#### Gliphe

##### Member
Actually, I found an alg for Z Perm.
R2 U R2 U' R2 F2 R2 U' F2 U R2 F2

#### xyzzy

##### Member
Can you list the algs for G perms and Z Perms for me, please???
If you're going to learn algs specifically for FMC, learn the more useful ones like 3e or 2e2e algs (especially the ones that aren't commutators).

Anyway, the RUuy G perms that used to be popular (R2 u R' U R' U' R u' R2 y' R' U R) are 12 moves long. If you already know the RUD G perms, that's almost the same alg:
R2 U R' U R' U' R U' R2 U' D R' U R D' ↔ R2 u R' U R' U' R u' R2 y' R' U R

For Z perm, there are these easy algs:
- M2 U F2 M2 F2 M2 U' M2 and mirrors/inverses
- M2 U M2 U F2 M2 F2 M2 and mirrors/inverses (you can also replace F2 M2 F2 M2 with M2 F2 M2 F2)
- S M2 S' u M2 u' M2 and mirrors/inverses
Choosing the appropriate one will always let you cancel at least two moves, but it's almost always better to do 2e2e insertions instead. The second alg above solves the U2 angle (where the corners are a U2 away from solved); the other two preserve the corners.

(Note: I'm writing these algs with slice moves and lowercase letters for wide moves. This is not accepted for FMC. When writing up your FMC solution, you have to use the standard WCA notation.)

Last edited:

#### hiraken

##### Member
hello, this is my first post.

I tried FMC of 6x6x6 cube and 7x7x7 cube for the first time.
It took 20 hours for 7FMC and 6 hours for 6FMC, it was so tough for me.

Instead of writing down here, I've posted scrambles and solutions on my website, so please take a look if you like. (written in Japanese).

By the way, can I claim myself as a new world record holder of 6×6 FMC and 7×7 FMC?

Last edited:

#### xyzzy

##### Member
By the way, can I claim myself as a new world record holder of 6×6 FMC and 7×7 FMC?
Ah, sorry, I still have the 7FMC record of 313 moves (168 centres + 117 edge pairing + 28 3×3). Nobody ever updated the wiki for that and I'm not vain enough to do it myself.

That's still pretty cool though—not too many other madmen are willing to even try big cube FMC.

When it comes to OLL parity / edge parity, you should fix that during the centres rather than edge pairing; by using a different number of 2Xw/3Xw quarter turns to solve the centres, you can affect the parity of the edge pieces. Parity algs are long, so doing this parity avoidance is absolutely worth it.

There's no similar counting trick to handle PLL parity. What I do is just try different ways of edge pairing until I find something that doesn't lead to PLL parity. Especially easy on the last two edges if you get this kind of case, since one way of doing slice-flip-slice will give you PLL parity and the other one won't.

#### ProStar

##### Member
Ah, sorry, I still have the 7FMC record of 313 moves (168 centres + 117 edge pairing + 28 3×3). Nobody ever updated the wiki for that and I'm not vain enough to do it myself.

That's still pretty cool though—not too many other madmen are willing to even try big cube FMC.

When it comes to OLL parity / edge parity, you should fix that during the centres rather than edge pairing; by using a different number of 2Xw/3Xw quarter turns to solve the centres, you can affect the parity of the edge pieces. Parity algs are long, so doing this parity avoidance is absolutely worth it.

There's no similar counting trick to handle PLL parity. What I do is just try different ways of edge pairing until I find something that doesn't lead to PLL parity. Especially easy on the last two edges if you get this kind of case, since one way of doing slice-flip-slice will give you PLL parity and the other one won't.
For the record, I was insane enough to do 7x7 "FMC".

(it actually wasn't really FMC but just a 7x7 example solve)

#### hiraken

##### Member
Ah, sorry, I still have the 7FMC record of 313 moves (168 centres + 117 edge pairing + 28 3×3). Nobody ever updated the wiki for that and I'm not vain enough to do it myself.

That's still pretty cool though—not too many other madmen are willing to even try big cube FMC.

When it comes to OLL parity / edge parity, you should fix that during the centres rather than edge pairing; by using a different number of 2Xw/3Xw quarter turns to solve the centres, you can affect the parity of the edge pieces. Parity algs are long, so doing this parity avoidance is absolutely worth it.

There's no similar counting trick to handle PLL parity. What I do is just try different ways of edge pairing until I find something that doesn't lead to PLL parity. Especially easy on the last two edges if you get this kind of case, since one way of doing slice-flip-slice will give you PLL parity and the other one won't.
Oh...I'm sorry, I wasn't aware that you already had the 7FMC record before (where can I access?)

Thank you so much for your advice. I was too tired to think about avoiding parity cases.
I should try again for my new record... maybe... someday.

#### xyzzy

##### Member
Oh...I'm sorry, I wasn't aware that you already had the 7FMC record before (where can I access?)
Scramble:
3Lw F2 Bw 3Bw2 3Rw2 Lw 3Fw2 Rw2 3Fw' B2 3Bw' 3Dw2 B L2 F' Uw Lw Uw' Lw' 3Fw2 3Uw2 L' 3Lw' Bw Rw' 3Dw Fw' Rw' 3Rw' Bw 3Rw Bw2 D2 Rw Dw L2 3Uw2 3Bw' F' B R' 3Lw' Uw2 Lw Rw Bw2 F2 U' B' F 3Uw B' D2 Uw' U 3Rw2 Dw Bw' R2 3Fw' Fw' Bw' U' 3Dw 3Lw 3Bw2 3Lw 3Fw' L2 3Uw B' Lw' L B Dw' F Bw2 3Fw' B2 3Dw' D F' R' F2 Bw R2 Rw' 3Fw2 L Rw2 3Bw2 3Lw Dw2 Rw' U2 R' 3Fw2 Bw D' U'

Solution (313):
3Fw 3Bw' D2 U Bw 3Dw2 F Uw 3Uw'
Bw' Rw' Bw 3Uw' 3Dw L 3Dw
B2 Rw' 3Rw2 R' F Uw2 F Uw'
3Lw2 U Rw 3Rw2 U Lw2 R2 Fw'

D2 B Lw' 3Rw2
L2 3Bw L2 3Bw' Bw' L' Bw
U' D' 3Lw2 3Rw' L' 3Uw' L2 3Uw
D 3Rw2 F B 3Lw2 U2 Rw L' Bw L2 Bw'
3Rw' 3Lw' Rw2 Fw2 D Fw2 3Lw2

D Rw D2 3Lw D2 3Lw' L U
3Bw Bw2 L2 3Bw' U 3Bw2
Rw2 B2 3Lw2 3Fw2
3Rw' F2 3Rw Fw2
D' B2 Rw' F' Rw' B2 3Lw D2 3Lw

D 3Lw' F U 3Lw Lw' U2 Lw
Bw2 U2 Bw2 F' D 3Lw' F2 U' 3Lw
Rw2 D 3Lw2 3Rw2 D' Rw2 3Rw' B2 3Rw'
U' D2 3Lw2 U2 Rw2 U' Rw' F2 Rw'

3Fw2 U 3Fw2 U' 3Rw2 U2 3Rw2 U 3Rw2 U 3Rw2
U Lw2 U2 Lw2 U Rw2 U2 Rw2 U' 3Rw2 U2 3Rw2
Rw2 U 3Rw2 U' Rw2 U Rw2 U2 Rw2 U2

D L2 D' Rw2 3Lw2 U' R2 U
D L2 D' 3Rw2 B L2 B' Rw2 U' L' U 3Lw2
U2 Bw2 D B' D' Bw2
R' U' R U 3Rw Rw B U L2 B' U' Rw' F' R F 3Rw'
3Fw2 L2 3Fw U B2 U' Bw F' L' F L D B' D' Bw' 3Fw' L2 3Fw2
B U' 3Fw2 Bw2 R2 U' R2 U Bw2
R B' R' 3Fw D F D' R B' R' 3Fw
D' 3Rw2 D B' D' B 3Rw2
U2 3Rw' Rw U R2 U' 3Rw D' R D Rw'
3Bw2 Bw2 D R' B D' R 3Bw2 Bw2
Lw B' R' B Lw' Uw B U2 B' Uw'

U F U' B' U F' U' D' R D
R2 D' R2 U' R2 B2 R F2 R' B2
R F' L F' L U L' U
// red centre
3f 3b' D2 U Bw 3d2 F Uw 3u'
Bw' Rw' Bw 3u' 3d L 3d
B2 Rw' 3r2 R' F Uw2 F Uw'
3l2 U Rw 3r2 U Lw2 R2 Fw'

// orange+blue centres
D2 B Lw' 3r2
L2 3b L2 3b' Bw' L' Bw
U' D' 3l2 3r' L' 3u' L2 3u
D 3r2 F B 3l2 U2 Rw L' Bw L2 Bw'
3r' 3l' Rw2 Fw2 D Fw2 3l2

D Rw D2 3l D2 3l' L U
3b Bw2 L2 3b' U 3b2
Rw2 B2 3l2 3f2
3r' F2 3r Fw2
D' B2 Rw' F' Rw' B2 3l D2 3l

// green centre
D 3l' F U 3l Lw' U2 Lw
Bw2 U2 Bw2 F' D 3l' F2 U' 3l
Rw2 D 3l2 3r2 D' Rw2 3r' B2 3r'
U' D2 3l2 U2 r2 U' Rw' F2 Rw'

// white, yellow centres
3f2 U 3f2 U' 3r2 U2 3r2 U 3r2 U 3r2
U l2 U2 l2 U r2 U2 r2 U' 3r2 U2 3r2
r2 U 3r2 U' r2 U r2 U2 r2 U2

// edge pairing
D L2 D' r2 3l2 U' R2 U // 2 in 1 out
D L2 D' 3r2 B L2 B' r2 U' L' U 3l2 // +2/4 in +3/4 out
U2 b2 D B' D' b2 // +2/6 out
R' U' R U 3r r B U L2 B' U' r' F' R F 3r' // +4/8 in +3/9 out
3f2 L2 3f U B2 U' b F' L' F L D B' D' b' 3f' L2 3f2 // +4/12 in +3/12 out
B U' 3f2 b2 R2 U' R2 U b2 // +2/14 in +2/14 out
R B' R' 3f D F D' R B' R' 3f // +3/17 in
D' 3r2 D B' D' B 3r2 // +2/19 in
U2 3r' r U R2 U' 3r D' R D r' // +1/20 in +4/18 out
3b2 b2 D R' B D' R 3b2 b2 // +4/24 in
l B' R' B l' u B U2 B' u' // +6/24 out

// 333
U F U' B' U F' U' D' R D
R2 D' R2 U' R2 B2 R F2 R' B2
R F' L F' L U L' U

My working notation for the first four centres was SiGN for triple-layer moves (3u) and WCA for double-layer moves (Uw). This kept the visual width of the multi-layer moves about the same, and also makes reading the moves faster while reducing ambiguity ("w" always implies two layers, "3" always implies three layers).

I think I had move count annotations on my rough paper but didn't bother to type them out.

---

The 333 part:

B' D' R D // square (4/4)
R2 D' R2 U' R' // p223 (5/9)
(U' L U' L2) // adjust, EO, F2L-1 (4/13)
(L F L' F') // ab5c (4-1/16)

Skeleton: B' @ D' R D R2 D' R2 U' R' # F L F' L U L' U
@ = B U F U' B' U F' U' // 2 cancel
# = R' B2 R F2 R' B2 R F2 // 2 cancel (16-4/28)

I didn't bother looking for 3c insertions in centres or edge pairing.
I probably still have UWR singles for megaFMC (107) and 5FMC (126). Don't have enough patience to try out 6FMC—no fixed centres, so the orientation could get very confusing.

#### xyzzy

##### Member
Weekly comp 2021-01 stuff. I pre-committed to using only DR for the first solve (before seeing the scramble), so I didn't bother looking for blockbuilding starts or other non-DR EO starts.

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

R F D' L // EO (4/4)
(L2 U2 F) // align blocks
R2 D U' F D F // DR (9/13)

Then after this DR was this junk with bad corners and a whole load of unsolved edges. Not really sure what to do with this, so I just went to solve the corners.

R2 U2 B2 U R2 U' B2 U2 R2 // 2c2c2e2e alg; based on the optimal alg for the F (R U R' U')3 F' ZBLL
U' // align layers I guess (10/23)

(Now I realise that the normal PBL alg actually solves the corners in only 8 moves rather than 10.) Loads of unsolved edges left: a 2-cycle and a 6-cycle.

skel: R F D' L R2 D U' F D F R2 U2 B2 @ U R2 U' B2 U2 R2 U' F' U2 L2
@ = B2 F2 L R U2 D2 L R // four 2-cycles; to ab3e (8−2/29)

The last insertion is just a basic M' U2 M U2 kind of 3-cycle.
skel: R F D' L # R2 U' …
# = F' B U2 F B' R2 // finish (6−2/33)

Meh.

IF finds some truly ridiculous 34−29=+5 insertions with everything selected:
Skeleton
R F D' R2 L U' D F D F R2 U2 B2 U R2 U' B2 U2 R2 U' F' U2 L2

B' D2 R2 B2 U2 F' B' U2 R2 F' (10-7)
Skeleton
R F D' R2 L U' D F D B' D2 R2 B2 U2 F' B U R2 U' B2 U2 R2 U' F' U2 L2

F2 B2 D L2 D' F2 B2 U R2 U' (10-10)
Skeleton
R F D' R2 L U' D F D B' D2 R2 B2 U2 F B' D L2 D' F2 U2 R2 U' F' U2 L2

F B' R2 F' B U2 (6-7)
Skeleton
R F D' R2 L U' D F D B' D2 R2 F B R2 D L2 D' F2 U2 R2 U' F' U2 L2

B2 D2 R2 U2 F2 U2 R2 D2 (8-5)
Final Solution
R F D' R2 L U' D F D B D2 R2 U2 F2 U2 F B R2 D L2 D' F2 U2 R2 U' F' U2 L2

Then again, even when using only 3-cycles, it finds a 31-move solution (32−24=+8 from the insertions). So much for my weird 2e2e2e2e insertion.
Scramble: R' U' F R D2 B2 R U2 L2 U2 L' B' D B L2 D R2 F' D L2 U' F2 R' U' F

Most people (myself included) went for the one-move 222 block; only Cale and Remo Pihel didn't.

L2 // 222 (1/1)
D' B2 D' B' // EO (4/5)
D' R D2 U R' # D @ R U' // ab3c3c (8/13)
@ = B' R F' R' B R F R' // 3-cycle (8−2/19)
# = R2 F' L' F R2 F' L F // 3-cycle (8−1/26)

The L2 start doesn't seem to have nice blockbuilding continuations, so I just went for EO and it just came together. Insertions were worse than typical; IF finds a better pair of insertions, with the first insertion cancelling only one move and the second insertion (nested in first) cancelling three.

Also found this:
L' B2 L' // 222 (3/3)
(B D B' R' B' U' R D' R' U) // F2L (10/13)
(F' D' F L D2 L' D' L D' L' D2 F' D F) // ZBLL (14/27)
The extra B2 move in the start makes a nice pair. I knew this ZBLL case is 10 moves optimal, but it's literally the only ≤10-move ZBLL case I don't know. Absolutely cursed. With the optimal ZBLL:
L' B2 L' // 222 (3/3)
(B D B' R' B' U' R D' R' U) // F2L (10/13)
(F B' U' L2 U B2 D' B' D F') // ZBLL magic (10/23)

I looked up the optimal alg, memorised it, and almost had a chance to use it in scramble #3 (it ended up one move worse than my other solution).
Nothing notable about my attempt for #3 besides the thing I already mentioned in the above spoilerboxes.

Last edited:

#### indev85

##### Member
Hello, everyone.

I have always heard of the well-known FMC, where NISS, Insertions, etc. are used. However, what about LINEAR FMC? Basically, it’s solving the cube in the fewest moves possible, but not being able to use NISS, insertions, or undo moves.

In this case, does edge orientation still matter? If so/not, what are the best methods that could be used to get the best results?

#### f96

##### Member
hello, this is my first post.

I tried FMC of 6x6x6 cube and 7x7x7 cube for the first time.
It took 20 hours for 7FMC and 6 hours for 6FMC, it was so tough for me.

Instead of writing down here, I've posted scrambles and solutions on my website, so please take a look if you like. (written in Japanese).

By the way, can I claim myself as a new world record holder of 6×6 FMC and 7×7 FMC?
yes you can for 6x6 at least
thats uwr excluding the 1 hour limit
good job

I probably still have UWR singles for megaFMC (107)
james macdiarmid has 100

Last edited:

#### Hssandwich

##### Member
Hello, everyone.

I have always heard of the well-known FMC, where NISS, Insertions, etc. are used. However, what about LINEAR FMC? Basically, it’s solving the cube in the fewest moves possible, but not being able to use NISS, insertions, or undo moves.

In this case, does edge orientation still matter? If so/not, what are the best methods that could be used to get the best results?
This is an interesting question! And I'm glad you use the word linear correctly, which a lot of people don't

I think EO is probably the best strategy, although freedom is restricted, it means that block building often comes together more easily and it's easier to get out of sticky parity cases at the end of the solve.

#### indev85

##### Member
Hssandwich,

Great! Thanks for the suggestion!

The reason why I’ve asked about EO was because in Sebastiano Tronto’s FMC “How To”, he specifically averred that “the more bad edges there are, the harder things will be.” However, I didn’t know if that applied for Linear FMC as well.

In that case, should EO be done in the beginning, as in ZZ? Or should it be done in a later step, as in Petrus?

#### Delta Phi

##### Member
With the rise of PDR, EO is often done right away in a solve unless something else obvious presents itself. However, I dont think PDR is the definite best linear FMC method, and instead you should use a low movecount method like Petrus which orients later, but of course as with regular FMC you shouldnt limit yourself to one method and should try to do the best solution regardless of method.

#### Kit Clement

Hssandwich,

Great! Thanks for the suggestion!

The reason why I’ve asked about EO was because in Sebastiano Tronto’s FMC “How To”, he specifically averred that “the more bad edges there are, the harder things will be.” However, I didn’t know if that applied for Linear FMC as well.

In that case, should EO be done in the beginning, as in ZZ? Or should it be done in a later step, as in Petrus?
I think it depends on the situation. When doing a normal FMC attempt, you generally list out a long list of EOs and screen them for ones that give pairs or other nice starts. A strong majority of them usually don't lead to easy to find solutions right away. That said, having EO done right away is still a boon early on - if you make pairs after EO, they will have their edges oriented if you make them, making them far more useful than if you don't have EO done. I think the context of the scramble will guide you on which route to take - is there a nice, easy to use pair on the scramble? Do you have 4 bad edges from one angle, or are most angles giving bad EO cases (i.e. 10 bad edges)? That will guide if you should do it early or late.

Of course, with domino solutions being the new trend in FMC, EO first is an absolute must. From what I can tell as someone who is not well versed in DR, it seems that this is actually a great method for doing linear solutions given how methodical and prescriptive it can be with a great deal of efficiency. I don't know how likely it is that you can produce an EO that leaves nice cases for finishing DR (e.g. 2e4c or an insert) and how bad it would be if you picked an EO that produces a difficult to finish DR, and then went on to a DR with bad corners, so I don't know for sure how good this would be, but I imagine it would still be very effective overall compared to standard EO/block approaches.

#### indev85

##### Member
I think it depends on the situation. When doing a normal FMC attempt, you generally list out a long list of EOs and screen them for ones that give pairs or other nice starts. A strong majority of them usually don't lead to easy to find solutions right away. That said, having EO done right away is still a boon early on - if you make pairs after EO, they will have their edges oriented if you make them, making them far more useful than if you don't have EO done. I think the context of the scramble will guide you on which route to take - is there a nice, easy to use pair on the scramble? Do you have 4 bad edges from one angle, or are most angles giving bad EO cases (i.e. 10 bad edges)? That will guide if you should do it early or late.

Of course, with domino solutions being the new trend in FMC, EO first is an absolute must. From what I can tell as someone who is not well versed in DR, it seems that this is actually a great method for doing linear solutions given how methodical and prescriptive it can be with a great deal of efficiency. I don't know how likely it is that you can produce an EO that leaves nice cases for finishing DR (e.g. 2e4c or an insert) and how bad it would be if you picked an EO that produces a difficult to finish DR, and then went on to a DR with bad corners, so I don't know for sure how good this would be, but I imagine it would still be very effective overall compared to standard EO/block approaches.
Oh yes, I have heard of DR before. However, I have found that reducing and solving the cube after reduction seems to take more moves than blockbuilding, probably because DR would work better with insertions and inverse scramble techniques, in order to exploit the reduced state. Due to experimentation since the time I have made my initial post, I have resolved my issue about EO.

Are there any other things that could help with Linear FMC? Like how to build the 2x2x3 block directly, as opposed to a 2x2x2 + 1x2x2?

#### Cubing Forever

##### Member
I average around 48-50 now and NISS and insertions(I usually do "insertions" at the end of skeletons due to my lack of knowledge) seem confusing. Any tips to approach it?
I always seem to finish with DR due to using the Mehta method. Any tips for sub 40?

#### hiraken

##### Member
yes you can for 6x6 at least
thats uwr excluding the 1 hour limit
good job
Thank you so much!!
I'm so glad to hear that.

1 hour limit ... I wonder if I'll try it someday.

#### trangium

##### Member
Around a month ago I tried 4x4 FMC.
Since Domino Reduction is such a good FMC method for 3x3, I extended it into a reduction-based method for 4x4, reducing from <U, D, L, R, F, B, Uw, Dw, Lw, Rw, Fw, Bw> to <U, D, L, R, F, B, Uw2, Dw2, Lw2, Rw2, Fw2, Bw2> to <U, D, L, R, F, B> then proceeding with 3x3 stage. It worked really well!

Scramble: D2 L2 U' R2 B2 L2 D R2 U B2 F' D' R U2 L R2 U' R2 D Fw2 D2 L' F' Rw2 B' L2 D2 Rw2 D2 Fw2 L' Uw Fw2 D2 R D' F' D Fw' Rw2 Fw Rw R2 F2
Rw Uw2 R U' Fw' // HTR U/D centers (5)
F2 B' Uw' R2 L F Uw // HTR centers (12)
F2 Rw2 U' Rw2 // 2e (16)
Fw2 // +1e (17)
U Rw2 // +1e (19)
U L2 D Fw2 // +1e, solve L/R centers (23)
D' U' R2 F Rw2 // +2e, create center bars (28)
B2 D' F' B' L B Rw' // solve centers and +2e, cancel into slice flip slice (35)
F' U L' F U' Rw' // reduction (41)
z2 y'
U2 L // EO (43)
D B2 U2 F' L2 [1] B2 U2 F' D // DR (52)
B L2 R2 D2 F L2 B' [2] F' // 5E (60)
[1] = U2 F2 D2 F2 U2 B2 (6-5/61) to 3E
[2] = B2 L' R D2 L R' (6-1/66) solves 3E

My second attempt was a DNF. I got to reduction in 38 but couldn't find anything for the 3x3 stage.

Scramble: L D' B' R U' B2 R' U' F2 R2 U' R2 D' L2 D L2 U R2 D L' D' Fw2 L2 Uw2 U' L D Rw2 Fw2 F2 D Rw2 Fw2 D2 Fw' U' R' D2 F Rw' F' Uw' F2 Uw2 R' D
Uw L' F2 Uw' Fw' R' Fw' // HTR LR centers (7/7)
B Rw2 F' D2 Rw // HTR centers + 1 edge (5/12)
Uw2 // 2 edges (1/13)
Fw2 // 3 edges (1/14)
U Fw2 // 4 edges (2/16)
F' U Fw2 // 6 edges (3/19)
F U2 D2 F' Uw R D2 R' Uw // 8 edges, force 2e2e (9/28)
R F L D F' L' Fw2 // reduction (7/35)
y2
(B' D2 L U F') // EO (5/5)
(D R' U L U L U') // DR (7/12)
L' F2 L D2 L' D2 (*) L U2 L' B2 F2 R' // HTR 2e2e + e-slice (12/24)
(*) D2 B2 D2 F2 U2 F2 // solve 2e2e (4/28)
DR part: L' [1] F2 L D2 L' B2 D2 F2 U2 F2 L U2 L' [2] B2 F2 R' [3] // solve e-slice (6-5/29)
// [1]: M' [2]: M' [3]: M2
R' D2 L B2 L' U2 B2 D2 F2 D2 L F2 R' F2 B2 R L2 // rewrite DR part without slices

Final solution: Uw L' F2 Uw' Fw' R' Fw' B Rw2 F' D2 Rw Uw2 Fw2 U Fw2 F' U Fw2 F U2 D2 F' Uw R D2 R' Uw R F L D F' L' Fw2 y2 R' D2 L B2 L' U2 B2 D2 F2 D2 L F2 R' F2 B2 R L2 U L' U' L' U' R D' F U' L' D2 B (64)

Scramble: B' L B2 R2 U D L' B L B2 D R2 L2 D B2 U2 R2 F2 R2 D' Rw2 B R2 Uw2 D' Fw2 D L2 B2 Uw2 L2 D' F' Rw Fw2 Rw R' D Uw R2 Uw B R' U
R Uw' R2 Uw // LR center HTR (4/4)
Rw2 U Rw U2 D2 Rw // Center HTR, avoid OLL parity, 2 edges (6/10)
Uw2 // 4 edges (1/11)
B Uw2 // 5 edges (2/13)
R2 D' B Uw2 // 7 edges (4/17)
R D' F2 B2 U' Fw2 // 8 edges, pre-reduction (6/23)
F2 L R B' D B' Uw2 // reduction (7/30)
z2 x
B2 L2 D' B' // EO (4/34)
(U' L2 D' F2 U' F2 B2 U L') // DR (9/43)
(D' F2 B2 U2 L2 U2 L2 D') // 3E (8/51)
3x3x3 stage portion: B2 L2 D' B' D L2 U2 L2 [*] U2 B2 F2 D L U' B2 F2 U F2 D L2 U
[*] U L R' F2 L' R U (7-1/57) solves 3E

Final solution:
R Uw' R2 Uw Rw2 U Rw U2 D2 Rw Uw2 B Uw2 R2 D' B Uw2 R D' F2 B2 U' Fw2 F2 L R B' D B' Uw2 z2 x B2 L2 D' B' D L2 U2 L2 U L R' F2 L' R U' B2 F2 D L U' B2 F2 U F2 D L2 U

My fourth attempt, 57 OBTM, is my best 4x4 FMC result so far.
The previous UWR was 65, according to the wiki. If that's accurate, I broke it by 8 moves!

#### Cubing Forever

##### Member
A typical but lucky FMC attempt

L2 D2 B D2 // 2x2x2 Block
U' B2 D' B U B' // 2x2x3 Block
R D R' D2 // F2L '1
F' R F R2 F' R2 F //F2L
R2 B' R' B' D B D' R B //OLL
R' F2 R U D' F2 U' D R F2 R2 //3e

41 ETM

Ending F2L differently,

L2 D2 B D2 // 2x2x2 Block
U' B2 D' B U B' // 2x2x3 Block
R D R' D2 // F2L '1
D R D' R' F D' F' D //F2L
D R2 D2 F D F' D R2 D' //OLL
D2 r' D R' D R D' r D2 B R' B' // 3c3e

43 ETM