

Thanks Electro Web,

I have now tried out your alg too. While it's a few moves longer, it does avoid the Back moves, so that's nice.

From the time when I posted my alg until now, I have since learned some LMCF and Roux.
So using that knowledge, these are now the shortest ones I can come up with...


17-Move Opposite U Edge Swap (UF <--> UB):
U r' U M U' M2 R U' R' U M U' R M' u2 M u2
U' M U2 M' U' R' F R U' M' U R' F' R' U2 M U2 M'

or this is one with 19 moves, but seems a bit smoother to execute:
M' U' M (U2 M' U2 M') U' M' (U2 M' U2 M') U M2 U' M' U2 M


-= Solvador Cubi