The New Method / Substep / Concept Idea Thread

[BenChristman1]

2x2 method: Flop there are reasons why I called this the flop method
1. First create the first layer (you can do this with look ahead)
2. Oll you can use the same Oll algs from 3x3 (some)
3. Pll case 1: standard V Perm alg case 2: T perm
that’s about what I can figure out with this method also this is actually for me the fastest method but nothing can beat the Orteg.

That’s called Layer by Layer, and CLL+EG is way better than Ortega.

 

[DNF_Cuber]

2x2 method: Flop there are reasons why I called this the flop method
1. First create the first layer (you can do this with look ahead)
2. Oll you can use the same Oll algs from 3x3 (some)
3. Pll case 1: standard V Perm alg case 2: T perm
that’s about what I can figure out with this method also this is actually for me the fastest method but nothing can beat the Orteg.

That is just layer by layer. That is possibly the most hilarious new method proposal ever.

 

[RPerm (2020GOME05)]

Ok

 

[WoowyBaby]

Here's a new method that I thought up today that's probably totally useless!!

Here's an example solve first (from 3x3 Example Solve Thread):

Scramble: B2 D' F R U' F2 R' D B2 R2 U2 F2 R2 U R2 B2 R2 B2 F R'
(z2 y)
B F L' D' R F2 U R2 // Step 1
F' U F2 R' F' R // Step 2
U D2 R L U L' U R' // Step 3
U F2 L2 U2 R2 B2 R2 // Step 4
(29)

The First Step is to form a "HTR F2L-1". Since you don't have to worry about the permuation of the corners yet, so you simply make a huge area of pure opposite colors. Calling it a kind of "F2L-1" limits your thinking about what you can even do with this as it's so fundamentally different from an actual F2L-1, although I will say that Petrus blockbuilding is a good framework. This first step can probably be planned in inspection and should take around 8 moves. (If you're ever using more than 11 moves, you're doing something wrong.)

The Second Step is to form an oppsite color square and orient edges, leaving only 4 corners and 3 (oriented) edges that are not in HTR. This can be achieved in a similar fashion to Tripod, but keep in mind that you do not have to preserve the opposite color area- Swapping around any edges or corners in the opposite color area from Step 1 does not mess anything up so there's some extra freedom. This is an intuitive step, doesn't take too many moves, probably one-lookable. Preferrably don't make a psuedo-opposite square/square that isn't in the back left, or else the next step's recognition could be a nightmare. The moveset is alright, but F/F2/F' moves could slow this step down a little bit.

The Third Step is the orient the remaining four corners, as well as double orient the edges and correct the corner permutation (for true HTR). This is an algorithmic step, unknown exactly how many algs but likely around 100. I still need to generate all of the algorithms, but from what I can tell, contains lots of RUD or RUL moves. Probably around 10-11 moves on average, and the example solve's case was only 7 moves long. Recognition isn't exactly great, definitely worse than something like CFOP OLL, but it's probably better than you think.

The Fourth Step is the long awaited magical Half-Turn state. In my example solve I included the one move before you actually enter this stage, because you can either do a clockwise or counterclockwise turn, depending on which one is better. The general way of reaching the finish line from here is either do a few moves to a pattern that you know and finish quickly, blockbuild your way to victory, solve the corners along with a bunch of edges and finish with a 3-cycle, or all of the above. If you have experience with FMC and DR/HTR solutions in FMC, then you will already have a pretty intuitive grasp of how to do this step efficiently and quickly. There are no algorithms in this final step, but some useful cases to be super familiar with are M' U2 M U2 and R2 U2 R2 U2 R2 U2 and U2 R2 F2 R2. If you're taking more than 13 moves (slice turn metric) or more than 4 seconds on this step, you're doing something wrong.

Here is another example solution for more comprehension of this method:
Scramble: B2 U B2 D' L2 U L2 R2 B2 R D' L' D' U' F U L2 R D

This method sucks // no (0/0)

REALIZATION: This sucks lmoa

Maybe would be better if I changed up the 2nd and 3rd steps to perhaps all corners -> all edges, would be two algorithmic steps instead of one which is probably better, and actually would be a pretty low number of algs due to the half-turn state's nature. Regardless, building the initial opposite color area is often bad, and this is just overall terrible for speedsolving even with this improvement.

 

[RPerm (2020GOME05)]

Here's a new method that I thought up today that's probably totally useless!!

Here's an example solve first (from 3x3 Example Solve Thread):

Scramble: B2 D' F R U' F2 R' D B2 R2 U2 F2 R2 U R2 B2 R2 B2 F R'
(z2 y)
B F L' D' R F2 U R2 // Step 1
F' U F2 R' F' R // Step 2
U D2 R L U L' U R' // Step 3
U F2 L2 U2 R2 B2 R2 // Step 4
(29)

The First Step is to form a "HTR F2L-1". Since you don't have to worry about the permuation of the corners yet, so you simply make a huge area of pure opposite colors. Calling it a kind of "F2L-1" limits your thinking about what you can even do with this as it's so fundamentally different from an actual F2L-1, although I will say that Petrus blockbuilding is a good framework. This first step can probably be planned in inspection and should take around 8 moves. (If you're ever using more than 11 moves, you're doing something wrong.)

The Second Step is to form an oppsite color square and orient edges, leaving only 4 corners and 3 (oriented) edges that are not in HTR. This can be achieved in a similar fashion to Tripod, but keep in mind that you do not have to preserve the opposite color area- Swapping around any edges or corners in the opposite color area from Step 1 does not mess anything up so there's some extra freedom. This is an intuitive step, doesn't take too many moves, probably one-lookable. Preferrably don't make a psuedo-opposite square/square that isn't in the back left, or else the next step's recognition could be a nightmare. The moveset is alright, but F/F2/F' moves could slow this step down a little bit.

The Third Step is the orient the remaining four corners, as well as double orient the edges and correct the corner permutation (for true HTR). This is an algorithmic step, unknown exactly how many algs but likely around 100. I still need to generate all of the algorithms, but from what I can tell, contains lots of RUD or RUL moves. Probably around 10-11 moves on average, and the example solve's case was only 7 moves long. Recognition isn't exactly great, definitely worse than something like CFOP OLL, but it's probably better than you think.

The Fourth Step is the long awaited magical Half-Turn state. In my example solve I included the one move before you actually enter this stage, because you can either do a clockwise or counterclockwise turn, depending on which one is better. The general way of reaching the finish line from here is either do a few moves to a pattern that you know and finish quickly, blockbuild your way to victory, solve the corners along with a bunch of edges and finish with a 3-cycle, or all of the above. If you have experience with FMC and DR/HTR solutions in FMC, then you will already have a pretty intuitive grasp of how to do this step efficiently and quickly. There are no algorithms in this final step, but some useful cases to be super familiar with are M' U2 M U2 and R2 U2 R2 U2 R2 U2 and U2 R2 F2 R2. If you're taking more than 13 moves (slice turn metric) or more than 4 seconds on this step, you're doing something wrong.

Here is another example solution for more comprehension of this method:
Scramble: B2 U B2 D' L2 U L2 R2 B2 R D' L' D' U' F U L2 R D

This method sucks // no (0/0)

REALIZATION: This sucks lmoa

Maybe would be better if I changed up the 2nd and 3rd steps to perhaps all corners -> all edges, would be two algorithmic steps instead of one which is probably better, and actually would be a pretty low number of algs due to the half-turn state's nature. Regardless, building the initial opposite color area is often bad, and this is just overall terrible for speedsolving even with this improvement.

I agree that method sucks some how...

 

[TheSlykrCubr]

do people use half centres for 5x5?

would it just be like solving middle + left bar for 3 centres, then finishing the rest Rw U gen?

 

[Platform_]

This guide contain my invention. I know i Maked mistake in this pdf. The idea is create block 1x1x3 and in the next solve cases Line.
Two steps.
One step have 25 algo but i supposed exist more algo for solve a block 1x1x3 example block three part red and three part yellow in LL. In the next step two cases of Line. In Wiki doesn't exist this method. I call my method as Platform.

I have never been able to catalog all the possible cases and algo of step one. In addition, the Line for the second step has optimized algorithms. This method I have created is unknown in the world for the moment.

 

[Nir1213]

This guide contain my invention (method alternative for LL)

This guide contain my invention. I know i Maked mistake in this pdf. The idea is create block 1x1x3 and in the next solve cases Line.
Two steps.
One step have 25 algo but i supposed exist more algo for solve a block 1x1x3 example block three part red and three part yellow in LL. In the next step two cases of Line. In Wiki doesn't exist this method. I call my method as Platform.

I have never been able to catalog all the possible cases and algo of step one. In addition, the Line for the second step has optimized algorithms. This method I have created is unknown in the world for the moment.

you can edit your posts into one big post instead by pressing the edit button under your posts, just saying. I understand your new here so welcome!

also i cant wait to see your method!

 

[DNF_Cuber]

do people use half centres for 5x5?

I doubt it would be very good, also since more people use redux than yau for 5x5 so the development would be slow. Interesting idea, but

 

[Platform_]

you can edit your posts into one big post instead by pressing the edit button under your posts, just saying. I understand your new here so welcome!

also i cant wait to see your method!

In this file my idea.

 

[Nir1213]

In this file my idea.

it is certainly interest, and you've been working on it for 12 years! wow

 

[Platform_]

it is certainly interest, and you've been working on it for 12 years! wow

Very difficult. I started for fun. In the past I didn't understand the guides. I didn't think I was going to develop something different.

 

[TheSlykrCubr]

I was just joking

calling a catholic in nothern ireland british is making you prone to be beat up by a gang like this

 

[BlueAcidball]

This clock method is similar to the normal clock method, just with block building incorporated into it. Here's an overview of it:
Step 1: Make a 2x2 block on any of the 4 parts on one side, preferably on UL, and move it to the 12 o'clock position, like this:

Step 2: Flip to the other side then create your first cross, then restore the corner from the 2x2 block. Your clock should now look something like this:

Step 3: Flip again, then expand the 2x2 block to a 2x3 block, then solve the second cross and move it to the 12 o'clock position. It should now look something like this:

Step 4: Flip one last time, then solve the corners as you normally would, completely disregarding the block, and it should be solved.

Now, since this is my first method I ever thought about, it's obviously not the most optimal method, so feel free to suggest things I could change in this method!

 

[abunickabhi]

Interesting concept yo.

 

[Cubing Forever]

Nice idea for petrus on clock yo

 

[Jam88]

I get the petrus aspect, not the clock aspect

 

[OreKehStrah]

It’s a cool method in terms of being novel, but there’s way too many flips for speed solving considering some people are starting to use the no-flip method.

 

[Cubing Forever]

I get the petrus aspect, not the clock aspect

It's like petrus but with EO(Flip side cross) before 223 but after 222

 

[GenTheSnail]

I'm not the best clock solver my any stretch, but I can solve one with a 'standard' method, and I'll bring up a few things I'm seeing.


It sounds like a cool method, but I'm a little confused as to why the first step is to make a 2x2 block if the corner is going to get 'broken' as soon as we do the next step. It seems like if you don't do the corner, then you won't have to solve it twice.

In addition, I'm unsure as to the reason you need to put the dials to the 12 o'clock position after each step. As I understand it, you don't ever really need to bother putting the dials into the 12 o'clock position until it's solved, and doing anything else just seems like a waste of moves.

And finally, as OreKehStrah pointed out, there are a lot of flips in this method. The normal clock method only has one flip, and it is possible to do no-flip solves (I'm not sure how this is done, but I would assume simply by looking ahead through the cross.)


If you wanna try to develop more methods, you can try posting ideas in the The New Method / Substep / Concept Idea Thread.