OLL vs OLL + PLL Parity Algs for 4x4x4

[AndyK]

What is the consensus about using a pure OLL parity fixer vs the double OLL + PLL parity alg?

I use these two algs:
OLL + PLL Parity - Rw2 B2 Rw' U2 Rw' U2 x' U2 Rw' U2 Rw U2 Rw' U2 Rw2 U2
Only OLL Parity - Rw2 B2 (U2 Lw U2 Rw')(U2 Rw U2)(F2 Rw F2 Lw') B2 Rw2
and much prefer executing the OLL + PLL alg so I use it exclusively. I reason that this is fine to do because PLL parity occurs at 50% frequency and so switching PLL parity blindly at the OLL stage with the OLL + PLL alg won't increase the chance of me seeing PLL parity when I get to it.

Do most recognize PLL parity at the OLL stage in the 4x4x4?

 

[darthyody]

Personally, if I have OLL parity at the OLL stage, I fix it and then the same thing at PLL stage. However, very often I feel as though I'm wasting time when both parities show up. I've been thinking of this very problem myself and this is what I think would be best:

Step 1: Perform OLL alg* (if there is parity, visualizing 1 of the edges as flipped gives you up to 4 different choices of OLL)
Step 2: If parity exists, choose which parity fix is needed. # (OLL, PLL, or OLL+PLL)
Step 3: Perform PLL alg and then you are finished.

*Execute an OLL alg whether or not there is OLL parity.
If there is OLL parity, then all you have to do is pretend that only ONE edge pair is flipped and execute that OLL alg. It is best if you visualize the case the would have the most edges correct, best case being all 4 (never leave zero edges flipped, it wastes too much time.) LL should be oriented now except for that ONE edge pair.
If there isn't OLL parity, then you must execute that OLL case.
#Now by looking around the LL, you can easily identify which parity case you have, if any.

 

[rachmaninovian]

unless you try syuhei omura's LL method for 4x4, where he tries to fix both parities at the same time, I guess using the OLL+PLL parity algorithm to fix the OLL parity is okay...
after all there's still a 50% chance of getting PLL parity...
and I don't think many would bother to check for PLL parity during OLL parity..

 

[JLarsen]

I think I'm going to start doing oll before solving parity too now. Its got to be a piece of cake for petrus. I don't look at edges for oll so it should be pretty easy. I'll go try.

Edit: Pll parity was easy to identify with one flipped edge, and oll more so. The only problem is I now need an OLL parity alg that perserves the rest of the last layer orientation.

Edit 2: Does anyone know an alg to preserve oll and only flip one edge?

 

[darthyody]

I think I'm going to start doing oll before solving parity too now. Its got to be a piece of cake for petrus. I don't look at edges for oll so it should be pretty easy. I'll go try.

Edit: Pll parity was easy to identify with one flipped edge, and oll more so. The only problem is I now need an OLL parity alg that perserves the rest of the last layer orientation.

Edit 2: Does anyone know an alg to preserve oll and only flip one edge?

r' U2 l F2 l' F2 r2 U2 r U2 r' U2 F2 r2 F2

 

[JLarsen]

That alg twists 2 corners. Maybe I'm wrong I'll try again. Nope still twisted. Perhaps its only slice turns? Checks* Yup. But that's incredibly slow is it not?

 

[darthyody]

For me, it's less than a 2 second difference in using the alg with slice moves vs not slice moves.

 

[rachmaninovian]

That alg twists 2 corners. Maybe I'm wrong I'll try again. Nope still twisted. Perhaps its only slice turns? Checks* Yup. But that's incredibly slow is it not?

sub 4.5 if you practise =) I do it pure for cage...=P and its not incredibly slow...it can be fast too...like 3.81 xD
the key is no/minimal pauses in execution and as little lockups as possible..

 

[deathmaths11]

What is the consensus about using a pure OLL parity fixer vs the double OLL + PLL parity alg?

I use these two algs:
OLL + PLL Parity - Rw2 B2 Rw' U2 Rw' U2 x' U2 Rw' U2 Rw U2 Rw' U2 Rw2 U2
Only OLL Parity - Rw2 B2 (U2 Lw U2 Rw')(U2 Rw U2)(F2 Rw F2 Lw') B2 Rw2
and much prefer executing the OLL + PLL alg so I use it exclusively. I reason that this is fine to do because PLL parity occurs at 50% frequency and so switching PLL parity blindly at the OLL stage with the OLL + PLL alg won't increase the chance of me seeing PLL parity when I get to it.

Do most recognize PLL parity at the OLL stage in the 4x4x4?

I use OLL + PLL Parity - Rw2 B2 Rw' U2 Rw' U2 x' U2 Rw' U2 Rw U2 Rw' U2 Rw2 U2 but after the OLL finished l have the PLL parity and the same alg won't work for some reason or didn't solve both parities at the same time because l looked at the website http://www.speedsolving.com/wiki/index.php/4x4x4_Parity_Algorithms says it solves them both simultaneously and l might not fully understand that?

 

[JonnyWhoopes]

Goodness, what's with all the super old bumps recently?

Anyway, the reason it doesn't work is because you may not have PLL parity in every solve. If you don't have it, and you use an algorithm that would "fix" the parity, then you are actually creating it instead.

 

[yoshinator]

Two really nice pure OLL parity lags are:
R U2 M' r U2 r U2 r' U2 l U2 r' U2 r U2 r' U2 r'
And
r U2 r U2 r' U2 r U2 l' U2 r U2 r' U2 M r' U2 r'

Waffles can sub-3 the second one, and I really like the first.

 

[deathmaths11]

Goodness, what's with all the super old bumps recently?

Anyway, the reason it doesn't work is because you may not have PLL parity in every solve. If you don't have it, and you use an algorithm that would "fix" the parity, then you are actually creating it instead.

So when l have OLL parity and use that alg, then l make a PLL parity? But isn't mine supposed to solve then both still? If not can l still use the same alg: Rw2 B2 Rw' U2 Rw' U2 x' U2 Rw' U2 Rw U2 Rw' U2 Rw2 U2 for PLL parity as l would use in OLL parity? Or do l need another PLL parity?

 

[JonnyWhoopes]

So I actually used that alg for the first time after your last post, and that is not a double parity (OLL parity and PLL parity at the same time) algorithm. It only solves OLL parity.

::EDIT:: So... I suck. It is a double parity alg.

 

[cubernya]

So I actually used that alg for the first time after your last post, and that is not a double parity (OLL parity and PLL parity at the same time) algorithm. It only solves OLL parity.

Not true. It swaps 2 corners without switching any edges

 

[Christopher Mowla]

So when l have OLL parity and use that alg, then l make a PLL parity? But isn't mine supposed to solve then both still? If not can l still use the same alg: Rw2 B2 Rw' U2 Rw' U2 x' U2 Rw' U2 Rw U2 Rw' U2 Rw2 U2 for PLL parity as l would use in OLL parity? Or do l need another PLL parity?

In every solve, there is a 50% chance that you will get OLL Parity, and a 50% chance that you will get PLL Parity. Getting either one is INDEPENDENT from the other.

The algorithm you have does solve both PLL Parity and OLL Parity at the same time.

If you do not have both OLL Parity and PLL Parity, then using that algorithm will be counterproductive.

Let your alg be called "OLL Parity + PLL Parity".

Scenario 1: Your cube has OLL Parity only.
OLL Parity + "OLL Parity + PLL Parity" = PLL Parity. So you must use a PLL Parity Algorithm to get rid of parities.

Scenario 2: Your cube has OLL Parity and PLL Parity.
OLL Parity + PLL Parity + "OLL Parity + PLL Parity" = no parity. You can solve as a 3x3x3.

Scenario 3: Your cube has no parity. (You obviously wouldn't do your alg for this scenario).

Scenario 4: Your cube has PLL Parity (You obviously wouldn't do your alg for this scenario, but you would need to use a PLL Parity algorithm to be able to solve as a 3x3x3.)

 

[deathmaths11]

In every solve, there is a 50% chance that you will get OLL Parity, and a 50% chance that you will get PLL Parity. Getting either one is INDEPENDENT from the other.

The algorithm you have does solve both PLL Parity and OLL Parity at the same time.

If you do not have both OLL Parity and PLL Parity, then using that algorithm will be counterproductive.

Let your alg be called "OLL Parity + PLL Parity".

Scenario 1: Your cube has OLL Parity only.
OLL Parity + "OLL Parity + PLL Parity" = PLL Parity. So you must use a PLL Parity Algorithm to get rid of parities.

Scenario 2: Your cube has OLL Parity and PLL Parity.
OLL Parity + PLL Parity + "OLL Parity + PLL Parity" = no parity. You can solve as a 3x3x3.

Scenario 3: Your cube has no parity. (You obviously wouldn't do your alg for this scenario).

Scenario 4: Your cube has PLL Parity (You obviously wouldn't do your alg for this scenario, but you would need to use a PLL Parity algorithm to be able to solve as a 3x3x3.)

Thanks! I really understand so l kinda like the algorithm you think it's good for me since it has chance to solve both? if not l can use for OLL then get another one for PLL? I think it's good since it's giving me the chance to solve both instead of one at a time the whole time right? Sounds reasonable?

 

[Christopher Mowla]

Thanks! I really understand so l kinda like the algorithm you think it's good for me since it has chance to solve both? if not l can use for OLL then get another one for PLL? I think it's good since it's giving me the chance to solve both instead of one at a time the whole time right? Sounds reasonable?

It all depends on your solving method. As others have already said in this thread, if you don't make the effort to recognize if you have PLL Parity before you do your OLL Parity algorithm, then you really have nothing to lose because you still have a 50% chance of getting PLL Parity anyway.

 

[Noahaha]

Does anyone use CLL ELL for 4x4 to recognize double parity?

 

[cubernya]

Does anyone use CLL ELL for 4x4 to recognize double parity?

That's actually an interesting idea. That would be kinda cool to be able to easily recognise the parities at speed. The problem is that the parity shouldn't effect the corners.

Side note: Any list of PLLs that have parity (even if it's only EP)

 

[JonnyWhoopes]

That's actually an interesting idea. That would be kinda cool to be able to easily recognise the parities at speed. The problem is that the parity shouldn't effect the corners.

How is that a problem? For those of us who use Roux for 4x4, unaffected corner permutation poses no problem at all.