Sub-20 with Less common methods

[cheiney]

I use Petrus and I personally feel it's easier to track pieces for a 2x2x3 while making the 2x2x2 block than tracking F2L pairs while making the cross. This allows me to start out with a much better advantage. I use OLL/PLL, however, with my Petrus. That way I only need to know like 6 or so algorithms for OLL. I'm not too fond of Roux method because although the M' move is easy for me, the M move just feels a little too awkward.

 

[emu]

I think that most people used to Petrus or Fridrich are slow at M moves and fast at R moves. But I think the gap can be closed by practice.

In theory, I believe Roux is superior to Petrus. The goal of both methods is to first reduce the cube to a state where two sides (or slices) can be turned. In Petrus, it is the 2x2x3 block, and Roux has the two opposite 1x2x3 blocks. In Petrus, there are 4 center pieces solved, 5 edge pieces solved, and 2 corner pieces solved. But in Roux, there are 2 center pieces solved, 6 edge pieces solved, and 4 corner pieces solved. What does that tell us? After the first two steps, Roux has three more pieces solved total. Petrus has two more center pieces solved, but center pieces are the easiest to solve since they need not be oriented, but only permuted. Corner pieces are much more difficult to solve.

So in conclusion, I just think that Roux is the better block-building method. There are probably a lot more people using Petrus than Roux. Even so, Roux users have managed to achieve sub-15s, even in competition (including Roux himself).

 

[fanwuq]

Emu,
You make some nice points, but Roux (1*2*3)*2 is not equivalent to Petrus (2*2*3). Petrus 2*2*3 takes less time and moves.
Roux (1*2*3)*2 is probably more like Fridrich without the last slot.

 

[JLarsen]

Wow Emu those are some nice points. Honestly I think Petrus takes the most thought, and therefore inhibits ones ability to look ahead. This is why a Petrus solve starts off really slow. I need to go slow during 2x2x2 and try my best to follow the pieces for 2x2x3. It really is tough to see a sub 6 2x2x2 with no rotation, and track all 3 pieces for 2x2x3, and put them all in without pausing. I mean really tough to do that. I personally agree when you look at time potentials Roux is as fast if not faster than Fridrich. Also the 1x2x3 blocks are easier and more ergonomic than the completely free style petrus type block building.

So in order of difficulty I firmly believe it goes

Petrus Toughest

Roux

Fridrich Easiest

However there is the fun argument that we can go into....but I'll spare it.

 

[spdcbr]

Excuse me...I'm a sub-17 Roux user. So don't say it's bad. I'm also a sub-25 petrus user so don't say I don't know anything about petrus. If you can (R,U) faster than you can (M,U) then go petrus but if can (M,U) faster than (R,U) like me then go Roux

What? Doing M,U faster than R,U is impossible. Maybe M',U would make more sense?

 

[JLarsen]

Excuse me...I'm a sub-17 Roux user. So don't say it's bad. I'm also a sub-25 petrus user so don't say I don't know anything about petrus. If you can (R,U) faster than you can (M,U) then go petrus but if can (M,U) faster than (R,U) like me then go Roux

What? Doing M,U faster than R,U is impossible. Maybe M',U would make more sense?

Regardless once you move away from classic 3x3 speed, MU is slower no matter what.

 

[StachuK1992]

That's an absolute lie.

 

[fanwuq]

That's an absolute lie.

False until you find some one who can do H-perm faster than Sune.

Pandaman,
You suck at block building. Try to do 2x2x3 in one look and that can probably reduce 3 seconds.
(Don't worry, I'm worse.)

Edit:

yes, but there are regrips in a 'normal 3x3 solve.
There's like no regrips in a MU 'solve'

Good point. It depends on one's fingertrick style. Sometimes regripping can be an advantage and can allow for really creative fingertricks. For me, RU is about 1.3 times faster than MU.

 

[StachuK1992]

yes, but there are regrips in a 'normal 3x3 solve.
There's like no regrips in a MU 'solve'

 

[ChaosWZ]

theres no real faster or slower method, if you start with petrus, thatll most likely be your best method same if you start with roux, or fridrich, or some other method like zz

 

[Paul Wagner]

Off topic: Nick, we have to talk about the competitions in that high school in your town.

On topic: I don't know fridrich and petrus have better time's in things other than 3x3. But Roux is fast but I don't think you can use two methods.

 

[blah]

On topic: I don't know fridrich and petrus have better time's in things other than 3x3. But Roux is fast but I don't think you can use two methods.

*cough* Commas! *cough*

Took me a while to understand that

 

[a small kitten]

ZZ is easily sub 20

 

[Ando]

learn fridrich f2l more
I found it excited.
When you sub10 on f2l, even with 4LLL you can reach sub20 average

 

[TMOY]

Weakness of Roux: You are screwed for big cubes

No.

and megaminx.

Yes, but nothing prevents you from learning a Fridrich-like method that you will use only on the megaminx. That's what I'm doing (I'm using corners first on 3^3, not Roux, but the problem is the same).

 

[nitrocan]

Weakness of Roux: You are screwed for big cubes

No.

How will you manage to do so many M's on a 7x7?

 

[puzzlemaster]

Weakness of Roux: You are screwed for big cubes

No.

Ok. Prove it. Do a fridrich solve on a 5x5 and then do a roux solve. If you want blockbuilding on a big cube then I'm sure that petrus would easily be the fastest.

 

[mazei]

I can average 15s with a kind of Petrus(After 2x2x3 block I skip EO and go on in a Fridrich way) and can still average sub-25 Roux after all these months of using Fridrich. Perhaps I'll change to Roux again slowly because I miss it.

Oh, and I was a Roux user from 31s up to 20s. That's where blah came in.

 

[TMOY]

How will you manage to do so many M's on a 7x7?

Ok. Prove it. Do a fridrich solve on a 5x5 and then do a roux solve. If you want blockbuilding on a big cube then I'm sure that petrus would easily be the fastest.

You obviously don't understand my point. Doing 3^3 solves on a 5^3 will prove absolutely nothing, whatever the result is.

 

[puzzlemaster]

How will you manage to do so many M's on a 7x7?

Ok. Prove it. Do a fridrich solve on a 5x5 and then do a roux solve. If you want blockbuilding on a big cube then I'm sure that petrus would easily be the fastest.

You obviously don't understand my point. Doing 3^3 solves on a 5^3 will prove absolutely nothing, whatever the result is.

Reduction ends with a 3x3 solve.. M' and M turns are so slow on big cubes. Maybe you don't understand our point. Doing a 3x3 solve on a 5x5 cube with the roux method will show that the roux methd is very slow for big cubes.