Theoretical Yau4 variations

[Bingus Ringus McDingus]

Let me preface this by saying that I probably shouldn't be the one talking about this, I am not great at 4x4. But, I think that I have three potential ideas that may improve Yau for 4x4.

Also, these ideas may previously have existed, and I apologize if I am repeating ideas that are already out there. That is unintentional and this is just some rambling.

1. Sandwich: Instead of making a T-cross on one side, make a T-cross on one side that is solved and a T-cross of edges that are paired and not solved on the opposite center. This opens up usage of M slices and r2/l2 moves instead of relying heavily on Rw2 moves, and gets a head start on 3-2-3 by making the 3-step faster or non-existent. This could also be useful because regripping for Rw moves could be reduced.

2. 2block: After doing the 2 step of 3-2-3, make a Petrus-style block of 2x2x3 solved pieces with corners and edges. This will not work with Sandwich but is more simple and easier to implement if 3x3 stage is slower for some people.

3. Targeted final redux: Only when the last 2 edges are going to be solved in F2L. 3x3 stage with two unsolved edges, similar to 2block, and during the last step, use the redux edge-pair algorithm either lefty or righty, depending on which makes the 3x3 stage cases easier. This may not be very viable but this is just a theory.

 

[Bingus Ringus McDingus]

For everyone who averages sub-1 or just thinks they're good at 4x4, can you try and implement one of more of these ideas for an AO5 to see if this has feasibility for widespread usage? I'm not good nor consistent at 4x4, so I feel as if someone who is more consistent could help me be the judge of these methods.

 

[LBr]

Ok. The first part I do not understand. There is already a sandwich method for big cubes but I don’t understand what you are saying it makes no sense. With the 223, that is not always possible and it ruins flexibility, if you are going to implement direct solving, do it but definitely do not integrate it with 4x4 323. The last thing seems useless to me, as it just makes you think more without really helping. However it can determine whether you get PLL parity, but I am not aware of any way of knowing how to harness that with Yau. I have experimented with K4, and obviously that eliminates PLL parity altogether. I suggest you look at it as it may open your eyes about 4x4 speedsolving

 

[Bingus Ringus McDingus]

Ok. The first part I do not understand. There is already a sandwich method for big cubes but I don’t understand what you are saying it makes no sense. With the 223, that is not always possible and it ruins flexibility, if you are going to implement direct solving, do it but definitely do not integrate it with 4x4 323. The last thing seems useless to me, as it just makes you think more without really helping. However it can determine whether you get PLL parity, but I am not aware of any way of knowing how to harness that with Yau. I have experimented with K4, and obviously that eliminates PLL parity altogether. I suggest you look at it as it may open your eyes about 4x4 speedsolving

The sandwich terminology was not something I knew of, I guess it could be called double-T. Basically, when making the first 2 centers plus the first 3 edges, solve another random 3 edges and make the 3/4ths of a cross on both centers, not just one. For the second set of 3, the edges don't have to be specifically the edges that make up a 3x3 cross. The best I can explain is this double-T case I improvised:

 Twizzle link 

 Setup 

F2 L B2 L' U2 F2 R D2 R' F2 R' F' R U2 F2 U R U R2 U2 R' U R Rw R' U R' Rw U Lw2 L2 Rw' R L' U' L U L' U' L U L' U' L U y2 R U R' U' R U R' U' R U R' U2 Rw R'

 Moves 

And 2block or 2x2x3 is possible after 3-2 of 323, pairing the final few edges using R U F moves with the solved 2x2x3 block at the back. Like this:

 Twizzle link 

 Setup 

U2 R2 U R U2 R' U' R U2 R' U' R2 U' R U' R U' R U' R2 U2 R' U R2 U y u' R U R' F R' F' R u R U R' U2 L F L' F' u' R U R' F R' F' R Uw

 Moves 

Sorry for the poor examples, I am not home and do not have my 4x4 for demonstration.

 

[randomness...]

"1. Sandwich: Instead of making a T-cross on one side, make a T-cross on one side that is solved and a T-cross of edges that are paired and not solved on the opposite center. This opens up usage of M slices and r2/l2 moves instead of relying heavily on Rw2 moves, and gets a head start on 3-2-3 by making the 3-step faster or non-existent. This could also be useful because regripping for Rw moves could be reduced."

this is how ap pairs edges in his big cube (11x11) solves.
i do this for Hoya 9, but its more limited due to the two centers.
i do agree with LBr though, look more into existing methods

 

[LBr]

Ok I understand now. I would argue this isn’t yau as edges are paired first. A similar idea was proposed in the 2021 method development which I will link here. It uses a reduction based approach to solve the edges and centers simultaneously and efficiently. I suggest you check it out

 

[Bingus Ringus McDingus]

Ok I understand now. I would argue this isn’t yau as edges are paired first. A similar idea was proposed in the 2021 method development which I will link here. It uses a reduction based approach to solve the edges and centers simultaneously and efficiently. I suggest you check it out

I think Bar4E is similar and definitely viable.

 

[Curunír]

I just dont do 4x4

 

[UncleFrank]

I just dont do 4x4

The Zero method.

 

[Christopher Mowla]

There is already a sandwich method for big cubes but I don’t understand what you are saying it makes no sense.

I saved Nicholas Ho's old website before it went down. I recently made this PDF (download to enable internal hyperlinks from the table of contents etc.) for those who aren't aware of that variant of Cage.

 

[JDAW6]

Let me preface this by saying that I probably shouldn't be the one talking about this, I am not great at 4x4. But, I think that I have three potential ideas that may improve Yau for 4x4.

Also, these ideas may previously have existed, and I apologize if I am repeating ideas that are already out there. That is unintentional and this is just some rambling.

1. Sandwich: Instead of making a T-cross on one side, make a T-cross on one side that is solved and a T-cross of edges that are paired and not solved on the opposite center. This opens up usage of M slices and r2/l2 moves instead of relying heavily on Rw2 moves, and gets a head start on 3-2-3 by making the 3-step faster or non-existent. This could also be useful because regripping for Rw moves could be reduced.

2. 2block: After doing the 2 step of 3-2-3, make a Petrus-style block of 2x2x3 solved pieces with corners and edges. This will not work with Sandwich but is more simple and easier to implement if 3x3 stage is slower for some people.

3. Targeted final redux: Only when the last 2 edges are going to be solved in F2L. 3x3 stage with two unsolved edges, similar to 2block, and during the last step, use the redux edge-pair algorithm either lefty or righty, depending on which makes the 3x3 stage cases easier. This may not be very viable but this is just a theory.

I avg sub50 and all these three are garbage . No offense btw

 

[Bingus Ringus McDingus]

I avg sub50 and all these three are garbage . No offense btw

Thank you for your unbelievably positive and absolutely not unnecessary feedback.

you dead built like a Vitamix blender on god

 

[Bingus Ringus McDingus]

you dead built like a Vitamix blender on god

I am not wrong

 

[randomness...]

bruh what has this come to