new method

[pjk]

I actually think you can avg. around 17 seconds with it, maybe even quicker. With practice, it seems like you could go extremely quick. What I would do is solve the middle all but one edge, separate white and yellow, insert the last middle layer piece, oriente all the pieces on both sides, and permute them. I can do this around 40 seconds right now with very little practice.

 

[Erik]

hey, this was one of my first own methods, but I did it a bit different and I always got stuck on the parities.
How I did it:
1. first get all the white cubies to their layer besides 1 corner (avg is very few moves)
2. solve M edges with keyhole principle and the last one 'normaly' (together with the last edge, (like F2L principle)
3. solve the U layer + parity
4. solve the D layer.

I just tried it again, but now with knowing the parities and got about low 30's.
Best time so far 27 seconds.

Edit: 25 seconds

 

[pjk]

Very nice. I think that method has some potential if someone really went and practiced it.