Rotationlessness

[Akash Rupela]

While doing some analysis of solves, i realised that even the 3-4 rotations i make slow me down by half a second each, mainly due to fail regrips. I saw videos about rotationless f2l, but they had some crazy M moves or inserts using back face or double turns of front face.

So I was thinking about something like this. While inspecting for the cross, if rather than our 1st pair, we look for the edge orientation after the cross, i mean there are 4 edges, so if i know that 3 of them are oriented 1 way and 4th one the other way, I will solve those 3 first, do a cube rotation, then do another one. By this way, My solve will be totally rotationless or have 1 rotation(I like solving f2l in L/R/U) , depending on the edge orientations, plus i will have full control when that rotation will occur.
True, some inserts might be non optimal, but I believe this is a good way to strike off 1-2 seconds and have consistency in solves with a little practice.
Has anyone thought of/tried this before?

 

[Escher]

This used to be an element in Breandan Vallance's old f2l system, back when he was just breaking sub 10 (before almost anybody else, iirc).

It used to go like:

- Solve cross+BL corner
- Solve BL edge
- Depending on EO, either
a) solve BR, y' and solve remaining pairs in one look
b) solve FL, solve FR and BR pairs in one look

Usually he would include some kind of EO manipulation before the last two pairs stage, but I remember he went through a bunch of last two pairs cases and had some cool solutions for non-oriented stuff. He also created a huge amount of stuff to solve BL efficiently.

Nowadays I don't think it's a necessary strategy. Seeing as you're a CFOP user, you should go on qqtimer and practice the 2nd block of Roux for a while (R, r, U scrambles are an option I think). If you practice and mess around with applying M and wide turns enough you'll realise just how simple most of these rotationless tricks for f2l pairs actually are. Most just start by doing an M' to solve the EO of the bad edge, and then fix the rest.

Btw, I will practice doing EO tracking in inspection myself again, I haven't tried for a good year or two, you reminded me


Edit: also to anybody else reading this, if you didn't already know, if you limit your cross-building to the ZZ group, that is, <R, U, L, D, F2, B2> (pretty easy to do usually) it preserves EO and you don't have to track anything...

 

[speedpicker]

Escher, I have heard you mention Breandons old and new approaches to F2L before. What is his new approach and why did he feel it needed changing? Also, having spent an age on Macky's site I would be interested in which (if any) cases you felt it was appropriate to rotate for rather than perform a potentially awkward or longer algorithm. What is established "good form" with regard to F2L and rotations?

I just posted a lengthy reply to a question in the one questions answer thread looking at ways to practice rotationless insertions http://www.speedsolving.com/forum/showthread.php?8710-The-One-Answer-Question-Thread&p=720540&viewfull=1#post720540and others have waded in with some other ideas (x turns instead of ys and the like). Can you clarify some of these issues please, your expertise would be an invaluable contribution to what I think is a widely misunderstood topic (not least by me).

 

[jskyler91]

Escher, I have heard you mention Breandons old and new approaches to F2L before. What is his new approach and why did he feel it needed changing? Also, having spent an age on Macky's site I would be interested in which (if any) cases you felt it was appropriate to rotate for rather than perform a potentially awkward or longer algorithm. What is established "good form" with regard to F2L and rotations?

I just posted a lengthy reply to a question in the one questions answer thread looking at ways to practice rotationless insertions http://www.speedsolving.com/forum/showthread.php?8710-The-One-Answer-Question-Thread&p=720540&viewfull=1#post720540and others have waded in with some other ideas (x turns instead of ys and the like). Can you clarify some of these issues please, your expertise would be an invaluable contribution to what I think is a widely misunderstood topic (not least by me).

I know I am not Escher, but I think I can chime in on this as well. I think people often misuse the term rotationless when they mean regripless. Cube rotations are definitely something that should be cut down on, but trying to eliminate them completely can actually be a detriment to your solve, for CFOP at least. If done correctly, cube rotations allow you to see different parts of the cube and help to eliminate pauses. Regrips, on the other hand, are something which should be cut out in as many places as possible. They are a useless movement that can almost always be avoided and done at the same speed without the regrip. What most people don't realize, is that almost all regrips are actually two regrips in one i.e. one regrip to get to new position and one regrip to get back to home position. I see people regrip over 10 times in a solve (20 times if your think of it in terms of up and back), this is a HUGE waste of time!!! I don't know about you, but 20 regrips takes almost 4 seconds for me to do consecutively. Only in rare instances should they be done, such as for OLL/PLL algs where there is no fast way to do the alg from home grip and in certain f2l cases where it is impossible to do the case quickly without one. If you look at the super fast cubers like Felik's and Cornelius in particular, they almost never leave home grip, yet they do like 2-3 cube rotations per solve. I have tried to mimic this and I recommend you do as well.

 

[Escher]

Escher, I have heard you mention Breandons old and new approaches to F2L before. What is his new approach and why did he feel it needed changing? Also, having spent an age on Macky's site I would be interested in which (if any) cases you felt it was appropriate to rotate for rather than perform a potentially awkward or longer algorithm. What is established "good form" with regard to F2L and rotations?

I just posted a lengthy reply to a question in the one questions answer thread looking at ways to practice rotationless insertions http://www.speedsolving.com/forum/showthread.php?8710-The-One-Answer-Question-Thread&p=720540&viewfull=1#post720540and others have waded in with some other ideas (x turns instead of ys and the like). Can you clarify some of these issues please, your expertise would be an invaluable contribution to what I think is a widely misunderstood topic (not least by me).

Just writing to say I've read this - I will try and get back to you soon, no time now. This might go on my list for more how-tos I want to write soon actually.

 

[speedpicker]

Thank you Escher, I am really looking forward to seeing what you come up with.

jskyler91: That is an interesting idea, thinking about regrips rather than rotations. Could you cite an example perhaps where the standard solution is inferior to another, perhaps longer solution due to this regrip concept? Having spent some time trying to mimic the style of the cubers you have mentioned i am starting to get an inkling of what you mean, but I think some specifics would really help me here.

 

[jskyler91]

Thank you Escher, I am really looking forward to seeing what you come up with.

jskyler91: That is an interesting idea, thinking about regrips rather than rotations. Could you cite an example perhaps where the standard solution is inferior to another, perhaps longer solution due to this regrip concept? Having spent some time trying to mimic the style of the cubers you have mentioned i am starting to get an inkling of what you mean, but I think some specifics would really help me here.

The problem of giving examples is that the move count will often be the same because you are just regriping, but the time of the solve will be slower because you are doing unnecessary regips. Watch Feliks solve in slow motion, he rarely regips to do most situations and often uses index pushes to avoid those places where most regrip. Also, see my "Advanced" fingertrick video, the f2l section, I talk about it there and i give some examples of where you dont need to regrip: http://www.speedsolving.com/forum/s...Advanced-quot-Fingertricks-(Cross-F2L-OLL-PLL)