[Help Thread] ZZ and ZB Discussion

AlphaSheep said:
I won't be able to record a video, but CriticalCubing has this video which shows his U perm executions. What I do is I think a little more subtle, but it's a combination of the same sort of thing in that video. The U perm is unusual because it has four consecutive R turns in the same direction, where as three is the most I usually do without a significant regrip. More than three consecutive turns in the same direction is really rare during F2L though.
Click to expand...
Ooh, that's interesting. I'll try that out for a while and see how I like it.
 
Hey guys. I'm having trouble learning EO line. My first try took me about 5 minutes to identify bad edges. (8). I reoriented them and tried to reorient 2 while making my line. Did not work, i then ended up having 4 more bad edges than last time. Any help on making the line or any simplified eoline rules?
 
weatherman223 said:
Hey guys. I'm having trouble learning EO line. My first try took me about 5 minutes to identify bad edges. (8). I reoriented them and tried to reorient 2 while making my line. Did not work, i then ended up having 4 more bad edges than last time. Any help on making the line or any simplified eoline rules?
Click to expand...
Have you watched/read any tutorials for EOLine? The second two ZZ links in my sig are to Phil Yu's youtube playlist tutorials and Conrad Rider's written tutorial. I learned a lot from watching/reading these, though practice is the only way you can really get better.

To address your post myself, I usually just trace my line edges during EO, and influence them as best I can. I can usually see my whole EOLine during inspection. If you're having trouble seeing the whole thing, don't worry about it. A stepping stone to EOLine is just EO+Line. If you're having trouble remembering which pieces are good and bad, just place your finger on each piece, and once you've examined the whole cube, you should be able to see what 'shape' the edges are in. With enough practice, once you see the shape of edges, you'll automatically know what type of strategy to use to orient them. With more time, you can incorporate the Line edges into your strategies for solving different 'shapes'.

Hopefully this made sense.
 
weatherman223 said:
Hey guys. I'm having trouble learning EO line. My first try took me about 5 minutes to identify bad edges. (8). I reoriented them and tried to reorient 2 while making my line. Did not work, i then ended up having 4 more bad edges than last time. Any help on making the line or any simplified eoline rules?
Click to expand...
For rules, you can't beat the tutorials by asmallkitten and Conrad Rider that @GenTheThief mentioned above. I remember making lots of mistakes when I started out. With lots of practice, the mistakes become rarer. I remember that out of my first dozen solves, I got EO wrong more than I got it right. Now I make a mistake once every couple hundred solves.

What helped a lot was ignoring mistakes in EO when building F2L. In other words, if you see an edge is bad, just insert it flipped. Then when you get to the last layer, it's much easier to see which edges you got wrong. Make a note of the edges you got wrong, then solve the cube and apply the same scramble again. Count the bad edges, this time paying special attention to edges you got wrong, and try figure out why you got them wrong.

There are five possible reasons for getting an edge wrong :
  1. The edge was bad, but you missed it in inspection
  2. The edge was good, but you incorrectly identified it as bad in inspection
  3. The edge was bad and you recognised it as bad, but the solution you came up with didn't flip the edge.
  4. The edge was good, and you recognised it as good, but the solution you came up with unintentionally flipped the edge.
  5. Your F2L solution used F, M or wide moves which can flip edges.
You will need to figure out which of these mistakes you are making. One suggestion is to stop and do the EO recognition after each step. This can help you figure out where you're going wrong (although if your recognition isn't right to start with, don't trust it too much)

Depending on how fast you are, you may find it worth doing edge only F2L solves to save time. Do EOLine, then solve the 6 F2L edges ignoring the corners they are supposed to connect to. This makes it incredibly easy to spot mistakes, and takes about a third of the time of a full solve.
 
To those of you who have learned ZBLL, how long did it take you? I've just started actually caring about finishing the algs and learned all the diag-swap H cases today. I also plan on learning ZBLS. Any tips to learn this many algs?
 
Correct me if I'm wrong: chances for Pi, T, U, L, Sune and AS cases are 4/27. Chances for H cases are 2/27.
There are 72 cases for each of Pi, T, U, L. So, odds for each case are 4/(27*72) = 1/486
There are 40 cases for H. So, odds for each case are 2/(27*40) = 1/540
Correct?

Or are some H cases more likely to appear than others?
 
Pyjam said:
Correct me if I'm wrong: chances for Pi, T, U, L, Sune and AS cases are 4/27. Chances for H cases are 2/27.
There are 72 cases for each of Pi, T, U, L. So, odds for each case are 4/(27*72) = 1/486
There are 40 cases for H. So, odds for each case are 2/(27*40) = 1/540
Correct?
Click to expand...
Not quite, Some H cases (the noswap/diagswap ones which can be executed from two AUFs; I believe there are 8 total) are less common than the other cases. There are 32 cases at 1/486, and 8 cases at 1/972 (half of 1/486).

Any case with rotational symmetry order 2 are half as likely to occur as cases with no rotational symmetry. Similar cases with rotational symmetry order 4 have a quarter of the probability of occuring (the only cases for this are H perm, N perms, and skip).
 
ZZ Method for 4x4 (and up?)​

So, I’ve always wanted a method for zz comparable to yau or Meyer, but there has never been one. There have been some suggestions such as NS4 method or ZZ Hoya, but these both have flaws (even if they’re pretty good suggestions), mainly movecount, although I don’t have any statistics for NS4.

I proposed an idea here: https://www.speedsolving.com/forum/...ncept-idea-thread.40975/page-227#post-1271658, but after solving with it a bit, the movecount is too high. Also, the centres are awkward (as are the NS4 ones).

I’ve had an idea of why not orienting each wing, but that ended up being bad, and I’ve alsp thought about doing my proposal in a different order, but it didn’t change much. After my experimenting, I felt that reducing to last 6 edges was the best, as it leaves you with, at max, 2-2-2 pairing, and eo is easy (petrus style). But how do we get to that efficiently? My best idea so far:

1. Solve L and R centres, then three random oriented edges like with yau, with one being the line edge or the corresponding cross edge (probably on L) which you place in between the two others.

2. Centres. Pretty self explanatory.

3. Solve the last edge on L, as either the first line edge or the cross edge, depending on which one you did. If you did cross in step one, then line, and vice versa.

4. Put the kind of pseudo cross on D (or keep it on L) and pair the final line edge and some other random edge, and place the line edge opposite the other one (on D), solve the line and then do an L2 to give a 3/4 cross on D. Place the other oriented edge on RD.

5. Do a y rotation for last six edges and then finish with petrus style eo and oll parity.

6. 3x3 stage with coll and epll+parity.

This (as you can probably tell) is based largely off yau, and will therefore have a similar-ish movecount, could be lower or higher. Any suggestions would be welcome, and I would happily race someone to sub x with the finalised method. Also, sorry for the long post, but I didn’t want to miss anything.


 
I have this account from a while ago but I recently got back into cubing... I use ZZ with COLL except for sune cases and average mid-20's.

I usually use the <M,U> U-perms because they're incredibly fast for me, but I recently learned regripless executions of the <R,U> algs for clockwise U-perm. Should I bother using these if I'm still using M moves for counterclockwise, or is it better to have continuity? I used both for a while and it wasn't horrible (although I sucked at predicting AUF) but I'm thinking it would just be faster to use M moves for everything.
 
Last edited:
I use <M, U> for front-solved ccw-u-perm, H and Z perms
<R, U> for back solved u perms
and for front-solved cw-u-perm I just do U2 and solve <R, U>
 
Thanks, it looks like it doesn't matter too much (I'm only solving TH) as long as I'm consistent, so I'll keep experimenting with what works best for me but I'll most likely be using the <M,U> algs.
 
I've noticed that ZZ solvers tend to solve left/right block in a certain order every solve. Is this a normal thing? I've always just solved whatever pieces I see next while using openslotting to help with lookahead, so I often have blocks on either side of the cube at some point during F2L.
 
I used to use this alg for Z-perm: M2 U M2 U M' U2 M2 U2 M'
Now I use M' U M2 U M2 U M' U2 M2
I recently saw a video by Critical Cubing saying he uses two different Z perm algs so he never has to pre-AUF for the PLL. Is this recommended? I've gotten used to just using the one alg and I didn't think it would be worth it to have to think more just to save one move 50% of the time.
 
You must log in or register to reply here.

Similar threads

iguana
  • Poll Poll
V Perm Discussion Thread
2
Replies
22
Views
397
iguana
iguana
B
CEOR help and discussion thread
Replies
7
Views
649
bcube
B
ikindalikeclock
General Hypercubing Discussion
Replies
1
Views
232
Martyt7777
Martyt7777
Martyt7777
SSL-S5 discussion
Replies
14
Views
928
Martyt7777
Martyt7777
abunickabhi
Speedtyping discussion
Replies
3
Views
259
Martyt7777
Martyt7777
Back
Top