Group solving, and case reduction.

[Jorghi]

Hahah xD I would never learn that. I was talking about ZZ's EOLine. But I think I will stay with Cfop.

The reason I made this thread was to see if the ZZ F2L has a low enough algorithm count to decrease the multi-slotting algorithms substantially. But I guess the answer was found.

 

[jskyler91]

Hahah xD I would never learn that. I was talking about ZZ's EOLine. But I think I will stay with Cfop.

The reason I made this thread was to see if the ZZ F2L has a low enough algorithm count to decrease the multi-slotting algorithms substantially. But I guess the answer was found.

Yea, honestly, it doesn't really and eoline is not very effective without full ZBF2l. You should probably try and work on learning full pll (I think you said you were only 4lll) as that will decrease your times and is only a few algs.

 

[uberCuber]

Oh? You sure? Sounded like no mirrors to me.

My wording on mirrors earlier in this thread was probably not worth listening to. Let me rephrase my thoughts into something that will actually work:

Disregard the idea of mirrors for a sec. During a CFOP solve, every single one of the 42 F2L cases (one being the solved case) can occur when the slot is in the FR position. During a ZZ solve, 21 cases (one being the solved case) can occur when the slot is in the FR position. This is a fact, because as I pointed out earlier, the other 21 cases in FR position cannot be solved with a y/d/F, none of which are necessary in ZZ. This remains true for any given slot; there are 42 CFOP cases for the BL slot, and 21 ZZ cases for the BL slot. This is the entirety of the point I ever wanted to make: for each and every slot, there are twice as many possible cases in CFOP as in ZZ.

To address the mirrors issue a tiny bit: You are trying to say that the mirrors in ZZ add to the alg count, but at the same time do not add to the 42 cases of CFOP because some of those 42 are already mirrors. Just look at it in the simplest way possible: Take case #1. If the slot is in FR, you will solve it either R' F R F' or U R U' R'. If the slot is in BL, you would solve it with L moves instead. Since you said you said you count mirrors as different algs, that is two different algs for a single case out of the 42 cases of CFOP (and of course, you could make it 3 by solving case #1 in the FL slot with F' L F L'). This kind of thing works for the rest of the cases as well, meaning that the number 42 goes up in CFOP just as the 21 goes up in ZZ. Ta-da, the difference between CFOP and ZZ is much more than 2.

If I failed to address something that I meant to in this post, by all means point it out.


EDIT: Wherever I mentioned the numbers 42 or 21 in this post, that includes the solved case. So really it is 41 cases that actually need an "algorithm" in CFOP, and 20 in ZZ.

 

[Jorghi]

Yeah, 1 solved pair, 19 cases + mirrors, and a total 3 of symmetrical cases. I already knew all about the edge orientation and things, but I thought the wiki's were ignoring mirroring.

So.. Because the Edges are all orientated in zz, that would mean that the multi-slotting algorithms would be much lower than normal Fridrich F2L?


With basic multi-slotting there are 82 algorithms, http://www.cubeloop.com/php/cube/expert.php?chapter=f2l&subchapter=multislot

If you used ZZ, you could maybe lower them.

 

[uberCuber]

Yeah, 1 solved pair, 19 cases + mirrors, and a total 3 of symmetrical cases. I already knew all about the edge orientation and things, but I thought the wiki's were ignoring mirroring.

So.. Because the Edges are all orientated in zz, that would mean that the multi-slotting algorithms would be much lower than normal Fridrich.

I honestly have no idea what the wiki's standpoint is on mirroring, or if it's actually consistent or not, simply because of the fact that it was written by so many different people.

As far as multi-slotting, the number definitely will be lower than Fridrich. I'm not sure if it is actually low enough to be worth trying to learn all of the cases, but a relatively small set of multi-slotting tricks would in theory* help more in ZZ than it would in Fridrich.

*I say "in theory" because in general, ZZ solvers use intuitive blockbuilding, making it an even less-algorithmic approach than Fridrich. But feel free to go ahead and prove that algs can be useful

 

[Pyjam]

ZZ is my main method. I don't use alg, except the very basic ones like R U R', and some for a couple of special cases, but no more than 3 or maybe 4.

Here are two samples I posted recently on another forum. I don't really use alg, instead I recognize patterns. The second one has an heavy use of multi-sloting.

SCRAMBLE 1 : B L2 R2 D' L' D B2 D2 R D B2 D2 B2 U D2 B F R' F2 B' L D B' D2 B2
L' F' R' B R B' (6)
R2 U L R2 U L2 R2 U2 R' U' L' (11)
U R U' R U' R' U R2 U R' U R U R' (14) = 31 (2 stages)

SCRAMBLE 2 : R2 U2 L' U' R2 U' R' F2 U2 F' R2 L' F L2 D2 F2 B' R' F' R B F2 D R2 U2
(L' B' R F U F) (6)
(L U R2 U R' U L2 R U R') (10 for 3 slots)
(U' L U2 L' U L U' L' : this is an alg !) (8) = 24 (2 stages)

 

[Jorghi]

ZZ is my main method. I don't use alg, except the very basic ones like R U R', and some for a couple of special cases, but no more than 3 or maybe 4.

Here are two samples I posted recently on another forum. I don't really use alg, instead I recognize patterns. The second one has an heavy use of multi-sloting.

SCRAMBLE 1 : B L2 R2 D' L' D B2 D2 R D B2 D2 B2 U D2 B F R' F2 B' L D B' D2 B2
L' F' R' B R B' (6)
R2 U L R2 U L2 R2 U2 R' U' L' (11)
U R U' R U' R' U R2 U R' U R U R' (14) = 31 (2 stages)

SCRAMBLE 2 : R2 U2 L' U' R2 U' R' F2 U2 F' R2 L' F L2 D2 F2 B' R' F' R B F2 D R2 U2
(L' B' R F U F) (6)
(L U R2 U R' U L2 R U R') (10 for 3 slots)
(U' L U2 L' U L U' L' : this is an alg !) (8) = 24 (2 stages)

WHAT! 10 for 3 slots? How advanced is that multi-slotting method?? How many algorithms? I really don't think triple slotting would work, that would require tracking so many pieces while ignoring other ones. But I think it sort of makes up for it by doing 3 at a time.

How many algorithms are in the 'heavy' triple zz multi-slotting?

Edit:"MultiBlocking"

Maybe for the first two layers you could solve entire sides using algs. The left f2l block, then the right f2l block. The cases could be, blocks: (Left, Right, Back, Front). That way you would only have to focus on 4 pieces.
This would be more manageable than the CFOP F2L Multi-Slotting because it would have half the number of cases(approximation) for solving 2 "slots" at a time.

About 41 Multi Slotting(2 slots) algorithms, similar to the amount of algorithms for CFOP F2L(1 slot).
The hard thing though would be tracking the 4 pieces to slot/block. But you would only need to do that step twice which would lower the "drag".

 

[Cubenovice]

WHAT! 10 for 3 slots? How advanced is that multi-slotting method?? How many algorithms? I really don't think triple slotting would work, that would require tracking so many pieces while ignoring other ones. But I think it sort of makes up for it by doing 3 at a time.

How many algorithms are in the 'heavy' triple zz multi-slotting?

Edit:"MultiBlocking"

You really don't get it don't you?

 

[Jorghi]

You really don't get it don't you?

What? That multi-slotting sets up the next pair to be inserted and can be done intuitively using different insertion methods like what Yu Nakajima did??? Or?

 

[tozies24]

What? That multi-slotting sets up the next pair to be inserted and can be done intuitively using different insertion methods like what Yu Nakajima did??? Or?

There are no slots in the examples posted above. He just solved the first two layers. I think you should stop worrying about this kind of stuff and either switch methods to where you can be more efficient or just stay with Fridrich and just deal with its inefficiencies.

 

[Pyjam]

How many algorithms are in the 'heavy' triple zz multi-slotting?

None !

Don't you see it was pure luck? I just looked at the pieces and took advantages of their good positions.

ZZ use block-building for the first 2 layers. With a good look-ahead, it's sometime easy to prepare 2 (or 3) pairs before insertion.

 

[Jorghi]

Okay. I guess anything is done in F2L after all. But now that I take the time, multi-slotting would be very difficult to lookahead for when you start moving up to 8 TPS solves. But maybe if people stop focusing on ZBLL which might shave 1 second from last layer time(If your OLL/PLL is 2 seconds in execution+recognition), and focusing on F2L(which is much more time consuming) and make their look ahead techniques to use by learning how to multi-slot(zz xD) then they could cut a lot of time from f2l! Why Slow DOWN using lookahead when you can just use recognition for normal f2l pair searching. Slowing down while looking ahead just seems to lower the potential gain.

But ITS VERY hard to go at high TPS and look ahead 3-4 TIMES for f2l pairs!

MultiSlotting(2 Pairs - 3 Pairs) = (Makes most use of Look ahead I think now, instead you only need 2 looks, and the second look would be much easier.)

Cross
2 Look F2L - First 2 Sides
1 Look OLL
1 Look PLL

CFSOP(Fridrich based - Shows how CFOP sucks!):
Cross
First 2 Sides (82 Cases)
Orientation Last Layer (57 Cases)
Permutation Last Layer (21 Cases)
160 Algorithms


When people are addicted to cubing and it becomes a life job and war comes in cubing:


EFOP(The best method, ultimate method ever.):
EOline - Intuitive
First 2 1x2x3 blocks each in 2 looks (Near 38-41 cases) - Partially Intuitive
OLL/PLL(28 Cases)

Once you learn 69 Algorithms you are a master of cubing and nobody can stop you.
There are special cases for F2S in when you do the slots diagonal order. But this is still included in the (near) 41 cases.
Edit:
This is random but I just got 2 Pll skips in a row lol

 

[Cubenovice]

Lol if you solve EO line and (first?????) 2 sides you have solved the full cube...

 

[Jorghi]

I mean First 2 Sides of F2L.

F2L is 2 looks. First 2 Sides sounds better. I'll chance the name to solving the first 1x2x3 blocks.


I think the CFOP version would be of more use if you were color neutral though.

 

[Cubenovice]

Calling it the first 2 1x2x3 blocks implies that you will be solving another 1x2x3 next...
If you do this you still have a chance of having two twisted corners left ;-)

You do realise "your" EFOP is still just ZZ?

 

[Jorghi]

Once you learn 69 Algorithms you are a master of cubing and nobody can stop you.
There are special cases for F2S in when you do the slots diagonal order. But this is still included in the (near) 41 cases.
Edit:
This is random but I just got 2 Pll skips in a row lol

I guess its a better variation of ZZ

 

[Cool Frog]

The point here Jorghi is that ZB limits the amount of cases you will encounter and the ones it leaves behind are dramatically easier. May I ask what made you switch to ZB? Were you stuck in Fridich or were you just interested in learning it? Because, from what i know, ZB is going to be harder to get faster at until you learn a great deal of algs. Plus ZB requires a lot of look ahead, which you have mentioned you dislike.

I am not saying that you shouldn't experiment, but you might get frustrated if you switch now as your times will go down for a while. ZB is one of those "I want to be faster in a few months or even years" type of things. If I were you I would concentrate on Fridich for a bit longer and get better at look ahead.

http://www.speedsolving.com/wiki/index.php/ZB

:|

 

[StachuK1992]

Please, post some example solves

 

[Cubenovice]

I guess its a better variation of ZZ

That is a pretty bold statement to make...

But what is actually different for the F2L stage?
In ZZ you typically freestyle through F2L so all that you can come up with is already being done.
Did you even propose anything new???

 

[tozies24]

CFSOP(Fridrich based - Shows how CFOP sucks!):

So you have been cubing for less than two months, and you have now stated that CFOP sucks and you take ZZ and try to play it off as a new method.
Do you even know full OLL or full PLL?

Also, I would like to see some example solves, too. I want to see how efficient your solves are compared to regular CFOP.