[Help Thread] Redi Cube Discussion and Help

[EntireTV]

Fun...

Stop Lq posting please

 

[Draranor]

Just ordered one; can't wait to get it! This looks like a puzzle I'll enjoy practicing, so probably gonna try to go for those UWRs

 

[EntireTV]

Just started practicing this and got a really lucky 18.91! Any tips on getting consistently sub 20?

 

[DGCubes]

Just got a couple UWRs! 11.74 average of 5 (pretty lucky; two counting 11s and a counting 12), and a 14.93 average of 12! This thing is really fun...

 

[DGCubes]

Sorry about the double-posting, but I just got a 15.50 ao50 and a 16.08 ao100! Both UWRs!

Just started practicing this and got a really lucky 18.91! Any tips on getting consistently sub 15?

Sorry, I didn't see this post until now. I'm not consistently sub-15 (don't think anyone actually is yet), but I'd recommend working on blockbuilding for the first layer and look-ahead during the second. Also, if you're not solving the top layer corners and the middle layer edges simultaneously, I'd recommend doing that.

 

[Gomorrite]

Just got a couple UWRs! 11.74 average of 5 (pretty lucky; two counting 11s and a counting 12), and a 14.93 average of 12! This thing is really fun...

I would like to see a video. I am curious to see how speedsolving of this cube looks like.

 

[DGCubes]

I would like to see a video. I am curious to see how speedsolving of this cube looks like.

I'll record one soon and post it here.

Spoiler alert: it's a LOT of table abuse.

 

[fabdabs1234]

Just did ~ 250 solves and broke all the UWRs

Single 8.83
avg5 11.62
avg12 12.91
avg50 14.42
avg100 14.69

I managed to drop around 0.5 seconds after modifying my puzzle slightly so that's cool.

 

[DGCubes]

Just did ~ 250 solves and broke all the UWRs

Single 8.83
avg5 11.62
avg12 12.91
avg50 14.42
avg100 14.69

I managed to drop around 0.5 seconds after modifying my puzzle slightly so that's cool.

Daaang, congrats! Any info on the mod? Is it just like regular Florian modding, or something completely different?

Either way, I'm going to need to practice a bit...

 

[fabdabs1234]

Either way, I'm going to need to practice a bit...

I disassembled the cube until only the screws and centre bases were left on, then i got some pliers and cut of around half a centimetre off of the three sticky outy bits(the ends of the traingular centre base) on each base (24 in total). After reassembling there was a lot less catching than before.

 

[I_<3_SCS]

Nice mod.

I disassembled the cube until only the screws and centre bases were left on, then i got some pliers and cut of around half a centimetre off of the three sticky outy bits(the ends of the traingular centre base) on each base (24 in total). After reassembling there was a lot less catching than before.

 

[xyzzy]

1. 6:17.952 B L' b f r b' r b' l B' R' B F' L' R' F

Spoiler

lol redi bld, is this uwr

I am so bad at memo and recall (2:17.548 memo, 4:00.404 execution). Sub-10 memo + sub-25 execution is probably easy for fast BLDers.

Spoiler: method

Edges are just 3-cycles with sledgehammers and sexy moves; I think at most two setup moves are needed in the worst case? Corner twists are done two at a time: solve UFL and UFR, then x' and repeat.

One-twist alg is same as on a pyraminx: F R F R' F R F R' (+ mirrors)

Every two-twist case is 10 moves optimal: X Y X Y X Y' X' Y' X' Y' where X twists the first corner to be solved and Y twists it the "wrong" direction.

Unrelated: Ran 10000 random states through an optimal solver (took a bit over an hour of CPU time), and got this distribution of move counts.

  0:  0
  1:  0
  2:  0
  3:  0
  4:  0
  5:  0
  6:  0
  7:  0
  8:  1
  9:  1
10:  13
11:  84
12:  477
13:  1848
14:  4310
15:  3013
16:  252
17:  1

Average is around 14.04 optimal, with God's number being at least 17. (Assuming there are no bugs in my solver.)

The "checkerboard"/cross pattern is the approximate equivalent of a superflip on a 3×3×3 (every piece is as far from solved as possible + lots of symmetry), but this one can be solved with only 16 moves: (F B l r) (f b L R) (F B l r)' (f b L R)'. If you replace all the moves with ccw turns, then you get the pure 8-twist, which is also 16 moves optimal. I wonder if there's a symmetric ≥17-move pattern.

 

[BABCGA]

Ok, just got an 8.25 single. Its already been updated on the wiki.

 

[Hucklebberry]

I'm here solving a pyraminx in 40 seconds...

 

[I_<3_SCS]

View attachment 8305 Ok, just got an 8.25 single. Its already been updated on the wiki.

Whoa that's fast... and whoa that's a big image

 

[Ghost Cuber]

I just got my Redi Cube the other day. It is a really fun solve and I also found out that it is tensionable. Just take the edges out and split the corners. There are screws under the corners. Five full screw turns seems to be the perfect tension. With the edges out, I put some SCS Galaxy on the ball core itself, as the pieces contact it directly. I also put a combination of Galaxy and DNM-37 on the pieces. Doing this setup instantly dropped my times by around 3 seconds (35-32). I strongly recommend it. I also got a 31.83 AO50 on a stream tonight using the random state scrambler by @xyzzy and I plan to do another AO50 on stream tomorrow or Saturday.

 

[I_<3_SCS]

I just got my Redi Cube the other day. It is a really fun solve and I also found out that it is tensionable. Just take the edges out and split the corners. There are screws under the corners. Five full screw turns seems to be the perfect tension. With the edges out, I put some SCS Galaxy on the ball core itself, as the pieces contact it directly. I also put a combination of Galaxy and DNM-37 on the pieces. Doing this setup instantly dropped my times by around 3 seconds (35-32). I strongly recommend it. I also got a 31.83 AO50 on a stream tonight using the random state scrambler by @xyzzy and I plan to do another AO50 on stream tomorrow or Saturday.

cool I'll do the mod as well then

 

[BABCGA]

Im currently creating "1LLL" for Redi Cube, where you solve the last seven pieces (5 edges + 2 centres) in 1 algorithm. Quite a lot of the cases would be 3-cycles but there will be some long algorithms.

 

[xyzzy]

Im currently creating "1LLL" for Redi Cube, where you solve the last seven pieces (5 edges + 2 centres) in 1 algorithm. Quite a lot of the cases would be 3-cycles but there will be some long algorithms.

I've found that you can usually get really close to optimal just by intuitively matching up the edges to the corners, although an "algorithmic" approach for some cases might be useful. (Maybe the ones where there's no ready-made corner-edge pair?)

 

[DGCubes]

1. 6:17.952 B L' b f r b' r b' l B' R' B F' L' R' F

Spoiler

lol redi bld, is this uwr

I am so bad at memo and recall (2:17.548 memo, 4:00.404 execution). Sub-10 memo + sub-25 execution is probably easy for fast BLDers.

Spoiler: method

Edges are just 3-cycles with sledgehammers and sexy moves; I think at most two setup moves are needed in the worst case? Corner twists are done two at a time: solve UFL and UFR, then x' and repeat.

One-twist alg is same as on a pyraminx: F R F R' F R F R' (+ mirrors)

Every two-twist case is 10 moves optimal: X Y X Y X Y' X' Y' X' Y' where X twists the first corner to be solved and Y twists it the "wrong" direction.

Unrelated: Ran 10000 random states through an optimal solver (took a bit over an hour of CPU time), and got this distribution of move counts.

  0:  0
  1:  0
  2:  0
  3:  0
  4:  0
  5:  0
  6:  0
  7:  0
  8:  1
  9:  1
10:  13
11:  84
12:  477
13:  1848
14:  4310
15:  3013
16:  252
17:  1

Average is around 14.04 optimal, with God's number being at least 17. (Assuming there are no bugs in my solver.)

The "checkerboard"/cross pattern is the approximate equivalent of a superflip on a 3×3×3 (every piece is as far from solved as possible + lots of symmetry), but this one can be solved with only 16 moves: (F B l r) (f b L R) (F B l r)' (f b L R)'. If you replace all the moves with ccw turns, then you get the pure 8-twist, which is also 16 moves optimal. I wonder if there's a symmetric ≥17-move pattern.

Very very cool. Has anyone calculated the amount of permutations yet? I just got 1,571,364,748,800, but my math is probably incorrect seeing as I did it in a couple minutes.

Also, side note: I just got a 14.34 avg50 UWR! I'm pretty close to the other average UWRs as well, but I still don't have a sub-9 single.

EDIT: Just got all average UWRs! 11.01 avg5, 12.74 avg12, 14.08 avg50, and 14.54 avg100. I also got an 8.98 PB single.