• Welcome to SpeedSolving.com — the world’s largest puzzle community!

    You’re currently browsing as a guest, which means you have limited access to discussions, resources, member profiles, and community features.

    Join 50,000+ cubers and puzzle enthusiasts from around the world to ask questions, share solves, improve faster, and be part of the community.

    Registration is free, fast, and easy — create your account today.

    Already a member? Log in here to hide this message and start participating.

ZZ speedcubing method

The only hard-seeming part to me is the first step, and that gets better with practice. Recognition's definitely no worse than ZB, and doesn't seem to be too much harder on top of the recognition needed for COLL cases.

Do I think ZZ is better than ZB? Daunting number of algorithms notwithstanding, I think ZB is still faster, at least in theory. For the last layer, it comes down to the finger-friendliness of the ZZ algorithms vs. those of ZB, and how the time for recognition of phasing + ZZLL combined with the execution time compares to that of ZB.

As for the F2L, I don't know if there is enough data to even make a comparison. If you can get the no-rotation-after-EOLine thing down, then ZZ might actually end up the same or faster than ZBLL.
 
Well ZZLL is just a subset of ZBLL.. the only difference is if ZZLL includes more good ZBLL's then bad ones.
In the end I think it comes down to the persons preference and what they are using it for. For me ZZ is better cause I do OH for 2H ZB might be better. Tho IMO the number of algs required for full ZB is a bit much.
I've been doing some solves and phasing isnt really that bad. EOLine causes me a lot of problems and often I just ende up getting a line and then doing the last 2-4 Petrus style. Also my block building isnt very fast. I could do it Fridrich style very fast but I would end up wasting moves. and then Phasing at the end was much easier then I expected it can get to be very fast as its just like 2moves average that are 2-gen.
The method seems like it has potential but it is daunting learning a completely new method when I am already so fast with Fridrich especially when there arent many list of the ZZLL algs...
And so for me ZZ is daunting enough ZB is just way out there.
Edit: In retrospect looking at when I was learning Fridrich I thought I would never be able to get my cross in 7 moves and see it in preispection. Now that part is pretty much instant. Maybe I should just learn it and get it over with cause unless something comes out better its my dream method with all the right mixes, LUR, part blockbuilding and a 1LLL.
Edit2: I have just noticed one thing. I was doing an F2L algorithm because I had a nasty case for my last pair. I know when I did Petrus for like 2 weeks my last pair was often not inserted fridich style as I woud think would be required for phasing. So anyway since I was doing the algorithm I missed the opportunity to phase how would this be avoided? loads of moves would have to be added to both inefficiently build the pair and then to phase.
 
Last edited:
I still don't understand the phasing step. How can you just switch 2 LL edges? Do you do it while doing the last 2x2x3 block? And even so, I still don't see how that would reduce the number of LL algs to learn...
 
You do it before inserting the last pair.

It reduces cases because now you have 2 last layer edges solved.
 
As I understand it, what you're doing with the phasing step is essentially solving two cross pieces that are across from each other. This leaves the other two either solved or opposite of their required place.

You also have all of the edges oriented. With those 2 pieces solved, the number of cases is reduces since, you know, the pieces are solved.
 
scramble: RUR' U' RUR' U2 RUR' U'

solve: U2 RUR'. normally you would do U R U' R' next. but then you will not have 2 across edges solved.
so instead you will trace the (relative) permutation of the edges.

then you see you don't want to insert like URU'R' but like:

UR U R' U'R U'R'
OR :
U2 R U' R' U R U2 R'

this is a nice start but 12 moves is a lot,

but if you can trace the permutation of the edges before connecting the corner/edge you could train yourself to see: ( perform just after scramble, pair is not connected): R U'R' U2 R U R' (7)

another example:
scramble: RUR' U' RUR' U2 RUR' U' + R U R'U R U2 R'

solve: U R U R' U RUR'
here you can see that UF and UB are solved. however, UR and UL are solved as well!! the only goal is to solve 2 edges.
now perform a U turn. you can see the edges are "solved" UB is always opposite of UF, and your BU and FU color is opposite as well.


http://fmc.mustcube.net/results.php ( fewest moves challange)
Zbigniew Zborowski has a lot of zz solves you can watch. take a look at the other rounds as well.

Remi
 
Some statistics for solving F2L in <R,U,L>:

Method 1
a) Solve DL and DR edges: average 3.47 moves
b) Solve four pairs: average 25.71 moves
=> average 29.18 moves

Method 2
Same as method 1 except don't start with DL and DR but start both left and right side with a 2x2x1 block
=> average 24.64 moves

Both methods tested with 100 solves, always choosing the shortest current subgoal to solve.

For comparison: "normal" Fridrich solving of the four pairs takes 23.03 moves. So solving four pairs in ZZ takes *more* moves. Although if you allow D turns, it drops a bit:

Method 1
3.34 + 24.38 = 27.72 moves average

Method 2
24.41 moves average

(highlighted numbers are for solving four F2L pairs in different ways, so they're somewhat comparable)
 
Last edited:
Hmm is it really more moves? I was thinking they would be about the same.
When I do ZZ I think I do method 2 tho. I solve a left 2x2x1 block then insert pair then right 2x2x21 then phasing and last pair so I use only one more move average not bad. I think the lack of cube rotations makes up for this one move. Plus how does the move count come in when you are doing an optimal cross and an optimal EOLine. Like if you started from completely scrambled to F2L done in fridrich and then completely scrambled to F2L plus EO in the ZZ method.
 
Lofty said:
Hmm is it really more moves? I was thinking they would be about the same.
When I do ZZ I think I do method 2 tho. I solve a left 2x2x1 block then insert pair then right 2x2x21 then phasing and last pair so I use only one more move average not bad. I think the lack of cube rotations makes up for this one move. Plus how does the move count come in when you are doing an optimal cross and an optimal EOLine. Like if you started from completely scrambled to F2L done in fridrich and then completely scrambled to F2L plus EO in the ZZ method.
Click to expand...

You'd also need to take into account the already-oriented top edges.

Edit: I've been practicing ZZ-VH, since I don't want to learn over 100 cases for the last layer just yet. 40 COLL cases is good enough for now. Anyway, the nice thing about this is that I have no need to insert a certain slot last, or even do them in any order whatsoever.

Could someone who is efficient at F2L take an average of a couple of solves doing the EOLine and then filling in whatever happens to be efficient without caring for the order in which the slots/blocks are done? I'd do it myself but my move count is very high, even when I'm going slow and trying for the least moves I can do.
 
Last edited:
Ah, ok thanks yurivish and cubeRemi. Didn't know "switch UF and UB" meant to solve them...
So only UF and UB are solved, or any 2 edges solved(or 4)?
 
Last edited:
2 and guys I made a video tutorial on ZZ so I'll post that if you want (though it might be a bit long for youtube (over 10 mins) so I'll make 2, 5min vids. It has my little tricks and tips on phasing and how I block build and stuff ok? I will do this if you guys answer me 2 questions....
1. how can I put windows movie maker things on youtube?
2. how do I get video from a camera which uses tape onto the computer?
 
Wow! Very nice algs! These must have taken a long time to generate.
Are they all optimal tho? I know of some cases that are 8moves but you only have a 18qtm for this case...
Edit: Nvm, i forgot to take into account that the corners must be correct relative to the edges not just the edges. So tho the COLL case is 8 moves this one is not.
btw i think most people count in htm not qtm, but im sure it would be a lot of work to add it now so its not a big deal.
 
Last edited:
fanwuq said:
Just did a few solves of
F2L with Oriented edges of LL:

Fridrich:

36, 37 (OLL also done), 36, 36, 40, 27

Petrus:

32, 32, 20 (whole solve in 30 moves, the ending was a COLL. This is tied as a PB, not as lucky as my other PB which was Heise with LL skip. This was 6min31seconds. http://www.ryanheise.com/cube/record.cgi?db=single&name=WF+30+moves I screwed up the LL badly at first, got back to the same position, then found the COLL on Jason Baum's site, and just did it), 32, I lost count-- low 20s (also amazing block building, LL was some easy OLL and PLL with nice cancellations. Total was in high 30s.)

ZZ: 45 (worse than usual, EO line in 14 moves, ugly F2L), 42 (EO line in 9), 26 (EO could have being done in 7, I continued with block building first, I think this might be a better approach.), 38 (EO line done in 6), 31 ( EO line could be done in 8, blocks first.)

All Fridrich F2Ls about 1min, Petrus are about 1-2 min.
ZZ also about 2 min each.
Click to expand...

Petrus ftw! After Heise.
 
I have a question about phasing.
Prior to ZZ the main blockbuilding I have done is in my attempts at the Petrus method. Often a case is encountered in which the standard pair-up/insert format is not used as is seen here: http://lar5.com/cube/s4b.html
When you encounter cases like these it would seem to be able to phase properly you would have to throw efficiency to the side and add moves to take apart the pieces, put them together properly and then phase them. Another alternative I was thinking of is phasing algorithms for certain nasty cases when it is not a standard pair.
 
Last edited:
What do you guys think would be a better idea for ZZF2L? Should I use regular F2L, or should I use blockbuilding? I know that they can both get very fast with practice, but in terms of move count etc., which would be better?
 
You must log in or register to reply here.

Similar threads

Caleb Valenzuela
How to make a “Super Duper Unique” Speedcubing Method (for all events)
Replies
1
Views
481
Jared Burkhart
Jared Burkhart
JackTriton
New 3x3 Method: (Named) JLA 3x3 method, based on Lau Method
Replies
19
Views
2,116
JackTriton
JackTriton
Ralph Laurence
Thoughts on Lau Method
Replies
11
Views
3,504
Ralph Laurence
Ralph Laurence
Michael Wimberly
Making A New Speedcubing Method
2
Replies
32
Views
3,833
Michael Wimberly
Michael Wimberly
Epinore
Gear Cube X Method: Basic Speedcubing Tutorial
Replies
3
Views
1,898
Epinore
Epinore
Back
Top