[Help Thread] Skewb Discussion

[Ranzha]

I decided to do an average of 12 and reconstruct for you all.

In the average, there was a pop. I deleted the solve. Haters come at me~

Scrambling notation will be WCA FCN, though solving notation will be in Rubikskewb notation because it accounts for turns of all 8 corners as well as rotations (and it makes sense!)

Rubikskewb notation:

Spoiler

For those of you who are not familiar with Rubikskewb notation and would like to ease the translation:
Meep FCN users: U R L D becomes B r l b
WCA FCN users: U R L B becomes B r l b
Acubist's method users: t r l d becomes B r l b
Sarah's method users: No change!
Ranzha's method users: F R L B becomes f r l b
1-alg method users: Switch methods, like seriously.

Reconstructions:
1. 5.56 B' R U B' L B R' B L B' U R'

Spoiler

y2 // inspection
f' r l' // Petrus Block
r' f' r // Welder's Mask
f' f // Started L4C, undid
r f' r' f // L4C forcing Pi CLL
y' f' r f r' f' r f r' // CLL

20 FTM / 5.66 seconds = 3.534 tps

2. 4.49 U' L R' U R' U' B L' U L U L

Spoiler

y' x' // inspection
r f r // Petrus Block
y2 f r f' // Welder's Mask forcing B
y' f' r f r' f' r f r' // L4C skip, CLL

14 FTM / 4.49 seconds = 3.118 tps

3. 4.87 B' R B L R L' U' L B L' B U'

Spoiler

z // inspection
l' r' f' // Petrus Block
y r f' r // Welder's Mask forcing R
y2' r' l' r' l r l // L4C+CLL - U Pi F

12 FTM / 4.87 seconds = 2.464 tps

4. 6.45 U L R U R B L R' L' R' L B

Spoiler

x2 // inspection
f r' l' // Petrus Block
r f r f' r' // Welder's Mask forcing R
y2' r' l' r' l r l // L4C+CLL - U Pi F

14 FTM / 6.45 seconds = 2.171 tps

5. 6.99 B L' R B U' R' B' R L U L U

Spoiler

y x' // inspection
r b' r' b // Petrus Block
f' r' f' r' // Welder's Mask
f' r f r' // CLL to force L3C - Ue-2
y x' f' r f r' y2 f' r f r' // L4C (L3C) with lockup =(

20 FTM / 6.99 seconds = 2.861 tps

6. 7.15 U' B R' U' R B' R U' R' L R' U

Spoiler

z2 x' // inspection
l' r f' r f // Petrus Block with some skips
y r // Realise L4Co case
y L y2 f' r f r' f' r f r' R' // L4Co

16 FTM / 7.15 seconds = 2.238 tps

7. 5.63 U L' U L' U' B L' U' B' L B' U'

Spoiler

x' // inspection
r y' r' f' r // Petrus Block
y' r' f' r' f' r f // Welder's Mask forcing B
r f' r' f r f' r' f // L4C skip, CLL

18 FTM / 5.63 seconds = 3.197 tps

8. (9.69) R L R' U B' U' R U' L U R' L

Spoiler

y2 // inspection
r' y' f r' f // Petrus Block
y2 f' r f' r f // Welder's Mask
y' f' r f r' f' r f r' // CLL
r' l' r l r f y' f' r' // L4C

25 FTM / 9.69 seconds = 2.580 tps

9. 5.43 L R' B' U' R' L' B U' B' U R' U'

Spoiler

y' z // inspection
l f r f' // Petrus Block
y r f r f // Welder's Mask forcing R
y r l r l' r' l' // L4C+CLL - U Pi B

14 FTM / 5.43 seconds = 2.578 tps

10. 5.81 L' U L U' B R' U R' L U' B L

Spoiler

y' // inspection
R F' R // Petrus Block
z y' r f r' f' r f r' f' // Welder's Mask maintaining L4C and forcing Pi CLL
y' f' r f r' f' r f r' // CLL

19 FTM / 5.81 seconds = 3.270 tps

11. (3.88) B' U B' U R' B' L' B L' R U' B'

Spoiler

y2 // inspection
r' f' // Petrus Block
y r' f' r' // Welder's Mask forcing R; notice F will solve, thus L4C will skip
f' r f r' f' r f r' // CLL

13 FTM / 3.88 seconds = 3.351 tps

12. 6.66 L R' U L R' L R B' L' B R' L

Spoiler

z y' // inspection
r' f' r // Petrus Block
y' f r' f r // Welder's Mask
f' r f r' // L4C
y f r l' b' y' r f' l' y' r l r' // CLL (optimal peanut)

21 FTM / 6.66 seconds = 3.153 tps

Average turns per solve: 17.167
Overall turns per second: 206 FTM / 72.61 seconds = 2.83 tps

 

[Sa967St]

I need help with Sarah's intermediate method.
Currently i use intuitive 1st layer + 1 look opposite center and corners(non intuitive-fixed alg for each case , ) + U perm(or two) and able to get about 15 seconds.
1. Is intuitive 2nd step better than non intuitive because you can influence centers in someway to get better cases?
2; What should be ideal splits for 10 seconds solve?
3. Can anybody make a good video for showing alg execution , i could not even follow Meep's slow exec video.
4. How fast should the basic sledgehammers and hedgeslammers be? (i saw Sarah's video. I can do a regripless sledgehammer but cant find anything cool enough for hedge).
5. Any other shortcuts/ methods you think working over which can help me get faster. I am very open to learning algs provided they can be performed faster than my current ones(obviously)

You should learn optimal (or near-optimal) algs for Z and H perms so you don't have to do U perm twice. Several of them have been posted in this thread.

1. Whichever is faster for you. I started out with an intuitive approach because it was easier, then I converted to just memorizing cases so I didn't have to think too much during solves.

2. 3-4 for first layer and 6-7 for the rest, for an average solve, I think.

3. I use the Z and H perm algs from http://meep.cubing.net/l5c.html. I can't do them as fast as Meep, but if you'd like I can make a video on how I execute them.

5. There are plenty of shortcuts. It's fun figuring them out. Experiment with sledgies and hedgies and see what you can come up with.

 

[TP]

3. I use the Z and H perm algs from http://meep.cubing.net/l5c.html. I can't do them as fast as Meep, but if you'd like I can make a video on how I execute them.

I would like to see that video.

 

[IanTheCuber]

Skewb is being added as an event in 2014 w00t! Better start practicing for Ann Arbor.

At the moment I use Chris Bird's simple one-algorithm method. I'll probably end up swapping over to Sarah's method.

 

[TiLiMayor]

Im finally making the switch for a LanLan Skewb, and I'm looking to buy a bunch of them..
Where is the cheaper you have seen them listed?

 

[Sa967St]

Im finally making the switch for a LanLan Skewb, and I'm looking to buy a bunch of them..
Where is the cheaper you have seen them listed?

Wallbuys.

http://wallbuys.com/Product/LL-WTS-211J-Turn-Oblique-Pre-packaging-Magic-Cube-Black-5808

 

[TiLiMayor]

Wallbuys.

http://wallbuys.com/Product/LL-WTS-211J-Turn-Oblique-Pre-packaging-Magic-Cube-Black-5808

Huh, now I know why it never showed up while searching.. Thank you..

 

[Sajwo]

Carrot, could you make a video with your sledge/hedge fingertricks?

 

[Izanagi]

So.... how long is an official WCA Skewb scramble? Is it 9?

 

[Lid]

So.... how long is an official WCA Skewb scramble? Is it 9?

So far when using TNoodle I have only seen 9 move scambles.

Spoiler: From jaaps puzzle page

Moves	Positions
0	1
1	8
2	48
3	288
4	1,728
5	10,248
6	59,304
7	315,198
8	1,225,483
9	1,455,856
10	81,028
11	90
Total	3,149,280

As you see upto 9 moves covers 97.42% of all states.

edit: I see that TNoodle-0.7.12 still hasn't been updated (to include Skewb), I've been using TNoodle-skewb-UNOFFICIAL btw.

 

[Methuselah96]

4b3) Specification for the scramble program: An official scramble sequence must produce a random state from those that require at least 2 moves to solve. The following additions/exceptions apply:
4b3c) Pyraminx and Skewb: The (random) state must require at least 7 moves to solve.

Sarah Strong
12/25/13
I suggest ≥7 for Skewb. That would filter 2.2% of the scrambles, which is more than how much is filtered for 2x2x2 and Pyraminx, but since Skewb is deep-cut and short solutions are harder to see, it seems like a better choice than ≥6, which only filters 0.04% of the scrambles.

It seems that it should generate any number of moves from 7 to 11.

 

[Izanagi]

So far when using TNoodle I have only seen 9 move scambles.

Spoiler: From jaaps puzzle page

Moves	Positions
0	1
1	8
2	48
3	288
4	1,728
5	10,248
6	59,304
7	315,198
8	1,225,483
9	1,455,856
10	81,028
11	90
Total	3,149,280

As you see upto 9 moves covers 97.42% of all states.

edit: I see that TNoodle-0.7.12 still hasn't been updated (to include Skewb), I've been using TNoodle-skewb-UNOFFICIAL btw.

I see. Thanks.

4b3) Specification for the scramble program: An official scramble sequence must produce a random state from those that require at least 2 moves to solve. The following additions/exceptions apply:
4b3c) Pyraminx and Skewb: The (random) state must require at least 7 moves to solve.

Sarah Strong
12/25/13
I suggest ≥7 for Skewb. That would filter 2.2% of the scrambles, which is more than how much is filtered for 2x2x2 and Pyraminx, but since Skewb is deep-cut and short solutions are harder to see, it seems like a better choice than ≥6, which only filters 0.04% of the scrambles.

It seems that it should generate any number of moves from 7 to 11.

Oh, so 7-11 would be ideal for practice solves then?

 

[Methuselah96]

I see. Thanks.



Oh, so 7-11 would be ideal for practice solves then?

Are you wondering about the number of moves you should use for hand scrambles or computer-generated scrambles?

 

[Ranzha]

I see. Thanks.



Oh, so 7-11 would be ideal for practice solves then?

I practice with qqTimer set to 12 moves. If I hand-scramble, I use 15.

 

[Methuselah96]

I practice with qqTimer set to 12 moves. If I hand-scramble, I use 15.

Prisma Puzzle Timer gives optimal scrambles.

 

[Izanagi]

I practice with qqTimer set to 12 moves. If I hand-scramble, I use 15.

Thanks, 9 scramble length felt weird lol. Hence the question.

Are you wondering about the number of moves you should use for hand scrambles or computer-generated scrambles?

I was wondering about both, actually.... Does it matter, really tho? I did a few solves on 9, and a few on 15.... Nothing really seemed all that different to me. Sorry for the confusion, but I am as of now.

 

[MatejMuzatko]

Anyone else uses a *cough* table abuse while doing some algorithms?

 

[TiLiMayor]

So besides prisma, which other timers offer optimal scrambles?

 

[Carrot]

I practice with qqTimer set to 12 moves. If I hand-scramble, I use 15.

I use 15 moves with qqTimer

 

[Izanagi]

I use 15 moves with qqTimer

Carrot y u no upload fingertrick video? XP