[Help Thread] Big Cubes Discussion

[Sub1Hour]

How do I get the last 2 edges on a 9x9?

Treat each layer of edges like 5x5 edges. Think about it as 3 5x5 edges stacked on top of each other. Just do slice flip slice and such until you get them solved. (parity is highly likely)

 

[ProStar]

How do I get the last 2 edges on a 9x9?

Same way as on a 5x5, just keep doing slice flip slice in different ways. It's nice to have an understanding of what that actually does though. Also you'll probably get parity, and it's like gluing together several layers together to make a 5x5

 

[zeruel85]

Hello!

I have a doubt about an edge parity configuration.

As you can see from the pics, it's simply the (Rw2 F2 U2 Rw2 U2 F2 Rw2) method, but with a flipped edge.
I'd like to know if there is a direct algorithm to solve this case, instead of flipping an edge and than proceed as per above.

Thanks!

 

[xyzzy]

Hello!

I have a doubt about an edge parity configuration.

As you can see from the pics, it's simply the (Rw2 F2 U2 Rw2 U2 F2 Rw2) method, but with a flipped edge.
I'd like to know if there is a direct algorithm to solve this case, instead of flipping an edge and than proceed as per above.

Thanks!

Flipping is generally considered the best way to solve this, but here are some alternatives:

r' U' R U 2R2 U R' U' r2 U' R' U r (from Rob Yau's L2E sheet)
r U R U' r2 U' R U r2 U R U' r' (variant of the above)
(x' z) R2 u' R2 u 4r2 F2 4r2 3u' R2 3u

Also, this isn't "parity"; it can be solved entirely using the basic cycling or slice-flip-slice methods.

 

[brododragon]

What makes Yau so much greater than redux on 4x4? It seems weird to just solve three edges. It majorly bandages the cube and turns F, U, B, and D moves into (adjust L layer) F/U/B/D. It just seems like it majorly shoots up the move-count for no real benefit.

 

[Nmile7300]

Good question. Let's go through every step of Yau to see why it is used by all the top 4x4 solvers. The first step is first two centers.This is the same as redux except that if you are not color neutral, you are limited to white and yellow. This might seem bad but because the centers on a 4x4 are so simple compared to other big cubes, the difference is almost nonexistent. The next step is to build 3 cross edges on the center. This is very efficient especially in comparison to free slice edge pairing. The next step is the last 4 centers. As you mentioned, you are kind of limited by the cross edges. But again, 4x4 centers are very simple. I would say you are only adding about 8 moves to the centers. Next is the last cross edge, which is basically just to finish up the cross for later. After that, you have the last 8 edges, which can be done quickly because you don't have to look at the bottom. Finally, you have 3x3 stage, which is way better (if you use CFOP). The cross is already done, and on big cubes, inspecting the cross and performing it quickly is one of the hardest parts of 3x3 stage. everything after is basically the same.

So, to recap-

• First two centers are barely worse if at all.
• First 3 edges are super efficient.
• Last four centers are slightly worse.
• Last cross edge is not anything super different.
• Remaining edges can be done very fast with 3-2-3 or other pairing methods, and you don't have to look at the bottom.
• 3x3 stage is much faster and easier for lookahead.

Also, you were wondering about the move count. I believe Yau has less moves than redux, maybe by about 5-10 less (depends on parity).

 

[Sub1Hour]

Good question. Let's go through every step of Yau to see why it is used by all the top 4x4 solvers. The first step is first two centers.This is the same as redux except that if you are not color neutral, you are limited to white and yellow. This might seem bad but because the centers on a 4x4 are so simple compared to other big cubes, the difference is almost nonexistent. The next step is to build 3 cross edges on the center. This is very efficient especially in comparison to free slice edge pairing. The next step is the last 4 centers. As you mentioned, you are kind of limited by the cross edges. But again, 4x4 centers are very simple. I would say you are only adding about 8 moves to the centers. Next is the last cross edge, which is basically just to finish up the cross for later. After that, you have the last 8 edges, which can be done quickly because you don't have to look at the bottom. Finally, you have 3x3 stage, which is way better (if you use CFOP). The cross is already done, and on big cubes, inspecting the cross and performing it quickly is one of the hardest parts of 3x3 stage. everything after is basically the same.

So, to recap-

• First two centers are barely worse if at all.
• First 3 edges are super efficient.
• Last four centers are slightly worse.
• Last cross edge is not anything super different.
• Remaining edges can be done very fast with 3-2-3 or other pairing methods, and you don't have to look at the bottom.
• 3x3 stage is much faster and easier for lookahead.

Also, you were wondering about the move count. I believe Yau has less moves than redux, maybe by about 5-10 less (depends on parity).

Yeah, that sounds about right. If you don't like yau there is another method called Hoya where its got similar principles as yau like making cross before finishing edges and 3-2-3 can also be used. Its a fun method but I prefer yau on 4x4 and redux for 5-7

 

[xyzzy]

What makes Yau so much greater than redux on 4x4? It seems weird to just solve three edges. It majorly bandages the cube and turns F, U, B, and D moves into (adjust L layer) F/U/B/D. It just seems like it majorly shoots up the move-count for no real benefit.

Psst this thread is for 5/6/7; you should've posted this in the thread for 4.

I believe Yau has less moves than redux, maybe by about 5-10 less (depends on parity).

Here's four scrambles with solutions up to reduction+cross with both redux and Yau; guess which one took 7.5 moves less on average.

Spoiler

Scramble 1: R' L' D F2 U2 F2 U' R U2 F2 D2 F' R2 F R2 D2 B U2 D2 B Rw2 F2 U Fw2 L Fw2 R' D2 Fw2 R F2 R Fw' Rw2 U2 Fw D F2 Uw' L' D Rw' Uw B
Yau, 79 moves / redux, 57 moves
Scramble 2: B D' R2 D F2 U' B2 L2 F2 R2 U' B2 U' B D L D' R2 F2 L F Rw2 Fw2 B' U' Rw2 U' B Uw2 D2 Rw2 F2 U2 Rw' F' B L U Rw' Fw Uw' B' U' F2 L'
Yau, 70 moves / redux, 74 moves
Scramble 3: D U R2 U B2 F2 R2 F2 R2 U' R2 F D' B' R2 U L B' L2 F2 U Uw2 B R F2 L Uw2 Rw2 D2 F' R' F2 Rw2 Uw' L' F' L' Fw2 L' Rw Fw L Uw' Rw' Fw2
Yau, 61 moves / redux, 61 moves
Scramble 4: L' D' L' D2 B2 U' R' L2 B D2 L F2 D2 L U2 L' U2 L2 F2 B2 L' Uw2 Rw2 L B' F L U2 Fw2 Rw2 U2 F R2 Uw' F2 Uw2 L' F' Uw Fw L2 B2 Rw' Fw' R' U
Yau, 75 moves / redux, 63 moves

Average move count:
Yau: 71.25
Redux: 63.75

F2C solutions are shared between the Yau and redux solutions because doing F2C differently might introduce some of my subconscious bias. I did try to minimise move count (for both methods), but only within reason; these are not linear-FMC-style solves. I did not backtrack, and I did not trace edge cycles ahead of time. Move counts are in OBTM (i.e. slices count double). Two of the scrambles were done with Yau first then redux, and the other two were done with redux first then Yau, again to try to minimise my subconscious bias.

There are huge error bars here, and a sample size of 4 really is too small to definitively determine which is more efficient (plus, to reiterate again, I might be biased), but I think it at least demonstrates that Yau is not 5-10 moves more efficient than redux.

PS: I average about 55 with Yau and 60-ish with this specific variant of redux that I'm using here, and 47 with my own method.

 

[Nmile7300]

Thanks for doing this, I'm actually glad to be proven wrong.
Also, I'm curious what your own method is.

 

[brododragon]

Psst this thread is for 5/6/7; you should've posted this in the thread for 4.

It says in OP's post you can talk about 4x4.

 

[Gan Air SM]

Getting faster at big cubes
Hey guys, i average around
4x4: 1:20
5x5: 2:15
6x6: 5:30.
How can i get faster?
i use the valk 4m, mgc 5x5,6x6
plz help
and i use redux for all of them

 

[BenChristman1]

I personally don't know Yau (yet), but most people would say to learn Yau for 4x4 and maybe 5x5.

 

[Cody_Caston]

Ok first use yau on 4x4 and learn 3-2-3 edge pairing.

 

[xxpaulwangxx]

Use Yau4 and try out Yau5. also, get used to lookahead on edge pairing

 

[Cody_Caston]

Wait before you learn yau, what method do you use for 3x3

 

[xxpaulwangxx]

Yau works for Cfop and zz varients

 

[Cody_Caston]

Yau works for Cfop and zz varients

There’s a method called 4z4 but i personally don’t know if it’s worth it.

 

[Owen Morrison]

Yau works for Cfop and zz varients

It would be very hard to do ZZ as you wouldn't get inspection time for EO.

 

[Gan Air SM]

Oh yeah i use yau on 4x4, and i use cfop on 3x3

 

[Gan Air SM]

How can i get faster on yau4?