The New Method / Substep / Concept Idea Thread

[bobthegiraffemonkey]

Okay then, but the Line last layer is literally an 18 case 1LLL, doesn't that give it some credit? think about it: we reduced the amount of pieces to solve by 1/3 and we reduce the amount of cases from over 1k to double digits. thats actually kind of incredible.

The case count is good, but that's not the only thing which affects how good a method is.

 

[Shiv3r]

The case count is good, but that's not the only thing which affects how good a method is.

true.
but lazy people want algorithm efficiency.

 

[wir3sandfir3s]

So I've recently found out abou ZZ-LoL and thought of a possibly viable idea for L2S.
1. Solve 2x2x3 in the back, EO, and last cross piece.
2. Create both pairs and put them adjacent to each other in the U layer.
3. Use an alg to insert each one, and while inserting, create a line in place of the pair. This will solve LL.
The alg count shout be much lower than the original L2S idea, as well as recog because the pairs are going in the front. The second alg may be a bit lengthy, though. Please share any thoughts, ideas or criticism

 

[wir3sandfir3s]

So what you're saying is that this method isn't one of the decent ones? Okay then

Interesting way to look at it...
Does that mean HK is decent???
(Just a joke don't get mad plz)

 

[Teoidus]

Okay, so I've been thinking about ZZ and I know that people say it's not popular because EOline can be very hard to get used to and there aren't results sufficiently convincing to back it up.

In an effort to resolve the former, I propose the following beginner's ZZ variant that someone can learn with no prior cubing knowledge (and give them a less painful way to transition into full ZZ later on).

The EZ Method (*rimshot*)
1. Cross (I was thinking on D, but perhaps on U would be possible as well).
2. Place all first layer corners without rotating. The process here could be similar to beginner methods, where you get the corner above its corresponding slot and use RUR'U' and its mirrors to solve.
3. F2L without rotating. Here introduce the idea of a front and front colors (shouldn't be too complicated, it's literally the color of the front center). For each edge piece (in top layer, like most beginner algs assume): get the piece to UF. If the top color is the same as the front color or its opposite, apply an algorithm to solve (e.g. U'RU'R'U'RU'R'URUR' and mirrors). If colors don't match, use FRUR'U'F' to orient the edge, and then solve it using the algorithm.
4. EO of LL using FRUR'U'F'.
5. CP using a similar recognition method to that described in DeeDubb's beginner Roux tutorial (match up corners to corresponding F2L slots, figure out which need to swap) and a Niklas to do the swapping. (I initially was convinced that people should use <R,U,D> Niklas to get used to D fingertricks, but perhaps learning to transition between R and L would be good as well)
6. CO using Sunes.
7. EP using an <M,U> U-perm.

This method would have 6 algorithms (sexy, F2L alg, fruruf, Niklas, Sune, Uperm), 5 of which remain useful later on.
F2L edge recognition is already partially built in, which provides an easier transition into ideas like EOline and orienting 4 edges at a time.
CP before CO might seem a bit odd, but I'm thinking that it would naturally open things up to more advanced CP recognition systems and provide an easier transition into COLL, since you'd be able to learn the algs while already having the recognition for it. This would let you actually understand what each COLL alg does to CP while learning it (personally I find this sort of concreteness very helpful).
I know people use <R,U> U-perms, but I suggest <M,U> here because I think <R,U> fingertricks can be developed during F2L, and <M,U> will be very useful later on when learning H/Z perms.
The only funky part that I see is in steps 2-3, where mirroring algorithms to the left and back might be a little hard to get used to.

Here's the sort of progression I'm imagining between EZ and ZZ (assuming the endpoint is ZZ-a, because that's always seemed the most hardcore to me):
EZ -> solve F2L by inserting pairs -> orient all edges before starting F2L, but after cross -> EOCross -> blockbuilding in the F2L, not just blindly solving cross pieces -> COLL -> PLL? (for when you OLL skip?) -> 2GLL? (for when you CP skip?) -> ZZLL -> ZBLL

Thoughts?

 

[gyroninja]

Okay, so I've been thinking about ZZ and I know that people say it's not popular because EOline can be very hard to get used to and there aren't results sufficiently convincing to back it up.

In an effort to resolve the former, I propose the following beginner's ZZ variant that someone can learn with no prior cubing knowledge (and give them a less painful way to transition into full ZZ later on).

The EZ Method (*rimshot*)
1. Cross (I was thinking on D, but perhaps on U would be possible as well).
2. Place all first layer corners without rotating. The process here could be similar to beginner methods, where you get the corner above its corresponding slot and use RUR'U' and its mirrors to solve.
3. F2L without rotating. Here introduce the idea of a front and front colors (shouldn't be too complicated, it's literally the color of the front center). For each edge piece (in top layer, like most beginner algs assume): get the piece to UF. If the top color is the same as the front color or its opposite, apply an algorithm to solve (e.g. U'RU'R'U'RU'R'URUR' and mirrors). If colors don't match, use FRUR'U'F' to orient the edge, and then solve it using the algorithm.
4. EO of LL using FRUR'U'F'.
5. CP using a similar recognition method to that described in DeeDubb's beginner Roux tutorial (match up corners to corresponding F2L slots, figure out which need to swap) and a Niklas to do the swapping. (I initially was convinced that people should use <R,U,D> Niklas to get used to D fingertricks, but perhaps learning to transition between R and L would be good as well)
6. CO using Sunes.
7. EP using an <M,U> U-perm.

This method would have 6 algorithms (sexy, F2L alg, fruruf, Niklas, Sune, Uperm), 5 of which remain useful later on.
F2L edge recognition is already partially built in, which provides an easier transition into ideas like EOline and orienting 4 edges at a time.
CP before CO might seem a bit odd, but I'm thinking that it would naturally open things up to more advanced CP recognition systems and provide an easier transition into COLL, since you'd be able to learn the algs while already having the recognition for it. This would let you actually understand what each COLL alg does to CP while learning it (personally I find this sort of concreteness very helpful).
I know people use <R,U> U-perms, but I suggest <M,U> here because I think <R,U> fingertricks can be developed during F2L, and <M,U> will be very useful later on when learning H/Z perms.
The only funky part that I see is in steps 2-3, where mirroring algorithms to the left and back might be a little hard to get used to.

Here's the sort of progression I'm imagining between EZ and ZZ (assuming the endpoint is ZZ-a, because that's always seemed the most hardcore to me):
EZ -> solve F2L by inserting pairs -> orient all edges before starting F2L, but after cross -> EOLine -> blockbuilding in the F2L, not just blindly solving cross pieces -> COLL -> PLL? (for when you OLL skip?) -> 2GLL? (for when you CP skip?) -> ZZLL -> ZBLL

Thoughts?

I don't see the difference from beginners method. Why not just teach them normal zz right off the bat (with the beginner ll)? It doesn't take that long to learn how to do eo or zzf2l, they just have to be shown a few cases.

 

[Teoidus]

Well, the difference is that you wouldn't rotate to solve F2L edges, and it's a bit of a lead-in to full EOLine. I don't think normal ZZ right off the bat is good because EO and ZZF2L are both pretty intense and I think that is what turns people away. I'm coming from the perspective of someone with literally no knowledge at all, and based on what I've seen from teaching friends how to solve, throwing things like EO and blockbuilding at newcomers right away isn't the best way to go.

I'm not sure what beginner LL is (i've seen many permutations of EO/EP/CO/CP). What do you consider it to be?

 

[gyroninja]

Well, the difference is that you wouldn't rotate to solve F2L edges, and it's a bit of a lead-in to full EOLine. I don't think normal ZZ right off the bat is good because EO and ZZF2L are both pretty intense and I think that is what turns people away. I'm coming from the perspective of someone with literally no knowledge at all, and based on what I've seen from teaching friends how to solve, throwing things like EO and blockbuilding at newcomers right away isn't the best way to go.

I'm not sure what beginner LL is (i've seen many permutations of EO/EP/CO/CP). What do you consider it to be?

EO is actually fun from a non speed focused perspective. It's a little puzzle that you have to solve. ZZF2L isn't too hard to comprehend if you are shown the basic ways of pairing up pieces.

Beginner ll is really ambiguous. It essentially means to me a 4+ looks last layer.

 

[Teoidus]

Interesting. I think you may have a unique perspective on EO--most people I've talked to find the step awkward or difficult to grasp without lots of practice. The recog isn't trivial either, so personally when starting out I found it frustrating when I messed up and had to spend a good 5-10 seconds redetecting edges. That's why I think if we introduce the EO in little pieces like this, it would make more sense to people when transitioning into EOLine.
That said I also don't think much time necessarily has to be spent on the beginner F2L--if someone really gets it, they could just go into EOLine.

I'm not sure about the F2L. Initially I was thinking that one could solve a line, then orient edges as necessary while blockbuilding the F2L, but later figured it might be overwheming with that much freedom and changed it. As far as I've seen, people like more structured beginners' methods over blockbuilding methods, but maybe it's possible to just start with the blockbuilding immediately.

As for LL, it technically could be done in any way, but I just think the way I suggest would be most beneficial down the line (EO CP CO EP). There's nothing particularly new about it, but I think that specific order is important

 

[AlphaSheep]

Well, the difference is that you wouldn't rotate to solve F2L edges, and it's a bit of a lead-in to full EOLine. I don't think normal ZZ right off the bat is good because EO and ZZF2L are both pretty intense and I think that is what turns people away. I'm coming from the perspective of someone with literally no knowledge at all, and based on what I've seen from teaching friends how to solve, throwing things like EO and blockbuilding at newcomers right away isn't the best way to go.

As someone who learnt EO and blockbuilding as a beginner, I wish I could get rid of this myth forever. EO for ZZ takes just a few hours to learn and blockbuilding is actually easier because you have more options.

ZZ is easy for beginners! Its only CFOP preconceptions that make it seem hard.

 

[Teoidus]

I think it's possible you two have just had an affinity for ZZ though. We should definitely push for dispelling any myths about difficulty, but CFOP has a lot more documentation, its first steps are simpler (they all involve directly solving pieces), and it has an even simpler beginner variant to ease people into F2L pairing. I just think that sort of simplicity is what would allow ZZ have a more general appeal. You may have found it not that bad, but it's possible some 7-year-old who just wants to learn to solve the thing might not enjoy it nearly as much

 

[adimare]

Here's the sort of progression I'm imagining between EZ and ZZ (assuming the endpoint is ZZ-a, because that's always seemed the most hardcore to me):
EZ -> solve F2L by inserting pairs -> orient all edges before starting F2L, but after cross -> EOCross -> blockbuilding in the F2L, not just blindly solving cross pieces -> COLL -> PLL? (for when you OLL skip?) -> 2GLL? (for when you CP skip?) -> ZZLL -> ZBLL

I think that the following is a more natural progression:

Beginner LBL -> F2L with pairs -> CFOP with 4LLL -> CFOP with 3LLL -> EO Cross (instant 2LLL without having learned full OLL) -> EO Line.

I think its best to switch to EO as the first step when you can already do proper F2L. Although I have no doubt that a beginner can learn EO very early on, I don't think he'll see any benefits from it until he's doing pair insertions during F2L, so why learn it before that?

 

[Teoidus]

Hm, true.

In that case, I think it'd be better even to just go with:
LBL with LL as proposed -> F2L with pairs -> EO Cross -> EO Line

 

[shadowslice e]

Tbh, ZZ isn't that bad if you just find a good tutorial like asmallkitten has on his channel. I actually learned basic CFOP, Roux, ZZ, PCMS, VHLS and Petrus (and heise; but i would not recommend that cause it can be pretty difficult) simultaneously when I was around 30 seconds. It actually really wasn't too bad and I got to see all the possibilities for EO.
Basically, asmallkitten's videos for ZZ and a few roux tutorials and you should be able to do it fairly easily given a week or less.

 

[Berd]

ZZF2L wouldn't be hard to teach in a beginner method, sexy move to insert corners, then the R U R U R U' R' U' R' for edges.

 

[Shiv3r]

Okay, so I've been thinking about ZZ and I know that people say it's not popular because EOline can be very hard to get used to and there aren't results sufficiently convincing to back it up.

In an effort to resolve the former, I propose the following beginner's ZZ variant that someone can learn with no prior cubing knowledge (and give them a less painful way to transition into full ZZ later on).

The EZ Method (*rimshot*)
1. Cross (I was thinking on D, but perhaps on U would be possible as well).
2. Place all first layer corners without rotating. The process here could be similar to beginner methods, where you get the corner above its corresponding slot and use RUR'U' and its mirrors to solve.
3. F2L without rotating. Here introduce the idea of a front and front colors (shouldn't be too complicated, it's literally the color of the front center). For each edge piece (in top layer, like most beginner algs assume): get the piece to UF. If the top color is the same as the front color or its opposite, apply an algorithm to solve (e.g. U'RU'R'U'RU'R'URUR' and mirrors). If colors don't match, use FRUR'U'F' to orient the edge, and then solve it using the algorithm.
4. EO of LL using FRUR'U'F'.
5. CP using a similar recognition method to that described in DeeDubb's beginner Roux tutorial (match up corners to corresponding F2L slots, figure out which need to swap) and a Niklas to do the swapping. (I initially was convinced that people should use <R,U,D> Niklas to get used to D fingertricks, but perhaps learning to transition between R and L would be good as well)
6. CO using Sunes.
7. EP using an <M,U> U-perm.

This method would have 6 algorithms (sexy, F2L alg, fruruf, Niklas, Sune, Uperm), 5 of which remain useful later on.
F2L edge recognition is already partially built in, which provides an easier transition into ideas like EOline and orienting 4 edges at a time.
CP before CO might seem a bit odd, but I'm thinking that it would naturally open things up to more advanced CP recognition systems and provide an easier transition into COLL, since you'd be able to learn the algs while already having the recognition for it. This would let you actually understand what each COLL alg does to CP while learning it (personally I find this sort of concreteness very helpful).
I know people use <R,U> U-perms, but I suggest <M,U> here because I think <R,U> fingertricks can be developed during F2L, and <M,U> will be very useful later on when learning H/Z perms.
The only funky part that I see is in steps 2-3, where mirroring algorithms to the left and back might be a little hard to get used to.

Here's the sort of progression I'm imagining between EZ and ZZ (assuming the endpoint is ZZ-a, because that's always seemed the most hardcore to me):
EZ -> solve F2L by inserting pairs -> orient all edges before starting F2L, but after cross -> EOCross -> blockbuilding in the F2L, not just blindly solving cross pieces -> COLL -> PLL? (for when you OLL skip?) -> 2GLL? (for when you CP skip?) -> ZZLL -> ZBLL

Thoughts?

This is the most brilliant thing I have ever heard. go on ahead, and I will support you every step of the way.
I suggest in the "intermediate" section of this you post the petrus ZBLL algset and idea, because it uses 40 algorithms to solve all ZBLL cases.

 

[Shiv3r]

Interesting. I think you may have a unique perspective on EO--most people I've talked to find the step awkward or difficult to grasp without lots of practice. The recog isn't trivial either, so personally when starting out I found it frustrating when I messed up and had to spend a good 5-10 seconds redetecting edges. That's why I think if we introduce the EO in little pieces like this, it would make more sense to people when transitioning into EOLine.
That said I also don't think much time necessarily has to be spent on the beginner F2L--if someone really gets it, they could just go into EOLine.

I'm not sure about the F2L. Initially I was thinking that one could solve a line, then orient edges as necessary while blockbuilding the F2L, but later figured it might be overwheming with that much freedom and changed it. As far as I've seen, people like more structured beginners' methods over blockbuilding methods, but maybe it's possible to just start with the blockbuilding immediately.

As for LL, it technically could be done in any way, but I just think the way I suggest would be most beneficial down the line (EO CP CO EP). There's nothing particularly new about it, but I think that specific order is important

for last layer, I taught my sister 4-look last layer(2Look OLL, 2Look PLL) and she totally gets it. so Last layer isn't a problem.
But I suggest that us, teodius, make a tutorial for this? I will definitely help!!!
ZZ-N, for ZZ-Newb

 

[Shiv3r]

we have to figure out though, how to get this method to be a natural progression to ZZ. I suggest:
-Line on top
-flip over and insert one cross edge with only R, U, and L moves.(this may help some people who can't bear destroying the stuff they've already done.) and if the cross is misoriented, then we do the alg.
-finish F2L on one side, then do the same thing on the right(this will help with EO detection, maybe fix last layer edges in the process, like in petrus)
-last layer

 

[Vexatious]

Here's an idea for a tutorial
The cube will be held with yellow on top and blue in front. Every time you find two bad edges, you place them in UF and RF (withiout using F and B) and then perform sledgehammer(or you could mirror the alg).

To find bad edges containing l/r, just find orange or red on top bottom, or e-slice front or back.
To find bad edges containing u/d, just find with or yellow on the first or third layer, or elide sides

And then make EO line with move restrictions
Then instead of making a cross, focus on building one block at a time( left block then right block or vice versa). Teach learners how to "hide pieces" to pair them up.

Last layer can be done in any way

 

[Shiv3r]

teodius's idea makes sense, more than yours. I think you think taht noobs can do more than you think they can.

Here's an idea for a tutorial
The cube will be held with yellow on top and blue in front. Every time you find two bad edges, you place them in UF and RF (withiout using F and B) and then perform sledgehammer(or you could mirror the alg).

To find bad edges containing l/r, just find orange or red on top bottom, or e-slice front or back.
To find bad edges containing u/d, just find with or yellow on the first or third layer, or elide sides

And then make EO line with move restrictions
Then instead of making a cross, focus on building one block at a time( left block then right block or vice versa). Teach learners how to "hide pieces" to pair them up.

Last layer can be done in any way