The New Method / Substep / Concept Idea Thread

[OtaMota]

Has anyone figured out a way to prevent PLL parity on 4x4 yet?

Not sure it is possible without cheating. From my understanding, it has a similar cause to Void Cube Parity. You can put your first center any side of the cube. You could prevent PLL parity by making it possible to tell sides apart, and then putting the centers correctly so that PLL parity does not happen. Cheating ofc though.

Edit: Now that I think about it, I think you could prevent it by doing Edges in a certain way. Not sure what type of way would yield parity free though.

 

[AhsanDaCuber125]

Not sure it is possible without cheating. From my understanding, it has a similar cause to Void Cube Parity. You can put your first center any side of the cube. You could prevent PLL parity by making it possible to tell sides apart, and then putting the centers correctly so that PLL parity does not happen. Cheating ofc though

Do cubers like Max Park find a way to prevent parity on 4x4?

 

[OtaMota]

Do cubers like Max Park find a way to prevent parity on 4x4?

No, for both OLL and PLL. The fastest cuber who currently uses OPA is WR12 iirc.

 

[AhsanDaCuber125]

Which method is easier to use for 3x3? CFOP, ZZ, or Roux? (I combine Beginner’s Method and CFOP together, still decent times)

 

[OtaMota]

Which method is easier to use for 3x3? CFOP, ZZ, or Roux? (I combine Beginner’s Method and CFOP together, still decent times)

Depends on what easier means. I think out of the 3, most will say that ZZ is the hardest to use.

 

[Garf]

Depends on what easier means. I think out of the 3, most will say that ZZ is the hardest to use.

Yeah, since it has the most concepts of EO. I think that the least concepts you have to learn the better. For all three you learn about orientation. I think the 2nd hardest thing would be cross(+1) since you have to scan for 4 or more pieces.

 

[bldcuber314]

Has anyone figured out a way to prevent PLL parity on 4x4 yet?

I'd rather prevent OLL parity... is there a way to do that?(Not that I should be learning stuff like that- I average 1:20 with Redux)

 

[OtaMota]

I'd rather prevent OLL parity... is there a way to do that?(Not that I should be learning stuff like that- I average 1:20 with Redux)

 

[Garf]

I'd rather prevent OLL parity... is there a way to do that?(Not that I should be learning stuff like that- I average 1:20 with Redux)

I already started doing OPA so I was wondering if there was a way to eliminate parity altogether.

 

[Silky]

I'd rather prevent OLL parity... is there a way to do that?(Not that I should be learning stuff like that- I average 1:20 with Redux)

There are two ways, that I know of, that avoid OLL+PLL parity. (1) is to use K4 / QTPI / other direct solving methods ( this is a technicality since you need to reduce to a 3x3 state to get parity, hence no deduction = no parity. In actuality you're really just solving parity directly ) or (2) is to BLD trace wings in inspection.

 

[bldcuber314]

You know what maybe I should learn Yau first

 

[OtaMota]

Might want to get @Christopher Mowla on this, I think he has good understanding on parities and how they work.

 

[Deleted member 36445]

I'd rather prevent OLL parity... is there a way to do that?(Not that I should be learning stuff like that- I average 1:20 with Redux)

cstimer > options > no parity scramble
next question please

 

[LBr]

I already started doing OPA so I was wondering if there was a way to eliminate parity altogether.

I don’t know how to track it, but theoretically PLL parity elimination is easy, basically let’s say you have an L2E case with edges in the same slice in adjacent F2L slots. If you do Uw’ then flip the right edge you will get a different PLL parity state compared to if you do Uw and flip the left edge. Or you could use a non reduction based method such as K4 where phenomena such as PLL parity don’t exist

 

[Garf]

cstimer > options > no parity scramble
next question please

Well that wouldn’t work since during the solve you still would need to do an even number of slices. It depends on the solver not the scramble.

 

[Deleted member 36445]

Well that wouldn’t work since during the solve you still would need to do an even number of slices. It depends on the solver not the scramble.

cstimer > options > 3x3x3 scrambles
there you go

 

[Sion]

I am not good with methods at all, let me make that clear, nor am I calling this *New*.

I just noticed that a lot of WR solves and otherwise super fast solves start off with an XX cross, or a 2x2x3 with the remaining D edge solved.

I'm wondering if brute forcing a 2x2x3 with block building and not doing an EO step until the last layer is a viable cubing strategy as opposed to making a cross and filling in all the F2Ls. This also pushes forth as much direct solving as possible, too.

It removes the need for a hard to recognize EO step, and the LL can be done in really any reasonable manner. I find that a lot of "new" methods create these large and hard to recognize algorithm sets that don't really work with a true brute force technique.

I ask this because, maybe asides of Roux, and OLL, EO steps don't really solve much of anything directly or even quasi-directly, and just make it easier to solve certain things and create skips later. However, sometimes I wonder if there really is a true net positive for Early-mid solve EOs as opposed to being an endgame type step when something could be otherwise solved directly and intuitively.

I'm sure there's a more technical reason, but part of me feels as if this particular variant of FreeFOP might just be a golden ticket.

 

[Hazel]

I am not good with methods at all, let me make that clear, nor am I calling this *New*.

I just noticed that a lot of WR solves and otherwise super fast solves start off with an XX cross, or a 2x2x3 with the remaining D edge solved.

I'm wondering if brute forcing a 2x2x3 with block building and not doing an EO step until the last layer is a viable cubing strategy as opposed to making a cross and filling in all the F2Ls. This also pushes forth as much direct solving as possible, too.

It removes the need for a hard to recognize EO step, and the LL can be done in really any reasonable manner. I find that a lot of "new" methods create these large and hard to recognize algorithm sets that don't really work with a true brute force technique

I'm sure there's a more technical reason, but part of me feels as if this particular variant of FreeFOP might just be a golden ticket.

I believe the main issues with this method are as follows: one, the ergonomics of forming a 2x2x3 often isn't worth it; two, smoothly transitioning from 2x2x3 blockbuilding to the remainder of F2L requires a very high level of look-ahead skill; and three, even at its best, 2x2x3 -> F2L just isn't faster than cross -> F2L. Planning a XXcross often requires a decently good scramble and can't be done every solve. Only the best cubers in the world can consistently plan cross+2 in inspection, I believe.

TL;DR: Being open to an XXcross while inspecting the cube is a good idea, if you have a good enough scramble. Brute forcing a 2x2x3 at the start of every, or even most solves, is likely not helpful.

 

[JDAW6]

Not sure it is possible without cheating

So ur saying there is a way.

 

[Deleted member 36445]

So ur saying there is a way.

Show me da wae