• Welcome to the Speedsolving.com, home of the web's largest puzzle community!
    You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

    Registration is fast, simple and absolutely free so please, join our community of 40,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!
Sorry, didn't see that part. I read it as
EOLine
solve Roux blocks with LUR
COLL
"Roux-style LSE", which if it was what I read would just be EPLL.

But what I said is still true. What I said there was actually about a method very similar to yours, except without the extra rotation. I think my misinterpretation could be slightly better, but only slightly, since it's only missing one rotation. I still don't think it's as good as Roux.

Yours is better, and has been suggested before. It's usually called EORoux, and its main flaw is that you basically lose your slice moves, which are one of Roux's greatest strengths, for little gain. Yo also need to build your first block on the fly, at least partly, which is difficult to do efficiently.
 
Yours is better, and has been suggested before. It's usually called EORoux, and its main flaw is that you basically lose your slice moves, which are one of Roux's greatest strengths, for little gain. Yo also need to build your first block on the fly, at least partly, which is difficult to do efficiently.
I know. It's not worth using for speedsolves (or, not as a main method at least).
 
About your whole "OLSLL" idea,
You could just solve with ZZ, and use Winter Variation, or use any method to solve your F2L and then orient LL edges, and then WV. your idea is not very different. The only thing you have that isdifferent is that you pair the pieces for the last F2L slot and then solve the pair, and so forth

Also, an Idea for a method that could be used for FMC I think, if put into practice.

Step one. Solve two columns on opposite sides of the cube

Step two. Expand those columns to 2x2x3 blocks each

Step three. Solve the cube using ? amount of algs.
 
Last edited:
About your whole "OLSLL" idea,
You could just solve with ZZ, and use Winter Variation, or use any method to solve your F2L and then orient LL edges, and then WV. your idea is not very different. The only thing you have that isdifferent is that you pair the pieces for the last F2L slot and then solve the pair, and so forth

Also, an Idea for a method that could be used for FMC I think, if put into practice.

Step one. Solve two columns on opposite sides of the cube

Step two. Expand those columns to 2x2x3 blocks each

Step three. Solve the cube using ? amount of algs.
Looks like too many slice moves to be good for FMC, but you could be solving it differently to how I'm doing it.
The number of cases for the last step is double ZBLL (NMZZLL + NMZLL with two flipped edges), so 335 not including solved. If you can force edges to be oriented then you can setup to a NMZZLL case, but imo it's not worth it.
 
Sure. Here is the example solve: here

Oh, and yeah, forgot slice moves count as 2 moves XD
BTW the alg that solved the cube after the blocks were made, was made in cube explorer (had to modify it for HTM so R R was R2)
 
So I've been thinking and I think I've derived a cool new method for solving the last slot on any Friedrich solve while also doing OLL. This will be a 2-look system, but isn't normal LS/OLL 2-look also?

OLSLL stands for "Orient Last Slot and Last Layer." Essentially, you do the following steps:
LSP - Last slot pairing: Pair up the two pieces for the last slot. They do NOT have to be oriented at all (disoriented edge and twisted corners are fine). This should be very quick.
L5EO - Last 5 edges orientation: Insert the pair and orient the last 5 edges on the puzzle in one algorithm (one of 16). Most of these are 3-gen, but some are 4-gen. :/
L5CO - Last 5 corners orientation: Orient the last 5 corners on the puzzle in one algorithm (one of 23). These are all <R, U> 2-gen sexiness.

To show you what a solve could look like, I have one here to show you guys:

Scramble: L2 U' B' U2 R' F2 L2 B' L F' L2 B2 L U2 R2 U B' R2 D' B2
Cross (on bottom): x' z D U' R L2 B' U' L2
F2L Pair 1: y' R U' R' U2 y R' U' R
F2L Pair 2: U' F' U F
F2L Pair 3: L' U L y' L' U' L
L5EO: U2 R U2 R' F' U' L' U2 L U' F
L5CO: U R U2 R' U2 R U2 R' U' R U R'
PLL: U' Dw2' R U R' F' R U R' U' R' F R2 U' R' U'

You may visually see the solve executed here. I do not currently have public algorithms available.

NOTE: This is not MGLS, EJF2L, VHLS, or anything like that. Similarities are intentional, but this method is not any that I've seen before.

ALSO NOTE: This method is purely an alternative to LS+OLL, similar to how CLL+ELL is an alternative to OLL+PLL.

What are you guys' thoughts on this? I think this is pretty cool and am currently learning the algorithms to put it in practice myself. Please keep this to a nice thread. I don't want this thread cancerous with all of this "YOUR METHOD SUCKS JUST USE NORMAL CFOP" stuff. This is a new idea. Also, if you want more example solves, just ask! I can do a few for you guys if you want.

This is just MGLS
 
This is a bit like a ZZ variant... what would you say the advantages of this over regular ZZ are? The second block has more moves than a normal SB and leaves you with something harder to solve, with less control over what you're able to do.

you can just do:

eoline

block 1

block 2 with double premove, while phasing to solve UF, UB

undoing the double premove

do some weird algs (same amount as ZZLL).


it might actually be LESS moves than ZZ, because you could always do the premoved block first or second, depending on when it seems easier.



but at that point its almost better to just leave the premoved block on D and just do regular ZZLL, undoing the premove at the end.
 
So recently I've been playing around with columns first methods and came up with this method. Its similar to PCMS; the other columns first method I came across but different in how the columns are formed.

Step 1) solve the e-slice
It doesn`t actually have to be the e-slice as long as the e-edges are permuted and oriented in relation to each other even if the centres are not solved. However I prefer to do the e-slice because I find it easier to plan out the moves so inspection time can be spent predicting the corner orientation cases.

Step 2) correct corner parity
1 in 2 chance of a skip. A corner parity is where the corners are oriented in such a way that cannot be corrected with standard roux algorithms as they belong to a slightly different, less restricted subset.

Step 3) orient all corners
I usually do this by orienting four corners first then bringing them to the bottom and then using roux algs for the top layer. This step can probably be made more efficient by using an alg that does a sune on the top and a sune on the bottom and then repeating or altering some of the orientation algs on a 2x2 and making them e-slice safe. This would create a new set of algs. However, this could be quite a large alg set. An alternative method for this step could be to orient three corners as in guimond then edit the algs to be e-slice safe .

Step 4) bring corners to the correct layers
Intuitive, takes no more than five moves

Step 5) permute all corners
This can be done using the algs for corner permutation on a square one.

From here you can either use the other columns first method or fill in the Rd and Ld edges as you would have in roux then do LSE.

Personally I prefer to do the last bit in three additional steps;
1) orient all edges in a manner similar to roux again.
2) pair the two Rd and Ld and insert in their respective positions (this is similar to the second last step on roux but upside down)
3) standard LSE with all edges oriented

NOTE: Although it is probably possible to combine steps 2 and 3 using algs that orient all corners on the u-face while also twisting a corner on the d-face so that it will reduce the corner space to that covered by COLL, i have not yet created the required algs for this. The permutation of these corners, which I have currently divided into 2 steps can likely also be combinded into one step though again I have no algs for this. In addition to not yet having the algs, I would reason that it is indeed better to keep each of the steps seperate as this would be unlikely to dramatically increase the move count (correcting the parity and bringing the corners to the correct layers will take no more that 5 moves) and would speed up recognition time.

This is dissimilar to the belt method as it deals with corners and edges seperately so is more like roux and the belt method is more like CFOP/fridrich.

I've done some research and not found anything exactly the same as this method. What does everyone else think?

UPDATE: I found this very helpful post on Ryan heise's website that helps with the corners; "Now for the corners, we can twist two corners at a time. Start with one
corner at LUF (corner A) and the other corner at FDR (corner B). Then
do:

R'D - twists the B corner
L2 - swaps the B and A corners
D'R - twists the A corner on the way back

Do this repeatedly until all corners are twisted the right way" this could be used either as suggested or be done to correct corner parity.
 
Last edited:
So recently I've been playing around with columns first methods and came up with this method. Its similar to pcmc? ( I think that's what its called) the other columns first method I came across but differentin how the columns are formed.

Step 1) solve the e-slice
Step 2) correct and corner parity (1 in 2 chance)
Step 3) orient all corners- I usually do this by orienting four corners first then bringing them to the bottom and then using roux algs for the top layer. This step can probably be made more efficient by using an alg that does a sune on the top and a sune on the bottom and then repeating
Step 4) bring corners to the correct layers- intuitive
Step 5) permute all corners (can be done using the algs on square one)

From here you can either use the other columns first method or fill in the rd and ld edges as you would have in roux then do lse.

I've done some research and not found anything exactly the same as this method. What does everyone else think?
Seems a lot like the Belt method...
 
So recently I've been playing around with columns first methods and came up with this method. Its similar to pcmc? ( I think that's what its called) the other columns first method I came across but differentin how the columns are formed.

Step 1) solve the e-slice
Step 2) correct and corner parity (1 in 2 chance)
Step 3) orient all corners- I usually do this by orienting four corners first then bringing them to the bottom and then using roux algs for the top layer. This step can probably be made more efficient by using an alg that does a sune on the top and a sune on the bottom and then repeating
Step 4) bring corners to the correct layers- intuitive
Step 5) permute all corners (can be done using the algs on square one)

From here you can either use the other columns first method or fill in the rd and ld edges as you would have in roux then do lse.

I've done some research and not found anything exactly the same as this method. What does everyone else think?
Very cool variation. Some PBL algs on 2x2 can be edited to be "E slice safe" if you're not into square 1.
 
4x4 method- Meyer variant
1) Solve a 1x3x4 block on the left side
2) Solve a 1x3x3 block on the right side
3) solve the centres using only U, R and M moves
4) Use the six edge method to solve other edges.
NOTES:
1) i break the first step into solving it as the outer edges first at the same time as two 1x1x3 blocks and then placing and the doing essentially the same technique for the centre-edge triplets. The second block is the same as in standard reduction/ Yau sort of stuff.
2) steps 3&4 are essentially the same as Yau but rotated 90 degrees so is in a similar orientation to the one that Yau himself uses.


I prefer this method for three reasons'
1) I prefer m moves to e moves as they are more ergonomic in my opinion (I use roux as my main method)
2) there is only one "hidden" edge as opposed to the two in Yau (small difference I know but I find it helps me)
3) it sets up to a roux 4x4 3x3 stage nicely- in the same way Yau sets up to CFOP by granting the cross, it gives almost the first entire two blocks.

overall this method is probably not any better (and may be slower) than Yau. But its the method i sometimes use for solving so i thought i might as well share it.

It's probably also not much use on anything bigger than a 4x4 because the m slices are too hard to pull off efficiently.
 
Last edited:
It just occurred to me that L5E would be a great way to finish off Petrus. The steps are as follows:
1) Build a 2x2x2 block
2) Extend to 2x2x3 block
3) Place fr and br edges (this is the orientation I use for Petrus)
4) Orient corners
5) Permute corners
6) y (because I don't want to use s moves in a speedsolve)
7) L5E

All steps from 3-5 are 2-gen.
Step 7 can be treated as 2-gen as well using only U and M moves.

This method can also be used in conjunction with Yau and and the fourth cross edge can essentially not be a required step for Yau and can be done along with the other edges.

What is the 6 edge method? And I assume that we are to solve the centers after the second block square thing, right?

Yes, you do solve the centres after the second block. Sorry I forgot to say

The six edge method is also used in Yau but uses e-slices rather than m-slices. It's kind of hard to explain through words but there are some good YouTube tutorials.
 
Last edited by a moderator:
It just occurred to me that L5E would be a great way to finish off Petrus. The steps are as follows:
1) Build a 2x2x2 block
2) Extend to 2x2x3 block
3) Place fr and br edges (this is the orientation I use for Petrus)
4) Orient corners
5) Permute corners
6) y (because I don't want to use s moves in a speedsolve)
7) L5E

All steps from 3-5 are 2-gen.
Step 7 can be treated as 2-gen as well using only U and M moves.

This method can also be used in conjunction with Yau and and the fourth cross edge can essentially not be a required step for Yau and can be done along with the other edges.
Step 5 is not 2-gen. Also, steps 4-5 can be done with CLL which isn't a very big alg set at all. Also you need to solve the DR corners at some point - maybe solve pairs in step 3.
 
Step 5 is not 2-gen. Also, steps 4-5 can be done with CLL which isn't a very big alg set at all. Also you need to solve the DR corners at some point - maybe solve pairs in step 3.

step 5 can be done as 2 gen- use the algs for square one (barring two cases which are still only three gwn so are still quite fast)
Steps 4&5, while they can be combined, cannot always be done using CLL (although CLL is a subset of the set required) as it requires the orientation and permutation of 6 corners rather than the 4 that CLL deals with.
 
Last edited:
Back
Top