Noah's CP-Block Method

[Kirjava]

It might be possible to do 2x2x2 + CP inspection, if you limit your movegroup while building the block it would be possible to calculate what type of case you'd end up with. I think this would be stupidly difficult to accomplish in <15s however.

CPBLOCK IS THE NEW EOLINE

Edit;

CP+1x2x3
Second block is the same
CLL+EO or LSCLL = not many cases & good algs
PLSE

 

[Noahaha]

These might be useful:
(block in back)
URF <> URB: R U R'
URF <> ULF: R U' R'
ULF <> ULB R U2 R'
And obviously:
URB <> ULB: U R U R'
URB <> ULF/URF <> ULB: F R U R'

 

[mariano.aquino]

Here's an example
Scramble
L' D L B' D' L2 U B' F U2 D2 B' R2 B' F' U' B2 L U' B R' D2 B R2 F'

Inspection x’
ULU’ R’ D2 2x2 block (5)
R2 F2 B’ R B R’ bottom corners and a 2x2x3 edge (11)
U R U R’ U’ F’ U’ F CP (19)
U z’ RU’RURURU’R’U’R2 z (The 2x2x2 block while making the bad edge count 0 instead of 2)(31)
y’ U R U’ R’ U R2 U’ R’ U’ R’ U’ R(F2L)(43)
U2 (RU2R’) U’ (RUR’) U’ (RU’R’) U2 (2GLL)(56)

This it's how I would've done it:

x'y' U2 R2 x' (cp setup)
cp skip
u' R' S r' u' R' U R u2 (left block)
U' r2 U M' U r' (middle block)
U2 R U R2 U' R' (right 2x2x1 block)
U R' U R2 r' U2 R' U R U M' (F2L+EO)
L2 U' L U L U L' U2 L U L2 U L2 (2GLL)
47stm

Or, roux version...

x'y' U2 R2 x' (cp setup)
cp skip
u' R' S r' u' R' U R u2 (left block)
U' r' U' r' U' r' U' R2 U' R' U R U2 r M' U2 r' U' R (right block)
l' U2 L U L' U l (2GCMLL)
M U' M U M U2 M U2 (LSE)
45stm

what do you think?

 

[A Leman]

This it's how I would've done it:

x'y' U2 R2 x' (cp setup)
cp skip
u' R' S r' u' R' U R u2 (left block)
U' r2 U M' U r' (middle block)
U2 R U R2 U' R' (right 2x2x1 block)
U R' U R2 r' U2 R' U R U M' (F2L+EO)
L2 U' L U L U L' U2 L U L2 U L2 (2GLL)
47stm

Or, roux version...

x'y' U2 R2 x' (cp setup)
cp skip
u' R' S r' u' R' U R u2 (left block)
U' r' U' r' U' r' U' R2 U' R' U R U2 r M' U2 r' U' R (right block)
l' U2 L U L' U l (2GCMLL)
M U' M U M U2 M U2 (LSE)
45stm

what do you think?

I was just giving an example of using a cycle to combine the 2x2x3 block and the EO Which I made steps toward in the solve. That solve was just done on the fly to somewhat fit within noahaha's steps,show my EO skip idea, and i reconstructeded it.Your method looks more like a CF method for edges to keep the cp. I still have a lot of doubts about cp first methods because they place a lot of restrictions on the solve and have recognition problems.As for Roux idea, I personally like the freedom roux offers and would freeze up on a restricted roux solve. I feel cp should be forced later on in a solve like kirjava to make better use of inspection.

 

[Noahaha]

With enough practice one could plan a CP block with one edge placed next to the oriented corner and be able to seamlessly transition into the 3 moves necessary for CP. Then just apply an algorithm to place the last edge combine with EO, and go 2 gen the rest of the way.

 

[mariano.aquino]

As for Roux idea, I personally like the freedom roux offers and would freeze up on a restricted roux solve. I feel cp should be forced later on in a solve like kirjava to make better use of inspection.

After doing cp, I find no restrictions at all, besides being restricted to just 7 kcmlls. What are the restrictions you see in doing this cp+left block?

 

[Noahaha]

How are people doing the Roux CP-block? Is CP planned in advance, or are you making the block and then doing R U R'? One idea I had, which is probably more of a Petrus/CFOP approach is making the back square of the second block and then doing F' U/U'/U2 F for CP (or the front square and B U/U'/U2 for CP), then filling the last slot. That way you can recognize it the same way as CPLS.

 

[5BLD]

With Roux it's not much of a benefit imo. The no swap cases are some of my worst CMLL cases.

I used to just use CPLS algs, but I only knew Im (which defeats the purpose) at the time. I realised it'd be pointless to learn CPLS because of the above reason. I think the good think about your method lies in the 2GLL finish which is why we do CP.

 

[A Leman]

After doing cp, I find no restrictions at all, besides being restricted to just 7 kcmlls. What are the restrictions you see in doing this cp+left block?

your solves were good. I just feel that since you had to setup and presrve cp you missed an easy Roux solve like below. It took me most of inspection to plan the first block, but it was worth it.(more than planing an 11 move cp left block).

What important benifit do you get from using cp control in Roux besides less CMLL?

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

Roux Solve
Inspection y2 z
D’F’B2D2 (4) Block 1
M U’M’ U2 R2 U’ R U2 R2 U R’ (15) Block 2
U2 r2 D’ RU2R’ D RU2R (25) CMLL
UM’UMU’M2U’MU2M’U (36) LSE

 

[mariano.aquino]

What important benifit do you get from using cp control in Roux besides less CMLL?

It could be a way of adding some more complex cases, such as edge orienting cmlls, or other stuff I may be missing right now =)
Still, I don't think it will be something worth learning as a main method, but maybe for cases with easy recognition, same way you would use some COLL, sometimes, in a Fridrich solve.
Maybe a preEO followed by cmll+EP algorithm? 4b, cmll+4a+c?

 

[AHornbaker]

eh, whatever.

I just really want someone to make a good/simple system for detecting CP after 2x2x3.

I came up with a quick system about a week ago, here goes.
1.) After first block, place the two remaining D corners in ULB and DRB. (2 moves max)
2.) Look at URF. Find the other two U corners with the same colors. They will make an "L" shape on either the U, R, or F face. Then, determine whether those two corners are correct or should be swapped. (I'll make some pics to illustrate this)
pics: http://www.quickmeme.com/meme/3vhcw6/
3.) Find case, apply your favorite alg. (First letter stands for face, second letter stands for U corners correct or swapped)

D Corners Correct CP (3 HTM)
Fc: (f R' f') or (d R' U' R d')
Fs: Correct CP
Rc: (f R f') or (d R' U R d')
Rs: (R F R F') or (R2 d R' U' R d')
Uc: (U' b' R b) or (d' L' U L d)
Us: (U f R f') or (d R' U2 R d')

D Corners Swapped CP (3 HTM)
Fc: Correct CP
Fs: (f R' f') or (d R' U' R d')
Rc: (R F R F') or (R2 d R' U' R d')
Rs: (f R f') or (d R' U R d')
Uc: (U f R f') or (d R' U2 R d')
Us: (U' b' R b) or (d' L' U L d)

*Speed-optimized algs. Hooray. <5 moves on average.
**Algs for D correct and D swapped are same, just flipped for each face. Only 5 algs total. I can go into more detail how the algs work and what they swap.
***DISCLAIMER: Don't get butthurt because the algs don't preserve EO. Working on that. Read below.

Doesn't placing D corners just to detect CP seem wasteful to you?
Why not do 2x2x3 -> EOCP -> F2L -> 2GLL?

I'm working real hard on EOCP and figuring out what makes things tick. Ideally, it would be one-look, and you could solve EO and CP in one step. That's pretty infeasible. Using my CP recognition method, you can determine 1 of 6 cases, then there are about 13 distinguishable EO cases (same with one face turn) for 7 edges as follows:

0 Edges Misoriented (1 case)
2 Edges Misoriented (21 unique cases)
---0 on U
------L shape on F
------Line on F
---1 on U
---2 on U
------L shape
------Line
4 Edges Misoriented (35 unique cases)
---1 on U
---2 on U
------L shape
------Line
---3 on U
---4 on U
6 Edges Misoriented (7 unique cases)
---3 on U
---4 on U

This might be a little off, but it's a start. Before I dive in head-first in developing a method, I need to find out the best scheme.

My first idea was to find ~5 algs that solve each CP case (swapping two specific corners) and flip two edges. These algs are like what Noahaha described earlier in the thread (ie. R U R', R U' R', L' U2 L, etc.) Let me introduce a concept. While solving corner PLL, you can use a 3 corner cycle to either solve CP, or keep the CP wrong, then use the 3 corner cycle again to solve corner PLL. Using my CP recognition and using these 5 algs intuitively, EOCP could be solved in this way:
0C/0E - solved EO/CP
2C/0E - Alg to flip 2 edges and keep CP swapped, turns into 2C/2E case
2C/2E - Place edges and 1 of 5 algs to solve CP and EO
2C/4E - Alg to flip two edges while keeping CP swapped, 2C/2E case
2C/6E - Alg to flip two edges and either solve or keep CP swapped, either 2C/4E or 0C/4E case
0C/2E - alg to keep 2 edges flipped (flip one correct and one wrong) and swap CP, 2C/2E case
0C/4E - Alg to flip 2 edges and swap CP, 2C/2E case
0C/6E - Alg to flip 2 edges and swap CP, 2C/4C case

This method would be alright, the only problem is keeping track of the CP case after initial recog. Then I had another idea. Place the edges and the corners at the same time. More setup moves, less cases. Ideally, 6 cases for CP and only 4 cases for EO (0,2,4, and 6 edges wrong, placed in specific spots) but this might be unfeasible due to high setup moves. I thought about it more and this would reduce setup to only R and U, making your solve less efficient. Silly, I know. The better idea is here somewhere... (hint- keep reading!) After the block, don't limit your moves to <U,R>, but use all sides to orient edges while you place the corners for CP recog. Much more efficient. The only downfall to this is it would be two looks (one for D corners/EO, then for CP). Let me know which you think is better. Can't wait until I have a good EO/CP system.

 

[MaikeruKonare]

I love the F2L with only U and R, it seems very speed efficient.