i met a chap named Keven on saturday who uses the method. there was another person as well but i cant actually remember his name :x im not sure in either of them use the forums, but the unnamed one said that he has written a program to assist with the muscle memory of algs. he will probably post it in the reddit subgroup(?) soon maybe... im not too sure.
 
Mr. Waffle's Web is down since like yesterday or the day before :( It was my main source of learning material and now I can't get my daily dose of Roux. I hope it goes up again soon. Anybody knows what happened?
 
I have a question about bigger cubes for roux especially the 4x4. I recently started practicing 4x4 again and learned Yau it helped me get my times down a lot from around 2 mins to a minute and a half. I'm wondering how other people use Roux for solving bigger cubes and how to tackle the M slices; I usually end up with inner-layers misaligned and the cube locking up on me. Also is there a variation of Yau optimized for Roux solvers. I really like the idea of building the first 4 edges before centers and keeping them on the bottom so all the other edge species are on top and improve recognition and reduce move count a little bit (long sentence). Although this does not work as well for Roux as much once on to the 3x3 stage. I was wondering if anybody has developed a similar technique that helps with building the first one or two blocks or if I would be just as well off to stick with CFOP for bigger cubes.
 
Back
Top