New 3x3x3 Method: 3JPB Method

[JackTriton]

Hi, I'm Jack Triton.

3 months ago, I made NJPB method which was advanced OBLBL method but this time, I made it into 3x3x3 method.
Although the steps are similar to LEOR and APB, the concept of it is different.

Steps

1. First Block: Make 1x2x3 block on your non-dominant hand
2. EO: Solve Edge Orientation using M, R, U and r.
3. WBLL (Wide Block & Last Layer): Solve rest pieces using several substeps, decide by how the rest pieces formed.

First Block and EO can be combined into FBEO, which is one of starters of LEOR.

Pros

• Ergonomic move sets after First Block
• Low move count.
• No rotations required.
• Easy selection of efficient solution.
• Can be used on OH as well.

About WBLL
There are several ways to do WBLL and these are the one that I used often but you can also make your own substep sets as well.

1. CFOP/ZZ/LEOR type

For this type, you solve middle block first, and then you solve second block, last layer.
There are several kinds of substeps such as: ( Inside () is what I use as decision to solve WBLL with these substeps.)

1. Middle Block >> Second Block >> ZBLL (Easy to solve Middle Block)
2. Petrus Block >> Slab Insertion >> ZBLL (Look-a-head-able 2x2x2 Block, if D is white, you can solve it with yellow layer as well)
3. Rest Cross >> F2L, ZBLL or WVLS, PLL (All dedges are placed other than FR and BR)
4. b-shape Petrus[Petrus Block + 1 dedge] >> WVLS, PLL (All dedges are placed either U or D)

2. Roux type

For this type, you solve second block first, and then you solve rest like roux method.

1. Second Block >> COLL >> LRM[LR and M Edge Permutation, EO solved Last Six Edges] (Easy to solve 3 dominant hand edges)
2. Second Block + WVLS >> CP, LRM (Easy to create Second Block)
3. Second Block >> Middle Block, ZBLL (Easy to solve Middle Block after solving Second Block)

3. FMC type
These are not for normal solving, but I'll document it.

1. 6CO + Middle Layer Belt - Domino + First Block >> 6CP, 7EP
2. (Solve middle layer while EO) Domino Reduction

Sample Solves
1. R2 L2 U B2 D' F L' B' D' R D2 R' F2 R' U2 F2 L' B2 U2 R2

Solution:
(z2 y) // Inspection
D U R U' r' U' L U2 L' R2 U R' U F' // First Block
U r U r U R U' r' // EO
U' R U' r U2 r // Rest Cross
R' U' R U' R' U R // dBR insert
R U R' U' R U2 R' // Pairing
U R U' R' U R' U' R U' R' U2 R // WVLS 5
R2 U' R' U' R U R U R U' R // PLL Ua
U // AUF

2. F' B2 D2 U2 L' D2 R U2 R D2 L2 F2 R F' U' B' F2 L B2 R' F2

Solution:
(x2) // Inspection
R B' R2' L U' D' F2 L2 // First Block
M' U M2 U R' U r // EO
U' r2 U' R U2 R' U' M U2 R2 r // Petrus Block
U R' U' R U' R' U2 R U' R U M' U2 r' // Front Slab
R2' F2 R U2 R U2' R' F' R U R' U' R' F' R2 // ZBLL T50
U' // AUF

3. L' U D' B R' D' L U2 B' D2 R2 D2 F' R2 D2 F' B' L' U2

Solution:
(y2) // Inspection
L' D' U' L' U R D L2 R U F' // First Block
M' U' M' R' U M' U R' U' M' // EO
R' U' R' U R' U2 R' U2 R' U R' U2 R U2 R' U R U R' U R // Second Block
U R' U' R U' L U' R' U L' U2 R // COLL B6
U M U2 M' U' M' U2 M2 U2 M' // LRM
U2 // AUF

Thanks for reading!
If you’re interested, please leave a comment and share your ideas!

 

[JackTriton]

Added:
Found some great steps along the method, so I’ll share it.

1. FBEO >> WBB(Wide Block Building) >> ZBLL
the fastest way to solve, hardest to remember all.

2. FBEO >> WBWV(Wide Block Winter Variant) >> PLL
one of second fastest way to solve, much easier to remember all and my current method to use.

3. FBEO >> SBWV(Second Block Winter Variant) >> PEPMLL(Paring Edge Permute Middle and Last Layer, PEPMiLL)
the other second fastest way to solve, recommend for Roux solvers.

Although FBEO is hard to look a head, separating to First Block and EO would give you better recognition

 

[cuberswoop]

Now that I'm out of the loop and I don't know or remember what makes a good method I judge how good a method is off of how old the op is.

 

[JackTriton]

Now that I'm out of the loop and I don't know or remember what makes a good method I judge how good a method is off of how old the op is.

Thanks for reply!

For me, how similar and how fast the method can execute compared with what you use right now is what makes a good method for speed solving.

And for fun, how much you can learn new patterns and figure the algorithms by yourself; not by guides is.

 

[oldstylemike]

Isn't this essentially identical to the "EOMR" variant of the LEOR method? The steps outlined in EOMR being building a 1x2x3 along with Edge Orientation, and then solving the rest of the puzzle by completing the DF and DB edge and the last 1x2x3 in any fashion (which seems similar if not idetnical to the "CFOP/ZZ/LEOR Type", and the "Roux type" essentially just being Roux but with Edge Orientation?

Please correct me if I'm wrong

 

[JackTriton]

Isn't this essentially identical to the "EOMR" variant of the LEOR method? The steps outlined in EOMR being building a 1x2x3 along with Edge Orientation, and then solving the rest of the puzzle by completing the DF and DB edge and the last 1x2x3 in any fashion (which seems similar if not idetnical to the "CFOP/ZZ/LEOR Type", and the "Roux type" essentially just being Roux but with Edge Orientation?

Please correct me if I'm wrong

Thanks for the comment!

If you follow one of them, then you’ll have EOMR but starting them with FBEO and decide which substeps to use depending on how the parts are formed is the main thing for this method just like Mehta method is.

For the Roux type, it’s basically Roux with EO solved but if you use some of PLL to pair opposite edges (ex. Yellow-blue and Yellow-green), you can insert them faster

 

[NukesAway]

I hope necroposting on-topic is allowed
I like the this "bag of tricks" idea of a method, both for the flexibility and for how widely used the alg sets mentioned are.
Yash Mehta has listed some of these on his YruRU site, you might have missed some (those that don't end in 2GLL), of which the most valuable in my opinion is square on top+z'-rotation.
However I think you may have to rethink some types, especially the roux types: doing EO before the last 4 corners and then just doing roux is kind of unecessary since it doesn't add much to ergonomics (compared to vanilla roux), nor is it any more efficient because RB isn't as flexible as when free, not to mention COLL is marginally worse than CMLL. Okay in real solves this doesn't matter since you only use them when there's an easy RB but I can't throw off the feeling I'm voluntarily opting for a roux variant with more steps and more algs (possibly).
The roux type with ZBLL seems the least favourable for the alg count, recog and lookahead .Have you considered putting the UR and UL edges in the DF and DB positions like NMP? This way you can just COLL then 4C (which is about 7 STM of brainless spamming). Movecount is about the same.
Now for normal EO223. Yes EOFB is a nightmare to plan but remember you can do normal roux FB with partial EO control then EODFDB can be done algorithmically (rUR gen). Doing EODFDB seperately with <M,R,U> is probably a bad idea.

 

[CR7GoatSiuuu]

eo FB Plus br pair plus wv permute corners and Dr edge and then finish with roux