# New 3x3x3 Method

#### trinity.db

##### Member
Hey guys,

I wanted to propose a viable 3x3x3 method.

There are 4 steps.

Build a First Block
Build a Second Block
Solve Last 4 Corners
Permute Last 6 Blocks

Yes, I am being very selective in wording this. This method is in fact different, than what most may think. Although it is truly inspired none other than Gilles Roux.

When you realize the distinction in between the methods, I believe that you will be quite pleased. I have prepared a series of videos and examples.

L2 R2 F2 D' F2 D' U B2 D L2 R F D2 B' R' F' D2 B' U B2 L'

Example 1 & 2:

Example 3:

B R2 L' F R F R F2 D2 R L2 U R2 B2 L2 U' R2 F2 U'

Scramble 2: Example 1

In this video, I give a simple explanation of nmCmLL configuration recognition using just case --

The CmLL -- R U R' U' R' F R F'

The Inverse: F R' F' R U R U' R'

Applied with pre-moves:
• (U) F R' F' R U R U' R'
• (R2) F R' F' R U R U' R'
• (R2 U) F R' F' R U R U' R'
• (U2 R2) F R' F' R U R U' R'
• (U2 R2 U) F R' F' R U R U' R'

I will share with you that there are 40 algorithms to permute and orient the corners. There is a recognition system that I have been using for over 10 years now. I was brushing up on the cases and I realized as I transitions from 2nd Block to Solving the Last 4 corners, I found a common case. Which then I experimented with alternative blocks and just realized that I save moves in the 2nd block.

I realized that my trick cost me a few moves to "fix" later. Then I realized that I could actually save those moves for later and get cancelations!

Basically, I took the 3x3x3 method and combined it with FMC concepts to reduce the move count even more. It was inspired from my parity elimination method for 4x4x4 that I have previously written extensively on.

I've already recorded a bunch of videos showing examples of steps and viable alternative solutions.

So -- I am going to post this, upload the videos, give links for the cube solutions on a virtual cube, share algorithms for all ~80 Corner Configurations and explain further the recognition system for that step. However, feel free to comment while I compile this thorough guide this week.

Inspiration Points:

• Roux Speedsolving
• Athefre nmCmLL Recognition System
• Fewest Moves Challenge Pre-moves and Cancelations
• CFOP Algorithm R2 U2 R' U' R U' R2
• Pretty Pattern R2 U2 R2 U2 R2 U2
• CFOP Color Neutrality
In my method, or perception of the method -- I would say that it is truly inspired by Gilles Roux and his methods. There is another hybrid of the method where you can actually apply a pre-move of F2 (even f2, or B2) and still solve the cube. The nmCmLL is different -- however there is a "parity" that is created by the F2 off-set. This will be explained in greater detail as videos dive into the 40 CmLL and the relative 80 configurations.

I have not developed a comprehensive algorithm set to eliminate AUF moves or influence LSE. The speed of CmLL and freedom thereof, combined with the simplicity of Last 6 Blocks leaves me to believe that making CmLL more robust would be redundant.

Lastly, it is also viable to save even more moves by freely permuting the FR, FL, BR, and BL in any of their oriented positions. Essentially, solving the DL 1x1x3 block and the DR 1x1x3 block with any possible 4c Permutation on the E slice edges.
• as a disclaimer: M2 E M2 E' -- is not ideal.

Last edited:

#### Burrito

##### Member
Seems interesting, maybe put this ona blog post with pictures and cubing.js code so you can have a reconstruction on the site itself