ZZ-CT Thread

[MethodNeutral]

What I was trying to say was that if you did the alg from the solved state you wont always get a ttll with headlights on right.

Can you send me a scramble starting with those moves in which you don't get one of those TTLL's? That doesn't sound right.


Sent from my iPhone using Tapatalk

 

[lillod]

D' R U R' D \\ setup alg
U R' U R' U' R U R' U' R2 U R \\ back block
U' R U2 R' U R U2 R' U' R U' R' \\ tsle
H perm ttll

 

[AlphaSheep]

I think @xyzzy has already explained it quite well but I'd like to add a little.

2-gen scrambles don't really preserve corner permutation. They actually only keep corner permutation into one of 3 groups that can be characterised by a swap of last layer corners (either no swap, adjacent swap, or diagonal swap). Having an unsolved F2L corner makes things a little more complicated.

Do R2 U R2 U R2 U2 R2 U2 on a solved cube, and look at the effect on corners. It swaps DFR and UFR, and swaps UBR and UBL. It does affect corner permutation, because it moves the F2L corner, but the corner permutation is still in the no-swap group.

TTLL isn't any different. All TTLL cases must fall into one of those three groups. We group them based on patterns that are easy to recognise (front headlights, right opposites, etc) because that's convenient, but that can't change the fact that there are still only 3 CP groups.

The three groups of TTLLs based on the CP groups are (assuming DFR corner is either solved or in UBL) :

No swap:

• EPLLs
• Opposites on front

Diagonal swap:

• E, V, Y, N PLLs
• Headlights on front

Adjacent swap

• A, F, G, J, R, T PLLs
• Headlights right
• Opposites right
• Headlights front and right
• Opposites front and right

D' R U R' D introduces an adjacent swap in PLL, so you can get any of the TTLLs in the adjacent swap group. Do a Y-perm followed by any RU TTLL and you are guaranteed to get a TTLL in the diagonal swap group only.

 

[Teoidus]

They actually only keep corner permutation into one of 3 groups that can be characterised by a swap of last layer corners (either no swap, adjacent swap, or diagonal swap).

Not very relevant, but I feel like I should point out that while permutations can be divided into no, adj, and diagonal categories, there are 4 distinct adjacent swaps, and you can't "transmute" between them <R,U> (so technically there are 6 groups).

This is because the size of <R,U> is 120 but S6 is 720, so there must be 6 distinct sets of 120 that <R,U> can reach at a time.

though, i'm actually not sure why a diagonal swap case wouldn't be reachable with D' R U R' D as well....

EDIT: Wait, okay, wait.

If you're doing D' R U R' D first thing before anything else every time, you will only get a specific subset of the adj swap TTLLs. You will also never ever get any of the 2gen TTLLs. This is because it's "equivalent" to swapping URF and URB, but there's no way to get, say, a ULF-ULB swap (or any of the other 3 adjacent swaps) by doing the D' R U R' D before anything else.

If you're doing <R,U> stuffs before D' R U R' D, then you can get literally any non-2gen TTLL. It's not restricted to adj swaps. Example of a diagonal swap TTLL: R U2 R' D' R U R' D R U R2' U' R U2 R' U' R2 U R' U R U2 R'. Yperm!

 

[AlphaSheep]

Not very relevant, but I feel like I should point out that while permutations can be divided into no, adj, and diagonal categories, there are 4 distinct adjacent swaps, and you can't "transmute" between them <R,U> (so technically there are 6 groups).

This is because the size of <R,U> is 120 but S6 is 720, so there must be 6 distinct sets of 120 that <R,U> can reach at a time.

Yes. I think what you're saying is very relevant.

What's really cool is that the 5 adjacent swap sets I mentioned above actually correspond to which of the 5 corners is in the DFR slot.

Starting from a solved cube with white top and green front, then doing D' R U R' D, then 2-gen scramble and then solve to TTLL, etc, you will always get the same colour cases.

If it's a PLL, it will always have orange headlights.
If its the TTLL case with headlights on the right, the headlights will always be green.
If it's the TTLL case with headlights on both sides, they will always be red in front and blue on the right.

You can obviously change the colours by choosing the AUF before the D' R U R' D. Adding an AUF switches which one of the 4 distinct adjacent swap groups you land up in.

 

[gateway cuber]

Let's start the zz-ct vs. cfop argument up again. For the sake of argument let's say that tsle=oll and ttll=pll in recognition and execution time. Now we have zz f2l-1 vs cfop f2l. Is there a clear winner in this ?

I will admit CT has more potential than CFOP, but nobody has mastered the method yet. You guys gotta get phil and andy in on this then there will be some faaaaassssst times.

 

[MethodNeutral]

Let's start the zz-ct vs. cfop argument up again. For the sake of argument let's say that tsle=oll and ttll=pll in recognition and execution time. Now we have zz f2l-1 vs cfop f2l. Is there a clear winner in this ?

Both of them are solving eight pieces, the main differences are that ZZ doesn't have rotations and that piece recognition is easier since edges are oriented. As far as CFOP F2L vs ZZ F2L-1, ZZ is the clear winner.


Sent from my iPhone using Tapatalk

 

[Zanaso]

Both of them are solving eight pieces, the main differences are that ZZ doesn't have rotations and that piece recognition is easier since edges are oriented. As far as CFOP F2L vs ZZ F2L-1, ZZ is the clear winner.


Sent from my iPhone using Tapatalk

This doesn't really consider the fact that CFOP F2L is capable of higher TPS. Also, the recognition thing isn't really significant, CFOP F2L recognition is fine lol. Another thing you didn't consider is that unless you want to mirror TSLE and do ADF (which is going to add time to you solve) you have to restrict your F2L to solving the FR slot last which will probably add time to your solve. Also I'm not quite convinced TTLL is as good as PLL tbh (I could definitely be wrong tho).

I personally don't think ZZ-CT is better than CFOP; they're probably about equal. I also don't really think ZZ-CT is better than ZZ-b.

 

[MethodNeutral]

This doesn't really consider the fact that CFOP F2L is capable of higher TPS. Also, the recognition thing isn't really significant, CFOP F2L recognition is fine lol. Another thing you didn't consider is that unless you want to mirror TSLE and do ADF (which is going to add time to you solve) you have to restrict your F2L to solving the FR slot last which will probably add time to your solve. Also I'm not quite convinced TTLL is as good as PLL tbh (I could definitely be wrong tho).

I personally don't think ZZ-CT is better than CFOP; they're probably about equal. I also don't really think ZZ-CT is better than ZZ-b.

I don't see why CFOP F2L would have any better TPS than ZZ. I do think that recognition is faster in ZZ because you only have to look at the top sticker to know if it's an F2L edge or belongs to the last layer, it's at least faster than CFOP, especially when checking the UB edge. I do agree about the last slot, but if you learn TSLE in terms of what it does rather than muscle memory it's not too bad to mirror it, and one rotation for TTLL is better than multiple rotations during F2L in my opinion. I do agree about the difference between ZZ-CT and ZZ-b, the only differences are alg count and recognition.

 

[Zanaso]

I don't see why CFOP F2L would have any better TPS than ZZ. I do think that recognition is faster in ZZ because you only have to look at the top sticker to know if it's an F2L edge or belongs to the last layer, it's at least faster than CFOP, especially when checking the UB edge. I do agree about the last slot, but if you learn TSLE in terms of what it does rather than muscle memory it's not too bad to mirror it, and one rotation for TTLL is better than multiple rotations during F2L in my opinion. I do agree about the difference between ZZ-CT and ZZ-b, the only differences are alg count and recognition.

CFOP has better TPS than ZZ because ZZF2L has way more regrips due to awkward insertions like R U R or switching between L and R moves. The recognition thing doesn't really matter because it's not like you pause for recognition; you do the recog while you're already solving pieces so it's not time wasted.

Also, the rotations in CFOP aren't that bad at all. y and y' are hardly worse than the regrips you will have to do for ZZF2L anyway.

 

[CubingGenius]

CFOP has better TPS than ZZ because ZZF2L has way more regrips due to awkward insertions like R U R or switching between L and R moves. The recognition thing doesn't really matter because it's not like you pause for recognition; you do the recog while you're already solving pieces so it's not time wasted.

Also, the rotations in CFOP aren't that bad at all. y and y' are hardly worse than the regrips you will have to do for ZZF2L anyway.

You forgot that CFOP also has regrips just like ZZ. Switching between L and R moves also happens in CFOP as well. You also F moves and especially rotations affect the TPS in a bad way.

I understand what you mean about the recognition. You don't have to worry about EO for ZZ, but you have to work out the permutation of the DL and DR edges, so it's kind of hard to tell which one is better.

Rotations are a lot worse than regrips. I'm not sure why you're trying to say regrips are so bad compared to rotations. They happen all the time in CFOP anyway, so it's like regrips vs regrips & rotations.

I think CFOP's move count is lower than ZZ's, but the TPS in ZZF2L would be faster than that of CFOP's F2L.

 

[AlphaSheep]

awkward insertions like R U R

I'm sorry, but I really have to nitpick here because I think this is one of the best F2L inserts... Maybe it's because you've developed sloppy fingertrick habits from always having to waste moves to restore the cross edge, but I do this as an R2 with a flick for the U in the middle. No regrip needed. I'm pretty sure I can do it faster than having to reverse direction half way through like in R U R'.

switching between L and R moves

But surely this happens in CFOP too? Or do you only solve slots on one side of the cube?

Anyway, many ZZ users work on one side first then the other unless there's a really obvious easy block, so the second half of F2L is pure 2-gen and practically regripless with really high TPS. During the first half, I typically minimize switching, so regrips are minimised. There are also plenty of ways to switch from R to L without a regrip, eg. Anti-slice moves (R L and R' L') rarely need a regrip.

There are valid weaknesses of ZZ F2L, but regrips and low TPS are definitely not one of them. ZZ is built specifically for ergonomics, and those ergonomics lead directly to a higher TPS.

 

[Teoidus]

I'm sorry, but I really have to nitpick here because I think this is one of the best F2L inserts... Maybe it's because you've developed sloppy fingertrick habits from always having to waste moves to restore the cross edge, but I do this as an R2 with a flick for the U in the middle. No regrip needed. I'm pretty sure I can do it faster than having to reverse direction half way through like in R U R'.

I don't think the issue is with the R U R itself but what comes before and after. An R2 is not easy to transition into and out of, as your wrist is forced into a more extreme position. This is where a regrip is often needed. Imagine R U R , and then U? What about an R U R U R U R' right block? Not exactly friendly..

But surely this happens in CFOP too? Or do you only solve slots on one side of the cube?

Yeah, I think the L/R regrip thing is overplayed a bit. There's an interesting thing that I see in videos of fast ZZ solvers, where in between switching between <R,U> and <L,U>, for example, the first couple <L,U> moves will actually be executed as <r,U/F> before the left hand becomes fully activated, easing transition between sides.

There's an argument to be made for things like L2 U2 R' U' R' never happening in CFOP, but an experienced ZZ user will force cross pieces to be in favorable positions during EOLine.

There are also plenty of ways to switch from R to L without a regrip, eg. Anti-slice moves (R L and R' L') rarely need a regrip.

Would you mind showing a video of you executing something that involves RL or R'L' and not having to regrip afterwards? I find this stuff awkward personally.

 

[AlphaSheep]

Would you mind showing a video of you executing something that involves RL or R'L' and not having to regrip afterwards? I find this stuff awkward personally.

I think it has a lot to do with grip. I tend to keep my thumbs on the FR and FL edges, at least during F2L, and I stabilise the M slice with my ring fingers on BU. It's fairly symmetrical and makes switching began L and R easier.

I will hopefully have a few spare minutes to film something tomorrow. Please remind me if I haven't done it by Friday.

 

[MethodNeutral]

Oh, I completely forgot to mention this, but ZZ-CT has a lower movecount than CFOP with 2LLL. Also, doing LSLL means there are less intuitive pieces to solve, so more of the solve is straight recognition and spamming TPS on algs, and TSLE is purely 2-gen save a few cases.


Sent from my iPhone using Tapatalk

 

[AlphaSheep]

Would you mind showing a video of you executing something that involves RL or R'L' and not having to regrip afterwards?

Moves:
R U' (L'R') U (L2'R) U' (L'R') U (LR) U R2' U R
U' R U R' U R U2' R'
There is a minor regrip needed after the L2

 

[Hammer]

Hey I think this method has great potential for me (cuz my tps is not that great and I use cfop) and I sorta know zz but im wondering whether i should stick to solving from one angle (so that I know which pieces to ignore) or i should be y axis color neutral, where there'll be more thinking involved about which pieces to avoid and secondly, should i finish the complete l block and then do the rb pair, or should i just do whatever pops up?

Thanks!

 

[Teoidus]

y neutral is nice, since it'll let you avoid bad cases more often. Most very fast people are y neutral I think.

Do whatever pops up, as long as you're leaving FR for last

 

[CubingGenius]

Hey I think this method has great potential for me (cuz my tps is not that great and I use cfop) and I sorta know zz but im wondering whether i should stick to solving from one angle (so that I know which pieces to ignore) or i should be y axis color neutral, where there'll be more thinking involved about which pieces to avoid and secondly, should i finish the complete l block and then do the rb pair, or should i just do whatever pops up?

Thanks!

Take a look at other things like CPLS/2GLL and especially ZZ-b before deciding.

 

[genericcuber666]

Take a look at other things like CPLS/2GLL and especially ZZ-b before deciding.

please dont do this cpls is horrible