It was exactly three years ago this weekend that I solved a Rubik’s Cube for the first time. Escaping the craziness that can come with too much family over Thanksgiving weekend, I spent an hour in front of YouTube, scribbled some notes, turned on my webcam, and recorded this six-minute solve. I thought I was awesome.

Fast forward three years, and I’m not longer impressed by being able to solve cubes, even big ones. Speed is impressive (certainly the sub-10 folks blow my mind), but that’s never going to be me. Three years in, I’m still hovering at 30 seconds for a 3×3, and I’m actually ok with that. What impresses and intrigues me even further is deepening my knowledge; I’m always interested in learning new techniques and approaches (in my very limited spare time).

For a little while, now, I’ve been focussed on using the M slice to make easier and reduce rotations in certain “tough” F2L cases. These two — #15 and #16 — boiled to the top:

I’ve identified 8 other cases that are interesting contenders for M slice F2L. But some would be just as clunky as the conventional approaches. As Cyoubx very clearly articulates in this video, M slice (as part of forced rotationless) F2L taken to its extreme starts to approximate Roux — and bad Roux at that. I’m interested in exploring Roux more. But, for now, I think there are certain key F2L cases that can be converted to M and used in Fridrich/CFOP without veering into bad-Roux-land. While I continue to explore the others, I think these two work quite well.