Random Cubing Discussion

[5BLD]

Hm... What kind of cube-pranks do you know? For example, I just found a way how to play a trick on a non-cuber (or not-so-experienced cuber):

Perform this on a solved 4x4: u2 M2 Uw2 M2 U2 M
Cuber: "If you can solve this for me in a few seconds, I pay you a beer!"
Non-cuber: "Yeah, free beer!" (Get the cube) "Wait, what?" (Doing R L R' L2 R etc.)

Another prank (whick I actually use... occasionally... sorry) is the next one:

Cuber: "If I scramble the cube with only ONE move, can you undo it?"
Non-cuber: "Sure!"
Cuber: (Execute R'UR' or a similar fingertrick - it is one move after all... even in its sound) "Your turn."
Non-cuber: (blinks)
Cuber: (Do the inverse with the same speed)

Assemble a ball core 4x4 without the sphere on the inside. Leave it on a table at work/school. Make the first person who touches it feel really guilty as you walk in and pretend to be furious.

Also, you can do the same thing with 3x3, with one turn from solved. This even makes noncubers tempted...

 

[Tao Yu]

Do R'U2RUR'zR2UR'DR (J perm without the last move)

Put your thumb over the FDR sticker and cover the BRD and BR stickers with your fingers to hide the J perm
I now looks like it needs only one move to be solved lol

I do this all the time lol

 

[5BLD]

Twist a corner then scramble, watch a sexy method user mess up at the very end...

 

[Julian]

1. Scramble a sq-1 with / 6,6 / 2,-2.
2. Hand it to them so they see F and R faces (or B and L).
3. Enjoy.

 

[Renslay]

Another one, which I used to fell of. :/
Put three Floppy cube on each other, make it look like a 3x3x3. Leave the upper one rotated (like a "U2"), and wait for the cubers. Someone will want to solve it...

 

[evogler]

In what way would this be different from any of the move-optimal solvers that already exist? It is really not that difficult to find optimal solutions in regular inspection.

I'm thinking about the process. If I try to describe my own cross-solving, I might describe it as "spotting familiar patterns that I know how to solve efficiently, then looking for how I can involve the remaining pieces to get them solved with a minimum of additional moves, and trying out 1 or 2 or occasionally 3 of these basic approaches". Even if a thorough brute-force search of all possible sequences of moves yields the same solution for the computer, we've gotten there by very different means.

It interests me to try to get a computer to do it more like I do, recognizing patterns and constructing a single solution. More out of curiosity than for any expected practical benefit. I'm curious, for example, how many rules you would need to get it to solve well. I don't really have a sense in myself of how much information my cross-solving process uses, and reconstructing my thought process in software seems like a way to find that out.

I guess conceivably it could be actually useful for improving at more challenging "intuitive" steps (e.g. xcross). Although I think new cubing ideas are usually not as useful as their inventors would wish.

 

[5BLD]

A good start may be to get it to solve two edges optimally and get it to find optimal inserted moves to solve the other two.

 

[qqwref]

Perform this on a solved 4x4: u2 M2 Uw2 M2 U2 M

= Uw2 M2 Uw2 M


I remember a cubing prank with a Square-1 I once did. The guy was an experienced solver so he was pretty familiar with his color scheme. So, I swapped the corner caps and the left and right center caps, and then solved it back to the mirrored color scheme. Of course, because of the way I did it, there was nothing wrong with the stickers or anything. Then I just had to wait until he tried to do a speedsolve and got confused

 

[uberCuber]

R2' U2' F R2 F' U2' R' U R'

An F2L edge flip that I can actually almost sub-1

Why did I not discover this a long time ago

 

[Escher]

R2' U2' F R2 F' U2' R' U R'

An F2L edge flip that I can actually almost sub-1

Why did I not discover this a long time ago

It's been on opticubes for years

 

[uberCuber]

And that's what I get for not ever really looking at any F2L algs :'(

 

[evogler]

A good start may be to get it to solve two edges optimally and get it to find optimal inserted moves to solve the other two.

Ya, that sounds reasonable.

 

[mrpotatoman14]

http://www.everbuying.com/product180538.html f*** yeah

 

[antoineccantin]

For all those people interested in clock, I've worked hard to make this great video for y'all!

Thanks to this amazing technique, I now have a 9.61 avg12 and 8.58 avg5!

(this video is a response to Womack's how to sub-20)

 

[bobthegiraffemonkey]

P = r2' F2 r U2 r U2' x U2 r U2' r' U2 r U2 r2' U2' x'.

F2 U' R U2' (P x) R' 3f (R 3l)
R' U R2 P U2 R2' U' R
R U' R2' U2' P R2 U R'
F' R U R' P U2 R U' R' F
R' U2' P R
2R U2 2R U2 2R' U2 2R U2 2L' U2 2R U2 2R' U2 3r' R2 2r' U2 2R'

When I learned that OLL parity alg recently, I experimented a little. Does anyone think this is a good idea? It could probably be improved a little and feel free to change the algs to ones you prefer which do the same thing, just throwing the concept out there. Idea is for OLL parity: OLL alg to leave one flipped edge, OLL parity + CP, PLL. I reckon I'll use this in conjunction with my 4x4 1LOLL system for some cases.

 

[qqwref]

4x4x4, WCA notation:

r U r U r' U' r' U' r U r U' r' U r' U r U2 r'

Unfortunately, I'm still having trouble solving <U,r>...

 

[Kirjava]

just wanted to post something, nothing special

 

[Julian]

Just realized that shooting to RBD with OP corners (R2 U' R' U' R U R' F' . R U R' U' R' F), which is R F' setup cancelled into Y-perm, is also the standard T-perm wrapped around the "."

 

[JonnyWhoopes]

Cubers are freaking obsessed with initialisms and acronyms. The most popular method is an acronym, with initialisms within it. Even our governing body has a sub body with an initialism containing the initialism of the main body within it.

 

[cubecraze1]

any one tried a triathlon sort of thing? MO3 with 3 different methods.
To do this the methods would probably be:
CFOP
Roux
Petrus or ZZ?
Results?