• Welcome to the Speedsolving.com, home of the web's largest puzzle community!
    You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

    Registration is fast, simple and absolutely free so please, join our community of 35,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!

Roux Method Creation (Interview with Gilles Roux)

pjk

Administrator
Staff member
Joined
Mar 13, 2006
Messages
6,499
WCA
2007KELL02
This interview was originally posted March 14th, 2012, by Sillas Tsutsui da Silva (@Sillas)

Well, this is an interview with Gilles Roux, who is primarily well-known for inventing an efficient 3x3 method and becoming very fast with it. The creator of Roux Method (wiki).

Location: France
Date: 14/03/2012


- What motivated you to developed a new method?
I wanted to find a personal way of solving the cube. As simple and efficient as possible. I was trying to learn the F2L+OLL+PLL algorithm. But because of my limited number of neurons, I thought I would never learn all the orientation sequences.

- When you had the idea to create your method?
Summer 2003.

- Which methods you knew before creating it?
As many others, I explored all possible ways of solving the cube, understanding the pros and the cons of every technique, trying to imagine as others all combinations of pieces for every step, reinventing the wheel, inventing things published later by others.

Before I decided to use my technique, I was rather fast following Lars Petrus' approach.

- You expect there may be some improvement in your method?
There are potential improvements, but it's hard to tell whether they can lead to faster times or not. Simplicity makes looking ahead easy. One day, perhaps, a crazy cuber will be able deal with random unmatching 1x2x3 blocks, early edges orientation, optimal 'last 6 edges' solving and other tricks... just not me.

- What challenges and obstacles you had to face to create it?
- Where and when you spent the most time developing your method?
- What are the flaws you had to correct or rework before finalizing?
- How long did it take you to develop your method from the time you started?

It was mainly a matter of having the basic idea. It's not a work that costed me 15 years and a leg, you know.

Let me just tell you how it came, if I can remember correctly.

As a beginner, I used to finish my solves with the edges of the last layer. Some sequences involved use a lot of inner slice moves I found rather understandable and ergonomical. I missed those sequences. And I saw that keeping 5 or 6 edges for the last part didn't make things much more complicated.

The idea of an "edges last" method was nothing but new of course. Before that, I had already analyzed "corners first" methods. In my opinion, they were relatively weak precisely because they were "corners first" -> Edges all around the cube, difficult to track.

I thought that before ending solves with M and U moves that make your cube look a "T", I could just build the first two 1x2x3 blocks in an efficient way, taking care of 2 or 3 pieces at a time.

I wanted to make the method as simple as possible and I was happy to realize how simple it is to orient the last 6 edges using very few tricks.

The last thing I needed to find in order to make the method elegant was to find a new way to solve the last corners. I hate to learn tons of stupid move sequences. But I failed. It took me 6 months to learn all corner sequences.

I remember when I discussed the limits of this algorithm with people on the forums. I was hoping for 16 seconds. The same limit I imagined for F2L+OLL+PL. :)

- How you systematized and developed the method?
Well, I thought you had the answer when I explained how I had the ideas of the different substeps.

There's not much to discribe for the first blocks, it's mainly efficient intuitive block building like in other methods for the first layers, 2 or 3 pieces at a time. I proposed sequences for many cases you can face while building the second block, but you can discover them naturally with practice.

For the last edges, orientations and solving the last L-edge and R-edge are just easy tricks.

And for the last corners, it's nothing more than sequences. Jelinek's ACube helped me to find good QTM-friendly sequences.

- Do you really expect that your method was considered one of the mosts efficient and popular?
I didn't expect it to be popular since it did not seem faster than the 'standard' speedsolving algorithm.

I told you I expected approximately similar times for this technique and F2L+OLL+PLL. But sometimes it's hard to tell, especially when you suck at solving fast.

Thank you so much for your patience to answer that interview.
"I believe in intuition and inspiration. Imagination is more important than knowledge. For knowledge is limited, whereas imagination embraces the entire world, stimulating progress, giving birth to evolution. It is, strictly speaking, a real factor in scientific research." Albert Einstein

Gilles Roux's tutorial
Waffle's Roux Tutorial
5BLD's and PandaCuber's Roux Tutorial
Forum Thread about that interview

By: Sillas Tsutsui da Silva
 
Last edited:

Rainger

Member
Joined
Apr 4, 2020
Messages
72
This interview was originally posted March 14th, 2012, by Sillas Tsutsui da Silva (@Sillas)

Well, this is an interview with Gilles Roux, who is primarily well-known for inventing an efficient 3x3 method and becoming very fast with it. The creator of Roux Method (wiki).

Location: France
Date: 14/03/2012


- What motivated you to developed a new method?
I wanted to find a personal way of solving the cube. As simple and efficient as possible. I was trying to learn the F2L+OLL+PLL algorithm. But because of my limited number of neurons, I thought I would never learn all the orientation sequences.

- When you had the idea to create your method?
Summer 2003.

- Which methods you knew before creating it?
As many others, I explored all possible ways of solving the cube, understanding the pros and the cons of every technique, trying to imagine as others all combinations of pieces for every step, reinventing the wheel, inventing things published later by others.

Before I decided to use my technique, I was rather fast following Lars Petrus' approach.

- You expect there may be some improvement in your method?
There are potential improvements, but it's hard to tell whether they can lead to faster times or not. Simplicity makes looking ahead easy. One day, perhaps, a crazy cuber will be able deal with random unmatching 1x2x3 blocks, early edges orientation, optimal 'last 6 edges' solving and other tricks... just not me.

- What challenges and obstacles you had to face to create it?
- Where and when you spent the most time developing your method?
- What are the flaws you had to correct or rework before finalizing?
- How long did it take you to develop your method from the time you started?

It was mainly a matter of having the basic idea. It's not a work that costed me 15 years and a leg, you know.

Let me just tell you how it came, if I can remember correctly.

As a beginner, I used to finish my solves with the edges of the last layer. Some sequences involved use a lot of inner slice moves I found rather understandable and ergonomical. I missed those sequences. And I saw that keeping 5 or 6 edges for the last part didn't make things much more complicated.

The idea of an "edges last" method was nothing but new of course. Before that, I had already analyzed "corners first" methods. In my opinion, they were relatively weak precisely because they were "corners first" -> Edges all around the cube, difficult to track.

I thought that before ending solves with M and U moves that make your cube look a "T", I could just build the first two 1x2x3 blocks in an efficient way, taking care of 2 or 3 pieces at a time.

I wanted to make the method as simple as possible and I was happy to realize how simple it is to orient the last 6 edges using very few tricks.

The last thing I needed to find in order to make the method elegant was to find a new way to solve the last corners. I hate to learn tons of stupid move sequences. But I failed. It took me 6 months to learn all corner sequences.

I remember when I discussed the limits of this algorithm with people on the forums. I was hoping for 16 seconds. The same limit I imagined for F2L+OLL+PL. :)

- How you systematized and developed the method?
Well, I thought you had the answer when I explained how I had the ideas of the different substeps.

There's not much to discribe for the first blocks, it's mainly efficient intuitive block building like in other methods for the first layers, 2 or 3 pieces at a time. I proposed sequences for many cases you can face while building the second block, but you can discover them naturally with practice.

For the last edges, orientations and solving the last L-edge and R-edge are just easy tricks.

And for the last corners, it's nothing more than sequences. Jelinek's ACube helped me to find good QTM-friendly sequences.

- Do you really expect that your method was considered one of the mosts efficient and popular?
I didn't expect it to be popular since it did not seem faster than the 'standard' speedsolving algorithm.

I told you I expected approximately similar times for this technique and F2L+OLL+PLL. But sometimes it's hard to tell, especially when you suck at solving fast.

Thank you so much for your patience to answer that interview.
"I believe in intuition and inspiration. Imagination is more important than knowledge. For knowledge is limited, whereas imagination embraces the entire world, stimulating progress, giving birth to evolution. It is, strictly speaking, a real factor in scientific research." Albert Einstein

Gilles Roux's tutorial
Waffle's Roux Tutorial
5BLD's and PandaCuber's Roux Tutorial
Forum Thread about that interview

By: Sillas Tsutsui da Silva
This is awesome!
 
Top