[Help Thread] FMC Discussion

[TheNextFeliks]

Ryan Pope FMC 32 Move Winning Solution Tree Town 2013

Here is my winning fmc solution from Tree Town 2013:

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

21 move skeleton Heise-style:
F R
B L B'
U @1 L' B2 L
U' F U2 @2 R2
U' F
D R2 D'
F' R' F

2 Insertions 5 moves cancel:
@1: U R U' L2 U R' U' L2
@2: U2 F D' F' U2 F D F'

Insertions were a pain without stickers but still good.

Btw: Chester came in 2nd with 36 and Nathan Dwyer came in 3rd with 39 iirc.

 

[antoineccantin]

Insertions were a pain without stickers

What? You normally have stickers on your cubes? That's bizarre, I always keep my cubes fully unstickered.

 

[TeddyKGB]

Insertions were a pain without stickers but still good.

Wow, I couldn't imagine doing insertions without stickers, Nice job. I came in 5th with 45 which I was extremely disappointed with but I usually fail pretty hard at comps so it's not surprising...

 

[TheNextFeliks]

Wow, I couldn't imagine doing insertions without stickers, Nice job. I came in 5th with 45 which I was extremely disappointed with but I usually fail pretty hard at comps so it's not surprising...

Thanks. Good job yourself.

Has anyone thought about doing, for insertions, take an unstickered cube, put the three stickers on the pieces that need to be switched? Then scramble and solve an look for insertions. Seems like it could be easier.

 

[MarcelP]

interesting...
Do you have an example where IF missed the optimal inertion?

Right now I am bug testing a first raw version and I have found some.

http://mf.qiyuuu.com/cube/if/7136f50742c252b3f1980464349c624e.cube

1 edge 3-cycle(s) insertion(s) needed
No proper insertion found
Runtime: 0 milliseconds


My program: Shortest solution :U2 D' F R2 U' D B2 U R' U2 D'

My program finds many solutions on that case. 'IF' is much much faster than my program however I have been in contact with BQ (developer of 'IF') and it seems I am missing a lot of commutators. So I will be focussing on that problem now. When I have fixed that I will have to find a way to make my program much faster. (Single 3 corner of 3 edge cycle is one second, but 4 corners or 5 corners takes a minute).

 

[okayama]

Right now I am bug testing a first raw version and I have found some.

http://mf.qiyuuu.com/cube/if/7136f50742c252b3f1980464349c624e.cube

1 edge 3-cycle(s) insertion(s) needed
No proper insertion found
Runtime: 0 milliseconds


My program: Shortest solution :U2 D' F R2 U' D B2 U R' U2 D'

My program finds many solutions on that case. 'IF' is much much faster than my program however I have been in contact with BQ (developer of 'IF') and it seems I am missing a lot of commutators. So I will be focussing on that problem now. When I have fixed that I will have to find a way to make my program much faster. (Single 3 corner of 3 edge cycle is one second, but 4 corners or 5 corners takes a minute).

It seems to work for me.
http://mf.qiyuuu.com/cube/if/2331dbd60639512dff45804be8a9581a.cube

You should check "edge 3-cycle" before your submission.

 

[Jakube]

Right now I am bug testing a first raw version and I have found some.

http://mf.qiyuuu.com/cube/if/7136f50742c252b3f1980464349c624e.cube

1 edge 3-cycle(s) insertion(s) needed
No proper insertion found
Runtime: 0 milliseconds


My program: Shortest solution :U2 D' F R2 U' D B2 U R' U2 D'

I just checked this with IF again, and I get a solution (the same as yours): http://mf.qiyuuu.com/cube/if/2331dbd60639512dff45804be8a9581a.cube
Maybe you forgot to search for edge insertions.

The only case I know, where the IF doesn't work, is when the optimal insertion is a 9-mover (even when there is a cancelation): http://mf.qiyuuu.com/cube/if/f360b2c7d4e77d89ab54ca9d85b38e1c.cube
Scamble: U2 B2 D F2 D' B2 D F2 U2 R2 (10)
Skeleton: R2 D' (2)

Using this skeleton there is no proper insertion (I mean 8-mover) possible, but you can insert a 9-mover, which cancels one move:
Scamble: U2 B2 D F2 D' B2 D F2 U2 R2
Skeleton: R2 # D' (2)
Insertion at #: D2 B2 D' F2 D B2 D' F2 D' (9)
So you result in a 10-move solution.

If my memory serves me right, Okayama had a solve once, where something like this occurred.

It would be really great, if your program is able to find these insertions. You could implement an option, that determines, if your program looks for 9-mover or not. It will be interesting to see the difference on certain skeletons.

 

[mycube]

If my memory serves me right, Okayama had a solve once, where something like this occurred.

Not just him, me too. But yes, you are right. When I posted my solution where the 9 move insertion is optimal some months ago Okayama replied with some other solutions.

 

[BQ]

If my memory serves me right, Okayama had a solve once, where something like this occurred.

Okayama's case: http://www.speedsolving.com/forum/showthread.php?13599-The-FMC-thread&p=778567&viewfull=1#post778567

I've no idea about finding such non-pure insertions. Maybe it can be solved by adding 9-moves and 10-moves algorithms.

 

[Bizarro]

Wow, I couldn't imagine doing insertions without stickers, Nice job. I came in 5th with 45 which I was extremely disappointed with but I usually fail pretty hard at comps so it's not surprising...

Haha. But you still got compliments on your solves, so there's that.

 

[MarcelP]

It seems to work for me.
http://mf.qiyuuu.com/cube/if/2331dbd60639512dff45804be8a9581a.cube

You should check "edge 3-cycle" before your submission.

Ah.. I see. I did not know that. I have used it before but I do not recall checking one of the options. In any case, my program does not need user input for that.

The only case I know, where the IF doesn't work, is when the optimal insertion is a 9-mover (even when there is a cancelation): http://mf.qiyuuu.com/cube/if/f360b2c7d4e77d89ab54ca9d85b38e1c.cube
Scamble: U2 B2 D F2 D' B2 D F2 U2 R2 (10)
Skeleton: R2 D' (2)

Using this skeleton there is no proper insertion (I mean 8-mover) possible, but you can insert a 9-mover, which cancels one move:
Scamble: U2 B2 D F2 D' B2 D F2 U2 R2
Skeleton: R2 # D' (2)
Insertion at #: D2 B2 D' F2 D B2 D' F2 D' (9)
So you result in a 10-move solution.

Yeah, my program does find these cases..

Spoiler

Scramble= U2 B2 D F2 D' B2 D F2 U2 R2
Commutator = U B' R' B L2 B' R B L2 U'
Moves cancelled = 0
Solution = U B' R' B L2 B' R B L2 U'R2 U B' R' B L2 B' R B L2 U'D'
Cancelled string= U B' R' B L2 B' R B L2 U' R2 U B' R' B L2 B' R B L2 U' D'

Scramble= U2 B2 D F2 D' B2 D F2 U2 R2
Commutator = D L' B' L F2 L' B L F2 D'
Moves cancelled = 1
Solution = D L' B' L F2 L' B L F2 D'R2 D L' B' L F2 L' B L F2 D'D'
Cancelled string= D L' B' L F2 L' B L F2 D' R2 D L' B' L F2 L' B L F2 D2

Scramble= U2 B2 D F2 D' B2 D F2 U2 R2
Commutator = D2 D' F2 D B2 D' F2 D B2 D2
Moves cancelled = 1
Solution = D2 D' F2 D B2 D' F2 D B2 D2R2 D2 D' F2 D B2 D' F2 D B2 D2D'
Cancelled string= D F2 D B2 D' F2 D B2 D2 R2 D F2 D B2 D' F2 D B2 D

Scramble= U2 B2 D F2 D' B2 D F2 U2 R2
Commutator = D F2 L' B' L F2 L' B L D'
Moves cancelled = 1
Solution = R2 D F2 L' B' L F2 L' B L D'D'
Cancelled string= R2 D F2 L' B' L F2 L' B L D2

Scramble= U2 B2 D F2 D' B2 D F2 U2 R2
Commutator = D2 F2 L' B' L F2 L' B L D2
Moves cancelled = 1
Solution R2 D' D2 F2 L' B' L F2 L' B L D2
Cancelled string= R2 D F2 L' B' L F2 L' B L D2

Scramble= U2 B2 D F2 D' B2 D F2 U2 R2
Commutator = U L2 B' R' B L2 B' R B U'
Moves cancelled = 0
Solution = R2 U L2 B' R' B L2 B' R B U'D'
Cancelled string= R2 U L2 B' R' B L2 B' R B U' D'

Scramble= U2 B2 D F2 D' B2 D F2 U2 R2
Commutator = D2 B2 D' F2 D B2 D' F2 D D2
Moves cancelled = 1
Solution = R2 D2 B2 D' F2 D B2 D' F2 D D2D'
Cancelled string= R2 D2 B2 D' F2 D B2 D' F2 D2

Scramble= U2 B2 D F2 D' B2 D F2 U2 R2
Commutator = D' B2 D' F2 D B2 D' F2 D D
Moves cancelled = 1
Solution = R2 D' D' B2 D' F2 D B2 D' F2 D D
Cancelled string= R2 D2 B2 D' F2 D B2 D' F2 D2

Scramble= U2 B2 D F2 D' B2 D F2 U2 R2
Commutator = U' L2 D' R2 D L2 D' R2 D U
Moves cancelled = 0
Solution = R2 D' U' L2 D' R2 D L2 D' R2 D U
Cancelled string= R2 D' U' L2 D' R2 D L2 D' R2 D U

 

[okayama]

It seems to work for me.
http://mf.qiyuuu.com/cube/if/2331dbd60639512dff45804be8a9581a.cube

You should check "edge 3-cycle" before your submission.

Ah.. I see. I did not know that. I have used it before but I do not recall checking one of the options. In any case, my program does not need user input for that.

Just in case you missed out: IF usually does not need the user input either.
"edge 3-cycle" is automatically checked after you input the skeleton, by
investigating the unsolved pieces. I'm not sure why that is not checked
in your case, though.

 

[JackJ]

An old scramble I know, but I just did the scramble from Saarland Open 2012.

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

2x2: U' D B' U B2 D2 R
2x3: L2 U' B' U' B
F2L: U' F U' F U' R U2 R' U' F U' F'
OLL: R B U' B' U' B U B' R2
PLL: U B' F R2 B F' U R2 U'

42 moves. I know my finishing F2L is bad, and my LL is not efficient, but I'm working on it. Hopefully get something around 42 moves at Worlds.

 

[ardi4nto]

Hi,

Haven't done FMC in a while, and I submitted classic FMC #459 in Per's FMC site, but it doesn't appear in the result.
Anyone know why? And How can I contact Per?

Here's my solution anyway:

Spoiler

premove R'

1x2x3: U' L2 * U' B L' B'
2x2x3 + some pairs: R' U L2 F' R' F
F2L: R2 D F' R F D
Correction: R'

pair 3 cycle at *:
B2 U' Fw2 U B2 U' Fw2 U (or without wide moves: B2 U' B2 D B2 D' B2 U)

Final solution: U' L2 B2 U' B2 D B2 D' B' L' B' R' U L2 F' R' F R2 D F' R F D R' (24)

 

[okayama]

Hi,

Haven't done FMC in a while, and I submitted classic FMC #459 in Per's FMC site, but it doesn't appear in the result.
Anyone know why? And How can I contact Per?

Here's my solution anyway:

Spoiler

premove R'

1x2x3: U' L2 * U' B L' B'
2x2x3 + some pairs: R' U L2 F' R' F
F2L: R2 D F' R F D
Correction: R'

pair 3 cycle at *:
B2 U' Fw2 U B2 U' Fw2 U (or without wide moves: B2 U' B2 D B2 D' B2 U)

Final solution: U' L2 B2 U' B2 D B2 D' B' L' B' R' U L2 F' R' F R2 D F' R F D R' (24)

http://www.speedsolving.com/forum/showthread.php?13599-The-FMC-thread&p=857076&viewfull=1#post857076

Currently-known workaround: write \' instead of ' in the comments.

 

[Sebastien]

Yes, it is very annoying. I had a 23 move solution for #459 and forgot to add the Backslashes (while I did this for #458). I kind of doubt that per is actively working on this. Is anyone in contact with him?

 

[PM 1729]

Nikhil Mande and me did the #459 on mustcube together. We got a pretty cool skeleton (15 moves with 5 corners left) and then insertions to a total of 24 in around 2 hrs. Here is the solution:

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

Solution: U L R' B R F R' B' R B' L' B R B' L B2 D2 R B' D B2 U F2 R

Explanation:
On scr: Blocks: U L F R (4)

Switch to inv with pre: R' F' L' U' (4)

On inv: Pseudo 2x2x3: R' F2 U' B'(4) +Pre: B' (1)

Back to scr with pre: B U F2 R (4/4)
On scr: 2x2x3: U L F R B (5/9)
F2L-1: D2 R (2/11)
Leave 5 corners: B' D B (3-1 cancels with pre=2/13)

So skeleton is: U L * F R @ B D2 R B' D B2 U F2 R (13)

Insert at *: R' B R F' R' B' R F' (8-3=5)
Insert at @: R' B' L' B R B' L B (8-2=6)
for a total of 24.

Sub Okayama! YAY!

 

[okayama]

Really well done. I should have investigated it more, in view of Sébastien's result (said 23 moves) as well.

 

[cubernya]

It (should) be a fairly simple fix in the coding, just switching ' to ", but it might be more than that. Either way, it should be a very simple fix.

(I'm assuming that the backslash is because it's not escaping the 's in a line of code)

 

[IQubic]

What exactly is FMC. It seems interesting. What is it? Can you get good at it? And do you practice it?
-IQubic