This is just a brief story about my experience trying more advanced methods for edges in blindfolded solving. I normally use M2 for edges and Orozco for corners. I decided I wanted to try out Orozco and Eka for edges, to see if I prefer one of these over using M2.
I should note before I continue, that I do use a lot of special cases when doing M2. Rarely will I do (setup target 1) M2 (undo setup), (setup target 2) M2 (undo setup). Instead, I tend to solve two pieces at a time. A lot of my cases involve setting up to U M’ U2 M U with R/L moves, or setting up to a commutator with E or S-slice interchange like R’ F’ R S R’ F R S’.
Anyway, after trying Orozco and Eka, I decided that I liked Orozco better (likely because I’m so used to the logic after using it for corners). I also had trouble remembering my setup moves for Eka, and although this would have gotten better with practice, I didn’t like the idea of setting up to a suboptimal commutator. So I continued trying it Orozco in solves, until I realized that I am much slower using Orozco rather than M2.
At first I had thought this difference was due to a lack of experience with Orozco commutators. Surely with more practice, I thought I would get faster. However, I realized that the problem was that Orozco is still solving one piece at a time, whereas with M2 (or rather, my variation on it), I was solving two pieces at a time for all but the easiest standard cases (mainly when one of my targets is A/UB).
All this to say that just because a method is more advanced, does not make it faster. Also, it is never too early to start incorporating commutators in your solves, especially if it’s for easier cases like the ones I mentioned above. I think familiarity with the letter scheme is much more important for becoming a fast blind solver.
I partially agree with what you said. I think it all depends on how fast you’re want to be. A lot of times an advanced method will be much slower for a while then the previous one just because you have to get used to it, it usually takes time. Also I wouldn’t call orozco a more advanced method than M2 (maybe in terms of the buffer it can be more advanced since UF is objectively better than DF, but other than that I think there’s not much difference). To be fair it all depends on how fast someone wante to become. If the goal would be for example to be sub30 or faster some day, learning good comms right from the start would be the best (many people learn orozco/M2/turbo and use the intermediate method when they don’t know the comm for some case, but sometimes the methods leave some habits that have to be changed the more a person knows 3style). Overall optimal 3style algs are the way to go to be fast.
Yes, learning 3style is definitely the way to go for those who want to get faster and are willing/able to put in the time to learn it.
I also agree that learning an advanced method can cause times to rise, only to decrease much lower after some practice. However in the example I was using, the “advanced methods” were either solving one piece at a time or had worse setups than what I was already doing, so that was not the same phenomenon.
I think my original post is somewhat misleading, as I said I was using M2 but the method I was using is closer to the Diadem method for edges, or 3style with a DF buffer. I say this because I was using commutators/setups to commutators for most cases, rather than using vanilla M2. This didn’t occur to me at the time, and I hadn’t heard of the Diadem method when I posted this so I didn’t realize what I was doing was very similar to this method.
A method that solves one piece at a time, is not an advanced method. Orozco is not more advanced than M2. Orozco is an intermediate method.
The best method is a freestyle method that incorporates many methods, the 3 style, 5 style, M2 and advanced M slice algs, BH, swaps, simple commutators and conjugates, and maybe even orosco and any other method that can solve two pieces at once with less than 10 half turns. What you are doing is the right thing in modifying M2 to work for you. The great thing is that most of these methods work well together. There are bad cases in every method, and great cases. If you think about solving two pieces at once as you are doing, you can avoid the bad cases by changing method for that single cycle if one method gives you a good case. It’s possible to solve any two faces with 10 half turns or less if you aren’t a slave to any one method. Heck, it’s possible to solve four edges at once with variations and setups to the simple 5 style algorithm M’UM U’.