The New Method / Substep / Concept Idea Thread

[TheSlykrCubr]

Its not worse

It looks to be bad, I don't know what zz-bridge is, but any 3x3 2H method that has parity (even really short) shouldn't be worth it

 

[Username: Username:]

Its not worse

firstly, let me compare the steps:
Random Cross > Random Cross + EO, not an insult to EOCross, but like, EO is not really direct solving, but combine that with random cross and you get bad lookahead, so RRLF wins this one.
random corners F2L without EO = random corner then random edges with EO, F2L is better than LBL but the very clear lack of direct solving makes lookahead terrible, both of them are in a draw in this step.
OLL < OCLL, we know OCLL is much better so ZZ-Bridge wins this one.
first PLL = PLL on top, PLL is nice so both of them wins here.
second PLL = PLL on bottom, another PLL is not good so they both lose here.

it's not really better or worse but both of them are bad for speedsolving. also, just wanted to say they're very indistinguishable with each other.

 

[PapaSmurf]

Solving orientation the same way as you solve permutation (ie doing F2L but not solving the pieces properly) is a bad idea. Cross and F2L is pretty much exactly the same time as random cross. Same with RCF2L. LL is same speed as normal LL THEN you have to another algorithm on D. ZZ Bridge has the same problem. Also I invented this method when I avereaged about 35 seconds, it's not really a new one.

 

[Silky]

Just a quick expansion on ZZ-LOL; meant for OH and FMC.

Steps:

(1) EO + Line on left
(2) 1x1x3 pseudo-block placed on LD.
(a) When placing 1x1x3 you simultaneously place any two U/D edges ( except RD edge relative to your pseudo-block ) in FD + BD. Usually you can influence at least one correct FD/BD edge relative to the pseudo-block, preferably on DB.
(3) Finish pseudo-F2L
(4) Perform COLL + L5EP/L6EP ( depending on what your FD/BD edges are ). Perform AUF + ADF.

This method is very similar to the Lin method on Square-1.

Benefits:

(1) Pseudo-block building is more efficient than standard block-building.
(2) RUD is more ergonomic for OH.
(3) COLL + L5EP/L6EP is more efficient than OLL + PLL ( on average ~4 moves based on comparison of average move counts on wiki )

Cons:

(1) ZBLL becomes extremely difficult unless you have correct FB/DB placement.
(2) Look ahead become significantly harder ( but since OH provides more time for look ahead this can be mitigated ).
(3) Pseudo-block building is hard.

Perhaps someone has already talked/expanded on this, if so forgive me.

 

[TheSlykrCubr]

seems good for OH, not for 2H

 

[PapaSmurf]

ZZ-LOL has already existed as EOLine on left (aka EOEdge from SSC), F2L RUD, z', ZBLL. You could also do EOLOL, FB on DL, SB on R, ZBLL. Normal ZZ is very good for OH anyway, so I don't see why you'd do this.

 

[Owen Morrison]

ZZ-LOL has already existed as EOLine on left (aka EOEdge from SSC), F2L RUD, z', ZBLL. You could also do EOLOL, FB on DL, SB on R, ZBLL. Normal ZZ is very good for OH anyway, so I don't see why you'd do this.

Normal ZZ is terrible for OH.

 

[Silky]

ZZ-LOL has already existed as EOLine on left (aka EOEdge from SSC), F2L RUD, z', ZBLL. You could also do EOLOL, FB on DL, SB on R, ZBLL. Normal ZZ is very good for OH anyway, so I don't see why you'd do this.

Well this is awkward.. It seems I have created a new method by complete accident ( help me name it i guess.. ).

As I said above RUD is more ergonomic than LUR. Pseudo-block building is more efficient than block building. You would also have lots of continuations after Pseudo-F2L.

If neither FD/BD are solved you can use COLL + L6EP.
If one edge is solved (FD/BD) than you can use COLL + L5EP. You could also use WV + HKPLL.
You can also permuted FD/BD after pseudo-F2L ( or pseudo-F2B ? ) and finish with ZBLL. If you choose to do this you technically have more time to recognize ZBLL since you can recognize CLL as you permute FD + BD.

Normal ZZ is terrible for OH.

Booo ! Go back to the OH Method Debate Thread

 

[Redcrosscfop]

Perform AUF +DUF

What does DUF mean.

 

[Silky]

What does DUF mean.

Mean to be ADF ( adjust D face ). Just edited it.

 

[RedstoneTim]

Normal ZZ is terrible for OH.

I don't want to get too off-topic here, but <RUD> is definitely worse than <RUz> or <RUF> for OH, virtually all good OH solvers (for example Antonio Kam) agree on this.
Also, ZZ is not terrible for OH. While it may not be as good as Roux, its potential has been proven by Phil Yu with his NAR and Simon Kalhofer, who currently is 22nd in the world and 2nd in Germany for OH average.

 

[Spacey10]

Well, posted this in the wrong thread lol
I don't even know if this so roux, but let's see, oh gosh.

Roux FB and SB like normal
COLL
Solve the white side only MU gen
Then there are 3 ways to solve this:
1. Beginner way would be like 3x3 beginner method, learn only one algorithm to solve the yellow side, but in this case it is M' U' M U' M' U' M U' M' U' M, which cycles through all the possible cases for the top layer.
Then, PLL with MU gen algorithms, because corners solved
2. Intermediate way is to memorize 3 algorithms for the yellow side, I don't want to write them down, remember, all MU gen. Then PLL all MU gen.
3. BIG BOI way is to memorize ALL EPLLs, and making sure they are all MU gen.
This probably sucks, or it may be another method, I do t know, I only know of CFOP

 

[PapaSmurf]

I think everyone "invented" some form of that method. If you were willing to read all 330 pages in this thread you would find that multiple times. It's even mentioned in the first post in this thread. Anyway, keep on thinking! Maybe you'll come up with the best method, who knows?

 

[ProStar]

Well, posted this in the wrong thread lol
I don't even know if this so roux, but let's see, oh gosh.

Roux FB and SB like normal
COLL
Solve the white side only MU gen
Then there are 3 ways to solve this:
1. Beginner way would be like 3x3 beginner method, learn only one algorithm to solve the yellow side, but in this case it is M' U' M U' M' U' M U' M' U' M, which cycles through all the possible cases for the top layer.
Then, PLL with MU gen algorithms, because corners solved
2. Intermediate way is to memorize 3 algorithms for the yellow side, I don't want to write them down, remember, all MU gen. Then PLL all MU gen.
3. BIG BOI way is to memorize ALL EPLLs, and making sure they are all MU gen.
This probably sucks, or it may be another method, I do t know, I only know of CFOP

Proper way to do it would be F2B->CMLL->White Side->ELL. It's been invented a lot, along with F2B->White->OLL->PLL and other variations. It's not great

 

[Metallic Silver]

Zim Method?? Idk what to call it, it's a 4x4 method.

1. 2 Opposite centers (white and yellow) + Line [like 2/4 cross in a line] (or you could do them separately)
2. Solve all centers
3. Pair up 4 yellow/white edges and place it in LD, RD, LB, RB
4. Edge pair the rest
5. EO/EO+Parity (I recommend using algs, however idk algs for orienting 3 edges on top or 5 bad edges parity algs)
6. 3x3 ZZ = ZZF2L ---> ZBLL or other variants

 

[Cubingcubecuber]

With Sarah’s Advanced on skewb, you avoid Z-perms and H-perms. What if you were to do the same on COLL for 3x3. 2 algorithms per case, each one solves a different ZBLL, you know which to do to give you a U perm or a skip. For example R2 D R’ U2 R D’ R’ U2 R’ and R2 D r’ U2 r D’ R’ U2 R’
Would be 80-84 algs including COLL depending on if the PLL cases count(2 for diag and 2 for adj). It seems like a nice stepping stone to ZZ, and if your U-perms are really fast, it could be quite viable, especially if you predict where they’ll be and which one you’ll get. I name this system UCOLL because it is COLL that forces U perms

 

[PapaSmurf]

It's definitely possible. Learning ZBLL (obviously) skips them. I would say though that doing it that way isn't any more useful than just doing OCLL/PLL when you have a z perm case and knowing which PLL you'll have.

 

[Athefre]

With Sarah’s Advanced on skewb, you avoid Z-perms and H-perms. What if you were to do the same on COLL for 3x3. 2 algorithms per case, each one solves a different ZBLL, you know which to do to give you a U perm or a skip. For example R2 D R’ U2 R D’ R’ U2 R’ and R2 D r’ U2 r D’ R’ U2 R’
Would be 80-84 algs including COLL depending on if the PLL cases count(2 for diag and 2 for adj). It seems like a nice stepping stone to ZZ, and if your U-perms are really fast, it could be quite viable, especially if you predict where they’ll be and which one you’ll get. I name this system UCOLL because it is COLL that forces U perms

Seems like you're essentially talking about what's currently known as CLL+1. It's an idea that has been around for about 10 years now. I'm currently developing this along with Louis de Mendonça (TDM). We will have both the oriented edges form (COLL) and the form for when edges aren't oriented (CLL). We are almost finished and it should be released soon.

I myself have spent over 100 hours developing this. From the system, to the algs, to the structure, and other things. I'll be glad to finally have this finished after having thought about this so much for 10 years.

 

[N's-cvt]

I had an idea for a ZZ variant method today and I don't know if it has already been invented but here it is. Start with EO-line but instead of forcing your DB edge to be the actual DB edge you use any of the 4 U-layer edges making the line more efficient, then solve ZZF2L normally, then do COLL and I'm not sure if having the freedom of the DB edge be "open" would allow for better coll's but I think it could improve 1 or 2 cases, then do a L5EP with the DB edge instead of the DF edge. There is only 16 algs (12 l5ep/4epll) and all are below 13 moves most algs being solved MU which allows for fast tps. I already made the DB L5EP algs today and wondered if anyone would be insterested in this varaint and the algs.

 

[TerryD]

I had an idea for a ZZ variant method today and I don't know if it has already been invented but here it is. Start with EO-line but instead of forcing your DB edge to be the actual DB edge you use any of the 4 U-layer edges making the line more efficient, then solve ZZF2L normally, then do COLL and I'm not sure if having the freedom of the DB edge be "open" would allow for better coll's but I think it could improve 1 or 2 cases, then do a L5EP with the DB edge instead of the DF edge. There is only 16 algs (12 l5ep/4epll) and all are below 13 moves most algs being solved MU which allows for fast tps. I already made the DB L5EP algs today and wondered if anyone would be insterested in this varaint and the algs.

I would be interested in the algs, however, why not let DF be any of the U-layer edges instead?