[Help Thread] 4x4 Discussion

[eeeeeeeee]

Anyone have a usable direct solution for [PLL parity + T-perm] ?

you might want to look at wong Chong wen's pll parity series on yt
@EPICGUYDUDE is his handle

 

[Renslay]

Okay, first of all, why all the Meyer threads are closed... I couldn't find any where I can ask my question!
Second, my actual question: does anyone have a recommended Meyer OLL-parity?

For those who don't know, Meyer is the Roux-equivalent of Yau: opposite centers, 1x3x4 block, remaining centers, edge pairing, finish as Roux with parities.
Since I check OLL parity after CMLL, I need an OLL parity that leaves the corners. But pure edge flips are long and complicated with lots of inner r and l moves.

This is my current one, which is a pure triple flip:
(Rw' U' Rw' U2') (Rw' U2 Rw' U' Rw U2) (r U2 Rw U' Rw' U2 Rw' U2') (Rw' U' Rw')

 Twizzle link 

(Rw' U' Rw' U2') (Rw' U2 Rw' U' Rw U2) (2R U2 Rw U' Rw' U2 Rw' U2') (Rw' U' Rw')

It's not bad, not much longer than a Yau-OLL, mostly wide moves. But is there any better? We have more freedom since the M layer can be scrambled.

 

[AJT17]

Okay, first of all, why all the Meyer threads are closed... I couldn't find any where I can ask my question!
Second, my actual question: does anyone have a recommended Meyer OLL-parity?

For those who don't know, Meyer is the Roux-equivalent of Yau: opposite centers, 1x3x4 block, remaining centers, edge pairing, finish as Roux with parities.
Since I check OLL parity after CMLL, I need an OLL parity that leaves the corners. But pure edge flips are long and complicated with lots of inner r and l moves.

This is my current one, which is a pure triple flip:
(Rw' U' Rw' U2') (Rw' U2 Rw' U' Rw U2) (r U2 Rw U' Rw' U2 Rw' U2') (Rw' U' Rw')

 Twizzle link 

(Rw' U' Rw' U2') (Rw' U2 Rw' U' Rw U2) (2R U2 Rw U' Rw' U2 Rw' U2') (Rw' U' Rw')

It's not bad, not much longer than a Yau-OLL, mostly wide moves. But is there any better? We have more freedom since the M layer can be scrambled.

This isn't a Roux/Meyer specific alg, but it does a single flip in the back without disturbing anything else, and it's pretty much equal to the standard parity alg.
Rw2' F 2R' F U2' F' 2R F U2' Rw U2 2R U2 Rw' F2 Rw2
Or it can be formatted like this for fingertrick purposes, eliminating the 2R moves in favour of Rw* R*
Rw2' F Rw' R F U2' F' Rw R' F U2' Rw U2 Rw R' U2 Rw' F2 Rw2
This alg also scales up to bigger cubes, and I originally found it as an inner-parity alg for 6x6 and 7x7.

 

[Christopher Mowla]

, my actual question: does anyone have a recommended Meyer OLL-parity?

For those who don't know, Meyer is the Roux-equivalent of Yau: opposite centers, 1x3x4 block, remaining centers, edge pairing, finish as Roux with parities.
Since I check OLL parity after CMLL, I need an OLL parity that leaves the corners. But pure edge flips are long and complicated with lots of inner r and l moves.
[...]

It's not bad, not much longer than a Yau-OLL, mostly wide moves. But is there any better? We have more freedom since the M layer can be scrambled.

As I wrote in this recent rCubers post, OLL parity algorithms which preserve the centers are 13 half turn moves (minimum). So we can only hope that structure (and inner slice turn count) are better in some algorithms than the standard ones.

You could convert 4 of the 7 inner slice turns in my 8-cycle (19,13) double parity alg from here, so that every other turn is a slice turn (and that the alg only contains 3 inner slice quarter turns):
r U2 2R' U2 r' D2 2R D2 r' B2 2R B2 r'

And although my (20,14) alg in that list contains a lot of inner slice turns, it's as simple in structure as it gets.
______________
Regarding algs with fewer inner slice turns (that aren't ridiculously long like Rw U2 Rw U Rw U2 Rw U2 Rw' U2 Rw' U 2R' U' Rw U2 Rw U2 Rw' U2 Rw' U 2R2 U2 Rw' U2 Rw' or Rw' U' Rw U' Rw U Rw2 U' Rw2 U2 Rw U2 Rw' U2 Rw U2 Rw U' Rw U Rw' U' Rw U2 Rw' U2), just yesterday I posted the following in a reply to the pinned comment of this video.
3Lw 3Rw U' 2R U' F2 U 2R' U' x U2 3Rw' U2 2R' U2 3Lw U2 3Rw2 x (23,17)

One odd nxnxn cube translation is:
4Lw 3Rw U' 2R U' F2 U 2R' U' x U2 3Rw' U2 2R' U2 4Lw U2 3Rw2 x
______________
Because I found the following (24,19) 3-flip double parity by hand (which somewhat resembles the structure of Kare's 3-flip that you use, which is 2 moves shorter, but contains an extra inner slice turn . . . from this list)
Rw' U2 2R U2 Rw' x' U2 2R' U' R' U' Rw' U2 Rw U R U' Rw R U2 x

I'm inclined to think that maybe @xyzzy may have (or may be able to find) something for you to try. (I recall that he was making algs for these types of cases with a custom solver he wrote some time ago.)

 

[Sub20Soon]

Around which time should i learn 3-2-3 edge pairing ? Right now im sub 2:10

 

[acrystal]

Around which time should i learn 3-2-3 edge pairing ? Right now im sub 2:10.

def rn bro

 

[xyzzy]

(I recall that he was making algs for these types of cases with a custom solver he wrote some time ago.)

pseudo-2-gen 1-flip OLL algs

pseudo-2-gen 1-flip OLL algs. GitHub Gist: instantly share code, notes, and snippets.

gist.github.com

I didn't gen 3-flips so it's possible that I missed something good in there, but the 1-flip algs that preserve CLL are pretty bad. (Caveat: these are meant for CFOP so they don't make use of the M slice freedom, and also they're RrU only.)

You could convert 4 of the 7 inner slice turns in my 8-cycle (19,13) double parity alg from here, so that every other turn is a slice turn (and that the alg only contains 3 inner slice quarter turns):
r U2 2R' U2 r' D2 2R D2 r' B2 2R B2 r'

This one seems okay, actually? Don't have a physical 4^3 with me right now to test though.

 

[Yann1qua12DagOat]

Around which time should i learn 3-2-3 edge pairing ? Right now im sub 2:10

I honestly would learn 3-2-3 and half centers immediately, because they're so instrumental at like sub-1:10 and beyond

 

[Renslay]

Thank you for the answers!

I re-wrote the 2R moves to (Rw R'), to make it easier for my eyes; and that's probably how I will execute them.

Rw2' F (Rw' R) F U2' F' (Rw R') F U2' Rw U2 (Rw R') U2 Rw' F2 Rw2
This one seems so far the best, I can execute it without regripping; the F moves a bit awkward. The most difficult part might be the U2' F' Rw, but I think I can make it work. My original alg requires 2 regrips, this one feels continuous, and even shorter!

Rw U2 (Rw' R) U2 Rw' D2 (Rw R') D2 Rw' B2 (Rw R') B2 Rw' R2
Although it is short, I find it hard to do D2 and B2 moves. Ring+pinky, maybe? But you have to do it multiple times...

Rw' U2 (Rw R') U2 Rw' x' U2 (Rw' R) U' R' U' Rw' U2 Rw U R U' Rw R U2 x
Not bad, but it requires at least one regrip at x', and another one after the last x if I want to continue the solve normally.

 

[Christopher Mowla]

Have you tried this alg too?

Lw Rw U' 2R U' F2 U 2R' U' x U2 Rw' U2 2R' U2 Lw U2 Rw2 x

 

[Renslay]

Have you tried this alg too?

Lw Rw U' 2R U' F2 U 2R' U' x U2 Rw' U2 2R' U2 Lw U2 Rw2 x

Looks promising, but I cannot do it fluently / without regripping. Somewhere around the middle I have to change.

 

[acrystal]

Have you tried this alg too?

Lw Rw U' 2R U' F2 U 2R' U' x U2 Rw' U2 2R' U2 Lw U2 Rw2 x

not that i am a part of this, but we can change the "Lw Rw" better

 

[Renslay]

Breaking down:

Lw Rw U' (Rw R') U' F2 -> this F2 bothers me and not sure how to execute it properly. I have to change grip just for this, and the following U 2R' is even more bothering.
U (Rw' R) U' x U2' -> here I have to regrip so I can fluently do the rest
Rw' U2 (Rw' R) U2 Lw U2 Rw2 x -> if I do the regrip, this part is really fast

 

[AJT17]

Rw2' F (Rw' R) F U2' F' (Rw R') F U2' Rw U2 (Rw R') U2 Rw' F2 Rw2
This one seems so far the best, I can execute it without regripping; the F moves a bit awkward. The most difficult part might be the U2' F' Rw, but I think I can make it work. My original alg requires 2 regrips, this one feels continuous, and even shorter!

I originally learnt this alg from this video. So it might help you to execute it better after watching someone go through the fingertricks.

 

[Christopher Mowla]

@Renslay , I can't claim that it's possible for someone to modify their custom solver to modify (and improve) on this, but algs like this do exist:

r' U M' U' M' U M' U M Uw2 l' Uw2 M' U M U' M' U M' U M U2 M' U2 r

• Post of origin
• Clearly that (double parity) alg can contain 1 inner slice turn, as making the first and last turns wide doesn't change anything.
• The first question is, are the moves Uw2 2L' Uw2 "speedsolvable".
• The next question is obviously if we can make this alg shorter (by allowing any permutation of the dedges).

(And for anyone who has little or no experience with using Cube Explorer, you can use Cube Explorer to find a bunch of alternate move sequences to replace the two MU portions, where you have to specify "Use Center Facelet Twists" from the Options menu.)
____________
My first guess is that this alg ^^ is unfortunately as good as it gets for this unconventional odd parity algorithm structure, but I hope I'm wrong!

 

[Renslay]

I originally learnt this alg from this video. So it might help you to execute it better after watching someone go through the fingertricks.

Ah, thank you very much! This is exactly what I was looking for.
Okay, so the main trick is to do the F' turn with your right index finger, and then you can do the alg fluently, without regrip. Thanks!

@Renslay , I can't claim that it's possible for someone to modify their custom solver to modify (and improve) on this, but algs like this do exist:

r' U M' U' M' U M' U M Uw2 l' Uw2 M' U M U' M' U M' U M U2 M' U2 r

• Post of origin
• Clearly that (double parity) alg can contain 1 inner slice turn, as making the first and last turns wide doesn't change anything.
• The first question is, are the moves Uw2 2L' Uw2 "speedsolvable".
• The next question is obviously if we can make this alg shorter (by allowing any permutation of the dedges).

(And for anyone who has little or no experience with using Cube Explorer, you can use Cube Explorer to find a bunch of alternate move sequences to replace the two MU portions, where you have to specify "Use Center Facelet Twists" from the Options menu.)
____________
My first guess is that this alg ^^ is unfortunately as good as it gets for this unconventional odd parity algorithm structure, but I hope I'm wrong!

I really appreciate this alg!
Not gonna use it for speedsolving, too many M moves; but it's quite interesting to see such a solution, where most of the moves just M and U, with so little exceptions, and it also does not insist on restoring the M layer.

 

[Christopher Mowla]

I really appreciate this alg!
Not gonna use it for speedsolving, too many M moves; but it's quite interesting to see such a solution, where most of the moves just M and U, with so little exceptions, and it also does not insist on restoring the M layer.

I hope no one ever considers using that for a speedsolving alg!

And its single parity counterpart is even longer! (by 3 STM)

r U' M' U M' U' M' U' M Uw2 r Uw2 U' M' U2 M' U M U2 M U M' U M U M' U' l'

(Same post of origin.)

But yeah, you got the point. It's just an alg to "illustrate the veracity of the 4x4x4 cube parity algorithm domain" (quoting myself from the 4x4x4 parity algorithms wiki page) and to motivate those who write custom solvers to maybe get some inspiration from that.

• This alg (and all others I mentioned in this conversation) are on that wiki page (FYI)

 

[Martyt7777]

Which is better, standard OLL parity alg from here, or U/U' first? (Ignore CO, CP, EP)

 

[AJT17]

View attachment 49046
Which is better, standard OLL parity alg from here, or U/U' first? (Ignore CO, CP, EP)

Deciding between line OLLs and L OLLs is pretty much just a matter of preference. If you like one set over the other, then choose that angle. If you are able to recognise what the OLL is gonna be from a specific angle easier than another, choose that angle as well.
Personally, I use this alg: Rw2' F Rw' R F U2' F' Rw R' F U2' Rw U2 Rw R' U2 Rw' F2 Rw2, which does pure parity, affecting only the edge in the back. This way I can look ahead to what OLL I'm gonna get, and by now I usually aim to reduce to line cases, since I find them slightly easier to recognise.

 

[Yann1qua12DagOat]

Is there a full OLL+ set? Because I know there's full 3 edges oriented, but for the single edge I can't find anything other than very easy, specific cases.

I know that it would probably be useless most of the time because the time that is spent to recognise the case as well as probably being super bad to execute would probably take longer than just parity from any angle and then the oll