L2L4 Documentation ("Finished") and Discussion

[BigSams]

Right handed people never use algs with L moves in.

What do you mean?

 

[StachuK1992]

What do you mean?

That was sarcasm. Plenty of right-handed cubers use L moves. 'Twas a rebuttal in a sense.

 

[BigSams]

Ah I see, can't tell sometimes when it's in writing.
Was't really warranted though. I mentioned a block-on-side variation because there are people that really hate non-dominant side moves and do just about everything with their dominant hand. Kirjava's just looking to pick a fight as usual, probably because I said something about many of his comments being useless a while back. Just because there are cubers with dominant hands that are ok with non-dominant side moves, it doesn't mean there are no cubers who prefer avoiding non-dominant moves. In fact, most fast cubers today who use RUF are faster than RUL. Don't ask me for a proof, it's just something I've noticed. I personally think that RUL has greater potential but even if that comes true, there will still be people who prefer RUF.

 

[cincyaviation]

In fact, most fast cubers today who use RUF are faster than RUL. Don't ask me for a proof, it's just something I've noticed..

Well, if you're going to state it as a fact, then i want proof. You had to notice it somehow? How did you come to that conclusion?

 

[riffz]

Kirjava's just looking to pick a fight as usual, probably because I said something about many of his comments being useless a while back.

No, you just don't understand.

 

[StachuK1992]

So let's get back on topic.
Which variation (original post or FL+FR, ELL) do you like more?

 

[Kirjava]

Kirjava's just looking to pick a fight as usual, probably because I said something about many of his comments being useless a while back.

No, I'm just pointing out stupid things you said because you say them a lot.

In fact, most fast cubers today who use RUF are faster than RUL. Don't ask me for a proof, it's just something I've noticed.

It's great that you can come up with thoughts like that all by yourself.

 

[oll+phase+sync]

Pseudosandwich method.



A layer is more awkward than a 2x2x2.



Why?

Pseudosandwich method. <- not meant as method but as help to simplyfy CO-Step (I haven't learnd all the algs jet)

A layer is more awkward than a 2x2x2. <- 4 against 8 thats not fair

Why? being colorneutral an a 2x2x2 is more common than on a 3x3x3 - because edge handling on a 2x2x2 is much simpler (really) - since edge juggling in L2L4 is algorithmic I think being colorneutral is easier, AFTER I have learned the methode.

@stachu

Yes I need a new username.

 

[Kirjava]

Pseudosandwich method. <- not meant as method but as help to simplyfy CO-Step (I haven't learnd all the algs jet)

I'm telling you that that's what that method is (sort of) called.

A layer is more awkward than a 2x2x2. <- 4 against 8 thats not fair

Then why was it fair when you made the original comparison? To add, a layer is more awkward than a 2x2x3.

More awkward doesn't have to mean higher movecount.

Why? being colorneutral an a 2x2x2 is more common than on a 3x3x3 - because edge handling on a 2x2x2 is much simpler (really) - since edge juggling in L2L4 is algorithmic I think being colorneutral is easier, AFTER I have learned the methode.

Algorithmic = easier to be colour neutral? Since when?

 

[Kirjava]

R D' B R F R F' R U B' F D R' B' R2 B F' D2 R' B' F2 D2 R' L' B'

FBD2L'BR'U2R2U

F L U' B' R F B2 U' B R U' D2 F' B2 U' L2 B D' L R2 B2 L2 U L2 D

LB'RFR2DU2R2U'B'

R2 L' B R2 F' R D' R' U' D F B L' B' D U' R L2 B2 F2 U F2 L2 B2 F'

LRU'D'F2UEL'U2LE' (lolmessy)

R' L2 B2 R2 L2 F2 R' D B2 L R F' U B2 L D' R' L D L U2 R U L R'

L2FL'UF'L2yU2RU'R'U

D' U R2 D2 L' D' B F2 R2 B L' B' D2 R' L2 B' U' R B' D' R2 D' B L R'

R'U'R2ULBy'M'UMy'M'U2M

L B2 F' R U R2 L2 F L R U2 L D' F R B2 D2 B2 U D L2 D' F U R2

B'D'F'R2UF'z2RU'r'U2M2U'

R B D B' F' R2 L D B2 R F' D R' L' F' U L' D2 B' D R2 B' D B L'

BL2D'Ryx'RU'r'UR'U'R'U

L D R U D2 R' L B L2 U2 R' F2 B L2 B F2 U' F2 U' B' U D2 R2 L U

DF2U'L2BL'U2R2URUR'U'

D F' L U2 F' U L R U2 B' R2 F2 B2 D L F2 U2 R2 F' U2 F R' U2 B' D

DR'B'R2DRU'R'UDR2D'

 

[HavoCentral]

I am defiantly putting this on my list of methods to learn.
I might start this as soon as i get better at block building,
I will also learn the ELL version first, then later on the pure version.

 

[oll+phase+sync]

Algorithmic = easier to be colour neutral? Since when?

Intuitiv LL scenario:
I remember my own experience when started with petrus. each time a different top color (mem clear after 6) , then the search for a pair (to do a niklas)... I had a complete color neutral vison of the cube, but was slow like hell.

Algorithmic LL scenario:
Learning CLL by pattern recognition (a lot to learn but all cases are now narrowed down now to 42) I never really trained doing CLL color neutral but I am.
Doing F2L for example I'm not more color neutral than I was before. So I belive fixed alg sets simplify pattern recognition.

A more interessting question might be, how much could be gaind by being color neutral in L2L4 on a average scramble?

Regarding awkwardness : I petrus step 1 and step 2 I have more freedom how to turn, right. But I never mastered to think of an 2x2x3 as one step , wich makes it awkward.

Obviously your reason to call a layer awkward must be a differnt .. wich?

 

[Kirjava]

Learning CLL by pattern recognition (a lot to learn but all cases are now narrowed down now to 42) I never really trained doing CLL color neutral but I am.
Doing F2L for example I'm not more color neutral than I was before. So I belive fixed alg sets simplify pattern recognition.

This is no real justification at all. It's all anecdotal.

Regarding awkwardness : I petrus step 1 and step 2 I have more freedom how to turn, right. But I never mastered to think of an 2x2x3 as one step , wich makes it awkward.

You're just replying to what I said because I replied to you and you can't drop something because you think it makes you look weak.

Try listening to my advice for once instead of trying to correct me, you can learn a lot from me.

I really don't like talking to you. Some of the things you say are odd, others just don't make any sense. It's hard to understand what you're saying a lot of the time.

 

[StachuK1992]

~solution examples~

Thanks!

Hrm, yet another idea (I'd like to find the best concept before dedicating myself to learning something.

First layer
EO
FL + CO
FR + CP
Finish​

EO is 18 cases
"CO" is 59
"CP" is 15
which adds up to 92 cases.

But how much is fin? I'm working on the math, but someone checking it would be awesome.

 

[mr. giggums]

6!/2 for edge permutation / 4 for AUF - 1 solved case = 89 cases

 

[StachuK1992]

Thanks!

 

[oll+phase+sync]

[
EO is 18 cases
"CO" is 59
"CP" is 15
...

I think in a layer (E or U doesnt matter) you can have 4 patterns with an even number of oriented edges 2 with an odd number -> 4*4 + 2*2 = 20 - 1 solved = 19. But I dont believe you care about one more.

There are also cases were doing y will swich from on case into another, but sadly that doesnt reduce the number of cases (Edit: but maybe help magic)

You should be able to preserve an solved edge during EO, right?

And maybe there is some kind of "E2/R2E2R2 - insertion magic" during or with less magic at the end or start of EO to get one more edge solved.

Even if success ratio for getting a solved edge during EO would just be sufficciently high, one could live with 2step LSE - You not if I'm right?

EDIT:
rethinking about EO+1 That would at first reduce the nuber of algs greatly,

but CO must preserve 1 EEdge and don't disturb orientation, algs may get longer
CP must preserver 2 EEedges ...

but there is one more concern:
I was first thinking about EO +1 just solve any edge and we just turn the cube afterwards in a way the edge gets to the slot we prefer ...
but Edges are oriented meanwhile, so we may not be able to do this ..
Instead we must add additional CO and/or CP algs to make up for this

I believe doubling CP cases would be enough.


P.S.
anecdotal knowledge is the source of wisdom

 

[Kirjava]

P.S.
anecdotal knowledge is the source of wisdom

http://en.wikipedia.org/wiki/Anecdotal_evidence

 

[StachuK1992]

Quick update:
1/4 of L2L4 memo'd.

Also, I found the following here

This method is an extension and variation on the keyhole method. There are 4 substeps to be learned, each can be learned individually without knowledge of the others and provide a fast method, or all of them can be learned for the most advanced version of the method. The goal is to learn the 4th substep, but knowledge of all the substeps would allow for the choice of the best startegy based on the given scramble. The strategy changes for each substep, but the overall strategy is to break up the F2L solve into various steps making sure to always perform useful affects on the last layer at each F2L step. This makes the F2L solve longer than a Fridrich style solve, but it drastically shortens the LL solve by building it and the F2L at the same time. The average number of moves for the L2L4 strategy is 53-55, which is about the same as Fridrich. 53 algorithms are required for the most advanced L2L4 version. This method also includes useful variations for the ZB method (see below).

Anyone have any idea what this 53-alg thing is?

Secondary source, currently reading:
http://web.archive.org/web/20080113153901/http://mysite.wanadoo-members.co.uk/cubertscubicle

 

[Cool Frog]

I was reading this this morning, I think im going to learn this also, Where would you go about learning this first?