The New Method / Substep / Concept Idea Thread

[Gappo]

Some L5C example solves:

1. R2 F' U' F2 U2 R U2 F' U'
x R' U B2 // V
y R U2 R2 U' F R' F' U R // L5C

12 HTM

2. R F2 R' F R' F' R F2 U
U' R' //V
y U2 F2 U F R' F R U2//L5C

10 HTM

3. U' F' R2 F' U2 F' U F2 U'
x' z U R2 // V
R U' R' U F' U' F U2//L5C

10 HTM

4. F' R2 U F2 U' R' F2 U R
R F' R //V
y U R U' F R U' F R F2 R' // L5C

13 HTM

5. F' R2 U' R U' R' U R2 U2
U F' // V
U' R2 U' F R2 F' U' R // L5C

10 HTM

Avg 11 HTM vs. CLL 14 HTM
Avg of V:2.4 moves

still need some optimize

 

[Duncan Bannon]

Of course the movecount is nuts. One looking is crazy and 1/5 the algs are already made. (TCLL). Might learn this... Half joking half not.

 

[M1306]

Yes, exactly that many.
486 is about as much as ZBLL though, which only a few has fully learned, so learning L5C / 1LLS (one look last slot) probably
has too many algs to be useful.

Except that you can one-look solves, so recognition isn't a problem.

 

[Duncan Bannon]

Except that you can one-look solves, so recognition isn't a problem.

And the algs for this are going to be close to 8? moves less than ZBLL

 

[Hazel]

CriticalCubing proposed LLOB (aka ZBRoux) for those who like Roux blockbuilding but dislike LSE, but what's the best method for those who dislike blockbuilding but want to do LSE? PCMS is the only thing I can think of.

 

[Neuro]

I guess you could do 1/2 cross (DLDR or UFUB) and make sure the other 2 cross slots aren't filled with F2L pieces (alternatively place ULUR or even UFUB edges in those slots) You can do this with either solved centers or M2 I suppose. Then you can do standard F2L and then OLLCP or CMLL, ending in LSE. Not sure why you'd want to do this though, seems very awkward

OLLCP Variant
CMLL Variant

 

[Spencer131]

CriticalCubing proposed LLOB (aka ZBRoux) for those who like Roux blockbuilding but dislike LSE, but what's the best method for those who dislike blockbuilding but want to do LSE? PCMS is the only thing I can think of.

If I were you I would at least consider LMCF, even if it sounds silly. I tried it for a couple weeks and really enjoyed it.

 

[Duncan Bannon]

Some L5C example solves:

1. R2 F' U' F2 U2 R U2 F' U'
x R' U B2 // V
y R U2 R2 U' F R' F' U R // L5C

12 HTM

2. R F2 R' F R' F' R F2 U
U' R' //V
y U2 F2 U F R' F R U2//L5C

10 HTM

3. U' F' R2 F' U2 F' U F2 U'
x' z U R2 // V
R U' R' U F' U' F U2//L5C

10 HTM

4. F' R2 U F2 U' R' F2 U R
R F' R //V
y U R U' F R U' F R F2 R' // L5C

13 HTM

5. F' R2 U' R U' R' U R2 U2
U F' // V
U' R2 U' F R2 F' U' R // L5C

10 HTM

Avg 11 HTM vs. CLL 14 HTM
Avg of V:2.4 moves

still need some optimize

Just did an Ao100 with making a V. Average of 100 was 2.2 moves. Also, does the 486 algs contain CLL or TCLL? Also, do you have an average of the movecount of the algs?

 

[Gappo]

Just did an Ao100 with making a V. Average of 100 was 2.2 moves. Also, does the 486 algs contain CLL or TCLL? Also, do you have an average of the movecount of the algs?

Algs are still generating , but I guess it'll be around 10 moves
CLL and TCLL are not included

 

[SomeRandomZZUser]

If I were you I would at least consider LMCF, even if it sounds silly. I tried it for a couple weeks and really enjoyed it.

Lmcf with l5c algs would probably be very fast when mastered.

 

[Gappo]

Just did an Ao100 with making a V. Average of 100 was 2.2 moves. Also, does the 486 algs contain CLL or TCLL? Also, do you have an average of the movecount of the algs?

I generated lots of algs recently, keep uploading
https://docs.google.com/spreadsheets/d/1-SH47hU3Kqr0zHY0QbnlLMp3jLmAC4N29m1Grb41zIQ/edit?usp=sharing

 

[Duncan Bannon]

I generated lots of algs recently, keep uploading
https://docs.google.com/spreadsheets/d/1-SH47hU3Kqr0zHY0QbnlLMp3jLmAC4N29m1Grb41zIQ/edit?usp=sharing

No Cll or Tcll included.... ouch. Algs are looking pretty good. Glad I got my name in there with the V movecount

 

[Gappo]

No Cll or Tcll included.... ouch. Algs are looking pretty good. Glad I got my name in there with the V movecount

Yeah... but I think you should learn CLL and TCLL before you learn this

 

[Duncan Bannon]

Yeah... but I think you should learn CLL and TCLL before you learn this

I know CLL now. And am finishing up Eg1. Just thought I might try a go at this. Going to finish up Eg and TCLL first I suppose

@WoowyBaby did HD algs. That should cover like 42 if the algs. They are nice algs. It will be when the V has all the corners orientated.

 

[Gappo]

I know CLL now. And am finishing up Eg1. Just thought I might try a go at this. Going to finish up Eg and TCLL first I suppose

@WoowyBaby did HD algs. That should cover like 42 if the algs. They are nice algs. It will be when the V has all the corners orientated.

I might take his algs as a reference (・∀・)

 

[WoowyBaby]

I might take his algs as a reference (・∀・)

I'm glad people are recognizing my stuff! I will work some more on the algs because they can always be better.

But my algs would only cover 6 of the algs (the Stollery-0 set) because the rest of the HD algs have non permuted bottoms, and L5C / 1LLS is permuted bottom V. Soo you can't just take my algs because they aren't very useful for L5C :/

I could help genning some L5C algs if you want, because 486 (or 358) algs is a lot. I want to help out with anything!

 

[Duncan Bannon]

I'm glad people are recognizing my stuff! I will work some more on the algs because they can always be better.

But my algs would only cover 6 of the algs (the Stollery-0 set) because the rest of the HD algs have non permuted bottoms, and L5C / 1LLS is permuted bottom V. Soo you can't just take my algs because they aren't very useful for L5C :/

I could help genning some L5C algs if you want, because 486 (or 358) algs is a lot. I want to help out with anything!

I forgot about the non permuted bottom layer. I suppose 6 algs is okay... Should "we" make a thread for L5C like they did for HD?

I might also be able to help do algs some.

 

[WoowyBaby]

Here's the thread for L5C: https://www.speedsolving.com/forum/threads/2x2-l5c-method.71231/

Don't post in this thread anymore as it's kinda cluttering it up.

 

[WoowyBaby]

So, I want to make a low movecount Skewb method, but first I want to see how low I can get with with Sarah’s real quick:

//WoowyBaby's Skewb Lowmovecount First Idea:
So I noticed people usually take a lot of moves
to solve a skewb by preforming lots of sledgehammers,
and so I want to make a lowmovecount method, but for
now I'm just optimizing Sarah's Intermediate.
Now I know that sledges are very quick so a low movecount method
probably won't be popular for speedsolves, and Sarah’s Int. and Advanced will probs always be the fastest way to go.

Here's my results for reducing Sarah's Intermediate:
AVERAGE: 18.500 MOVES.
Under 20 already.
God’s number for skewb is 11,
and average movecount for optimal solves is 8.3636 (according to Jaap’s page)
Probs could get under 16 or something.

I'm pretty sure I factored in probabilty correctly,
but I need more data on moves it takes for the first layer/side.

Probs ao100 would be good.

////////////////////////////////////
1) Make layer/side - avgMoves= 4.400?
2) Orient/solve other side - avgMoves= 6.600 (15 cases, 11 algs)
3) Permute final centers - avgMoves= 7.500 (12 cases, 3 algs)

Making first layer: (NOT optimal)
I just looked at ten random scrambles
and this is what I got:
1- 5moves 6- 6
2- 2 7- 4
3- 5 8- 3
4- 2 9- 6
5- 6 10- 5
AVERAGE MOVES- 4.400?
As I said before I would need like 100 solves for this number to be accurate.

For the algs I just used Grubiks skewb solver which was pretty slow.
Orienting(optimal algs, I think)
(Its same as sarahs interm. step 2)
(all the “Solved, C [side]” cases are the same but I repeated it so all the cases would be equal probability)
Solved, C Top- 0 (done)
Solved, C Front- 8
Solved, C Left- 8
Solved, C Back- 8
Solved, C Left- 8
Pi, C Top- 8
Pi, C Front- 4
Pi, C Left- 8
Pi, C Back- 4
Pi, C Right- 8
L, C Top- 8
L, C Front- 8
L, C Left- 7
L, C Back- 8
L, C Right- 8
AVERAGE MOVES- 6.600 MOVES

Permuting Last Centers:
(all the u perms are the same but I repeated it so all the cases would be equal probabilty)
U-perm#1- 8 U-perm#5- 8
U-perm#2- 8 U-perm#6- 8
U-perm#3- 8 U-perm#7- 8
U-perm#4- 8 U-perm#8- 8
Z-perm#1- 9 Z-perm#2- 9
H-perm- 8 Solved- 0
AVERAGE MOVES- 7.500 MOVES
///////////////////////////////////

I just want to put this out there before I develop a method to get the movecount even lower.
Also I need to figure out ksolve+ if I want to make low movecount method but I’m no programmer but it doesn’t look *too* complicated so I could igure it out

Tell me if there’s already a low movecount/FMC method for Skewb, beause the wiki doesn’t mention any.

 

[Neuro]

This'd only be worth it if the algs proved significantly faster than what people already do for those cases. I don't know of nay low movecount skewb methods, might be cool to see one though. I guess you might look into improving Sarah's advanced with advanced 2x2 techniques (face, solve the rest with EG 1/2, TCLL variants, etc.)