Edges First Keyhole

[Cride5]

The basic concept: After cross place 3 E-slice edges, and then use the remaining 'slot' to solve three D-layer corners.

The idea certainly isn't new. I searched the forum for references to this idea and and found these, there may be others though:
http://www.speedsolving.com/forum/showthread.php?p=143534#post143534
http://www.speedsolving.com/forum/showthread.php?p=310149#post310149

The reason for the thread was just to draw attention to it because it seems like a better option than corners first. I've found while trying to speedsolve with 8355, that edge placement using regular keyhole is pretty bad for lookahead because (although technically not necessary), I do find myself looking at the D-layer quite a lot because I loose track of the working corner.

With edges-first, I can just insert 3-edges without bothering about a 'working corner'. I then rotate the cube so that the unsolved 'slot' is in FR. Finally a series of D-moves and regular corner-insertions completes the F2L-minus-slot. I don't need to look at the D-layer because all the information I need is on the sides of the cross edges.

Does anyone else use this with 8355 or LBL, and do you find it easier/faster?

 

[Lucas Garron]

Have you tried working corner with cross on left?

 

[Cride5]

That does improve it a little, but feels kinda awkward. I guess practice would remedy that though.

Do you think doing keyhole corners-first is actually better?

 

[riffz]

I actually prefer doing u moves when solving corners first as opposed to D moves for edges first. I Find looking at the middle layer to determine which edge should be inserted is easier for me than looking at the cross pieces.

However, I do like the fact that the U layer is not being rotated by u moves with edges first. I think this would give a slight edge in lookahead.

I don't think order really matters. It's probably just a matter of taste and I can't see there being any clear reasons why one is better than the other.

 

[Ranzha]

I found edges first LBL instead of corners first LBL was faster. It's 2-gen algorithms, so it's easy to finger trick.
I wouldn't personally think that edges first keyhole would be worth the trouble, but recognition would be better with it than with corners first keyhole.
The biggest downside is inserting the last E-layer edge after the FL is keyholed. For this reason, I find solving the entire E layer and inserting the corners once they're placed is faster.

That's the method I teach usually. Once I teach the corners first method to someone who I taught ECross LBL (the method described above) to, they find it easier to recognise the corners easier than the edges, especially in CN solves.

 

[Forte]

I thought this was normal keyhole ;-;