Random Cubing Discussion

[uberCuber]

I don't really know where else to say this, but do people who are not color neutral do FMC on white cross always?

I would hope they aren't starting on a cross at all. I'm not color neutral for speedsolving, but for FMC I just blockbuild whatever I see.

 

[TheMachanga]

I would hope they aren't starting on a cross at all. I'm not color neutral for speedsolving, but for FMC I just blockbuild whatever I see.

Ok, then last layer on yellow?

 

[uberCuber]

Ok, then last layer on yellow?

No. Just because I can't recognize F2L cases on other colors quite as fast as I can on white when speedsolving doesn't mean it's at all difficult to consider the cube from a color-neutral standpoint during FMC, when you don't have to recognize a case and act in a fraction of a second.

 

[TMOY]

And if solving color neutral for FMC still causes you trouble, you also have the option of scrambling with a different orientation.

 

[vcuber13]

What is the shortest 2 or 4 cycle within the UF and UB dedge in <r,U2>?

I found one by hand that's 25 htm.

 

[Kirjava]

 

[5BLD]

What is the shortest 2 or 4 cycle within the UF and UB dedge in <r,U2>?

I found one by hand that's 25 htm.

I don't see how HTM is a good metric for this?

 

[uberCuber]

I don't see how HTM is a good metric for this?

Just a guess, perhaps he means that in SiGN notation, with the r being a wide turn rather than slice move? That would make HTM more reasonable than for slice moves.

 

[5BLD]

:/
So much persuasion
So little information
So much time

 

[EVH]

Good find, very interesting, and I think it is viable; I did feel like he was trying to sell me something, not explain a new method though.

 

[vcuber13]

I don't see how HTM is a good metric for this?

how isn't it? if it were qtm each U2 would be two moves

and i meant r to be just the slice, i think qq showed that <Rw, U2> can't be done.

 

[Julian]

how isn't it? if it were qtm each U2 would be two moves

and i meant r to be just the slice, i think qq showed that <Rw, U2> can't be done.

What about STM?

 

[vcuber13]

forgot about that, and that slices are two in htm

 

[evogler]

Random question: Has anyone worked on software to simulate fingering for / estimate finger-friendliness of sequences of moves?

 

[Athefre]

Random question: Has anyone worked on software to simulate fingering for / estimate finger-friendliness of sequences of moves?

I've been working on a program that sorts a list of sequences based on their potential feel.

I doubt I'll finish it before the end of the year because something has come up.

 

[evogler]

Neat. May I ask roughly how you approach the problem?
I've thought about either searching for desirable fragments, or trying to simulate where the cuber's fingers would actually be, and what moves they would be ready to make without re-positioning. That plus branching when there are multiple ways to execute something (like when there are half turns that could go either way).
I've just been toying a tiny bit with the latter idea. I kept track of R turns, and considered it a negative to do a U (but not U') if the sum of R turns wasn't 0. Just a first stab, but it seemed to select better than average sequences from a list.

 

[Robert-Y]

I calculated 3916 "unique" LL cases. I just counted the number of repeated cases in Kirjava's huge 2alg LL table and subtracted it from 4128 (43 rows * 96 columns).

Does this sound reasonable? Can anyone back this up?

 

[irontwig]

Sounds about right, should be something between (492/176)*1211 and 4*1211.

 

[StachuK1992]

Evogler,

An old one by badmephisto, works pretty well.
http://www.cubing.net/software/
See "gripper"


I also made an inefficient one shown here
http://ideone.com/mfVk3

I had a better one, but I can't find my code anywhere.

 

[cuBerBruce]

I calculated 3916 "unique" LL cases. I just counted the number of repeated cases in Kirjava's huge 2alg LL table and subtracted it from 4128 (43 rows * 96 columns).

Does this sound reasonable? Can anyone back this up?

Yes, it's exactly correct, if you're including the solved case as a case.