2x2x2 Optimal Perm

[philkt731]

Does anyone know what the optimal solution to swtiching two corners while preserving orientation is? Is it 10 moves, or is there less?

 

[Lofty]

I'm pretty sure the A pll i use is only 9 moves that will switch two corners next to each other.

 

[magicsquares]

Don't A PLLs cycle 3 corners? Anyway, I think you should do the T-perm it's a pretty fast algorithm.

 

[Lofty]

They do cycle 3 corners.
But with no edges and from a certain perspective they switch two quite well.

 

[hdskull]

What do you guys use for the diagonal switch? I found Y on the 2x2 to be very slow.

 

[masterofthebass]

I got this from Yu Nakajima's page. It's actually not his, but from what I can tell a friend.

R U' R' U' F2 U' R U R' D R2

it switches URF and UBL. A good algorithm.

As for the 2 switches, J perm is probably the fastest, unless you can do As faster. Lofty was right. If you have the 2 piece switch, just do a U or U' and you'll end up with 1 corner solved and the others in a 3 cycle.

 

[philkt731]

yeah A is nine moves but you have to AUF so that it switches two. I use a variation of a J perm: x U2 R' U' R U2 L' U R' U' R2. That alg from nakaji is a variation of one I saw on erik's website, but I haven't gotten used to it. I use V Perm

 

[TimMc]

My bottom layer is intuitive, that is to say: I haven't bothered to convert the movements I commonly use into standard cube notation

For the Last Layer I just use:
- Niklas & Anti-Niklas
- J Perm

I'm also using the above algorithms primarily for a 4LLL on my 3x3x3

Tim.

 

[TobiasDaneels]

I got this from Yu Nakajima's page. It's actually not his, but from what I can tell a friend.

R U' R' U' F2 U' R U R' D R2

it switches URF and UBL. A good algorithm.

As for the 2 switches, J perm is probably the fastest, unless you can do As faster. Lofty was right. If you have the 2 piece switch, just do a U or U' and you'll end up with 1 corner solved and the others in a 3 cycle.

Man, that's a super algoritme.
That case is my worst at 2x2.
Now, it's ok.
Thanks

 

[Kenneth]

The one I use is also pretty fast I think:

R U' R' D' L2 D' L U L' D L2

Or mirror if you like that one better:

L' U L D R2 D R' U' R D' R2

The use of D-truns is to avoid "RUF", using D you can go "RDR" instead.

 

[Erik]

I got this from Yu Nakajima's page. It's actually not his, but from what I can tell a friend.

R U' R' U' F2 U' R U R' D R2

That algorithm probably comes from me. As far as I remember I was the first one to notate the last 2 moves as D R2 instead of U F2, though that alg was known among a little number of cubers already.

 

[cuBerBruce]

It requires 10 turns to swap two adjacent corners in the U layer (preserving orientation) with all 6 other cubies fixed. Of course, 9 turns (A-perm alg, for example) are sufficient if you don't care about keeping alignment between U and D layers.

For swapping two diagonally opposite corners in the U layer (preserving orientation), the maximum number of turns (11) is required, regardless of alignment between U and D layers.

For QTM, 13 quarter turns are required for adjacent corners, but 12 suffice if you don't care about keeping alignment between U and D layers. 13 quarter turns are also required for diagonally opposite corners, but the maximum of 14 quarter turns are required if a quarter turn of the U or D layer is also required along with the swap.

 

[philkt731]

I think I'm going to start using that alg instead of v perm. Also on the last R' I would probably do a z' at the same time so the R' changes into a U' as you roate it and the last two roations are R U2

 

[philkt731]

actually I think its better to turn z' while doing the F2 and finish off with a fast L' U L U' R U2