The New Method / Substep / Concept Idea Thread

[Renox]

1) 2x2x3 block
2) Pair/2x1x1 block on D face
3) Solve the last 6 edges( no algerinos because I'm on my phone and not sure if you can solve intuitively like roux l6e
4)L5C
By the way, is there a more efficient way to solve the last five corners instead of the 8355 way?

 

[shadowslice e]

1) 2x2x3 block
2) Pair/2x1x1 block on D face
3) Solve the last 6 edges( no algerinos because I'm on my phone and not sure if you can solve intuitively like roux l6e
4)L5C
By the way, is there a more efficient way to solve the last five corners instead of the 8355 way?

L5C=comms or similar

Also, M-CELL (see sig)

 

[Aiminer357]

Method Proposal: Aim Method

Hello! This post is to explain a method I have came up with. For now I will call it the Aim Method.

I know there is bound to be a lot of flaws in this so tell me ASAP so I dont get carried away learning it. Also tell me if this has been thought off before

This method was originally based on the idea that I was too lazy too learn full OLL. Long story short, I combined the Petrus Method with CFOP. This is what I came up with.

1. Make a cross and insert three F2L pairs.
2. Orient all edges.
3. Insert last F2L pair.
4. Orient all corners.
5. Permute.

It was okay but not enough. I realised I could do better by doing Winter Variation for step 3. I was happy until I realised I could go even more advance. More algorithms must be memorised but not that much. This is my final result.

1. Make a cross and insert Two F2L pairs.
2. Insert third pair using VHLS (or ZBLS if you're really ambitious!). Special recognition but not too complicated.
3. Insert last pair with Winter Variation (or Summer Variation if you memorised it). Last layer should be oriented.
4. Permute Last Layer.

You will learn 44 algs excluding mirrors. 16 for VHLS and 28 for Winter Variation. You could learn Summer variation(28 algs) or ZBLS(125 algs) but it would defeat the purpose of not learning full OLL. All the algs are easy to find on the internet

Pros:
Less algs than CFOP.
Easy for a CFOP user to learn
1 less look than CFOP (I think)

Cons:
To pair the last F2L pair can be frustrating
Small chance of not able to orient the edges

I have no way of calculating the average move count but it should be less than CFOP. If not, then tell me.

If this is truly as good as it sounds, it might go mainstream in a few years. If it is a total dud, than I have been wasting a whole bunch of time.

 

[Berkmann18]

Hello! This post is to explain a method I have came up with. For now I will call it the Aim Method.

I know there is bound to be a lot of flaws in this so tell me ASAP so I dont get carried away learning it. Also tell me if this has been thought off before

This method was originally based on the idea that I was too lazy too learn full OLL. Long story short, I combined the Petrus Method with CFOP. This is what I came up with.

1. Make a cross and insert three F2L pairs.
2. Orient all edges.
3. Insert last F2L pair.
4. Orient all corners.
5. Permute.

It was okay but not enough. I realised I could do better by doing Winter Variation for step 3. I was happy until I realised I could go even more advance. More algorithms must be memorised but not that much. This is my final result.

1. Make a cross and insert Two F2L pairs.
2. Insert third pair using VHLS (or ZBLS if you're really ambitious!). Special recognition but not too complicated.
3. Insert last pair with Winter Variation (or Summer Variation if you memorised it). Last layer should be oriented.
4. Permute Last Layer.

You will learn 44 algs excluding mirrors. 16 for VHLS and 28 for Winter Variation. You could learn Summer variation(28 algs) or ZBLS(125 algs) but it would defeat the purpose of not learning full OLL. All the algs are easy to find on the internet

Pros:
Less algs than CFOP.
Easy for a CFOP user to learn
1 less look than CFOP (I think)

Cons:
To pair the last F2L pair can be frustrating
Small chance of not able to orient the edges

I have no way of calculating the average move count but it should be less than CFOP. If not, then tell me.

If this is truly as good as it sounds, it might go mainstream in a few years. If it is a total dud, than I have been wasting a whole bunch of time.

Lol, I've been doing that for a while !! And it's really an awesome method especially when combined to CLS.

 

[crafto22]

Hello! This post is to explain a method I have came up with. For now I will call it the Aim Method.

I know there is bound to be a lot of flaws in this so tell me ASAP so I dont get carried away learning it. Also tell me if this has been thought off before

This method was originally based on the idea that I was too lazy too learn full OLL. Long story short, I combined the Petrus Method with CFOP. This is what I came up with.

1. Make a cross and insert three F2L pairs.
2. Orient all edges.
3. Insert last F2L pair.
4. Orient all corners.
5. Permute.

It was okay but not enough. I realised I could do better by doing Winter Variation for step 3. I was happy until I realised I could go even more advance. More algorithms must be memorised but not that much. This is my final result.

1. Make a cross and insert Two F2L pairs.
2. Insert third pair using VHLS (or ZBLS if you're really ambitious!). Special recognition but not too complicated.
3. Insert last pair with Winter Variation (or Summer Variation if you memorised it). Last layer should be oriented.
4. Permute Last Layer.

You will learn 44 algs excluding mirrors. 16 for VHLS and 28 for Winter Variation. You could learn Summer variation(28 algs) or ZBLS(125 algs) but it would defeat the purpose of not learning full OLL. All the algs are easy to find on the internet

Pros:
Less algs than CFOP.
Easy for a CFOP user to learn
1 less look than CFOP (I think)

Cons:
To pair the last F2L pair can be frustrating
Small chance of not able to orient the edges

I have no way of calculating the average move count but it should be less than CFOP. If not, then tell me.

If this is truly as good as it sounds, it might go mainstream in a few years. If it is a total dud, than I have been wasting a whole bunch of time.

Movecount:

CFOP:

Cross: 6
F2L: 6 per pair * 4 = 24
OLL: 9
PLL: 11
AUFs: ~4

Total: 54

Aim:

Cross: 6
F2L: 6 per pair * 2 pairs = 12
VHLS: 8 + ~1 (because the edge of the other pair must be neutral, forcing you to occasionally have to swap two edges) = 9
WV: 11
PLL: 11
AUFs: ~4

Total: 53

Your method saves about 1 move, but sacrifices the good recognition that comes with oriented the last layer with OLL. WV recogntion sucks, and ZBLS and VHLS recog great either. Your making a 7-look system (CFOP) into an 8-9 look system (cross, pair, pair, create pair, sometimes replace edge, VHLS, create pair, WV, PLL).

Yeah, not worth it.

 

[shadowslice e]

Hello! This post is to explain a method I have came up with. For now I will call it the Aim Method.

...

This is very similar to MGLS which by itself is a passable method though is only really worth it when combined with ZZ as the EOLS isn't really worth it.

However, the recognition isn't very good and essentially adds another step which is less efficient than the 1 step OLL as well as being less efficient than the 2-step MGLS.

 

[Sion]

I created a new method based on pure flexiblility and intuivity. I dub it the TAC method. It shares elements from CFOP and Roux, but very different from either of them. I'm not sure if such method exists yet, but I switched to this method as my main and find ways to make it faster.

I will start gathering algorithms once I get the clear that the method is truly original.

it works like this:


[COLOR="#FF000"] b4p: this matches an edge with it's corresponding corner. this is pretty much building four 1x1x2 blocks, and unlike CFOP, the fact that there is no cross makes the whole step more flexible, as you can use the S and M slices more frequently, making the step much easier to learn intuitively. Solving the corners can be used for beginners, but this habit should be dropped once you get familiar to the method. [/COLOR]

C&DLE: Centers and down layer edges are solved here. Wait- why solve the centers? using S and M slice moves puts the centers in place; using a similar pattern you can solve the down layer edges, which at first looks menacing, but it's really a quick and easy step to do since everything is set up from the start of the solve. This method is related to Roux in the fact that slice moves are frequent in PCE and C&DLE.


OLL: orient last layer

Pll: permute last layer.


*solved cube*

Even if this method isn't as efficient as I hope, It should probably serve as a great transitional method for someone deciding to learn CFOP or Roux. But who knows?

 

[crafto22]

I created a new method based on pure flexiblility and intuivity. I dub it the TAC method. It shares elements from CFOP and Roux, but very different from either of them. I'm not sure if such method exists yet, but I switched to this method as my main and find ways to make it faster.

I will start gathering algorithms once I get the clear that the method is truly original.

it works like this:

S1/2C :Solve D or U layer corners (I solve all U layer corners since it gives me better recognition; I will switch to D layer once I get faster)


PCE: this matches an edge with it's corresponding corner. this is pretty much building four 1x1x2 blocks, and unlike CFOP, the fact that there is no cross makes the whole step more flexible, as you can use the S and M slices more frequently, making the step much easier to learn intuitively

C&DLE: Centers and down layer edges are solved here. Wait- why solve the centers? using S and M slice moves puts the centers in place; using a similar pattern you can solve the down layer edges, which at first looks menacing, but it's really a quick and easy step to do since everything is set up from the start of the solve. This method is related to Roux in the fact that slice moves are frequent in PCE and C&DLE.


CLL: Corners of Last Layer

ELL: Edges of Last Layer


*solved cube*

Even if this method isn't as efficient as I hope, It should probably serve as a great transitional method for someone deciding to learn CFOP or Roux. But who knows?

This is just PCMS but instead of solving pairs you solve the corners then the edges.

 

[Sion]

This is just PCMS but instead of solving pairs you solve the corners then the edges.

Ah, I was expecting that. There are some differences.

PCMS solves last layer corners after pairs. TAC goes directly to solve the last layer; TAC can have the pairs solved many wayd, and glides with easier intuition than pcms, which is more of intuition from memorization. The relation between TAC and PCMS is similar to the relation between Pika********t and Petrus.

 

[crafto22]

Ah, I was expecting that. There are some differences.

PCMS solves last layer corners after pairs. TAC goes directly to solve the last layer; TAC can have the pairs solved many wayd, and glides with easier intuition than pcms, which is more of intuition from memorization. The relation between TAC and PCMS is similar to the relation between Pika********t and Petrus.

I think you fail to see how the fact that you've swapped steps around doesn't make the method different. Also, this "freedom" you speak of when building pairs is not greater with your method. If you build the corners first, you've know make edge placement very difficult compared to pair building. It's like saying beginner's F2L (corners then edges) is more efficient/has more freedom than CFOP F2L.

Here are two example solves using the same scramble. The first is with PCMS, and the second is with your method.

Scramble: D L2 D2 L2 F2 D' U2 R2 U' L' D' U' B L' U B2 R' F U2

PCMS:

z' R2 D2 // First Pair
U' M2 l U l' // Second Pair
U M U M' F // Third Pair
y' R U' R' M2 U2 r U r' // Fourth Pair
R U2 R' U2 R' F R2 U R' U' F' // CLL
U S' M S' U S' M U M' U2 M U' M' U2 y M' U2 M y' M2 U M U2 M' U M2 U2 // L8E

Total moves: 56

TAC:

x2 U' L U2 L' U' L R' U R // Bottom Layer corners
U2 R U' R' r U r' // First Pair
L U' L' l U l' // Second Pair
S2 U2 L' U L R' F' R // Third Pair
d' R U' R' r U r' // Fourth Pair
M' S M' S2 U' M2 S2 M U2 M' // Centers and D layer

At this point you could do CLL/ELL or OLL/PLL. I prefer the latter, but either is fine. So the last part doesn't have to be ELL/CLL. but since that's what you suggested (striking resemblance to PCMS where CLL is necessary) so that's how I'll finish.

U2 R U R' U R U' R D R' U' R D' R2 // CLL
U' M' U' M U2 M' U' M // ELL

Total moves: 69

So obviously your method is insanely inefficient. I got a lucky low-move ELL, so I'd estimate the average movecount would be around 75.

Your method is honestly just a weird inneficient F2L that is a mix of beginner's method and PCMS with a standard LL that you chose to use CLL/ELL for.

 

[Sion]

Another thing about TAC is flow:it's not designed for lowest move count. You can use the u slice which makes it faster. 8f there are any ways I can make this method a little better without making it an existing method, let me know.

Edit: I removed the s1/2e and merged it with pce, to make b4p, build four pairs.

 

[crafto22]

Another thing about TAC is flow:it's not designed for lowest move count. You can use the u slice which makes it faster. 8f there are any ways I can make this method a little better without making it an existing method, let me know.

Edit: I removed the s1/2e and merged it with pce, to make b4p, build four pairs.

Flow? Do you mean ease to lookahead? Consider this. What has better lookahead? Beginner's method F2L, or CFOP F2L? Beginner's method obviously has better lookahead, but because the number of steps there are, there's no point in having good lookahead. Also, if you make the first step just pairs, that LITERALLY makes your method into PCMS but the L8E are done in two seperate steps with CLL in between making it less efficient. I'm sorry to say this method doesn't bring anything new to the table, it just simplifies an old concept and makes it worse.

 

[crafto22]

I was thinking a lot about solving EO with pseudo pairs and came up with some random Roux/CFOP variant thing that solves EO with the F2B through pseudo pairs. Here are the steps:

1. 3x1x1 block on DL

2. Complete a Roux FB and SB and solve EO through pseudo pairs

At this point there are two options. You can either solve FD and BD, then solve COLL and EPLL, or you can solve COLL and solve a super easy L6E. I prefer the first option, but only because I am a CFOP solver. I know the whole “Solve EO through pseudo pairs” thing might sound confusing, so here is an example solve:

Scramble: R2 U2 F2 U2 F2 D L2 U B2 L2 U' B L F' L' D R U2 F2 U' B

Solution:
z2 U F2 L U' L' // 3x1x1
R' U' R u R' // DR + some EO
r2 R' U2 R' U' R U R' U' R U R' // FR Pseudo-Pair
u2 U r2 U' R2 U R2 U' R' U R M2 U' y R U2 R' d' L' U L // FL Pseudo-Pair + lots of EO
U' M' U2 M' U2 // F2L
r U R' U' r' F R F' // COLL
U F2 U' M' U2 M U' F2 // EPLL

High movecount cuz I don't have a cube with me, I had to use alg.cubing.net, and it's hard to be efficient.

 

[Renox]

New(I think) Roux variant:
1. FB
2.SB
3. Place the UL and UR edges on the U layer
4.CMLL
5.Permute and Orient the UL and UR edges
6. EOPMS(Edge Orientation and Permutation of Middle Slice)? Not sure how many algs

Pros:
No idea lol
Cons:
High Alg Count?
Isn't Roux

 

[shadowslice e]

New(I think) Roux variant:
1. FB
2.SB
3. Place the UL and UR edges on the U layer
4.CMLL
5.Permute and Orient the UL and UR edges
6. EOPMS(Edge Orientation and Permutation of Middle Slice)? Not sure how many algs

Pros:
No idea lol
Cons:
High Alg Count?
Isn't Roux

This is just sandwich and it sucks (both for alg count and movecount and recognition)

Also,

Isn't Roux

How is this a pro?

 

[sqAree]

Well, he told it is a con actually.

 

[Renox]

This is just sandwich and it sucks (both for alg count and movecount and recognition)

Also,


How is this a pro?

Of course it sucks lol, I made it purely for fun and not for actual speedsolving(you could, but it would be really bad).
And "Isn't Roux" is in the cons?...

 

[shadowslice e]

Of course it sucks lol, I made it purely for fun and not for actual speedsolving(you could, but it would be really bad).
And "Isn't Roux" is in the cons?...

Oh yeah I didn't see the cons cause I thought you would have broken the pros and cons apart. My mistake.

 

[crafto22]

L10P - A new alternative to LS/LL

So everyone knows that CFOP, Petrus and ZZ, among other popular methods use an F2L/LL system. Essentially, one solves the first two layers of the cube, and then solves the final layer most often using two algorithms. Here is what I propose:

1. F2L-1
2. L5E
This style of last five edges would use mainly F, R and U moves to quickly solve the last five edges. This would require quite a few algs, so the best method I could think of was to quickly orient then permute the edges using two short algs.
3. L5C
Last five corners is quite a few algs, and you can learn it entirely if you want, but I propose using an orient-permute method like the edges.

Why is this better than LS/OLL/PLL?

I don't know if it is, but it might be. This method is essentially a 4-look LS/LL, compared to CFOP's 3-look LS/LL, but if you were to learn many algs, you could have a 2-look LS/LL. Honestly I think it would be best to use only some very fast cases.

I'll update this after I've generated the algs. I'm generating these for fun, and because some cases might be useful.

 

[shadowslice e]

So everyone knows that CFOP, Petrus and ZZ, among other popular methods use an F2L/LL system. Essentially, one solves the first two layers of the cube, and then solves the final layer most often using two algorithms. Here is what I propose:

1. F2L-1
2. L5E
This style of last five edges would use mainly F, R and U moves to quickly solve the last five edges. This would require quite a few algs, so the best method I could think of was to quickly orient then permute the edges using two short algs.
3. L5C
Last five corners is quite a few algs, and you can learn it entirely if you want, but I propose using an orient-permute method like the edges.

Why is this better than LS/OLL/PLL?

I don't know if it is, but it might be. This method is essentially a 4-look LS/LL, compared to CFOP's 3-look LS/LL, but if you were to learn many algs, you could have a 2-look LS/LL. Honestly I think it would be best to use only some very fast cases.

I'll update this after I've generated the algs. I'm generating these for fun, and because some cases might be useful.

How is this different from M-CELL?

Edit: oh I see you don't have the last edge in DF but it's still pretty much MGLS.

Also, it's better to do Corners->edges because you have more symmetries as he corners have worse algs and recognition.