The New Method / Substep / Concept Idea Thread

[AJT17]

Oh yeah i remember looking at that method when i was gonna switch.
Is this the method with YruRU cp tracing or is that LEOR ?
Might look into both methods again, maybe something in them could help me improve my concept.

Thanks !

CEOR has YruRU, LEOR does essentially the same thing except it doesn't care about CP. Hoping you can gather some useful information from both methods to help you improve yours!

 

[That0newe1rdo]

I found way to do LXS that only has 11 algs !

I present We1rdo's LXS :

step 1 : FR DR (for real drake ?????!??!?!?!)

use 1 of 8 algs that permute edges FR and DR. They are short and have an average STM of 5, and a average MCC of 5.95

step 2 : LS

use one of 5 last slot algs that solve the cube, my personal favourite algs have an STM of 8.6 , and an MCC of about 9

adding these STMs and AUF together, you have an estimated average movecount of 15.1 STM.
The reason this is estimated as Its more accurate to account for the likehood of each case and the movecount, which right now i havent got the time to do.

Pros :

Much more likely to get a LS skip of any form in LXS
Less algs

Cons :
roughly 1 move less efficient then standard beginner LXS, and about 7 moves less effiecient then advanced

ALG SHEET :

We1rdo's LXS + My APB algs.

docs.google.com

 

[Martyt7777]

Guys is this method good:

So first you solve Cross+3, and then you solve OLL and the last F2L corner in a single step, then you solve PLL and the last F2L edge also in a single step

It might have already been made before so just let me know

 

[6 by 6]

HELLO GUYS!

This is my new method of solving big cubes (Edge Anchored Centers), in this method the idea is same as my idea of petrus method blockbuilding.

Here are my playlists of both the methods on big cubes.

Edge Anchored Centers- https://youtube.com/playlist?list=PLQgHknWGK7VnYCYrycrEy8aOaX4ZkVKuv&si=s_NQ1NUBqkVfYzTN

Petrus method on big cubes- https://youtube.com/playlist?list=PLQgHknWGK7VmJG_caUJXjM5HtTY38nUZa&si=NURDjIafoWBhdlql

In petrus method solves i try to solve cube more logically, like choosing an edge then attaching the required center pieces with it then inserting the edge, and i gradually increase the block. And in EAC method i again do the same thing expanding the edge by attaching the required center pieces then inserting it.

But the structure of EAC method is different.

Please give me the feedbacks on these methods of solving big cubes logically, specially EAC method, is it already exists or i just discovered a new approach.

 

[Caleb Valenzuela]

Guys is this method good:

So first you solve Cross+3, and then you solve OLL and the last F2L corner in a single step, then you solve PLL and the last F2L edge also in a single step

It might have already been made before so just let me know

how are you supposed to solve OLL when there’s one yellow edge missing? I assume you mean something like the entire top face is solved except the UR edge. This seems similar to BLE (brook’s last edge). You might wanna research some things on the Speedsolving.com wiki for more info

here’s BLE:

 

[Michael Wimberly]

So recently I had this idea of solving a 3x3 by first solving cross + first layer and then instead of solving the middle and top layer, you would solve one adjacent layer (which makes the solved parts look like an L-shape) and then the last pieces. I'll share images later because I know this might sound a little confusing.

 

[NukesAway]

So recently I had this idea of solving a 3x3 by first solving cross + first layer and then instead of solving the middle and top layer, you would solve one adjacent layer (which makes the solved parts look like an L-shape) and then the last pieces. I'll share images later because I know this might sound a little confusing.

There're many reasons why this is a probably a bad for speedsolving (supposing this is its purpose)
1.During the second face, you've limited blockbuilding because there aren't free faces anymore
2.The best way to build a first block is either a Roux block or a Petrus block, because they 1. are move efficient 2. can be planned easily 3. leave many faces free to turn
4. After the second face, any intuitve steps are inefficient because they have to destroy one CEC line or more; similarl for any algs which is why the 42 method was invented for Roux
5. Algsets for last few pieces may be of low quality; movecount-algcount-ergonomics-recognition balance may be hard to maintain; I used a solver and the solutions were kind of unappealing. All 2look finishes I came up with were worse than CFOP in at least 2 of the 4 gauges listed above
I might not have understood your idea fully but sorry I cant see how it can be better than any speedsolving method anyone would use as a main

 

[Michael Wimberly]

There're many reasons why this is a probably a bad for speedsolving (supposing this is its purpose)
1.During the second face, you've limited blockbuilding because there aren't free faces anymore
2.The best way to build a first block is either a Roux block or a Petrus block, because they 1. are move efficient 2. can be planned easily 3. leave many faces free to turn
4. After the second face, any intuitve steps are inefficient because they have to destroy one CEC line or more; similarl for any algs which is why the 42 method was invented for Roux
5. Algsets for last few pieces may be of low quality; movecount-algcount-ergonomics-recognition balance may be hard to maintain; I used a solver and the solutions were kind of unappealing. All 2look finishes I came up with were worse than CFOP in at least 2 of the 4 gauges listed above
I might not have understood your idea fully but sorry I cant see how it can be better than any speedsolving method anyone would use as a main

It's not for speed solving, just a different way of solving the cube.

 

[efattah]

So recently I had this idea of solving a 3x3 by first solving cross + first layer and then instead of solving the middle and top layer, you would solve one adjacent layer (which makes the solved parts look like an L-shape) and then the last pieces. I'll share images later because I know this might sound a little confusing.

Back in 1981-1983 this was an actual method, one of the most popular methods, and it totally sucks (sorry). Back then people had a limited understanding of the cube. Solving corners first was only a thing for people who had 'advanced' knowledge of cube geometry. Most of the population would focus on 'getting one side.' For most, one side meant one side of stickers, not pieces. For more advanced solvers, they could solve one side correctly (i.e. the U-slice). Logic followed that if you could solve one side, the next thing you should do is solve a second side, which was usually R-slice. This was usually done by breaking off a top corner and putting back in a different way until two bottom corners were solved. Then intuitive use by rotating the cube and using U2 M/M' U2 M'/M to 3-cycle the missing 3 edges with other cube edges, and hope they were correctly oriented which was rare. Once in a rare while it would all line up and you would get the second slice solved. Now that you had 2 'sides' done, the next thing was to get '3 sides', i.e. the U-slice, the R-slice and the F-slice. This was now almost a lost cause because you can't do a single move on the cube without totally destroying what you have created. Nonetheless in if you lucked out and all the bottom corners were solved it was still possible with extreme luck and it did happen to me once and once only, before I switched to corners first/keyhole which was vastly better. Some less advanced cubers would solve 1 side of stickers and then get a second 'side' of stickers as well. The cube is nowhere near solved but to ignorant people it looks like you have made tons of progress, I once saw a cube that had 3-sides of 'stickers' solved and the person hadn't move the cube in weeks because they couldn't figure out how to progress without breaking what they had created.

 

[Michael Wimberly]

Back in 1981-1983 this was an actual method, one of the most popular methods, and it totally sucks (sorry). Back then people had a limited understanding of the cube. Solving corners first was only a thing for people who had 'advanced' knowledge of cube geometry. Most of the population would focus on 'getting one side.' For most, one side meant one side of stickers, not pieces. For more advanced solvers, they could solve one side correctly (i.e. the U-slice). Logic followed that if you could solve one side, the next thing you should do is solve a second side, which was usually R-slice. This was usually done by breaking off a top corner and putting back in a different way until two bottom corners were solved. Then intuitive use by rotating the cube and using U2 M/M' U2 M'/M to 3-cycle the missing 3 edges with other cube edges, and hope they were correctly oriented which was rare. Once in a rare while it would all line up and you would get the second slice solved. Now that you had 2 'sides' done, the next thing was to get '3 sides', i.e. the U-slice, the R-slice and the F-slice. This was now almost a lost cause because you can't do a single move on the cube without totally destroying what you have created. Nonetheless in if you lucked out and all the bottom corners were solved it was still possible with extreme luck and it did happen to me once and once only, before I switched to corners first/keyhole which was vastly better. Some less advanced cubers would solve 1 side of stickers and then get a second 'side' of stickers as well. The cube is nowhere near solved but to ignorant people it looks like you have made tons of progress, I once saw a cube that had 3-sides of 'stickers' solved and the person hadn't move the cube in weeks because they couldn't figure out how to progress without breaking what they had created.

Didn't know that was an actual method, cool

 

[CR7GoatSiuuu]

is this a viable 3 look last layer? it is similar to platform:
1. orient last layer corners while solving 2 adjacent corners. maybe use cll algs? auf so the solved corners are on the left
2. orient last layer edges while solving the UL edge. ELL algs?
3. U perms, J perms, or F perm

 

[CrossPetrus]

I’m not sure that’s viable. The recognition for orienting the corners while solving two adjacent ones is not suited for a beginner. Plus, if you just orient the edges (and solve UL) afterward, then you add on some not nice algs. That just doesn’t seem well suited for someone who would need a 3 look last layer. Plus it wouldn’t help them learn a better last layer technique unless they were going to use CFCE… I would say the idea needs some work. But I may be wrong. I am a newb to method development.

 

[CR7GoatSiuuu]

I’m not sure that’s viable. The recognition for orienting the corners while solving two adjacent ones is not suited for a beginner. Plus, if you just orient the edges (and solve UL) afterward, then you add on some not nice algs. That just doesn’t seem well suited for someone who would need a 3 look last layer. Plus it wouldn’t help them learn a better last layer technique unless they were going to use CFCE… I would say the idea needs some work. But I may be wrong. I am a newb to method development.

isn't it just easier cll recog?also I don't know any algs for ell, but if the algs aren't good then I will abandon this

 

[CrossPetrus]

For the recognition of the corners, it isn’t bad, just not really good for a beginner to recognize orientation and permutation. For ell, the algs aren’t very fingertrick friendly for a beginner and are just weird. They have awkward move gens (like the subset of moves the algs use). Most of them are also long and overall, it seems easier just to do eo-> co-> pll, as recog would be a bit more beginner friendly and there are more friendly algs (and tons of videos with fingertricks). I’m not necessarily saying it’s a bad idea, I’m just skeptical of its practicality compared to other 3lll methods.

 

[CR7GoatSiuuu]

for hexagonal Francisco what is the point of starting with a first block then solving E layer + CLS? why not build a bell like in nautilus then do WV, L5EOP, PLL? assuming 20 moves for the bell, 5 moves to build the pair, 7 moves for modified WV without EO, and then 10 moves each for PLL and L5EOP, that gives you 52 moves vs the wiki estimate of Hexagonal Francisco plus you save 80 algs. Or am I just describing another crappier version of nautilus?

 

[PapaSmurf]

is this a viable 3 look last layer? it is similar to platform:
1. orient last layer corners while solving 2 adjacent corners. maybe use cll algs? auf so the solved corners are on the left
2. orient last layer edges while solving the UL edge. ELL algs?
3. U perms, J perms, or F perm

2 look OLL into PLL is just better. The recog is better, the algs will probably be better.

 

[Caleb Valenzuela]

is this a good alternative to ZBLS?
after solving cross+3, do the following
solve EO
solve the last pair

after some thinking, maybe solving F2L then doing EO after would be better?

 

[ruffleduck]

is this a good alternative to ZBLS?
after solving cross+3, do the following
solve EO
solve the last pair

Isn't this just 2-look ZBLS?

after some thinking, maybe solving F2L then doing EO after would be better?

FRUR'U'F' into ZB?

I could be misunderstanding, but from what I think you are describing, this is just the usual "intuitive" approach to 2-look ZBLS

 

[That0newe1rdo]

is this a good alternative to ZBLS?
after solving cross+3, do the following
solve EO
solve the last pair

after some thinking, maybe solving F2L then doing EO after would be better?

i believe i sort of made this, it is more effecient.
i think i have a thread on it

 

[NukesAway]

is this a viable 3 look last layer? it is similar to platform:
1. orient last layer corners while solving 2 adjacent corners. maybe use cll algs? auf so the solved corners are on the left
2. orient last layer edges while solving the UL edge. ELL algs?
3. U perms, J perms, or F perm

Look up CLL+1 on the wiki。It's a bit different but you'll see what I mean