• 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.

[Help Thread] ZZ and ZB Discussion

yeah i realized quickly how pointless making the cross was, since i would be moving that face around alot anyway it was alot easier and freed my mind a bit. also realizing my experience from CFOP for doing f2ls essentially taught me how to blockbuild 1x1x2s anyway and doing an insertion was like making a 1x2x2 helped me get my head around the concept of the block building. getting the hang of this more and more :)

and yes i know eo-line will take quite awhile, i realize now why they say the concept is easy to learn but hard to master, slowly getting used to it though just gotta practice it alot. thanks though :)
 
down to 35~ish seconds with ZZ, better than the almost 1 minute average when i started LOLOLOLOL. EOLine is so hard to pick up QQ
 
Some new NMLL example solves. I wasn't satisfied with the others. This time the blocks are realistic for what I would do in a solve if I was a ZZ user.

Scramble: F2 R U2 R2 D2 L B2 L D2 U2 L' U B' L D' B F R' F L2

EOLine: U' B' U2 R L' F' B2 D2
Block 1: R' U' L' U' L' R2 U R U' L'
Block 2: U' R U2 R U' R U' R' U R
Separate: U l' U' L U R U' L' U x'
Permute: U R' U2 R U R' U R U R U R' U R U2 r'

Scramble: B2 U B2 R2 B2 D' L2 R2 D' F2 U L D' B2 U' F' R2 D L' F2 D2

EOLine: U L B R' D'
Block 1: L R' U2 R U L U2 R U L U L'
Block 2: R2 U2 R2 U R'
Separate: U2 L' U2 L U2 R U' L' U M' x'
Permute: L U' R U2 L' U M' B' R B2 L' x' U M'

Scramble: F U2 B' L2 U2 B2 F' L2 D2 F' L B2 U L B' R D' F' L' R F2

EOLine: F' L' U D' B' D L' D
Block 1: u2 L2 u2 U' L' U2 L R' U R
Block 2: L' U L U' L2 U' L' U L U' L'
Separate: U' R' U2 R U L U2 R' U' R U2 L'
Permute: U2 x L2 D2 L' U2 L D2 L' U2 r

Scramble: D2 L' B2 R U2 L R2 D2 U2 B2 L2 F' R B2 D2 B2 L2 D B D'

EOLine: U' R L' D B' D
R1x2x2: U R' U' R'
L: U2 L' U' L' R' U2 L' U2 L U2 L'
R: R U R' U R U R'
Separate: U R U2 R2 U' R2 U' R2 U2 R
Permute: r' U r' U2 l U' l' U2 r2 B' U' r

Scramble: R2 U2 R' D2 B2 D2 R2 F2 R D2 B2 F' L U B2 U2 L' D' B' L B2

EOLine: B D L U F D L2 D
Block 1: R U' R' L U2 L2 U' R U2 R' U' L'
Block 2: U R' U' R' U' R U R U2 R' U2 R U2 R'
Separate: U' R U' L' U R' U' L
Permute: U r' U r' U2 l U' l' U2 r2 B' U' r M'
 
Was thinking of something atm; However, as it's pretty sure everyone and their mother has already thought of it, I'd like to know what's wrong with it.

Imagine you've made your EOline, followed with two 2*2*1 and a Corner-Edge pair: you're then left with a CE slot, let's say RF pair, and the last layer to solve. From there, you can follow in a lot of different ways:
- Last CE pair + ZBLL
- Last CE pair + OCLL + PLL
- Last CE pair + COLL + EPLL
And so on (I could continue with Winter Variation, ZZLL etc, but you've understood what I mean)

However,the idea would now be to place and orient DRF corner, while orienting all remaining corners. To do this step, you need to learn 104 algorithms (81 if DRF corner is in Upper layer + 23 if he's already in DRF)
The recognition here seems at first glance pretty easy, as you only need to recognize DRF corner + the orientation of 3 other corners.

After this, you eventually finish the cube in a second step, which will place UR edge, while solving the rest of the cube (Which only consists of four edges and corners all oriented). Here you'll have 93 algorithms to learn (21 PLL if FR corner is already placed + 72 if it isn't). As for the recognition, it's exactly the same to recognize than a 2-sided PLL recognition system, which is, if I'm not mistaking, pretty easy to recognize.

Basically, just think of it like MGLS, except you don't care about the last slot edge before the very last step (It's really just a MGLS variation)

So, as we have a method that seems to have a good recognition and a relatively low amount of algos (197 isn't very big as compared to other methods which solve the LL in one step like ZZ-a or ZZ-b), I suppose the reason I haven't seen you guys talking about it is the movecount: It's right I suppose the second step to have a big movecount. However, I will give you an average movecount once I'll have generated all the algs.

Any thoughts?
 
The alg count would give anyone pause. And I personally don't like that I have to commit to always leaving a certain slot until last. Besides that it sounds pretty reasonable.
 
Le CaRiBoU said:
Any thoughts?
Click to expand...

I once calculate movecount for Last-5-Edges (wich is a subset of what you need), so statistical average is 9.8
When just taking movecount for Edge-5-Cycles mouvecount average is 11.3 (the longes alg I use has 13 moves)
I dont see a reason why the other LEPLL (naming suggestion for that step?) cases should be much different.

Compared to MGLS (= ELS+CLS+PLL) you even reduce the method to really tow steps CLC + LEPLL (naming suggestion 2, CLC = CornerSLastCorner )

I wonder how much faster CLC is compared to CLS - If many algs were the same ...
 
Last edited:
oll+phase+sync said:
I once calculate movecount for Last-5-Edges (wich is a subset of what you need), so statistical average is 9.8
When just taking movecount for Edge-5-Cycles mouvecount average is 11.3 (the longes alg I use has 13 moves)
I dont see a reason why the other LEPLL (naming suggestion for that step?) cases should be much different.

Compared to MGLS (= ELS+CLS+PLL) you even reduce the method to really tow steps CLC + LEPLL (naming suggestion 2, CLC = CornerSLastCorner )

I wonder how much faster CLC is compared to CLS - If many algs were the same ...
Click to expand...
Well, Ok for the name ^^.

I'm happy for the movecount, and it would be really good if the remaining algs had such a low movecount (Even if I really doubt, because it seems too nice).
I have generated all the algs (Only Optimal, and without any restriction for the moves used, even if most of them are in < R , U , L > ) for the CLC step, and the movecount is 7,89 HTM, which seems pretty high as compared to pure/normal F2L... However, I haven't had the time to compare the list to CLS to see common algs, but I have just remarqued that the 7 cases where the last corner was already solved are pure OLLs.

The alg count would give anyone pause. And I personally don't like that I have to commit to always leaving a certain slot until last. Besides that it sounds pretty reasonable.
Click to expand...
While being big, the alg count isn't that big, at least when compared to other methods that solves last slot + Last Layer in only 2 steps like ZZ-a or ZZ-b (It's "only" 197 less 21 PLL and 7 OLL that you may already know, and maybe you could even reduce the count with some CLC being CLS, assuming you already know CLS (Which isn't my case ^^) ).

As for the idea of having to leave a certain slot until last, I don't think I understand what you mean here: You can solve any last slot if you use symmetries/y2. However, it's right that having to learn symmetries for such a big amount of algs may be a big task...
 
Le CaRiBoU said:
Well, Ok for the name ^^.

While being big, the alg count isn't that big, at least when compared to other methods that solves last slot + Last Layer in only 2 steps like ZZ-a or ZZ-b (It's "only" 197 less 21 PLL and 7 OLL that you may already know, and maybe you could even reduce the count with some CLC being CLS, assuming you already know CLS (Which isn't my case ^^) ).

As for the idea of having to leave a certain slot until last, I don't think I understand what you mean here: You can solve any last slot if you use symmetries/y2. However, it's right that having to learn symmetries for such a big amount of algs may be a big task...
Click to expand...

The alg count is definitely manageable (I went the ZZ-b route myself), I'm just saying it's not something one enters into lightly.

You did address my last slot concern. Learning that many symmetries sounds painful. d or y turns would work, though it's a bit ironic after 3-gen (no cube turn) F2L ;)

But by all means, continue developing the method. I'm following the topic with great interest.
 
Le CaRiBoU said:
Was thinking of something atm; However, as it's pretty sure everyone and their mother has already thought of it, I'd like to know what's wrong with it.

Imagine you've made your EOline, followed with two 2*2*1 and a Corner-Edge pair: you're then left with a CE slot, let's say RF pair, and the last layer to solve. From there, you can follow in a lot of different ways:
- Last CE pair + ZBLL
- Last CE pair + OCLL + PLL
- Last CE pair + COLL + EPLL
And so on (I could continue with Winter Variation, ZZLL etc, but you've understood what I mean)

However,the idea would now be to place and orient DRF corner, while orienting all remaining corners. To do this step, you need to learn 104 algorithms (81 if DRF corner is in Upper layer + 23 if he's already in DRF)
The recognition here seems at first glance pretty easy, as you only need to recognize DRF corner + the orientation of 3 other corners.

After this, you eventually finish the cube in a second step, which will place UR edge, while solving the rest of the cube (Which only consists of four edges and corners all oriented). Here you'll have 93 algorithms to learn (21 PLL if FR corner is already placed + 72 if it isn't). As for the recognition, it's exactly the same to recognize than a 2-sided PLL recognition system, which is, if I'm not mistaking, pretty easy to recognize.

Basically, just think of it like MGLS, except you don't care about the last slot edge before the very last step (It's really just a MGLS variation)

So, as we have a method that seems to have a good recognition and a relatively low amount of algos (197 isn't very big as compared to other methods which solve the LL in one step like ZZ-a or ZZ-b), I suppose the reason I haven't seen you guys talking about it is the movecount: It's right I suppose the second step to have a big movecount. However, I will give you an average movecount once I'll have generated all the algs.

Any thoughts?
Click to expand...

i think a better idea to solve the last pair whil simeltaneously orienting the last layer. kinda like winter variation and cls, but for all f2l cases. i imagine the algcount would be pretty painful but still worth it.
 
ottozing said:
i think a better idea to solve the last pair whil simultaneously orienting the last layer. kinda like winter variation and cls, but for all f2l cases. i imagine the algcount would be pretty painful but still worth it.
Click to expand...
It's about 27*20=540 algs. Personnaly, I give up ^^

@nickvu2:
So I didn't really understand what you meant, but now it makes much more sense. And you're totally right.
Well, learning symmetries is obligatory, because you can't solve a LF or RB last slot if you only know the alg for RF/LB, even if you allow y/d turn (Unless if the algorithm used manages not to affect the last slot edge, which will be the case only for a few algs). However, as most algs are < R , U , L >, I don't think it'll be so painful to learn symmetries.
As for the y2, I don't think it's really a big problem to make one y2/d2 during the solve (Even if now, it's not really anymore 3gen F2L).

And yes, it's right that it won't be that easy to learn those algorithms. But for the moment, I have to generate optimal algs for the 2nd step, to be sure there will be an interest to learn this method after all my exam.
 
Le CaRiBoU said:
However,the idea would now be to place and orient DRF corner, while orienting all remaining corners. To do this step, you need to learn 104 algorithms (81 if DRF corner is in Upper layer + 23 if he's already in DRF)
Click to expand...

I was thinking about RU - FU mirroring (which can be used in CLS alot ) does not work for CLC.

Exsample: RUR' is a valid CLC (it is no CLS because FR edge moves around to UF )
but the F'U'F - while solving the mirrored corner case - is no CLC because it would generate misoriented edges.

This does not increase alg count but makes them more nasty, I don't know an optimal alg for F'U'F, but the corresponding CLS alg always works for this cases.

But if BR is the last Slot (as nickvu2 mentioned I need to learn all the mirrored versions like R'U'R) (CSL algs will work again).

But it would be 2 X 104 - 7 = 201 cases for CLC but also LEPLL would bcome 2 x 93 - 21 = 165 algs ?
 
thatkid said:
its called ZBF2L
Click to expand...

i was refering to corner orientation, not edge orientation.

(slightly off topic considering its a zz/zb thread), i think the idea i "proposed" would work better for petrus because you wouldn't have the chance of you last pair being in the back left or front left slots (assuming you only learned algs for the front right slot) which would give more chances to skip oll.

back on topic sorta, just a quick question for zz users, do you allway do the left or right block first or just whats more obvious at the time during a solve?
 
Last edited:
what's more obvious to me at the time

i need to work on not being as biased to making cornet edge pairings, i basically still do pretty much all corner pair with a middle slide edge rather than a bottom slide edge.

that and lookaheads are still my bane but it's getting more and more fluid as i go.
 
I'm begining to lose the will to live with ZZ. I've been at it for about a month now and it's still significantly worse than my CFOP solves. I wante to know wht the typical breakdown for a fast zz solve would be. Say for someone who averages around 15 seconds.

my cfop solves: 3 second cross 10 seconds for slots and edge orientation in LL. then about 6 seconds for completing the last layer.

with zz: 5 second EOLine, 12 seconds for finishing f2l and again 6 seconds for LL

I also want to know, do any of the top top cubers use zz? and if so who are they? I'm just starting to wonder whether its worthwhile continuing to learn this method.
 
CubeRoots said:
I also want to know, do any of the top top cubers use zz? and if so who are they? I'm just starting to wonder whether its worthwhile continuing to learn this method.
Click to expand...

If by that you mean, are there any top cubers who use full ZZ as their main method? or used ZZ to get their top times? if so, then no. There aren't (unless I'm mistaken). The reason is because they are simply more used to CFOP, Petus, Roux, etc. not because the method itself is bad.

I would LOVE to see someone learn ZZ as their first method of solving, and then extrapolate to the future to see their speeds, as I'm sure that this plays a big factor in ZZ's usage.
 
It seems so good in theory, yet the Potential of it doesn't seem to have been realized by anyone. i'm wodering if it ever will be by anyone. Never mind me. I would really love to see that aswell. it really is a beautiful method.
 
pdilla said:
If by that you mean, are there any top cubers who use full ZZ as their main method? or used ZZ to get their top times? if so, then no. There aren't (unless I'm mistaken). The reason is because they are simply more used to CFOP, Petus, Roux, etc. not because the method itself is bad.

I would LOVE to see someone learn ZZ as their first method of solving, and then extrapolate to the future to see their speeds, as I'm sure that this plays a big factor in ZZ's usage.
Click to expand...

Phil Yu uses ZZ exclusively and is one of the best OH solvers in the world.
 
You must log in or register to reply here.

Similar threads

B
CEOR help and discussion thread
Replies
7
Views
631
bcube
B
ikindalikeclock
General Hypercubing Discussion
Replies
1
Views
221
Martyt7777
Martyt7777
Martyt7777
SSL-S5 discussion
Replies
14
Views
910
Martyt7777
Martyt7777
abunickabhi
Speedtyping discussion
Replies
3
Views
249
Martyt7777
Martyt7777
acrystal
Ziyu Wu Former Megaminx WR Average Discussion
Replies
1
Views
693
Martyt7777
Martyt7777
Back
Top