[Help Thread] ZZ and ZB Discussion

[lorki3]

That was very helpful, thank you! I will try to integrate this in my solves.
ZZ has so much potential - with only a week or so of practice I average 20-21 with OH, nearly sub-20.

Another question:
What do you do when you have 4 bad edges which are all located in the E slice? I have a lot of trouble with this case. I guess the same goes for 4 in the M-slice. Thanks!

 

[Petro Leum]

That was very helpful, thank you! I will try to integrate this in my solves.
ZZ has so much potential - with only a week or so of practice I average 20-21 with OH, nearly sub-20.

Another question:
What do you do when you have 4 bad edges which are all located in the E slice? I have a lot of trouble with this case. I guess the same goes for 4 in the M-slice. Thanks!

something like (x') Rw (U' D) F2 R2 F('), but that depends on the case. something similar for the 4 on the m slice, but if that happens nowadays i tend to check the second orientation after a y rotation

 

[supercavitation]

effiency wise, its best to be flexible and able to go back and forth through both blocks

for onehanded, it's a different story; it's usually better to stay in one rotation as long as possible, even if it's just a z/z'.


Also, if the right and left block take roughly equal as many moves, it would be wise to solve the one with the worse move set first (for example, my [L,U] is much worse than my [R,U], so i would prefer [L,U] first), because the first block usually takes more moves than the second one, and you want as many of the moves you have to make on your preferred move-set.

i hope that was understandable, i'm very tired

Didn't Conrad Rider calculate that on average you save a move by doing one block, then the other?

 

[porkynator]

Didn't Conrad Rider calculate that on average you save a move by doing one block, then the other?

https://www.speedsolving.com/forum/...F2L-Move-Count&p=342121&viewfull=1#post342121
1x2x3 -> 1x2x3 is better than 1x2x2 -> pair -> 1x2x2 -> pair and 1x2x2 -> 1x2x2 -> pair -> pair, but in both cases you are restricting yourself in some way. Also, building an optimal 1x2x3 is not so easy in a speedsolve.

 

[Petro Leum]

Didn't Conrad Rider calculate that on average you save a move by doing one block, then the other?

while that may be true, the THIRD option, which i was talking about, is always the best: be flexible. chose, based on the scramble, if youre gonna do 1x2x3 -> 1x2x3 or 1x2x2 -> 1x2x2 -> pair -> pair. and the difference is significant, although i can't tell you any numbers, i once calculated it.

 

[mDiPalma]

Has anyone mastered this advanced EOLine strategy?

Solving the EOLine while isolating the pieces from 1 F2L block into either the correct side for that F2L block, or into the U layer.

In other words, you solve the EOLine while making sure that none of the left block pieces require an R move to be solved.

(Kind of like ZZ-rainbow back in the day, except the isolation step takes place during the EOLine, not while placing the DR corners.)

Is there a significant efficiency loss in solving the first F2L block 2-gen instead of 3, assuming you solve 1 F2L block at a time? I know it prevents "open slotting" which is a great move-saver.

Spoiler

For those who don't know, "open slotting" is the act of leaving the 2x2x1 F2L block 1 move away from being correctly placed, such that the next F2L pair can be created in the U layer and inserted into the "open" slot, to line up with the F2L block. It instantly reduces your ZZF2L movecount by 2 canceled moves, every solve, no matter what (assuming you solve F2L in blocks and pairs, not all at once like computers and Pyjam used to).

 

[lorki3]

I just got a 17.76 official OH average with ZZ! I'm very happy with this.

Out of principal, I did not want to be added to the list of ZZ users until I had got an official time with it. So if you can, please add me to the list my real name is Thijs Engel!

 

[adimare]

This method is amazing. Learned it a couple of weeks ago and got addicted to it. I average ~23 sec with it (~16 with CFOP) but with a large variance, so I'll get stuff like a 17 followed by a 31 regularly. Hardest part has been to figure out the EO line in less than 15 sec, a lot of times I still can only figure out EO during inspection and need to solve the line edges on the fly. F2L lookahead seems a lot harder than it is on CFOP since you're often solving 3 pieces at once rather than 2, so going into auto-pilot while looking for the next pieces to solve becomes difficult (hoping that will become less of an issue with more practice). Definitely switching, seems like a much more elegant approach to solving a Rubik's cube than CFOP. I just hope I don't forget the OLLs since I still need them for larger cubes

 

[Petro Leum]

This method is amazing. Learned it a couple of weeks ago and got addicted to it. I average ~23 sec with it (~16 with CFOP) but with a large variance, so I'll get stuff like a 17 followed by a 31 regularly. Hardest part has been to figure out the EO line in less than 15 sec, a lot of times I still can only figure out EO during inspection and need to solve the line edges on the fly. F2L lookahead seems a lot harder than it is on CFOP since you're often solving 3 pieces at once rather than 2, so going into auto-pilot while looking for the next pieces to solve becomes difficult (hoping that will become less of an issue with more practice). Definitely switching, seems like a much more elegant approach to solving a Rubik's cube than CFOP. I just hope I don't forget the OLLs since I still need them for larger cubes

i can tell you; if you already know them and you still solve a big cube once a month, you wont forget them. you just wont want to learn new ones (but i guess you already know them all?)

good luck with zz, it's worth the hassle

 

[mDiPalma]

zz, it's worth the hassle

now THAT is quote of the year.

 

[lorki3]

This method is amazing. Learned it a couple of weeks ago and got addicted to it. I average ~23 sec with it (~16 with CFOP) but with a large variance, so I'll get stuff like a 17 followed by a 31 regularly. Hardest part has been to figure out the EO line in less than 15 sec, a lot of times I still can only figure out EO during inspection and need to solve the line edges on the fly. F2L lookahead seems a lot harder than it is on CFOP since you're often solving 3 pieces at once rather than 2, so going into auto-pilot while looking for the next pieces to solve becomes difficult (hoping that will become less of an issue with more practice). Definitely switching, seems like a much more elegant approach to solving a Rubik's cube than CFOP. I just hope I don't forget the OLLs since I still need them for larger cubes

It took me about a month of practice to get to 16 seconds with ZZ. Now I'm basically sub-16 . I recommend lots of slow solves.

Btw, I still plan merely EO most of my solves unless the line edges are placed very easily.

 

[pinser]

I just hope I don't forget the OLLs since I still need them for larger cubes

You won't.
zz, worth the hassle.


Does anyone know the average movecount for zz-d porky v2?
Also, how helpful is y-axis neutrality?

 

[lorki3]

You won't.
zz, worth the hassle.


Does anyone know the average movecount for zz-d porky v2?
Also, how helpful is y-axis neutrality?

I think y-axis neutrality is basically the same as opposite cross in CFOP. It opens possibilities for better EOlines. I guess it would be best to start learning y-axis neutrality as soon as possible so you have more time to get used to it.

I totally made that quote my signature

 

[MrMan]

I'm having an issue with ZZ, I love it, I personally found the F2L much more easier to look ahead and just generally solving them, but it takes me too much time to orient edges it demotivates me.

 

[lorki3]

I'm having an issue with ZZ, I love it, I personally found the F2L much more easier to look ahead and just generally solving them, but it takes me too much time to orient edges it demotivates me.

It doesn't take too long to get the hang of it. Just keep at it, it's worth the hassle

 

[MrMan]

Well not learning full OLL but just COLL has gotten me too !
I'm going for 50-100 solves right now and if I can get any improve on EO I'll definitly switch !

 

[lorki3]

Well not learning full OLL but just COLL has gotten me too !
I'm going for 50-100 solves right now and if I can get any improve on EO I'll definitly switch !

You can stick to 7 OCLL's + 21 PLL's for a while and once you're more comfortable with the method, you can slowly learn COLLs.

 

[AlphaSheep]

I have a question about y-axis neutrality.

If I'm looking at a scrambled cube with blue/green front and yellow/white top, I notice only 2 things that change with a y rotation.
1. red/orange-blue/green edges in the U and D layers change from good to bad or bad to good, but edges in the E slice remain unchanged.
2. edges with yellow/white in the E layer change from good to bad or bad to good, but edges in U/D remain unchanged.

Is it better to use this information to figure out which edges will change with a y rotation, or is it better to just turn the cube and learn to recognise the bad edges based on the new orientation? My instinct says the second would be better in the long term, but I find the first easier.

 

[MrMan]

You can stick to 7 OCLL's + 21 PLL's for a while and once you're more comfortable with the method, you can slowly learn COLLs.

Mistyped OCLL ^^

 

[TDM]

I have a question about y-axis neutrality.

If I'm looking at a scrambled cube with blue/green front and yellow/white top, I notice only 2 things that change with a y rotation.
1. red/orange-blue/green edges in the U and D layers change from good to bad or bad to good, but edges in the E slice remain unchanged.
2. edges with yellow/white in the E layer change from good to bad or bad to good, but edges in U/D remain unchanged.

Is it better to use this information to figure out which edges will change with a y rotation, or is it better to just turn the cube and learn to recognise the bad edges based on the new orientation? My instinct says the second would be better in the long term, but I find the first easier.

(I only use yellow top and blue/red front)
I use your first method to count whether there will be more oriented edges on red or blue front. Here's an example scramble:
F' U B2 R B L B' R B' U R2 F L2 F' L2 F R2 U2 B U2 F2

First, I look in the U/D layers for F2L edges. If it has red/orange facing U/D I'll think 'red', and if it has blue/green facing U/D I'll think 'blue'. Here I see one 'red' and one 'blue'; if the number is the same for both I'll just forget them because it doesn't matter which orientation I choose. I usually look at the U layer first, then do an x2 so I'm in normal orientation, then look at the new U layer (yellow).

Then I look in the E layer for white/yellow stickers. If they're facing L/R (on the red/orange faces), then I count them as 'red', and if they're on B/F (the blue/green faces) I count them as 'blue'. Here were have two 'red' edges, so if I use red front then I'll have two extra oriented edges. So I would use red front this solve.

tldr: I count which orientation is best before planning anything, then I plan EOLine in that orientation