[Help Thread] FMC Discussion

[G2013]

With scramble B2 U R2 F2 R2 D' B2 F2 U R2 D2 F D2 F2 U2 R' B' L' B D2 F
Pre: R2
F' R //2x2x2
L B2 D2 //2x2x3
D L D' //F2L-1 (I noticed that my F2L-1 was the same that EMI posted, but that is just a coincidence. Many of you may find the same solution for F2L-1, because it is too easy)
B2 L' B L B2 U B U' B //Edges
R2 //Correction

IF:

F' R L B2 D' [@1] L D' B2 L' B L B2 U B U' B R2
Insert at @1: D' R D' R D' L D2 R2 U' B2 U D L'
After the 1st insertion: F' R L B2 D2 R D' R D' L [@2] D2 R2 U' B2 U B2 L' B L B2 U B U' B R2
Insert at @2: L' D R' D' R L U' R' D R U D'
Fewest moves: 29. 13 moves cancelled
The final solution: F' R L B2 D2 R D2 R L U' R' D R U D R2 U' B2 U B2 L' B L B2 U B U' B R2

 

[Torch]

Scramble: F' U' R2 F2 U' F2 D R U' F B2 L2 D B2 U2 B2 R2 U' D2 R2 B2

B R D' B D' B'//2x2x2
F' U2 L F U2 L U//3x3x3
F L2 U' F U//F2L-1
F' D F' R F R' F' D' L' F2 L//L3C

Skeleton: B R D' B D' B' F' U2 L F U2 L U F L2 U' F U F' D F' [1] R F R' F' D' L' F2 L (29)

[1]F' D2 F U F' D2 F U'

Solution:B R D' B D' B' F' U2 L F U2 L U F L2 U' F U F' D F2 D2 F U F' D2 F U' R F R' F' D' L' F2 L (36)

IF says 36 is optimal for that skeleton

 

[AJ Blair]

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

Solution:

R2 F' U' F'//1x2x2 Block (4/4)
U' F D2 U L'* U' L U' //2x2x3 (8/12)
F D F2 D L' F2 L F' //3 Corners (8/20)

Insert at *: U2 R2 U L2 U' R2 U L2 U

3 move cancelation

Solution: R2 F' U' F' U' F D2 U L' U2 R2 U L2 U' R2 U L' U' F D F2 D L' F2 L F'

26 moves

That makes my last three attempts: 28, 26, 26 or 26.67 Mo3 which is most certainly a PB and one that I am quite proud of.

 

[ottozing]

Weird FMC I did in a Skype call with yearofthewalker and anonymouse. Optimal insertions would have given me a 29 but I had less than 10 mins to do insertions :/

L2 F2 U2 L2 B' U2 R2 B' U2 L2 F2 L' B' F R F2 L2 R' U B R

Inv
i. F' L' (Square)
Switch
n. U L U (2x2x2)
Switch
i. R2 F' R' F (2x2x3)
Switch
n. U' B U (EO)
B R B R2 (F2L-1)
B R' B' R B' R' B R (AB4C)

Skeleton: U L B U "B R B R2 B R' B' R B' R' B" R F' R F R2 L F (22)

22 to 4 corners = lame
hmm ima solve the 2gen spam minus corners and see what happens to corners

Scramble: B R B R2 B R' B' R B' R' B

B' R' B2 R2 B R B' R' (stuff)
then I inverted it (R B R' B' R2 B2 R B) and put it into the skeleton

Skeleton: U L B U * R B R' B' R2 B2 R B R F' R F R2 L F (19 to 5 corners)

* U F D' F' U' F D F'

New skeleton: U L B U2 F D' ^ F' U' F D F' R B R' B' R2 B2 R B R F' R F R2 L F (26)

R2 F' L2 F R2 F' L2 F (Cancels 2)

Solution: U L B U2 F D' R2 F' L2 F R2 F' L2 U' F D F' R B R' B' R2 B2 R B R F' R F R2 L F (32)

 

[Prakhar]

Can anyone please explain what is meant by domino reduction. Is is better than leaving 3-5 corners and inserting them earlier?

 

[porkynator]

Can anyone please explain what is meant by domino reduction. Is is better than leaving 3-5 corners and inserting them earlier?

I have just answered you on facebook lol...
Copy-paste:

It is a completely different approach; you reduce the cube to be solvable with only <U, D, R2, L2, F2, B2> moves (like a 3x3x2 or "Domino" cube). To do so you need to:
-Orient edges (like in Petrus or ZZ)
-Orient corners (with the U/D color belonging to U/D)
-Place the E layer edges in the E layer (not necessarily in their exact place, just in that layer)
After that you usually look for a "brute force" solution
Example solves: http://fmcsolves.cubing.net/?tag=domino

 

[guysensei1]

Is it advisable to look for insertions by splitting up half turns?
For example, when there is a U2, do you split it into U U or U' U' to find insertions?

 

[porkynator]

Is it advisable to look for insertions by splitting up half turns?
For example, when there is a U2, do you split it into U U or U' U' to find insertions?

No, you would get the same cancellations withou splitting it up.
At least it never happened to me and I can't think of any case where it would be better.

 

[bobthegiraffemonkey]

scramble: L2 R2 B' D2 F L2 R2 B' R2 B F2 U' B2 L B2 D2 L F' L' U'

U L2 F D2 R F D2 F2
D' R D
on inverse
F2 B2 R'

I tried continuing after bolded part and found almost straight away a 22 move skeleton up to L3C:

U L2 F D2 R F D2 F2 @
D' R D
R D B2 D' F2
U B R B' R' U'

Insertion finder says optimal is 28 with a 10 move 3-cycle and 4 moves cancelled: @=F2 D F2 R2 U F2 U' F2 R2 D'. Maybe I'm being slow, but that inserted alg looks weird, can someone explain it to me?

 

[tseitsei]

No, you would get the same cancellations withou splitting it up.
At least it never happened to me and I can't think of any case where it would be better.

Well I'm not good at FMC or anything but at least in theory that's not true I think.

Example:

Starting scenario: Your last 3 corners to solve are in the cycle UBL->RDB->FRD->UBL in some point of your skeleton.
And now let's assume that the next move of your skeleton is R2.

Now if you do R2 and look for insertions you can see that there are now simple 8-move insertions and move on to the next move of the skeleton. Am I right here?

HOWEVER if you do R instead you can easily see that there is an 8-move insertion now starting with R' so that means 2-moves cancel but 1 extra move for the remaining R part of the R2 move at the end of your insertion needs to be added. So overall splitting this R2 to be R R and inserting in between gives a 7-move insertion effectively.

Of course the insertion could be done by doing R2 and then inserting a A9 type commutator after that(R2 R2 cancels out 2-moves so also 7-moves insertion) but (at least me) many people only check for 8-move insertions I think...

 

[porkynator]

Well I'm not good at FMC or anything but at least in theory that's not true I think.

Example:

Starting scenario: Your last 3 corners to solve are in the cycle UBL->RDB->FRD->UBL in some point of your skeleton.
And now let's assume that the next move of your skeleton is R2.

Now if you do R2 and look for insertions you can see that there are now simple 8-move insertions and move on to the next move of the skeleton. Am I right here?

HOWEVER if you do R instead you can easily see that there is an 8-move insertion now starting with R' so that means 2-moves cancel but 1 extra move for the remaining R part of the R2 move at the end of your insertion needs to be added. So overall splitting this R2 to be R R and inserting in between gives a 7-move insertion effectively.

Of course the insertion could be done by doing R2 and then inserting a A9 type commutator after that(R2 R2 cancels out 2-moves so also 7-moves insertion) but (at least me) many people only check for 8-move insertions I think...

You can solve that cycle with [D', R U2 R'], 8-1=7 moves, the same.

 

[tseitsei]

You can solve that cycle with [D', R U2 R'], 8-1=7 moves, the same.

Well that's true. And I'm an idiot

 

[tseitsei]

Ongoing competition from here: http://www.ocf.berkeley.edu/~dadams/fmc/

Nice solve for me the last insertion probably isn't optimal because I had to hurry up
Criticism and tips are most welcome

Spoiler

scramble: R2 U' D2 L R2 B' D2 F2 L2 R' F' B' D' L F U' L' F2 D' B'

solution: B2 R2 L F R2 U2 B R' B' U L' B R2 B' L B R2 U2 B R B2 R' B' D L' D' R D L D' R2 U

explanation:
B2 R2 L F //2x2 4/4
R2 U2 //square 2/6
B R' //pseudo2x2x3 2/8
B' U B U2 //pseudo F2L-1 4/12
Now I use premoves R' U to correct pseudoness
B R B2 R' B' //solve edges. L5C 5/17
R' U //undo premoves 2/19

skeleton:B2 R2 L F R2 U2 B R' B' U (1) B U2 B R B2 R' B' (2) R' U 19-moves

(1)=[L': B R2 B'] 2-moves cancel
(2)=[D L' D' : R] 1-move cancels

19+16-3moves =32moves

 

[ch_ts]

I tried continuing after bolded part and found almost straight away a 22 move skeleton up to L3C:

U L2 F D2 R F D2 F2 @
D' R D
R D B2 D' F2
U B R B' R' U'

Insertion finder says optimal is 28 with a 10 move 3-cycle and 4 moves cancelled: @=F2 D F2 R2 U F2 U' F2 R2 D'. Maybe I'm being slow, but that inserted alg looks weird, can someone explain it to me?

It looks like an A-perm shifted. I shifted that alg around a little until it came to this one: R2 U F2 U' F2 R2 D' F2 D F2 where the corners are all in the U layer. It looks like it might be the 10th one down in the wiki. (I suspected it might be there because of their orientation.) I'm not exactly sure how IF works, it might just have a list of algs that it runs through and tries them one by one. If that's the case, the programmer must have filled its database with a bunch of commutators, other short algs, their shifts, etc

(Aside: does IF work for 2 corners and 2 edges? I couldn't figure how to do this.)

 

[bobthegiraffemonkey]

It looks like an A-perm shifted. I shifted that alg around a little until it came to this one: R2 U F2 U' F2 R2 D' F2 D F2 where the corners are all in the U layer. It looks like it might be the 10th one down in the wiki. (I suspected it might be there because of their orientation.) I'm not exactly sure how IF works, it might just have a list of algs that it runs through and tries them one by one. If that's the case, the programmer must have filled its database with a bunch of commutators, other short algs, their shifts, etc

(Aside: does IF work for 2 corners and 2 edges? I couldn't figure how to do this.)

[U R2 U' R2, f2 y2 U'], slightly awkward comm format (with rotations) but not really helpful, basically it cycles 3 edges by (U F2 U' F2), does R2, then solves the edges differently (D' F2 D F2), then undoes the R2. Weird.

IF doesn't seem to have an option for 2C2E, don't know why it wouldn't though.

 

[AJ Blair]

Okay...wtf

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

Premove: U

B' U2 L' // 1x2x3 (3/3)
R' U' B* U2//2x2x3 (4/7)
D F' //F2L-1 (2/9)
D2 F' D' F D //3 Edges (5/14)
U //Undo Premove (1/15)

Insert at *: B2 R L' D2 L R'

Cancels 1 move

Final Solution: B' U2 L' R' U' B' R L' D2 L R' U2 D F' D2 F' D' F D U

20 moves.

That was ridiculous...

 

[ottozing]

More FMC on Skype. This time with Kit Cement and Mr. 69. Got a 12 move skeleton to 7 corners which was kinda wtf

R2 U' L2 D L2 R2 B2 L2 R2 D R2 B U' L B2 F' D U' F' D2

Spoiler: gj

R2 U D B' (EO)
L2 D' U2 R' F2 U2 L U2 (AB7C)

Skeleton: R2 U D B' L2 D' U2 R' F2 U2 & L U2 (12)

& F R F' L F R' F' L' (Cancels 2)

New Skeleton 1: R2 * U D B' L2 D' U2 R' F2 U2 F R F' L F R' F' U2 (18)

* R' U R D2 R' U' R D2 (Cancels 2)

New Skeleton 2: R U R ^ D2 R' U' R D' U B' L2 D' U2 R' F2 U2 F R F' L F R' F' U2 (24)

^ R' B R F' R' B' R F (Cancels 2)

Solution: R U B R F' R' B' R F D2 R' U' R D' U B' L2 D' U2 R' F2 U2 F R F' L F R' F' U2 (30)

Optimal insertions would have given me a 25 but I was running out of time (Still wouldn't have gotten the optimal insertions with more time because that involved using 10 move comms).

Here's the first thing I did on the scramble which I kept for lols

Spoiler: bj

L2 D U F (EO I think)
R U R U' R' F2 (Not quite a 2x2x3)
Switch
B2 U' L U' L' (Not quite an F2L-1)
R B R' B2 R B2 R' (Getting somewhere maybe)
B' D2 U2 F2 D2 U2 (All but 2 twisted corners)

Skeleton: L2 D U F R U R U' R' "F2 U2 D2 F2 U2 D2" B R B2 R' B2 R B' R' L U L' U B2 (28)

"F2 U2 D2 F2 U2 D2"
U2 D2 B2 U2 D2 B2 does the same thing except cancels into the skeleton

Skeleton: L2 D U F R U R U' R' U2 D2 B2 U2 D2 B' R B2 R' B2 R B' R' L U L' U B2 (27)

 

[Kit Clement]

From the new Weekly FMC. I don't force L3C often, and this one took me a lot of moves to do. Great cancellations saved the attempt in the end.

Spoiler: Final Solution

R' U' L' U R' U2 L U L' U' L U' F U2 L' B L B' U2 F' L2 U' L2 B' L2 B D L D' L' B' D2 R2 B2 (34)

Spoiler: Explanation

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

(on inverse)
PM: R2
B2 R2 D2 //2x2x2 (4/4)
B L D L' D' B' L2 B //2x2x3+blocks (8/12)
L2 U L2 F U2 B L' B' L U2 F' //F2L+blocks (11/23)
U L' U L U' L' U L U //L3C (9/32)

Skel (on norm): R2 U' L' (@1) U' L U L' U' L U' F U2 L' B L B' U2 F' L2 U' L2 B' L2 B D L D' L' B' D2 R2 B2

@1: L U R U' L' U R' U' (6 moves cancelled)

 

[tseitsei]

My solution for the new weekly comp. Porkynators guide really motivated me to practise and actually learn how NISS work.
So here is my first sub-30 (27-moves)

Spoiler

scramble: L' B U2 F' R D' F' R' U2 D' B U' R2 D2 B R' L2 B' R2 L2
solution: D2 F' L2 F R' D' L D R D' U' B2 U F U' B2 U D F D' F' L2 D U2 L2 F B2 (27)

On inverse:
B2 F' L2 U2 //2x2

Switch to normal (premoves: U2 L2 F B2)
D2 F' L2 //pseudo 2x2x3

Fix pseudoness with premove D (premoves D U2 L2 F B2)
F D' L //pseudo F2L-1 + pair

Again we fix pseudoness with premove L2 (so premoves L2 D U2 L2 F B2)
F D F D' F' //solve edges (and 1 corner) leave 3-cycle+1twistedc

skeleton: D2 F' L2 F D' [1] L F D F D' F' L2 D U2 L2 F B2 (17)

[1]=D R' D' L D R D' [2] L'
[2]=U' B2 U F U' B2 U F'

solution: D2 F' L2 F (D' D) R' D' L D R D' U' B2 U F U' B2 U (F' L' L F) D F D' F' L2 D U2 L2 F B2 33moves

6 moves cancel so 27 moves
Second time ever I used NISS and first time ever I inserted commutator inside another commutator

 

[TDM]

Got bored, did FMC. Scramble from qqTimer; not from a weekly comp or anything.

Scramble: L' F2 U' B2 D F' B' U2 R' L2 U' F2 U D2 L2 B2 D R2 U F2
Solution: U2 F D' L R2 U' L2 U F2 U' B U F2 U' R B F D U2 F' D F U2 F' D' F' R' F D' F2 D' B2 D F2 D' B2 D' (38 moves)
alg.cubing.net

U2 F D' L U' L2 // 222 (6/6)
L2 U R2 U' L2 * B R B // 223 (8/14)
F D @ F'
F' R' F // EO (6/20)
D # // AB6C (1/21) (yeah, this was luck)

Insert at *: [U F2 U', B]
Insert at @: [U2, F' D F]
Ran out of time and couldn't do the last three corners by an insertion, so insert at #: [D2: [F2, D' B2 D]], which is just a ZBLL.

Final solution: U2 F D' L U' L2 L2 U R2 U' L2 U F2 U' B U F2 U' B' B R B F D U2 F' D F U2 F' D' F F' F' R' F D D2 F2 D' B2 D F2 D' B2 D D2
= U2 F D' L R2 U' L2 U F2 U' B U F2 U' R B F D U2 F' D F U2 F' D' F' R' F D' F2 D' B2 D F2 D' B2 D'

After just inserting the first commutator, it's AB5C in 22 moves. This was quite lucky...

That's my fifth FMC solve, so I finally have an Ao5 without a DNS: 48, 48, 48, 35, 38 = 44.67.