[Help Thread] FMC Discussion

[xyzzy]

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

X2 Y2
RW' R' U RW' D' U2 R' U' B U B' (11 Moves f2l-1)

I don't get how to get a good sleketon after this.

Y'
R U' R' F' U F R U' R2 F R2 U' R' U' R U R' F' (18) (I somehow found this to 2 twisted corners)
Y2 F2 U' R F2 D' F2 D R2 U' R U F2 (11)

40 total. But I used Cube explore to get the twisted corner case.

Protip: don't use rotations or wide moves. Don't think of the letters used in the notation as being short for "up"/"front"/"right"/etc., but as being short for "white"/"green"/"red"/etc. (Assuming you use WCA orientation, you get U=white and F=green for free, along with R=red and B=blue having obvious mnemonics, which leaves L=orange and D=yellow you have to explicitly memorise.)

So let's rewrite your solution first, and then work on that.

R' L' B R' D' U2 L' U' F U F' // F2L-1 (11)
B U' B' R' U R B U' B2 R B2 U' B' U' B U B' R' // ab2c (18/29)
L2 U' F L2 D' L2 D F2 U' F U L2 // corner twists (12/41)

>B U' B' R' U R B U' B2 R B2 U' B' U' B U B' R' // ab2c (18/29)
You can use B U' B' to form the pair, then insert with a sledgehammer for edge control as well. For this ZBLL case (J perm + two twisted corners), you can use either a Sune (L U L' U L U2 L') or an Antisune (B U2 B' U' B U' B') to get you to L3C. (Fun fact: this ZBLL case came up in Feliks's 38.52 555 WR, which he solved with Antisune + L3C.) This overall isn't very good, especially if you don't do insertions (7 for F2L + 8 to set up to L3C + 9 for L3C = 24 moves), but it's probably the first option I'd try.

Backing up a bit:
R' L' B R' D' // 223 - corner (5)

From here the obvious thing is to do what you did, but that led to a bad continuation so let's try something else.
L B L2 B' // F2L-2 (4/9)

It's tempting to just insert the last pair (since it's already formed!), so let's see where that leads us. We can do a normal insert (L U2 L'), AUF then normal insert (U' L U' L'), or AUF then sledgehammer (U2 L' B L B'). Let's evaluate all of them.

L U2 L' / B' U' R' U R B U' // leaves 2+2+1 corners (10/19)
U' L U' L' / F' L' U' L U' F U F' U F U // leaves 3+1 corners (15/24) (note: this ten-move alg is, I believe, a common OLL alg)
U2 L' B L B' / L U L' U L U2 L' U' // leaves 3+1 corners (13/22)

First one is shortest, but leaves five corners in a bad state (you need three 3-cycle insertions), whereas the second and third options leave the same "type" of corners left (needs two insertions), but the third is shorter and so is more likely to be useful. But what else can we try?

B' U2 B // EO (3-1/11)
L U' L' U L U' L' U2 // F2L-1, edges; leaves 2+2+1 corners (8/19)

That turned out to be useless. What about going to the inverse scramble? I don't see anything nice with standard F2L solving, so let's try EO then edges.

R' L' B R' D' // 223 - corner (5)
L B L2 B' // F2L-2 (4/9)
(L U2 L' U' B' U B U2 B' U' B) // leaves 2+2 corners (11-1/19)

If you don't feel like doing insertions, you can also solve these last four corners with a triple sexy setup ([B2: (B' L' B L)3]), and that cancels a few moves.

R' L' B R' D' // 223 - corner (5)
L B L2 B' // F2L-2 (4/9)
(L U2 L' U' B' U B U2 B' U' B) // leaves 2+2 corners (11/20)
(B L' B L B' L' B L B' L' B L B2) // L4C (13-2/31)

What algs should a FMCer want to learn? If any?

In no particular order, with a focus on last layer:

Spoiler

• Optimal algs for all last layer corner 3-cycle cases. The 9- and 10-move cases (e.g. A perm) are very rarely useful for FMC since you can usually use an insertion and do a 3-cycle in 8 moves, but if you're very short on time and you're slow at insertions, this can save you from getting a DNF. The 8-move cases are all just commutators, so you don't really need to learn them as algs per se, although you should know which ones have multiple 8-move solutions (e.g. R B R' F R B' R' F' = F' L' B L F L' B' L).
  
• Sune for edge permutation. (Also keep in mind that Sune preserves a pair.)
• "Super Sune" (R2 U R2 U R2 U2 R2), also for edge permutation, when one of the F2L slots has its edge solved and corner not solved. Preserves two pairs and corner orientation.
  
• R' U2 R' D' r U2 r' D R2. This preserves a 2×2 square while swapping the other two edges. (This is just the standard headlights OLL alg but with some wide moves.)
• Bruno (R U2 R2 U' R2 U' R2 U2 R) and optimal U perm (F2 U M' U2 M U F2) can also be used for edge permutation, but are generally less likely to be useful than the above options. (In particular, U perm can sometimes be replaced with a shorter edge insertion.)
• L' R U R' U' L U2 R U2 R' (and inverses etc.) when there's an oriented corner and you want to swap the two edges around it while leaving the other two edges alone. There's also an RUF version of this alg, which has a different AUF: F U R' U' R F' U' R' U2 R. (Very rarely useful because of its length.)
• r U' r U2 R' F R U2 r2 F. This is optimal for swapping opposite edges with no AUF. (With AUF, optimal T perm is better. Also note that this is related to an optimal J perm alg: r2 F r U' r U2 R' F R U2. You can get a whole bunch of optimal 2c2e algs just by cycling this one.)
  
• F R U R' U' F' and mirrors/inverses for edge orientation and permutation. Avoid using these if you're only orienting the edges, because you can usually use edge control instead. Note that sexysledge orients and permutes the edges in the same way, so you can try both and see whether you can force AB3C instead of just AB4C.
• Likewise, but for fat Sune (r U R' U R U2 r') and longsexysledge (R U2 R' U2 R' F R F' / r U2 R' U2 r' F R F'), which also flip two edges but permute them differently from F R U R' U' F'. (Note that fat Sune, longsexysledge, and their mirrors and inverses all solve the same LLEF case.)
  
• Optimal J perms and T perm.
• The 5-move VHLS algs (F R U' R' F' and F R U R' F').
• F2 R U R' U' F2 for last slot. (Preserves 2-gen reduction; depending on AUF, this could either be the same move count as the usual intuitive F2L solution or be one move shorter.)
• The six-move edge 3-cycle: M' U2 M U2. Writing this as L R' F2 L' R U2, we can also shift one or both of the L and R' moves to the end, e.g. L F2 L' R U2 R', to get another alg for exactly the same 3-cycle. This sometimes allows for more move cancellations.
• [F' R D2 R' F, U2] for twisting two corners. (Can be used from multiple angles to increase the chances of cancelling moves.) If you have time to spare, doing two 3-cycle insertions is usually better than using this alg because you can force a ton of moves to cancel.

(I have used almost all of the algs in this list at least once in a "serious" FMC attempt. Usage of the "alternative EP" algs is a very case-by-case thing; going from AB4C to AB3C saves you one insertion (~6 moves), so Sune + AB4C (7+6+6=19) is likely to be worse than, say, Bruno + AB3C (9+6=15), but if we're comparing Sune + AB4C to Bruno + AB4C, obviously the former wins. It's still possible that the latter ends up being shorter due to better cancellations, but it's very unlikely to make up for it in cancellations.)

 

[theos]

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

X2 Y2
RW' R' U RW' D' U2 R' U' B U B' (11 Moves f2l-1)

I don't get how to get a good sleketon after this.

Y'
R U' R' F' U F R U' R2 F R2 U' R' U' R U R' F' (18) (I somehow found this to 2 twisted corners)
Y2 F2 U' R F2 D' F2 D R2 U' R U F2 (11)

40 total. But I used Cube explore to get the twisted corner case. What would you guys do after F2l-1? What algs should a FMCer want to learn? If any?

I strongly suggest learning to write your move sequences without rotations or wide moves. This will make it much easier to insert into your skeleton and find cancellations. Your solution so far removing rotations and wide moves:
R' L' B R' D' // 2x2x2 (5/5)
U2 L' // F2L-2 (2/7)
U' F U F' // F2L-1 with 2 edges, 5 corners remaining (4/11)

The easiest skeletons to finish with inserts are skeletons with only corners remaining. So the obvious way to proceed is with finishing the edges. 2-Look edges:
R U B U' B' R' // orient edges (6/17)
B' U2 B U B' U B U // permute edges (7/24)
This leaves you with 4 corners unsolved (twisted 3-cycle + 1 corner twisted in place) which can be solved with 2 corner cycle insertions. From IF this can be done in 8 moves (2x8 move cycles with 8 moves cancelling) for a total of 32 moves.

If you know 1-look edge algorithms:
F2 B' R F' R F R2 F' R F' B // solve remaining 2 edges (1 move cancels so 10/21)
This leaves you with 5 corners unsolved (5-cycle) which can be solved with 2 corner cycle insertions. From IF this can done in 10 moves (2x8 move cycles with 6 moves cancelling) for a total of 31 moves.

Personally I'd prefer to do edge orientation a lot earlier in the solve as it makes finishing the skeleton a lot easier - no need to know any last layer algorithms. For example, a quick attempt following from your 2x2x2...
R' L' B R' D' // 2x2x2 (5/5)
L B L2 // 2x2x3 (3/8)
U2 L' B L // EO (4/12)
U B' U B U2 B2 // F2L-1 (6/18)
U2 B' U2 B U // Solve remaining edges and 1 corner, leaving 4 corners (5/23)
From IF this can be done in 10 moves for a total of 33 moves.

 

[Duncan Bannon]

Spoiler

Protip: don't use rotations or wide moves. Don't think of the letters used in the notation as being short for "up"/"front"/"right"/etc., but as being short for "white"/"green"/"red"/etc. (Assuming you use WCA orientation, you get U=white and F=green for free, along with R=red and B=blue having obvious mnemonics, which leaves L=orange and D=yellow you have to explicitly memorise.)

That's simple, I will start doing that.

So let's rewrite your solution first, and then work on that.

R' L' B R' D' U2 L' U' F U F' // F2L-1 (11)
B U' B' R' U R B U' B2 R B2 U' B' U' B U B' R' // ab2c (18/29)
L2 U' F L2 D' L2 D F2 U' F U L2 // corner twists (12/41)

>B U' B' R' U R B U' B2 R B2 U' B' U' B U B' R' // ab2c (18/29)
You can use B U' B' to form the pair, then insert with a sledgehammer for edge control as well. For this ZBLL case (J perm + two twisted corners), you can use either a Sune (L U L' U L U2 L') or an Antisune (B U2 B' U' B U' B') to get you to L3C. (Fun fact: this ZBLL case came up in Feliks's 38.52 555 WR, which he solved with Antisune + L3C.) This overall isn't very good, especially if you don't do insertions (7 for F2L + 8 to set up to L3C + 9 for L3C = 24 moves), but it's probably the first option I'd try.

How did you get ab2c? After forming the pair, why did you do what you did to get ab2c? Also, how would the ZBLL leave your l3c? wouldn't it leave a solved cube? Why is "this overall isn't very good"

Backing up a bit:
R' L' B R' D' // 223 - corner (5)

From here the obvious thing is to do what you did, but that led to a bad continuation so let's try something else.
L B L2 B' // F2L-2 (4/9)

What makes this a bad continuation? It leaves a pair.

It's tempting to just insert the last pair (since it's already formed!), so let's see where that leads us. We can do a normal insert (L U2 L'), AUF then normal insert (U' L U' L'), or AUF then sledgehammer (U2 L' B L B'). Let's evaluate all of them.

L U2 L' / B' U' R' U R B U' // leaves 2+2+1 corners (10/19)
U' L U' L' / // leaves 3+1 corners (15/24) (note: this ten-move alg is, I believe, a common OLL alg)
U2 L' B L B' / L U L' U L U2 L' U' // leaves 3+1 corners (13/22)

First one is shortest, but leaves five corners in a bad state (you need three 3-cycle insertions), whereas the second and third options leave the same "type" of corners left (needs two insertions), but the third is shorter and so is more likely to be useful. But what else can we try?

You just do EO and EP right? In one step, that's what B' U' R' U R B U'/ F' L' U' L U' F U F' U F U/ L U L' U L U2 L' U', are right?

B' U2 B // EO (3-1/11)
L U' L' U L U' L' U2 // F2L-1, edges; leaves 2+2+1 corners (8/19)

That turned out to be useless.

Just EO then doing a pair, did you do anything special to get all edges done? Is it bad cause it would take 2 cycles or like 20 moves?

What about going to the inverse scramble? I don't see anything nice with standard F2L solving, so let's try EO then edges.


R' L' B R' D' // 223 - corner (5)
L B L2 B' // F2L-2 (4/9)
(L U2 L' U' B' U B U2 B' U' B) // leaves 2+2 corners (11-1/19)

I seem to struggle to get a skeleton like - (L U2 L' U' B' U B U2 B' U' B) why did you write it as 2+2 corners.

If you don't feel like doing insertions, you can also solve these last four corners with a triple sexy setup ([B2: (B' L' B L)3]), and that cancels a few moves.

R' L' B R' D' // 223 - corner (5)
L B L2 B' // F2L-2 (4/9)
(L U2 L' U' B' U B U2 B' U' B) // leaves 2+2 corners (11/20)
(B L' B L B' L' B L B' L' B L B2) // L4C (13-2/31)

For some reason I couldn't get L4C to work. How did you know triple sexy would work? When you say "if you don't feel like doing insertions" do you mean corner comm as a insertion?

In no particular order, with a focus on last layer:

• Optimal algs for all last layer corner 3-cycle cases. The 9- and 10-move cases (e.g. A perm) are very rarely useful for FMC since you can usually use an insertion and do a 3-cycle in 8 moves, but if you're very short on time and you're slow at insertions, this can save you from getting a DNF. The 8-move cases are all just commutators, so you don't really need to learn them as algs per se, although you should know which ones have multiple 8-move solutions (e.g. R B R' F R B' R' F' = F' L' B L F L' B' L).
  
• Sune for edge permutation. (Also keep in mind that Sune preserves a pair.)
• "Super Sune" (R2 U R2 U R2 U2 R2), also for edge permutation, when one of the F2L slots has its edge solved and corner not solved.
• R' U2 R' D' r U2 r' D R2. This preserves a 2×2 square while swapping the other two edges. (This is just the standard headlights OLL alg but with some wide moves.)
• Bruno (R U2 R2 U' R2 U' R2 U2 R) and optimal U perm (F2 U M' U2 M U F2) can also be used for edge permutation, but are generally less likely to be useful than the above options. (In particular, U perm can sometimes be replaced with a shorter edge insertion.)
  
• r U' r U2 R' F R U2 r2 F. This is optimal for swapping opposite edges with no AUF. (With AUF, optimal T perm is better.)
  
• F R U R' U' F' and mirrors/inverses for edge orientation and permutation. Avoid using these if you're only orienting the edges, because you can usually use edge control instead. Note that sexysledge orients and permutes the edges in the same way, so you can try both and see whether you can force AB3C instead of just AB4C.
• Likewise, but for fat Sune (r U R' U R U2 r') and longsexysledge (R U2 R' U2 R' F R F' / r U2 R' U2 r' F R F'), which also flip two edges but permute them differently from F R U R' U' F'.
  
• Optimal J perms and T perm.
• The 5-move VHLS algs (F R U' R' F' and its inverse)

I may need help using these algs later, but for now they should work

Thank you ever so much for the help, I know I asked about another million questions, so sorry for that. Thank you very very much.

@theos - Ill get you a reply later. Thanks for the help. It may be in a PM.

@Robert Marik - Would you then finish it as a 4 corner 2 edge comm?

 

[xyzzy]

How did you get ab2c? After forming the pair, why did you do what you did to get ab2c? Also, how would the ZBLL leave your l3c? wouldn't it leave a solved cube? Why is "this overall isn't very good"

The ab2c is your solution (with the weird sledge then cancel into OLL thing), lol. I meant that if you did this:
R' L' B R' D' U2 L' U' F U F' // F2L-1 (11)
B U' B' // pair (3/14)
U B' R B R' // VHLS (5/19)
you'll be left with a last layer case that can be solved by doing Sune or Antisune, then a 3-cycle. (A decent proportion of adjacent-EP ZBLL cases are like that, actually.) Since we're doing FMC, it's better to do the 3-cycle as an insertion, rather than at the end of the solve. (Is this clearer?)

L U L' U L U2 L' U // edges; ab3c (8/27)

From Insertion Finder:
L' [@1] R' B R' D' U2 L' U' F U F' B U' B' U B' R B R' L U L' U L U2 L' U
Insert at @1: L U' R2 U L' U' R2 U
Fewest moves: 33. 2 moves cancelled
The final solution: U' R2 U L' U' R2 U R' B R' D' U2 L' U' F U F' B U' B' U B' R B R' L U L' U L U2 L' U

I said that it's "not very good" because overall it would have used 27 moves to get to the last three corners. (I guess it's a matter of what your goal is; if you're aiming at sub-35, it's fine, but it's not good enough for sub-30 unless you get very lucky with insertions.)

>What makes this a bad continuation? It leaves a pair.
There's a pair, but it led to an awkward last layer, so that's bad.

>You just do EO and EP right? In one step, that's what B' U' R' U R B U'/ F' L' U' L U' F U F' U F U/ L U L' U L U2 L' U', are right?
Yes. (Well, the last one is a Sune, which doesn't affect EO.) I'm not entirely sure which cases are worth learning, other than those I listed in my previous post.

>Just EO then doing a pair, did you do anything special to get all edges done? Is it bad cause it would take 2 cycles or like 20 moves?
Nothing special. You can kinda track the edges to see if the EP will end up being solved after doing your F2L pair. It's bad because you need three 3-cycles when the corner cycles have length 2, 2, 1 (counting a corner twist as a "1-cycle"), and that takes 18-ish moves depending on how much you can cancel. Combined with a 19-move skeleton, you're looking at 37 moves total.

>I seem to struggle to get a skeleton like - (L U2 L' U' B' U B U2 B' U' B) why did you write it as 2+2 corners.
Parenthesised moves are done on the inverse scramble. The solution up to (but not including) that part is: R' L' B R' D' L B L2 B'. Do the inverse of these moves on a solved cube (i.e. B L2 B' L' D R B' L R), then do the inverse of the scramble sequence (F' U R …).

(L U2 L' U' B' U B U2 B' U' B) // leaves 2+2 corners (11-1/19)
The L U2 L' brings the green-white-red pair onto the white face, while orienting the edges. (Doing multiple things at once is good!) U' B' U B extends that pair into a square, and U2 B' U' B solves the last three edges. This leaves four corners.

Also, now that I'm looking at it again, if you do the inverse of [B2: (B' L' B L)3] instead (which permutes the corners in the same way), you get to cancel more moves and get sub-30, without even using insertions:
R' L' B R' D' // 223 - corner (5)
L B L2 B' // F2L-2 (4/9)
(L U2 L' U' B' U B U2 B' U' B) // leaves 2+2 corners (11/20)
(B2 L' B' L B L' B' L B L' B' L B') // L4C (13-4/29)

Insertion Finder finds a 28, if you do L4C with insertions instead of triplesexy:
R' L' B R' D' L B L2 B2 U B U2 B' U' B [@1] U L U2 L'
Insert at @1: D L D' R2 D L' D' R2
After the 1st insertion: R' L' B R' D' L B L2 B2 U B U2 B' U' B D L D' R2 D L' D' R2 U L [@2] U2 L'
Insert at @2: L' U' R2 U L U' R2 U
Fewest moves: 28. 7 moves cancelled
The final solution: R' L' B R' D' L B L2 B2 U B U2 B' U' B D L D' R2 D L' D' U L U' R2 U' L'

 

[Duncan Bannon]

Thank you sooooo much. I understand it much much better. If I have further questions do you mind answering them? If so, should I post em here (vs PM)?

 

[xyzzy]

Thank you sooooo much. I understand it much much better. If I have further questions do you mind answering them? If so, should I post em here (vs PM)?

If you post the questions here, other people can answer them too! (My free time isn't unlimited, unfortunately.)

 

[theos]

To build on what xyzzy said about solving on the inverse scramble...

If you have a scramble X, then you can take a scramble X' (the inverse scramble) and find a solution S. Then S' is a solution to the original scramble. This basically means you have double the possibilities for your solve.

But where it gets really powerful is that you can switch back and forth between solving on the original scramble and the inverse scramble repeatedly throughout the solve. If you have a scramble X and a partial solution P, you can do a scramble P' X' and proceed from there. If you then add more moves Q, you can switch back to continuing on Q' X P. Continue until you have a skeleton. What makes this especially powerful is that you have double the possibilities at every step of your solve!

For your scramble, switching between original and inverse scrambles can get us this solution:
Scramble: R' U' F D B' U R D2 R B U' F2 B2 U' L2 F2 D R2 B2 D B2 F R' U' F
L' B R2 // pseudo 2x2x2 (3/3)
Switch to inverse scramble, so doing R2 B' L as pre-moves to the inverse scramble...
(D) // 2x2x2 (1/4)
(L2 U2 L F' U2) // 2x2x3 (5/9)
Switch back to original, so doing U2 F L' U2 L2 D' as pre-moves to the original scramble...
F U L U' F L // Solve edges leaving a corner 5-cycle (6/15)

Spoiler: Aside

You could have also found this by staying with the inverse scramble as (L' F' U L' U' F').

To do inserts we want a skeleton on the original scramble without any pre-moves. All the pre-moves can be shifted to the end of the solution.
Skeleton: L' B R2 F U L U' F L ... U2 F L' U2 L2 D' (15 moves)

Spoiler: Optimal insertions...

L' B R2 * F U @ L U' F L U2 F L' U2 L2 D' (15 moves)
* = R B2 R' F' R B2 R' F (6/21)
@ = U' F2 U B U' F2 U B' (4/25)
Solution: L' B R' B2 R' F' R B2 R' U B U' F2 U B' L U' F L U2 F L' U2 L2 D' (25 moves)

What algs should a FMCer want to learn? If any?

In no particular order, with a focus on last layer:

• Optimal algs for all last layer corner 3-cycle cases. The 9- and 10-move cases (e.g. A perm) are very rarely useful for FMC since you can usually use an insertion and do a 3-cycle in 8 moves, but if you're very short on time and you're slow at insertions, this can save you from getting a DNF. The 8-move cases are all just commutators, so you don't really need to learn them as algs per se, although you should know which ones have multiple 8-move solutions (e.g. R B R' F R B' R' F' = F' L' B L F L' B' L).
  
• Sune for edge permutation. (Also keep in mind that Sune preserves a pair.)
• "Super Sune" (R2 U R2 U R2 U2 R2), also for edge permutation, when one of the F2L slots has its edge solved and corner not solved.
• R' U2 R' D' r U2 r' D R2. This preserves a 2×2 square while swapping the other two edges. (This is just the standard headlights OLL alg but with some wide moves.)
• Bruno (R U2 R2 U' R2 U' R2 U2 R) and optimal U perm (F2 U M' U2 M U F2) can also be used for edge permutation, but are generally less likely to be useful than the above options. (In particular, U perm can sometimes be replaced with a shorter edge insertion.)
  
• r U' r U2 R' F R U2 r2 F. This is optimal for swapping opposite edges with no AUF. (With AUF, optimal T perm is better.)
  
• F R U R' U' F' and mirrors/inverses for edge orientation and permutation. Avoid using these if you're only orienting the edges, because you can usually use edge control instead. Note that sexysledge orients and permutes the edges in the same way, so you can try both and see whether you can force AB3C instead of just AB4C.
• Likewise, but for fat Sune (r U R' U R U2 r') and longsexysledge (R U2 R' U2 R' F R F' / r U2 R' U2 r' F R F'), which also flip two edges but permute them differently from F R U R' U' F'.
  
• Optimal J perms and T perm.
• The 5-move VHLS algs (F R U' R' F' and its inverse).

Thanks for this xyzzy. The stuff for last layer edge orientation and permutation is really useful. I consciously tried to incorporate it into my latest timed practice solve and it helped to get me one of my best solves ever within the 1 hour competition limit...
Scramble: R' U' F L U' F' U2 R2 L2 D2 B F' R F2 R2 F' L' U D2 R' F' R2 B2 R' U' F
(U2 B' D' U' L') // 2x2x2 (5/5)
(F2 R' B' R' B) // 2x2x3 (5/10)
(R2) // F2L-1 (1/11)
F2 R U' R' U // EO using sledgehammer as discussed a bit earlier (5/16)
U F U' F' U F2 U' F // AB3C, permuting edges and keeping the last layer pair using Sune as mentioned in the quote above (7/23)
Skeleton: F2 R U' R' U2 F U' F' U F2 U' F R2 B' R B R F2 L U D * B U2
Insert at *: D2 B U2 B' D2 B U2 B' // corner cycle (3/26)
Solution: F2 R U' R' U2 F U' F' U F2 U' F R2 B' R B R F2 L U D' B U2 B' D2 B (26 moves)

 

[Duncan Bannon]

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

Or the inverse- U' L' U' D B' D R D F D2 R2 D B2 U F2 L2 F2 U' R2 U

and then I have
(R D' U L F )-2x2x2 (5/5)
(R D' R2 B-)2x2x3 (5/9)

R' F' R' F D' R D'- F2L-1 (7/16)
R D' R' D-EO (4/20)
R2 B R' B' R B R' B' F2L (8/28)
F R2 F' R' F R' F'- OLL + CO (7/35)



D R' B' F D2 B F' R' D2 (the u perm at the end) Where to insert it?


What would the final solution be? moves in ( ) done on inverse.

 

[theos]

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

Or the inverse- U' L' U' D B' D R D F D2 R2 D B2 U F2 L2 F2 U' R2 U

and then I have
(R D' U L F )-2x2x2 (5/5)
(R D' R2 B-)2x2x3 (5/9)

R' F' R' F D' R D'- F2L-1 (7/16)
R D' R' D-EO (4/20)
R2 B R' B' R B R' B' F2L (8/28)
F R2 F' R' F R' F'- OLL + CO (7/35)



D R' B' F D2 B F' R' D2 (the u perm at the end) Where to insert it?


What would the final solution be? moves in ( ) done on inverse.

The partial solution needs an R' at the end to turn it into an AB3E skeleton. Your skeleton is your moves on the original scramble followed by the inverse of the moves on the inverse scramble, so:
R' F' R' F D' R D' R D' R' D R2 B R' B' R B R' B' F R2 F' R' F R' F' R' ... B' * R2 D R' F' L' U' D R' (36 moves)
There isn't a nice [M, U2]-style edge insertion into this skeleton. IF says the best insertion (among many other options) is at *: B' R' U' D B2 U D' R' B (8/44)
Solution: R' F' R' F D' R D' R D' R' D R2 B R' B' R B R' B' F R2 F' R' F R' F' R' B2 R' U' D B2 U D' R' B R2 D R' F' L' U' D R' (44 moves)

Looking over your solve, I immediately see that there's no need to complicate things with an OLL step after the F2L step. Ending immediately with R' leaves us with a 29-move AB4C skeleton which is almost certainly better.

Spoiler: 29 move AB4C skeleton

(R D' U L F) // 2x2x2 from original solve (5/5)
(R D' R2 B) // 2x2x3 from original solve (4/9)
R' F' R' F D' R D' // F2L-1 from original solve (7/16)
R D' R' D // EO from original solve (4/20)
R2 B R' B' R B R' B' // F2L from original solve (8/28)
R' // AB4C (1/29)
Skeleton: R' F' R' F D' R D' R D' R' D R2 B R' B' R B * R' B' R' ... B' R2 D R' F' L' U' D R' (29 moves)
(Insertions from IF)
Insert at *: B L B' R' @ B L' B' R // Corner cycle (4/33)
Insert at @: R B R' F' R B' R' F // Corner cycle (4/37)
Solution: R' F' R' F D' R D' R D' R' D R2 B R' B' R B2 L R' F' R B' R' F B L' B2 R' B' R2 D R' F' L' U' D R' (37 moves)

And finally, completing F2L doesn't usually give you a good skeleton. You can get to a decent AB5C skeleton by just permuting the remaining edges after EO in your solution.

Spoiler: 25 move AB5C skeleton

(R D' U L F) // 2x2x2 from original solve (5/5)
(R D' R2 B) // 2x2x3 from original solve (4/9)
R' F' R' F D' R D' // F2L-1 from original solve (7/16)
R D' R' D // EO from original solve (4/20)
R B R B' R2 // AB5C by permuting edges (5/25)
Skeleton: R' F' R' F D' R D' R D' R' D R B * R B' R2 ... B' R2 D R' F' L' U' D R' (25 moves)
(Insertions from IF)
Insert at *: F L2 F' @ R2 F L2 F' R2 // Corner cycle (7/32)
Insert at @: F' L F R2 F' L' F R2 // Corner cycle (4/36)
Solution: R' F' R' F D' R D' R D' R' D R B F L2 F2 L F R2 F' L' F2 L2 F' R' B' R2 B' R2 D R' F' L' U' D R' (36 moves)

 

[Duncan Bannon]

Perfect! That is really useful! I will try to do a few solves and get back to you on how I did. Thank you very much for your help.

 

[G2013]

Just got PB mo3! Rolled a 33, 25, 26 into a 25, 26, 27

25:

Spoiler

Skrembol: R' U' F L D B2 D' F L B' R' D F U2 B' U2 F2 B' R2 U2 D2 B D2 R' U' F
Inverse: F' U R D2 B' D2 U2 R2 B F2 U2 B U2 F' D' R B L' F' D B2 D' L' F' U R

(R' D R2 U') //2 X 2x2x1
(B' D F' D' R) //2x2x3 extension
(F B' L F B' U') //6 unsolved pieces

Backup: U B F' L' B F' R' D F D' B U R2 D' R (U' L U R U' L' U R' + B L' S' L2 S L' B') (32)

Skeleton: U B F' L' B F' R' D F D' B U R2 D' (F' L2 F R F' L2 F R') R (15+8-2=21)
Next skeleton: U B F' L' B F' R' D F D' B (F U' F' U2 D2 B U B' D2 U2) U R2 D' F' L2 F (F' R L' U2 L R' F') R F' L2 F (21+7-3=25!!!!)

Final solution: U B F' L' B F' R' D F D' B U R2 D' F' R L U2 L R' F' R F' L2 F (25) HTM!!!!! PB SINGLE!

26:

Spoiler

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

R B' R' B D F' //Pseudo 2x2x3
[D2 L2 //f2l-2 pseudo
B D B' D' B D B' D' //pair
F L' F' L D' L D L' F' //some sort of VLS?
R F R B2 R' F R B2 R2 //to Backup in 34 moves]

Skeleton: R B' R' B D F' D2 L2 B D B' D' B (F2 D B' D' F2 D B D') D B' D' F L' F' L D' L D L' F' (25)

Final solution: R B' R' B D F' D2 L2 B D B' D' B F2 D B' D' F' L' F' L D' L D L' F' (26!!!!!! 7 moves cancelled!!!!), giving PB mo3 = 28.66

27:

Spoiler

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

D L F D R' //Pseudo 2x2x3
U' F2 R' F R2 //Pseudo F2L-1
(R2) //Premove
(B L' B') //EO
U' L U L' U' L2 U L //L3C

Skeleton: D L F D R' U' F2 R' F R2 U' L U L' U' L2 U L B L B' R2 (22)
Backup: D L F D R' U' F2 R' F R2 U' L U L' U' L2 U L B L B' R2 L2 B R' B' L2 B R B' (30)
Insertion: D L F D R' U' (R B2 R' F2 R B2 R' F2) F2 R' F R2 U' L U L' U' L2 U L B L B' R2 (27!!!)
PB mo3 26.00!!!!!!!!!!!!!!!!

Giving PB mo3 (25, 26, 27) = 26.00!!!!! :0

First time I ever get 3 sub30s in a row, and given that I don't really practice FMC often, this is super good for me. I'm content now hahaha

Any tips are absolutely welcome and I'm grateful for them

 

[Shymon]

R' U' F U F2 U2 F2 R2 B2 U2 L2 U R F R F2 L2 R2 D L' D' F2 R' U' F
Could anyone help me with this scramble?

I couldn't find anything worth sticking to and finally gave up after an hour. The best thing I found was ~25 to l3c but that was while doing some random moves mid solution and I couldn't reconstruct it.

 

[xyzzy]

R' U' F U F2 U2 F2 R2 B2 U2 L2 U R F R F2 L2 R2 D L' D' F2 R' U' F
Could anyone help me with this scramble?

I couldn't find anything worth sticking to and finally gave up after an hour. The best thing I found was ~25 to l3c but that was while doing some random moves mid solution and I couldn't reconstruct it.

Spoiler: ab3c, 26 moves

F D2 L B' // eo (4/4)
U' D R' // squares and pairs (3/7)
L U L U2 F2 // blox (5/12)
L U2 L U2 L2 // ab4c2e (5/17)
L2 F U2 F' L2 D2 F' R2 F D2 B' // T perm; ab3c (11-2/26)

Spoiler: ab3c, 26 moves

F D2 L B' // eo (4/4)
U' D R' // squares and pairs (3/7)
L U L U2 F2 // blox (5/12)
B2 L B R B R' L' // F2L (7/19)
L' B' L B' L' B2 L B' // edges; ab3c (8-1/26)

Spoiler: ab2c2e, 20 moves

F D2 L B' // eo (4/4)
U' D R' // squares and pairs (3/7)
L U L U2 F2 // blox (5/12)
(D' B2 D) // insert edge (3/15)
(U2 L U) // blocks (3/18)
U' L // ab2c2e (2/20)
skeleton: F D2 L B' U' D R' L U L U2 F2 U' L U' @ L' U2 D' B2 D
@ = F2 R' D2 R F2 U2 R B2 R' U2 L // T perm (11-4/27)

Can't really find anything nice either, other than the 2c2e skeleton. (20 moves for 2c2e isn't good, but cancellations make up for that.)

 

[G2013]

Looking for some critiques, I'm a n00b

30 mover, found in 27 minutes, gave PB avg5 (33 25 26 27 30 = 27.66)
S: R' U' F R2 U' B2 R2 U2 F2 R2 U L2 D2 B2 F' D L D L2 R' D2 F2 L U2 R' U' F
I: F' U R U2 L' F2 D2 R L2 D' L' D' F B2 D2 L2 U' R2 F2 U2 R2 B2 U R2 F' U R

Spoiler

R2 F L D U L' D' //2x2x3
B D' R' D2 B2 D' //EO + Blocks
R' B2 //F2L-1
(R B' R' B) //Pairs n stuff
R B' R2 B R2 //L3C

Skeleton: R2 F L D U L' D' B D' R' D2 B2 D' R' B2 R B' R2 B R2 B' R B R' (24)
Backup: skeleton + R2', F' L F = 31 moves
Insertion: R2 F L D U L' D' B D' R' D2 B2 D' (U R' D2 R U' R' D2 R) R' B2 R B' R2 B R2 B' R B R'
Final solve: R2 F L D U L' D' B D' R' D2 B2 D' U R' D2 R U' R' D2 B2 R B' R2 B R2 B' R B R' = 30

 

[theos]

R' U' F U F2 U2 F2 R2 B2 U2 L2 U R F R F2 L2 R2 D L' D' F2 R' U' F
Could anyone help me with this scramble?

I couldn't find anything worth sticking to and finally gave up after an hour. The best thing I found was ~25 to l3c but that was while doing some random moves mid solution and I couldn't reconstruct it.

Got this pretty good skeleton on the inverse scramble.
F2 U2 R U L2 F' // 2x2x2 + couple pairs (6/6)
D2 R' D' // F2L-2 for either R or B last layer (3/9)
B2 R' B R B2 // Fill in a slot to end with AB5C (5/14)

This wasn't the first 2x2x2 block I tried and I rarely have time in the hour to investigate more than the best 1 or 2 starting blocks. This took me well over an hour to find, though in retrospect I can see why this is a very promising first block start.

 

[Shymon]

Got this pretty good skeleton on the inverse scramble.
F2 U2 R U L2 F' // 2x2x2 + couple pairs (6/6)
D2 R' D' // F2L-2 for either R or B last layer (3/9)
B2 R' B R B2 // Fill in a slot to end with AB5C (5/14)

This wasn't the first 2x2x2 block I tried and I rarely have time in the hour to investigate more than the best 1 or 2 starting blocks. This took me well over an hour to find, though in retrospect I can see why this is a very promising first block start.

Wow, this skeleton looks great! Canceling 4 moves on 2 insertions gives 26 moves total. That's quite impressive for such a scramble.

 

[Bubtore]

Swisscubing Cup I - 2018 attempt 2

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

(F2 U' R) // EO
D F D2 // 122 block + 2 pairs
(R2 F L2 F2) // 223
B' U2 B2 U2 B' // 2E3C

Skel // D F D2 + B' U2 B2 * U2 B' F2 L2 F' R U F2 - 14

* B2 U B' U B U2 B' U B' // reduce to L3C (9-3=6)
+ D B' U' B D' B' U B (8-6=2)

Solution // D F D' B' U' B D' B2 U B U2 B' U B' U2 B' F2 L2 F' R U F2 - 22 Moves

 

[Robert Marik]

Swisscubing Cup I - 2018 attempt 2

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

(F2 U' R) // EO
D F D2 // 122 block + 2 pairs
(R2 F L2 F2) // 223
B' U2 B2 U2 B' // 2E3C

Skel // D F D2 + B' U2 B2 * U2 B' F2 L2 F' R U F2 - 14

* B2 U B' U B U2 B' U B' // reduce to L3C (9-3=6)
+ D B' U' B D' B' U B (8-6=2)

Solution // D F D' B' U' B D' B2 U B U2 B' U B' U2 B' F2 L2 F' R U F2 - 22 Moves

Nice. How did you find B2 U B' U B U2 B' U B' ? Is it an alg which you remember? Or is it just reordering edges using B-U moves? In the second case, did you know before the execution of this alg which corners remain intact?

 

[mazh]

24.33 personal ao5 PB! (24,27,24,25,24)=24.33
Last 3 scrambles are also my mo3 PB. (24,25,24)=24.33

Spoiler: 24moves, solve it without insertion

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

NISS:
U2 B2 R L U' // 222(5/5)
D R2 // pusedo 223(2/7)
NISS:
R' // 223(1/8)
NISS:
D F' D2 F // pusedo TEC(4/12)
R' D' R // block(3/15)
F2 L2 U2 B' U' B U' L2 F' // finsh(9/24)

Spoiler: 27moves, 17L4C

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

NISS:
U' D' B L' R U' D R' # B2 // TEC(9/9)
U2 B' U B U2 L * U' L' // L4C(8/17)

* L' U R U' L U R' U' (8-4/25-4)
# L' F2 L B2 L' F2 L B2 (8-2/29-2)

Spoiler: 24moves, 21L3C

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

NISS:
U L' U2 F' U F2 // 222(6/6)
B' L2 B' L' B // pusedo TEC(5/11)
NISS:
L2 F' D' F D L B' D' B D2 // L3C(10/21)

Skeleton: L2 F' D' F D L B' D' B D2 B' L B * L2 B F2 U' F U2 L U'
* B R B' L2 B R' B' L2 (8-5/29-5)

Spoiler: 25moves, 20L3C

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

D' F U2 L' D' // 222(5/5)
F' U' // EO(2/7)
F' R F' R' // block(4/11)
NISS:
R' F2 B' R F' R' B R' U2 // L3C(9/20)

Skeleton: D' F U2 * L' D' F' U' F' R F' R' U2 R B' R F R' B F2 R
* U R' U' L' U R U' L (8-3/28-3)

Spoiler: 24moves, 19L3C

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

L2 * D2 B' L F U' // 222(6/6)
U' B U2 R' U' // pusedo 223+blocks(5-1/11-1)
D R' D2 B R2 // T-TEC(5/15)
D R' D' R // L3C(4/19)

* L U' L' D2 L U L' D2 (8-3/27-3)

Nice. How did you find B2 U B' U B U2 B' U B' ? Is it an alg which you remember? Or is it just reordering edges using B-U moves? In the second case, did you know before the execution of this alg which corners remain intact?

It just is a sune case, and ues it with set up.
B (B U B' U B U2 B' U) B'

 

[Bubtore]

If you do B at that point of the skeleton you will have all unsolved pieces on the same layer, and both edges are oriented. This means that it is a ZBLL. So I first set this up to get that ZBLL, but I didn't know it (now I do ). Instead, I saw that doing a sune using B U moves will lead to 3 corners, and cancels 3 moves on top of that.

So actually you have to look at the insertion like mazh described it