The New Method / Substep / Concept Idea Thread

[crafto22]

So I've been posting a lot here lately, but whatever. I guess I just have a lot of ideas.

New LL system

Here's what you do: recognize if your edges are phased or unphased during your last slot insertion. Next, apply one of two algorithms for the OLL you have that will either keep the edges phased or that will phase them for you. You then end up with twelve different PLL possibilities: T perm, F perm, E perm, H perm, A perm or solved. EDIT: N perms as well (say goodbye to G perms)

Pros:

- Recognition is just as fast for OLL as you recognize EP during LS
- Recognition for PLL is faster as you have less algs to choose from
- 1/12 chance of PLL skip
- Alternate OLL algs can be used to phase edges, so no need to generate new ones

Cons:

- 57 new algs to learn
- IMO F perm and E perm kinda suck
- Having to choose from 2 different algs during OLL makes execution slightly slower

 

[PenguinsDontFly]

So I've been posting a lot here lately, but whatever. I guess I just have a lot of ideas.

New LL system

Here's what you do: recognize if your edges are phased or unphased during your last slot insertion. Next, apply one of two algorithms for the OLL you have that will either keep the edges phased or that will phase them for you. You then end up with twelve different PLL possibilities: T perm, F perm, E perm, H perm, A perm or solved. (say goodbye to G perms)

Pros:

- Recognition is just as fast for OLL as you recognize EP during LS
- Recognition for PLL is faster as you have less algs to choose from
- 1/12 chance of PLL skip
- Alternate OLL algs can be used to phase edges, so no need to generate new ones

Cons:

- 57 new algs to learn
- IMO F perm and E perm kinda suck
- Having to choose from 2 different algs during OLL makes execution slightly slower

another con:
u perms are gone

 

[supercavitation]

So I've been posting a lot here lately, but whatever. I guess I just have a lot of ideas.

New LL system

Here's what you do: recognize if your edges are phased or unphased during your last slot insertion. Next, apply one of two algorithms for the OLL you have that will either keep the edges phased or that will phase them for you. You then end up with twelve different PLL possibilities: T perm, F perm, E perm, H perm, A perm or solved. (say goodbye to G perms)

Pros:

- Recognition is just as fast for OLL as you recognize EP during LS
- Recognition for PLL is faster as you have less algs to choose from
- 1/12 chance of PLL skip
- Alternate OLL algs can be used to phase edges, so no need to generate new ones

Cons:

- 57 new algs to learn
- IMO F perm and E perm kinda suck
- Having to choose from 2 different algs during OLL makes execution slightly slower

That seems less worthwhile than just using different OLL algs to avoid Diag cases, since I'll take a G perm over an N perm any day.

 

[crafto22]

That seems less worthwhile than just using different OLL algs to avoid Diag cases, since I'll take a G perm over an N perm any day.

Damn, i forgot about n perms... Okay, this is useless. I hope no one learns this

 

[sqAree]

What about this method for 2x2:

1. Bring 4 pieces of one face to the correct spots (aka building a face without paying attention to orientation).
2. OBL (orientation of both layers)
3. PBL

Step should be possible in 3 moves max, if cn probably only 2. Easy look-ahead to the next steps.
Step 2 requires 35 algs, 7 of them are already known OLLs. As I don't know how to generate algs, I have no clue how long / fingertrick-friendly they are.
Step 3 just like in the Ortega method (or whatever it is called now).

Maybe cn has other benefits, like forcing a specific subset of those 36 cases, making one-look possible for the whole method.

Just a random thought, I usually don't invent methods so it's probably gonna be bad, and EG is better anyway (although for this less algs are required).

 

[shadowslice e]

What about this method for 2x2:

1. Bring 4 pieces of one face to the correct spots (aka building a face without paying attention to orientation).
2. OBL (orientation of both layers)
3. PBL

Step should be possible in 3 moves max, if cn probably only 2. Easy look-ahead to the next steps.
Step 2 requires 35 algs, 7 of them are already known OLLs. As I don't know how to generate algs, I have no clue how long / fingertrick-friendly they are.
Step 3 just like in the Ortega method (or whatever it is called now).

Maybe cn has other benefits, like forcing a specific subset of those 36 cases, making one-look possible for the whole method.

Just a random thought, I usually don't invent methods so it's probably gonna be bad, and EG is better anyway (although for this less algs are required).

This is essentially what Stefan proposed when he researched RoFL(ROtten First Layer). It also eventually led to the creation of TCLL which I think is a pretty good method with some potential.

 

[sqAree]

Alright, I see my mistake, there are more than 35 algs required.
But nice to know it has been useful somewhen.

 

[PenguinsDontFly]

this has definitely been thought of before but I really think it has sub 8 potential.

First bock
Second block
EO + F2L
ZBLL

I just got a 8.82 ao 50 with this but using COLL+EPLL

 

[2180161]

this has definitely been thought of before but I really think it has sub 8 potential.

First bock
Second block
EO + F2L
ZBLL

I just got a 8.82 ao 50 with this but using COLL+EPLL

Or do EO+ F2l, COLL, EPLL, Less algs, same speed I think personally.

 

[molarmanful]

this has definitely been thought of before but I really think it has sub 8 potential.

First bock
Second block
EO + F2L
ZBLL

I just got a 8.82 ao 50 with this but using COLL+EPLL

I smell a Roux-CFOP hybrid lol

 

[2180161]

I smell a Roux-CFOP hybrid lol

Exactly. And coming from PDF no less.

 

[PenguinsDontFly]

Exactly. And coming from PDF no less.

what?

 

[2180161]

what?

Most roux/ anything hybrids are made by beginners
you are not a beginner correct?

 

[PenguinsDontFly]

Most roux/ anything hybrids are made by beginners
you are not a beginner correct?

lol thats true
but still, this is way faster than the usual suggestion which is FB, SB, F2L, OLL, PLL
sub 9 ao 50 with it is pretty decent

edit: also, the movecount is pretty nice. 35 STM + LL

 

[TDM]

lol thats true
but still, this is way faster than the usual suggestion which is FB, SB, F2L, OLL, PLL
sub 9 ao 50 with it is pretty decent

edit: also, the movecount is pretty nice. 35 STM + LL

Still not very fast. No faster than CFOP w/ ZBLS or ZZ imo. Movecount isn't anything special either.

 

[PenguinsDontFly]

Still not very fast. No faster than CFOP w/ ZBLS or ZZ imo. Movecount isn't anything special either.

not faster than cfop, but maybe faster than roux. F2L+EO is definitely faster than CMLL, ZBLL is about the same as my (terrible) LSE.

 

[TDM]

not faster than cfop, but maybe faster than roux. F2L+EO is definitely faster than CMLL, ZBLL is about the same as my (terrible) LSE.

I think LSE can definitely be faster than ZBLL, and if your TPS is good then CMLL is fast too.

Also, doesn't your post imply you think CFOP is faster than Roux? If so, why use Roux?

 

[PenguinsDontFly]

I think LSE can definitely be faster than ZBLL, and if your TPS is good then CMLL is fast too.

Also, doesn't your post imply you think CFOP is faster than Roux? If so, why use Roux?

cmll execution is fast, but recognition kills it. my cmll recog + exec is about 1.7, and I have been doing it for a year. i started practising eo+f2l this morning and its already quite a bit faster. lets not even start about my lse...

yeah, cfop is still faster than roux. I only picked roux because I hate being mainstream

 

[TDM]

cmll execution is fast, but recognition kills it. my cmll recog + exec is about 1.7, and I have been doing it for a year. i started practising eo+f2l this morning and its already quite a bit faster. lets not even start about my lse...

ZBLL recognition is even worse, and that's a fact. I think that if your LSE is bad then you may want to focus on that a bit, since I think LSE can be very fast.

 

[PenguinsDontFly]

ZBLL recognition is even worse, and that's a fact. I think that if your LSE is bad then you may want to focus on that a bit, since I think LSE can be very fast.

well you would know the COLL case and the corner swap because f2l+eo doesnt affect corners. then just look at edges.

yeah, thats what i've been doing.