The New Method / Substep / Concept Idea Thread

[Existential Shrimp]

I came up with something similar before, solving a cross (edges don't have to be permuted), doing F2L (corners can be in wrong position, edges have to be in the right position), doing OLL, then do PBL.

Advantages are lower movecount, more freedom during F2L so it could be more efficient, disadvantages are too hard to recognize the PBL case fast.

that actually sounds good. kinda like pseudo-slotting but not really. so like, pseudo-pseudo-slotting, into PBL. also are there actually PBL algs on 3x3?

 

[Thom S.]

I came up with something similar before, solving a cross (edges don't have to be permuted), doing F2L (corners can be in wrong position, edges have to be in the right position), doing OLL, then do PBL.

Advantages are lower movecount, more freedom during F2L so it could be more efficient, disadvantages are too hard to recognize the PBL case fast.

Don't think you are saving a lot of moves. You are essencially using moves to solve the first layer corners twice.

 

[Existential Shrimp]

Maybe there are, but I didn't find them anywhere (not that I searched a lot)
I can gen them though, I'm still trying to figure out how to make batch solver generate PBL algs

i tried j/j and it worked

my guess is that anything that only has 3 and -3 moves will work, because its essentially just a quarter turn

 

[OreKehStrah]

idk if this is 100% original, but i came up with this one called square-3 for 3x3. you solve one side, use any method you want to solve the edges on the opposite side, solve the oll, do pll on both sides, solve the middle 4 edges in any way you want.

note: you can get single or double parity when trying to do your plls. do M2 U2 M2 U2 for double parity. if you have single parity, rotate the cube so your - usually top face - is in the front. then do an e-perm but do wide R moves instead of the regular ones. i can record an example solve if you want, bc i suck at explaining things while typing.

These types of methods are not new or good lol. I have genned random pbls and they are also quite bad

 

[Deleted member 36445]

your favorite EPLLs are Z and H and you simply hate U perms?

then this is your lucky day!

do phasing at the last pair of a ZZ or petrus solve and use COLL. you are guaranteed to have 1 out of 3 cases: z, h or skip

:3

of course this a joke method

 

[Thom S.]

your favorite EPLLs are Z and H and you simply hate U perms?

then this is your lucky day!

do phasing at the last pair of a ZZ solve and use COLL. you are guaranteed to have 1 out of 3 cases: z, h or skip

:3

of course this a joke method

I'll buy 3 of those.

 

[Existential Shrimp]

your favorite EPLLs are Z and H and you simply hate U perms?

then this is your lucky day!

do phasing at the last pair of a ZZ or petrus solve and use COLL. you are guaranteed to have 1 out of 3 cases: z, h or skip

:3

of course this a joke method

b-but i l-love u perms (((

 

[NigelTheCuber]

Sais the one who likes 2x2.

i don't see a problem in that...

 

[NigelTheCuber]

People who only like big cubes (5, 6, 7) because of Kevin Hays and Feliks Zemdegs.

and Max Park

 

[Hazel]

idk if this is 100% original, but i came up with this one called square-3 for 3x3. you solve one side, use any method you want to solve the edges on the opposite side, solve the oll, do pll on both sides, solve the middle 4 edges in any way you want.

note: you can get single or double parity when trying to do your plls. do M2 U2 M2 U2 for double parity. if you have single parity, rotate the cube so your - usually top face - is in the front. then do an e-perm but do wide R moves instead of the regular ones. i can record an example solve if you want, bc i suck at explaining things while typing.

I'll go into more detail as to why this is flawed, to hopefully give you a better understanding of methods if you want to come up with more in the future.

Step 1 is fine, it's even made better if you combine step 1 and 2. There's the possibility of OLL parity, but not if you're careful. The problem comes with PLL on both sides. Doing a PLL followed by a z2 followed by another PLL is slow. Also, due to the nature of the E-slice not yet being solved, you only avoid parity 25% of the time, which is really bad (25% of the time you need an M2 U2 M2, 50% of the time you need an R2 U2 R2 U2 R2 because it can be needed on either side, 25% of the time there's no parity). You can build your first side with special care to bring the odds of parity from 75% to 50%, but still, not great. Then you have the E slice to solve, which isn't great since the edges aren't necessarily oriented. I opened CubeExplorer to take a look at what the solution algs might look like, and they aren't good.

Expanding on what Thom said, solving something incorrectly (e.g. solving a face instead of a layer) is usually a bad idea. Solving a full layer isn't that much slower than solving one side, and it avoids having to do PLL later, but at that point you're just doing a worse version of the beginner's method, so not ideal.

I don't mean to be too negative, if you're interested in method creation then definitely keep at it, and your ideas will become better over time!

 

[Deleted member 36445]

I just invented 2 methods in like 20 seconds!

The hope method:

1. hand scramble the cube hoping it will be solved.
2. if not solved, go to step 1
3. solved!

The zeroing method:

1. solve the method you want, but be assured to make a mistake that will lead you to an even faster solution
2. solved!

 

[bulkocuber]

I just invented 2 methods in like 20 seconds!

The hope method:

1. hand scramble the cube hoping it will be solved.
2. if not solved, go to step 1
3. solved!

The zeroing method:

1. solve the method you want, but be assured to make a mistake that will lead you to an even faster solution
2. solved!

There's already a method called Hope, but it's an OLL parity avoidance method

anyway, your hope method seems really inefficient compared to the zeroing method.
(Also, doesn't zeroing have to do with inserting the last F2L corner with a commutator getting an LL skip or an easy PLL?)

 

[IsThatA4x4]

There's already a method called Hope, but it's an OLL parity avoidance method

anyway, your hope method seems really inefficient compared to the zeroing method.
(Also, doesn't zeroing have to do with inserting the last F2L corner with a commutator getting an LL skip or an easy PLL?)

The z(z)eoring made by @ruffleduck does that, yes. You go up to LS with ZZ, then do OCELL but leave the DFR (assuming FR as last slot) corner misoriented (leaves another corner on top misoriented too, the corner at DFR MUST be an LL corner)
Then you do a variant of twisted TTLL that can just be solved with 2 commutators max every time (because the only cases you can get are pure 3 and 5 cycles). It's a cool method!
Pretty sure zeroing was a joke term for a long time about what feliks did that made his solves so fast, although I don't know the history that well.

 

[hyn]

Method: Reduction to RU-gen z2 RU finish a.k.a. RRUz2RU
1. Orient all edges and place the 5 FL, DF, DL, BL and DB edges in the FL, DF, DL, BL and DB slots, in any order(not necessarily solved).
2. Use Old Pochmann corners(or 3-style?) to permute all corners. Other methods to do this can also be used, given they don't unorient edges or move the 5 edges from step 1.
3. Solve the 2x2x3 block from UF to BR using R and U moves. This can be done by moving the UR edge to DR and solving the two slots while phasing
4. z2
5. Solve the rest

 

[CornerTwisted]

It is an interesting method I have been developing. I have been told that this method is very bad, but here is the tutorial! https://docs.google.com/spreadsheets/d/19SgE0Rnjys9ga0w01g7hFER1HhdpK10r05lEA5DcKd4/edit?usp=sharing

 

[Deleted member 36445]

isn't this just a slightly worse version of petrus?

 

[Timona]

It is an interesting method I have been developing. I have been told that this method is very bad, but here is the tutorial! https://docs.google.com/spreadsheets/d/19SgE0Rnjys9ga0w01g7hFER1HhdpK10r05lEA5DcKd4/edit?usp=sharing

This is actually very cool. Just saying...

 

[PiKeeper]

It is an interesting method I have been developing. I have been told that this method is very bad, but here is the tutorial! https://docs.google.com/spreadsheets/d/19SgE0Rnjys9ga0w01g7hFER1HhdpK10r05lEA5DcKd4/edit?usp=sharing

This seems like slightly worse LEOR

 

[OreKehStrah]

This seems like slightly worse LEOR

Yeah it's just worse petrus because you loose out on efficiency from following a restricted path to 223, which is part of why mehta also is bad

 

[CornerTwisted]

I no longer trust your judgement

part of why mehta also is bad