An alternative way of solving (edge first)

[eo]

While learning ZZ, i played with eo+line for a few days, and feel it pretty interesting in a fun play, rather than speed solving... Here's a strange way i tried in recent a few studies of solving(about 20~30...)

(apologize if this method has already been discussed before, not mean to be garrulous...)

1. EO-Line
   
   
2. Build cross on all 6 surfaces
   - - - - A trivia job since all edges are oriented already, just pull them blindly into position, the top cross is automatically shaped up. Well, may need 1 or 2 "sune" to finalize...
   
   
3. Place all corners
   - - - - This is the part with most fun, and really can be seen as a training towards color-neutralizing... 8 corners, 3d-cube planning, 2 or 3 algs. (need more analysis to confirm it, during my trial and tests, looks there always be a way to achieve it within 3 algs...)
   
   
4. Orient one face
   - - - - This is another fun part, can be seen as a "orientation algs recognition" training... Just check each of the 6 faces, look for one can be oriented with a known algs. (if not, just break one face with a sune, it makes things back to track... however, this may just because I don't know how to deal with the special case... need more analysis here again)
   
   
5. Restore the edges
   - - - - Orienting the face breaks the edge position, so need to restore it.
   
   
6. Flip the cube(put the solved face down) and finish top layer. (as the ZZ-F2L is done ... without any paring up and slots insertions... )

Hopefully this is not a nonsense post (if not again...)

 

[chris w]

May have been previously discussed somewhere. I've been messing around with something similar ie edges first but not doing eo first. And for solving the corners, by using commutators combines the other steps into 1 or 2 steps (if you have twisted corners left at the end).
Pretty fun to do non speedsolving solves with but could give fairly efficient solutions with lots of oppertunities to cancel moves

 

[eo]

yes, it's more like an eye/brain training rather than finger race. I totally agree that it can be optimized a lot.

I do not quite understand the "commutators" yet, definately need some time to look at that...

 

[RNewms27]

EOline
Crosses
BH corners

Unless I am not quite understanding your concept

 

[chris w]

L F' L2 F' D L' F L' U D2 B' F' R D2 L2 B D' L' F D2 L2 R2 F' L2 R2

z' R2 L F L' (4/4) // block
y U R' F B U' B' F U F' (9/13) //some edges
R U R' U R U2 R' U2 (8/21) // last edges
y' D2 R U' R' D2 R U R' (8/29) // some corners
y2 F R B R' F' R B' R' (8/37) // other corners
x' D' L U2 L' D L U2 L' (8/45) // last corners

This was a solve I did a couple of weeks ago in the "Example solve" thread. Same basic idea, just there was a nice block to start it with. Haven't really tried any with and eo-line start yet but that could be quite nice also

 

[eo]

EOline
Crosses
BH corners

Unless I am not quite understanding your concept

Nice! just checked the page, looks exactly what I want when fumbling around cube these days...

Thanks for pointing it out.

 

[eo]

...
This was a solve I did a couple of weeks ago in the "Example solve" thread. Same basic idea, just there was a nice block to start it with. Haven't really tried any with and eo-line start yet but that could be quite nice also

Checked, quite unique... saw those 4 corner were solved to 3, and then all alligned in last steps... neat and cool.

Here's the garron.us's show url i used to check it: http://tinyurl.com/algz-R2LFL-yUR

 

[chris w]

Checked, quite unique... saw those 4 corner were solved to 3, and then all alligned in last steps... neat and cool.

Here's the garron.us's show url i used to check it: http://tinyurl.com/algz-R2LFL-yUR

Thanks, yeh I only know 3 cycle corner comms so there's probably a better solution to the 2 2cycles at the end. Not really the best speedsolve method, but a good mostly intuitive method. I suck at eo, but with efficient eo and edge placement you might get so pretty good FMC solutions.

 

[Yuxuibbs]

I came up with solving edges first then using A perm and sexy to solve corners during the summer. What do you average with it?
I still don't really get commutators.

 

[chris w]

I came up with solving edges first then using A perm and sexy to solve corners during the summer. What do you average with it?
I still don't really get commutators.

If this is directed at me, just did and avg12 26.46, 30.73, 21.55, 28.19, 25.24, (38.59), (15.89), 36.84, 30.40, 31.69, 36.06, 27.98 =29.51
The 15 had only 5 corners to solve after edges.
Im pretty bad at commutators still, but how I learnt was like this. set up with R U R' D R U' R' D' . In the form of ABA'B' so: [A] D (solves corner at DFR) R U R' (solves other corner, (D' off)) [A'] D' (solves corner) [B'] R U' R' (fixes last 2 corners)

 

[eo]

L F' L2 F' D L' F L' U D2 B' F' R D2 L2 B D' L' F D2 L2 R2 F' L2 R2

x2   FD2LB			//EO
     LDL'DRU'R'U'R'URLU'L2UL
y    RUR'URU2R'U2		//All edges done (6 crosses)

y2x' LU'R'UL'U'RU		//A perm
xy   R'ULU'RUL'U'		//A perm
x    LU'R'UL'U'RU		//A perm
x2y  R'ULU'RUL'U'		//A perm
y    R'ULU'RUL'U'		//A perm (all corners arrive on position)

y'zx L'U'LU'L'U2LU2		//Sune
     L2USU2S'U			//U perm (green face done)
z    RUR'URU'R'URU2R'U2		//Double Sune
     F2U'M'U2MU'F2		//U perm (blue face done, all finished)

I tried playing with your scramble, it comes as more than doubled of your moves, pretty discouraging... Please show your kind not to laugh too much at this...

 

[Kirjava]

An optimisation is doing orientation and permutation for edges and corners at the same time. Maybe that was obvious though.

Solving all edges give a restriction that doesn't seem to produce nice results for speedsolving based methods.

There are currently none that are close to decent that use this approach.

 

[chris w]

x2   FD2LB			//EO
     LDL'DRU'R'U'R'URLU'L2UL
y    RUR'URU2R'U2		//All edges done (6 crosses)

y2x' LU'R'UL'U'RU		//A perm
xy   R'ULU'RUL'U'		//A perm
x    LU'R'UL'U'RU		//A perm
x2y  R'ULU'RUL'U'		//A perm
y    R'ULU'RUL'U'		//A perm (all corners arrive on position)

y'zx L'U'LU'L'U2LU2		//Sune
     L2USU2S'U			//U perm (green face done)
z    RUR'URU'R'URU2R'U2		//Double Sune
     F2U'M'U2MU'F2		//U perm (blue face done, all finished)

I tried playing with your scramble, it comes as more than doubled of your moves, pretty discouraging... Please show your kind not to laugh too much at this...

Nah it's okay, I liked the eo and edge placement, you're just loosing a lot of moves from all the A perms, maybe you try solve some corners as you go. I tried some solutions going off your eo, but couldnt find anything too good. Just keep trying more scrambles/solves and I'm sure you will find some short cuts or tricks that will reduce the overall movecount

 

[eo]

Just rethought the edges and felt I can optimize it a bit (of course it still clumsy, i'm in learning... Thanks a lot for all of your precious helps!)

x2  FD2LB           //EO
    DL              //Do a careless cross
    UR'U'R
    U2RUR'
    LUL'U2L'U'L     //Do a careless F2L

x2  RUR'URU2R'      //Sune
y2  L'U'LU'L'U2L    //Antisune
y   R2US'U2SUR2     //U perm (OLL/PLL for first layer)

x2  R'ULU'RUL'      //A perm
y   RUR'URU2R'      //Sune
    L'U'LU'L'U2L    //Antisune
y'  R2US'U2SUR2     //U perm (OLL/PLL for last layer)

    D2U'            //align the 3 layers.

where:
--- "careless cross" means cross in any order (can be shuffled, since EO is done, so any pulling together will still looks as a cross...)

--- "careless F2L" means just join any first layer corner with any middle layer edge and insert to the slot(according to the edge), no orientation cared (for the corner pieces... edges orientation are promised already)

And by the way, the two layers' OLL/PLL dealing looks can be switched in any order... Just realized that.

I really need to put down the cube for a while... even my eyes feel hurt now...

 

[eo]

after a short sleep, my eyes are feeling much better now, hopefully so does my mind...

now I see that the issue in my first post is obvious: making all 12 edges on position is a big mistake. We have some edge turns not affecting corners at all, but lack of the same freedom vice versa. Preserving all edges generally makes any further steps very unefficient.

I found later a better way: always place edges that won't affect the next steps.

so, please allow me to repropose the edge-first method again here:

1. EO
2. Careless Cross
3. Careless F2L
4. ZBLL on top layer
5. ZBLL on bottom layer

Where:
*) Careless Cross means: just send those 4 edges to bottom layer, no care about the specific alignment.
*) Careless F2L means: pare up any bottom corner with any middle layer edge, and insert to the slot for that edge.

Notes:
*) Need to peserve EO during the first 3 steps
*) The last 2 steps can be done in any order. and infact, the step 4 has even more freedom than strict LL. Any tricks not affecting middle layers should be fine.



Well, about my previous solve(2nd one), please do forgive my jokable-millions-looking-LL moves, it definately can be optimized with those amazing 1-look zbll algs... and for twice!

 

[eo]

Played it a while, looks those "careless" steps really convenient... However, I am so clumsy and lack of imaginations so can't make it happen creatively enough (like using D turns to send in a slot piece... just did it once...)

Besides, I noticed that sometimes the last "careless" slot insertion creates a result need parity moves between U/D layers. That's not ideal, but can be delt with a single "cared" insert.

 

[Cubenovice]

Hi eo from reading your two threads I think you might like HTA too

here's an example on Chris' scramble L F' L2 F' D L' F L' U D2 B' F' R D2 L2 B D' L' F D2 L2 R2 F' L2 R2

R' D2 U' L
D2 U' F' D' B D B'
R U' L2 U R
B2 D L2 D L2
D2 F2 D F2 D2 R2 D R2
B' F R2 F' B
D' U2 L' R B2 R' L
U' B D2 B2 D2 B2 D2 B U

Although in casual solving I would offcourse build on the nice blocks

 

[eo]

R' D2 U' L
D2 U' F' D' B D B'
R U' L2 U R
B2 D L2 D L2
D2 F2 D F2 D2 R2 D R2
B' F R2 F' B
D' U2 L' R B2 R' L
U' B D2 B2 D2 B2 D2 B U

Thanks a lot for your help! i'm checking this out... http://tinyurl.com/algR-D2U-LD2U-F
Looks very smooth... but may need some comments from you on the plans ...

 

[xabu1]

(need more analysis to confirm it, during my trial and tests, looks there always be a way to achieve it within 3 algs...)

how detailed are your algs? I am sure you can solve all corners with one alg if you wanted to learn that many

3 of how many algs would you be using?

 

[eo]

A perm / U perm / sune, (with mirrors) looks always can solve it...

anyway, this method(12 edge first) is approved rather like a "eye/plan exam", than in a solving purpose...