Fewest move challenge for the 5x5 cube

[adragast]

Hello !

Sorry to disturb you but I had a very stupid question. Has anyone ever tried to solve a 5x5 cube with a very small moves count ?

It seems to me that the reduction method uses really a lot of moves (we basically have to add the full 3x3 solve moves count at the end...).
Is there any Petrus-like method for 5x5 cube ? What would the skeleton of a 5x5 cube look like ?

Regards,

Kiyoshi

 

[masterofthebass]

K4 would probably be a little less moves. I wonder how much straight commutators would be.

 

[Johannes91]

I've experimented with a few block building methods and they work quite nicely, especially on even cubes. I don't have my records at the moment, but IIRC, good solves for 5x5x5 were below 200 moves.

 

[masterofthebass]

Ok... just did a normal solve. reduction in 162 moves and 3x3 in 63... This was just one of my normal solves, and if I tried to minimize moves I could probably cut off a bunch more. The count was done in HTM.

 

[Mike Hughey]

Straight commutators would be outrageously many moves for the center pieces. But commutators for the edges might help - that would basically cut out the 3x3x3 solve part with what should be a similar number of moves (except for the corners).

 

[Kenneth]

I average just about 200 moves using my, earlier, at start, Roux alike, now, coloumns first method (the two are about the same but has got the stages in a diffrent order). After first blocks + corners I compleate F4L and end using my 4-step ELL.

My FMC record is 172, try to beat that boysngals =)

masterofthebass: K4 won't solve MES, there are no MES on a 4x4x4, better use K5

 

[Mike Hughey]

After thinking about this, I realize now that straight commutators probably won't help on the edges either relative to reduction. In another thread we were analyzing both Arnaud's and the bigcubes.com methods, and the typical number of moves for matching the edges for both is somewhere around 70 - 90 moves. To solve just the wing edges with straight 8-move commutators, assuming two pieces solved per commutator, would take 8 x 12 = 96 moves. Figure on a little optimization there, and you still would wind up with taking as many moves to solve the wing edges in place as you take to match up the edges. Since a CFOP then is just 50 - 60 moves for the 3x3x3 part, it's unlikely that straight commutators would compete with reduction. Some sort of optimizations would be needed in order for it to even be as few moves as reduction.

Kenneth, that's an impressive record!

 

[masterofthebass]

I'm pretty sure that the method is still called K4... K5 is really just a name people give for the adaption of K4 to the 5x5... the method is still called K4 though.

 

[Kenneth]

And I'm pretty sure I have seen Thom "Kirjava" Barlow using the term "K5", but I can be wrong.

Easiest way to save turns is: build two opposite centres, use the free slice to pair 6 edges (no restoration turns needed), place these edges at RB/RD/RF and LB/LD/LF after you compleate them. Then build the rest of the centres using M/m/r/l + U turns. From there continue as normaly.

 

[masterofthebass]

After just talking to Thom, he now calls it KP, Kirjava Pembo. That's his name for the general method, but K4 and K5 are the optimized versions for their respective cubes.

 

[Kenneth]

Ok, thank's. My method is pretty much in the same style as he's and we invented them in the same time (1.5 years ago). But I'm pretty sure I got shorter LL-algs on average than he got (if he's not using my algs of course =)

I just made a solve counting the turns, got 206 STM but did an error in ELL and had to redo a step, a 10 turn commutator so otherwise 196 STM, a compleatly normal solve (exept the error).