The New Method / Substep / Concept Idea Thread

[qwr]

Ok guys I am keep my opinion to my self now since everyone keeps on saying this is a CFOP method

1. Make cross
2. Solve corners leave 1 behind
3. Use the unsolve corner to solve the edges
4. F2L 1% chance of being skipped
5. Oll
6. Pll

 

[OreKehStrah]

The whole point of a belt method is to form the belt as the first step no? And this certainly is not that. This is just a worse version of CFOP.

 

[Sub1Hour]

Keyhole F2L - Speedsolving.com Wiki

Compare your method to some of the variations of keyhole stated on the wiki and tell me what exactly makes it belt instead of keyhole

"You start, like in normal LBL from a cross, then you fill in three of the first layer corners. From that point you solve three of the middle layer edges using the empty corner position as the keyhole "
Your steps 1-2
"

• First position the mid-layer edge in the U-layer such that it is above the centre matching the edge's up-facing colour.
• Using a D-Turn position the bottom layer so that the unsolved corner is directly below the mid-layer edge's position.
• Now execute one of R U' R' or F U F' to insert the edge.

"
Your step 3
"After three corners and edges are solved you complete F2L by first solving the last corner and finally the last edge "
Your step 4
And step 5 and 6 are just OLL and PLL

The keyhole method that's closer to an actual belt method is Edge First keyhole, and even then it's hardly a belt method.

 

[the dnf master]

Ok guys I am keep my opinion to my self now since everyone keeps on saying this is a CFOP method

Well at least you made this thread into a decent argument

 

[ElyasCubing]

Exactly its the same thing tbh

 

[Cubing Forever]

A lolmethod originally intended to be a beginner's method:
1) Cross+3c: Solve cross+3 corners - ~10-12 moves (mostly rotationless)
2) F2L-1: solve the 3 edges that belong to the slots whose corners are solved using keyhole: ~20-25 moves (rotationless) <RUUwF> gen
3) F<RU>F' LS + EO: Keeping the last slot on FL, solve it by doing an F and then solving the pair 2 gen(Jayden McNeill has a good video series on it) and cancelling into EO. ~11-12 moves with some Dw move to bring the unsolved slot to the FL position
4) LL: ZBLL (12-14 moves on average) or OCLL/PLL (20-25 moves on average) or COLL/EPLL(~18 moves on average)
Total movecount: 53-54 with ZBLL, 60+ with other variants
(NOTE: All movecounts are calculated guesses and for the keyhole, using Uw moves instead of D moves makes this method entirely rotationless)

Pros:
Entirely rotationless
Doesn't require EO at the start of the solve to achieve the above.
Easier lookahead since one has to look only for edges.
One can use ZBLL everytime
<RUUwF> gen F2L with no rotations therefore superfast F2L

Cons:
Higher movecount. High algcount to lower it.
F<RU>F' LS is hard to learn.
Cross+3c in inspection is hard blockbuilding the same is just harder

This is an idea I came up with as a beginner method to teach my little brother how to solve a 3x3 and I just expanded it with EO and F<RU>F' LS just for fun.

Someone please come up with a good name for it.
This method is not a replacement to CFOP.

Maybe it can be a good OH method???

 

[PapaSmurf]

It’s basically keyhole into ZB. Not a new idea, but one with potential.

 

[crazykitten499]

do coll-like algorithms that orient the edges and orient the corners, as well as permutation of the corners, exist?

 

[Cubing Forever]

It’s basically keyhole into ZB. Not a new idea, but one with potential.

Keyhole into ZB but no rotations without EO at the start.
<RUuF> F2L with no rotations without EO at the start is superfast.
Somehow if anyone can come up with some intuitive EO process for LS with low movecount then overall movecount will go down to high 40s
e.g F<RU>F' LS(that's the only method that can cancel into EO easily as of now) etc.

do coll-like algorithms that orient the edges and orient the corners, as well as permutation of the corners, exist?

OLLCP is the only algset that I know of that fits this criterion

 

[TheSlykrCubr]

do coll-like algorithms that orient the edges and orient the corners, as well as permutation of the corners, exist?

ollcp

 

[PapaSmurf]

do coll-like algorithms that orient the edges and orient the corners, as well as permutation of the corners, exist?

OLLCP

Keyhole into ZB but no rotations without EO at the start.
<RUuF> F2L with no rotations without EO at the start is superfast.
Somehow if anyone can come up with some intuitive EO process for LS with low movecount then overall movecount will go down to high 40s
e.g F<RU>F' LS(that's the only method that can cancel into EO easily as of now) etc.

I definitely think that this style of F2L has potential. I made a post somewhere back in this thread in context of Zipper, but ZB works too.

 

[Silky]

So I've been experimenting around with my Souxp method from a while back. This is what I've come up with so far.

(1) First block
(2) Belt + corner separation ( placing DFR + DBR corners in D with no specific orientation or permutation )
(3) DR corner orientation + permutation + U layer corner orientation
(4) U layer corner permutation + EO
(5) L7EP
=> Either through DF DB placement + L5EP or
=> Edge separation with a SSC finish (?)

 

[kevinlin]

I have been testing out a last slot method and it sometimes gives quite nice movecounts.

(1) Use a method that does EO before last slot
(2) Make a 2x2 block on U layer like the one in Tripod method
(3) Insert last edge
(4) Solve a corner using commutators
(5) Finish by L3C or L3C+L2E

 

[Seth1448]

The belt method SUCKS. It’s just a way worse version of cfop so stop wasting you’re time.

 

[Deleted member 55877]

The belt method SUCKS. It’s just a way worse version of cfop so stop wasting you’re time.

It's true that pure belt is worse than CFOP, but it IS possible that a modified version of belt could turn out to be good. Besides, the whole point of this thread is to experiment and come up with new methods, regardless of whether they are good for speedsolving or not.

 

[qwr]

I wish I could unlearn how to solve a 3x3 so I could invent my own method.
Now that I know the basic ideas of CFOP, Roux, and Petrus, I'll never be able to create something truly original.

 

[Athefre]

I wish I could unlearn how to solve a 3x3 so I could invent my own method.
Now that I know the basic ideas of CFOP, Roux, and Petrus, I'll never be able to create something truly original.

While that is a real psychological effect, the truth is that it has always been difficult to think of something new. Gilles Roux was aware of Corners First, LBL, Edges First, Columns, CFOP, Petrus, and others. Yet he was able to think of something new. The same can be said for Zbigniew Zborowski. It isn't easy and shape options are running out as time goes on and the community grows, but you can do it.

 

[qwr]

While that is a real psychological effect, the truth is that it has always been difficult to think of something new. Gilles Roux was aware of Corners First, LBL, Edges First, Columns, CFOP, Petrus, and others. Yet he was able to think of something new. The same can be said for Zbigniew Zborowski. It isn't easy and shape options are running out as time goes on and the community grows, but you can do it.

oh I'm not looking to create something new, just something that's entirely my own creation.

 

[Platform_]

Hello guys! I created a guide to my personal alternative method for LL.
Good vision!

 

[Cubing Forever]

Ok so this is just reduction to line/flipped line 1LLL right?