Could there be a more efficient/faster 2 look LL method?

[peterbone]

For layer by layer methods there are currently two 2 look methods that I know of. These are OLL-PLL and CLL-ELL. However, there are many other ways to split the last layer into 2 steps. Has anyone looked into these to see if one of them is more efficient / has faster algs than either OLL-PLL or CLL-ELL?
One example I can think of is first permuting and orientating 2 adjacent corners and the adjoining edges in a single step (a C shape) and then solving the remaining 2 corners and edge in one step (I shape). One advantage here would be that you have 4 options for choosing the first 2 corners and so can choose the most efficient. However, recognition may be a problem and I guess that the number of algs would be more that OLL-PLL.
Another option could be to orientate all corners while permuting all edges followed by orientating all edges while permuting all corners.
There are several other options for splitting up the last layer into 2 stages so I'm just wondering how much they've been investigated? Who knows, one of them may have a low average move count with algs that are good for high TPS.

 

[RNewms27]

http://www.speedsolving.com/wiki/index.php/Snyder_Method

Small amount of moves used + decent LL execution, but recognizing the cases is very difficult. Solve 4 edges + 1 corner -> solve remaining corners.

For a human, Oll-Pll seems most efficient overall.

 

[Godmil]

first permuting and orientating 2 adjacent corners and the adjoining edges in a single step (a C shape) and then solving the remaining 2 corners and edge in one step (I shape).

Wait, how many edges are you solving in the first stage and second stage? Cause it sounds like 3 and 1, which isn't possible.

 

[peterbone]

Good point Godmil. I hadn't really thought it through. I guess you just solve the 2 remaining corners in the last step then.
RNewms27, according to the Snyder method, it solves one LL edge while placing the final F2L pair so it's not exactly a LL method is it? You say that OLL-PLL seems the most efficient overall. Does that mean that all other ways of splitting the final layer have been shown to be inferior?

 

[Godmil]

Good point Godmil. I hadn't really thought it through. I guess you just solve the 2 remaining corners in the last step then.

Fraid that's still only possible if the last two corners are twisted... you can't permute them. You could be getting at a two corners/two edges... another two corners/two edges, but I'd imagine that's a ton of algs (I think Aaron posted something recently with some of those algs).

One of Snyders things is to do the edges + one corner, then do L3C... but that's a ton of algs too.

 

[mariano.aquino]

First you orient all edges while you insert last pair. You also make sure one edge ends up in its correct place. Then you solve remaining three edges, making sure one corner is solved as well. Finally, you permute last three corners. I think it's called fish and chips =)

 

[peterbone]

The problem with the two corners/two edges is that it's better for the first stage to solve more since it can make use of the fact that it doesn't need to preserve the second stage.
I'm starting to see that most of the variations have been considered though. How about the other idea I mentioned (OE+PC / PE+OC or visa-versa)?
I'm not sure that recognition is a big issue. I do full CLL-ELL which has harder recognition than OLL-PLL but once you've learnt it, it can be just as quick.

 

[irontwig]

RNewms27, according to the Snyder method, it solves one LL edge while placing the final F2L pair so it's not exactly a LL method is it?

No, you only need to avoid to get four flipped edges, which is easy. The recog for the first LL look is a pain though when you're speeding. It's pretty much the perfect thing to do though when you're FMCing.