The New Method / Substep / Concept Idea Thread

[TuanNgoTN]

Hi everyone, I'm Tuan and I created the TN4 method.
TN4 (short for Tuan Ngo 4x4) is a 4x4 speedsolving method proposed by Ngô Anh Tuấn in 2025. TN4 method solves up to F2L-1, starting the 3x3 stage with LSLL.
Speedsolving wiki page: https://www.speedsolving.com/wiki/index.php?title=TN4
View attachment 41892
STEPS

1. Solve 2 opposite centers.
2. Solve a 1x3x4 block on the left using one of the solved centers (like Meyer method).
3. Solve a 1x3x3 block on the opposite back right, except the top BR edge piece.
4. Solve D center and 4 F3L pairs on the F and B faces (like F3L method, each F3L pair consists of 1 edge piece and 2 center pieces).
5. Rotationless edge pairing using FR slot (the top BR edge piece will be placed correctly after this step to complete F2L-1).
6. LSLL (+ parities).

PROS

• Good lookahead and high TPS are possible in steps 5 and 6.
• F2L-1 is completed prior to the 3x3 stage.
• The blockbuilding and intuitive nature.
• Mostly rotationless.

CONS

• Requires a lot of lookahead and M-slice moves in steps 2, 3, and 4.

EXAMPLE SOLVES
Example solve #1
Example solve #2
Example solve #3
Example solve #4
Example solve #5

Here is TN4 method Tutorial video

 

[That0newe1rdo]

also where can i find more sources about 4x4 methods because i can't seem to find a lot on the speedsolving wiki

Just search "4x4 methods" here.

Youll get met with alot, which is great as if you go through them and only like certain bits you can try implementing them into another method that combines everything you like in 4x4 methods you find .

For example lets say i found these methods:

- Yau :
I like the edges and 3 2 3
I dont like the centers and restrictions when solving them.

- Hoya :
I like the centers and edges
I am aware edges can be shorter.
I dont like 6-2 (I dont even use it anymore lol i use 3 2 3 with hoya now)

- F3L :
I like the combined efficiency of edges/centers
Look ahead is hard as there are lots of scrambled pieces when doing F3L.

 

[That0newe1rdo]

View attachment 41892

yo how did you do this ?
i want to do this so its easier to understand my method

thanks !

 

[TuanNgoTN]

yo how did you do this ?
i want to do this so its easier to understand my method

thanks !

Do you mean the solution steps image?

 

[That0newe1rdo]

Do you mean the solution steps image?

Yeah

 

[TuanNgoTN]

Yeah

I used ibisPaint X app

 

[That0newe1rdo]

I used ibisPaint X app

Thanks

 

[Twisted101]

Yeah

VisualCube Editor 2.0

 

[That0newe1rdo]

Yo so i came up with a last layer method !
I call it WLL (We1rdos last layer)
So far I am currently unaware of the number of algs for the mpre advanced parts of this method, but I am 90% sure the most advanced of this method has less algs then OLL + PLL
I also think this might already exist lmao

Heres how it goes :

- Orient and Permute Edges
- Orient and Permute Corners
Estimated number of algs : 68

Here are the 3 Look versions :

1 :
- Orient and Permute edges
- Orient corners whilst preserving edges
- PLL (A perm, E perm, H perm, PLL skip)

2 :
- Orient Edges
- Permute Edges
- Orient and Permute Corners

Estimated Number of Algs - 1 : 32, 2 : 47

Here is the 4 look version (Currently the only complete version tmk) :
- Orient Edges
- Permute Edges
- Orient Corners while preserving Edges
- PLL (A perm, E perm, H perm)

Thanks !
Please give your opinions on this, and if this isnt already a thing/this doesnt get any attention I will make a thread as I know itll get more attention then.

 

[Fred]

Yo so i came up with a last layer method !
I call it WLL (We1rdos last layer)
So far I am currently unaware of the number of algs for the mpre advanced parts of this method, but I am 90% sure the most advanced of this method has less algs then OLL + PLL
I also think this might already exist lmao

Heres how it goes :

- Orient and Permute Edges
- Orient and Permute Corners
Estimated number of algs : 68

Here are the 3 Look versions :

1 :
- Orient and Permute edges
- Orient corners whilst preserving edges
- PLL (A perm, E perm, H perm, PLL skip)

2 :
- Orient Edges
- Permute Edges
- Orient and Permute Corners

Estimated Number of Algs - 1 : 32, 2 : 47

Here is the 4 look version (Currently the only complete version tmk) :
- Orient Edges
- Permute Edges
- Orient Corners while preserving Edges
- PLL (A perm, E perm, H perm)

Thanks !
Please give your opinions on this, and if this isnt already a thing/this doesnt get any attention I will make a thread as I know itll get more attention then.

LLEF - Speedsolving.com Wiki

www.speedsolving.com

Last Four Corners - Speedsolving.com Wiki

www.speedsolving.com

99 algs total.

 

[That0newe1rdo]

Here is a comparison for the 4 look version compared to standard 4 look.

Average Move Count (w/AUF) :

WLL - aMC is 45.38 (2dp)
Standard - aMC is 44.85 (2dp)

Alg Count :

WLL - 16
Standard - 16

PLL Skip chances :

WLL - 1/12 ?
Standard - 1/72

There are probably more ways to compare them, but these are the ones i think are most important.

 

[That0newe1rdo]

Just saw @Fred s post, ignore everything but the comparison bit lol.
Sad to see its already a thing, but im probably gonna learn it lol.

 

[efattah]

LLEF, CFCE and other different ordering of solving last layer always ends the same, the recognition time is much worse than standard OLL/PLL which has fast recognition. So it is not useful, to count moves, number of algs, etc, the main thing is the recognition time.

 

[Fred]

CFCE is fine as far as recognition goes. The problem with CFEC is that ELL first requires a peek at a third side, kind of like PLL on a 4x4. There's not always enough info for two-sided (plus U) recognition unless you also look at the corners, at which point you're doing recognition for 1LLL. The main problem with alternate last layer orders is that the algs are much less developed. An enormous amount of time has been spent finding the best OLL and PLL algs and figuring out the best fingertricks.

 

[ILCubing]

you do the puzzle

 

[TuanNgoTN]

Hi everyone, I'm Tuan and I created the TN4 method.
TN4 (short for Tuan Ngo 4x4) is a 4x4 speedsolving method proposed by Ngô Anh Tuấn in 2025. TN4 method solves up to F2L-1, starting the 3x3 stage with LSLL.
Speedsolving wiki page: https://www.speedsolving.com/wiki/index.php?title=TN4
View attachment 41892
STEPS

1. Solve 2 opposite centers.
2. Solve a 1x3x4 block on the left using one of the solved centers (like Meyer method).
3. Solve a 1x3x3 block on the opposite back right, except the top BR edge piece.
4. Solve D center and 4 F3L pairs on the F and B faces (like F3L method, each F3L pair consists of 1 edge piece and 2 center pieces).
5. Rotationless edge pairing using FR slot (the top BR edge piece will be placed correctly after this step to complete F2L-1).
6. LSLL (+ parities).

PROS

• Good lookahead and high TPS are possible in steps 5 and 6.
• F2L-1 is completed prior to the 3x3 stage.
• The blockbuilding and intuitive nature.
• Mostly rotationless.

CONS

• Requires a lot of lookahead and M-slice moves in steps 2, 3, and 4.

EXAMPLE SOLVES
Example solve #1
Example solve #2
Example solve #3
Example solve #4
Example solve #5

Here is TN4 method Example Solves video

 

[Bttyx]

So while searching for the pros and cons for 2x2 methods, I found out that one looking cll is pretty difficult compared to eg. However, cll only has 42 algs. So my goal was to create a method with less than 60 algs but can still be one looked easily.
Steps:
1. separate the colours while making sure the U and d layer can be orientated using oll.(for example sune case on top and pi case on bottom)
2. orient all the edges in one alg while making sure the permutation stays the same(you may think making algs that make sure the permuatation stays the same is stupid, but stay with me)
3. pbl
61 algs total
So by now, many of you are confused and are probably rushing to the comments to say this is not one look, but if you think about it, if you take not of the orientation and permutation at the start, you do not have to pause to recognize pbl after orientating. This makes this a one look method
pros-
easy to one look(beginners could probably do it too)
cases are easy to recognise
cons-
high move count(15 stm)
generally not as good as other methods
2nd step algs(incomplete)(will complete at home):

TSOL

docs.google.com

 

[That0newe1rdo]

So while searching for the pros and cons for 2x2 methods, I found out that one looking cll is pretty difficult compared to eg. However, cll only has 42 algs. So my goal was to create a method with less than 60 algs but can still be one looked easily.
Steps:
1. separate the colours while making sure the U and d layer can be orientated using oll.(for example sune case on top and pi case on bottom)
2. orient all the edges in one alg while making sure the permutation stays the same(you may think making algs that make sure the permuatation stays the same is stupid, but stay with me)
3. pbl
61 algs total
So by now, many of you are confused and are probably rushing to the comments to say this is not one look, but if you think about it, if you take not of the orientation and permutation at the start, you do not have to pause to recognize pbl after orientating. This makes this a one look method
pros-
easy to one look(beginners could probably do it too)
cases are easy to recognise
cons-
high move count(15 stm)
generally not as good as other methods
2nd step algs(incomplete)(will complete at home):

TSOL

docs.google.com

I mean you can increase the number of algs while also making it better, as in any 2x2 scramble you can actually have this (the layer you are talking about) right at the start, and its not even rare.
However, there are parity cases that appear occasionally.
Of course, you can look at a different layer thing, as you are pretty much guaranteed this is a 2 step method.
Or in the event every layer combination has this "parity" you could make a alg that fixes this.

Also, hopefully you can clear up the fact that your research was literally asking me in a DM if 1 looking CLL is hard or not - btw, I dont know full CLL so I dont know fully, so all is said is its harder to one look then EG, but didnt specify if its barely harder or not since i didnt really know.

The 15 MC is probably actually 13/12 MC since step one literally almost never isnt skipped.

Im also not suprised if this isn't like the belt method on 3x3 as I for one have already "made" this method personally like a year ago when I was fiddling with a 2x2.

Edit - ngl this method isnt very goood at all as i have realised the layer is made differently to how I thought it was.

 

[Bttyx]

I mean you can increase the number of algs while also making it better, as in any 2x2 scramble you can actually have this (the layer you are talking about) right at the start, and its not even rare.
However, there are parity cases that appear occasionally.
Of course, you can look at a different layer thing, as you are pretty much guaranteed this is a 2 step method.
Or in the event every layer combination has this "parity" you could make a alg that fixes this.

Also, hopefully you can clear up the fact that your research was literally asking me in a DM if 1 looking CLL is hard or not - btw, I dont know full CLL so I dont know fully, so all is said is its harder to one look then EG, but didnt specify if its barely harder or not since i didnt really know.

The 15 MC is probably actually 13/12 MC since step one literally almost never isnt skipped.

Im also not suprised if this isn't like the belt method on 3x3 as I for one have already "made" this method personally like a year ago when I was fiddling with a 2x2.

Edit - ngl this method isnt very goood at all as i have realised the layer is made differently to how I thought it was.

It's not parity, in the first step you make sure you get a solvable oll cases on both sides, however I plan to make a advanced variant which does this. And I did do more research than that(doing cll solves to see how many I can one look), I just asked you for confirmation.
And uh this method isn't meant to be good it is just a joke method for skill issues who find one looking cll hard

 

[notjamestack]

Well, since we are making new methods for 4x4, let me introduce to you...

Ok, just pretend I put an actual name here.

This is just a concept but I think it's cool

1. Solve a center.
2. Solve two adjacent centers next to the one you already made. (For example, Green and Orange.)
3. Pair up 3 edges and solve them with their corner, making a block.
4. Solve the last 3 centers.
5. Solve the rest of the edges.
6. 3x3 stage.

Pros:
Last 3 steps have better lookahead.
Reduces edge pairing to just R, U, and F moves.
Guaranteed XCross if you use CFOP or ZZ.
Rotations are almost completely unneeded for the last 3 steps.

Cons:
Lookahead can be tricky on the first few steps.
Roux doesn't really work well with this method.
Can be difficult to find the right edges to create a block.