Idea for CFOP users learning ZBLL

[Kirjava]

Kirjava: I was referring to how a 1-look system utilizing COLL and EPLL would work, not to a normal solve using COLL then EPLL in two looks.
Sorry, what do you mean about the Y-perm?

A 2 look COLL/EPLL system would probably be faster than a 1 look system.

The Y perm is two OLLs, but you don't recall each OLL separately.

 

[Anonymous]

That's a good point. All I can say is that IMO, the reason why the Y-perm is different from my example is simply that it's part of a much smaller algorithm set. The whole point of learning a 1-look 2-alg system is so that each case can be easily memorized as two "pieces" of information. I think that the reason why people don't do that for the Y-perm is because that's simply not how they memorize it. Am I wrong?

 

[Kirjava]

When someone remembers a 2 alg combo for a case, they don't execute the first and then work out the second - they already know what the second will be before executing the first.

This may become an issue with a larger algorithm set, but this is because the number of cases is large - not because of the technique.

 

[Anonymous]

To your second point: Yes, but to me that's exactly the point- two OLL combinations could be harder to remember than COLL EPLL combos when you talk about nearly 500 cases.

As for your first point, I realize that, but is it relevant? I know that you'd recall both algorithms before starting to perform them, but the fact remains that recalling them is easier (I think it would be, anyway) when dealing with COLL and EPLL pairs.

 

[Kirjava]

two OLL combinations could be harder to remember than COLL EPLL combos when you talk about nearly 500 cases.

This is clearly incorrect.

As for your first point, I realize that, but is it relevant? I know that you'd recall both algorithms before starting to perform them, but the fact remains that recalling them is easier (I think it would be, anyway) when dealing with COLL and EPLL pairs.

As is this.

I don't think you're quite sure of what you're talking about. Did you read this and this?

 

[Anonymous]

I'm not exactly following...

Yes, I read both of those posts, but I'm not quite sure how what I'm saying relates. This thread is meant to be discussing the possibility of using two OLLs in a row to solve a ZBLL case, right? Someone earlier made a suggestion to instead use a system which uses COLL and EPLL combos instead of two OLLs. All I'm saying is that I think that that's easier, for reasons I outlined in another post. Could you explain why you disagree with that?

The second thing you quoted was essentially me saying the same thing, but trying to express that what I was saying addressed both of your earlier points. Could you tell me why you think I'm wrong?

 

[Kirjava]

Find me any COLL+EPLL combination that's less moves than an OLL+OLL ZBLL combo.

The reason why I'm telling you that you don't understand things is that it's impossible for an OLL+OLL system to cover every ZBLL case (efficiently). Because of this, OLL+OLL is only used for a few cases because it's fast for them. For some of these cases, it's the optimal solution.

Therefore, you can't really argue against using OLL+OLL combos for ZBLL. Saying 'COLL+EPLL is probably easier than OLL+OLL' is simply a silly argument to make since you cannot comare the two.

 

[Anonymous]

I know that 4Chan mentioned that some of his algs are OLL combinations and optimal, but I was under the impression that whatever portion of ZBLL these OLL combinations could cover, it wasn't necessarily meant to be optimal; it was meant to be an easy alternative to actually learning new algorithms. In that context, I was just saying that COLL + EPLL was more appealing.

 

[Kirjava]

It isn't.

 

[Anonymous]

Are you saying I still have some fundamental misunderstanding, or do you just disagree?

 

[Kirjava]

I disagree.

 

[4Chan]

I know that 4Chan mentioned that some of his algs are OLL combinations and optimal, but I was under the impression that whatever portion of ZBLL these OLL combinations could cover, it wasn't necessarily meant to be optimal; it was meant to be an easy alternative to actually learning new algorithms. In that context, I was just saying that COLL + EPLL was more appealing.

Yep! Those were a coincidence actually. xD

Another coincidence is that on the sune cases with

subset, and double edge swap in UF and UR.

An optimal solution is actually a sune+optimal J perm with cancellation. LOL

 

[rubiknewbie]

I think there might be coincidence cases where the optimal solution happens to be a COLL + EPLL? But maybe isolated cases.

However as a system, I don't think 1-look COLL+EPLL is better than even 2-look COLL+EPLL because:
1. recognition wise, 1-look COLL + EPLL is much worse than 2-look COLL + EPLL if you are already narrowed down to U, Z, and H on the 2nd look which are all super nice for recognition. Even with no pause in between, the 1-look may take longer than the 2 looks COMBINED, if you have to recognise and recall the combination.
2. move wise you are not saving up in number of moves and fingertrickiness, since you are combinining the same algorithms?

I think the selling of point of Lars Petrus system is that combining nice algorithms you can sometimes solve last layer in 1 look. You will reject combinations if the algorithms are not nice. So you make up for poor recognition with guaranteed good execution. Not all COLL+ EPLL combinations are nice for execution.

 

[Rpotts]

How does anonymous not get what kirjava is saying?

There are so many ZBLL algs you need to consider all different kinds of alg combos to cover all of them. OLL/OLL is good for some, COLL/EPLL may be good for some too. But why try to cover all of them with one strategy? I mean, you're learning almost 500 algs, put forth the effort.

 

[Anonymous]

My point was simply that overall, I think that in most situations OLL + OLL is not as good as COLL + EPLL. I understood fully what he was saying. I happen to disagree, but I understand what I'm disagreeing with. I don't think that COLL + EPLL could cover every case better than two OLLs, all I'm saying is it's likely to be better able to cover a larger portion of ZBLL efficiently in terms of memorization and recall.

Edit: Rubiknewbie, I actually don't think that COLL + EPLL could ever be optimal. Both COLLs and EPLLs are mostly pretty long.
And to your point on 1-look vs 2- you might be right, but I'm not so sure. All you have to do after quick recognition of the COLL case is look at two edges, right? If I understand correctly, ZBLL is theoretically superior to a normal LL system composing of either OCLL/PLL or COLL/PLL because it does two things: it lowers movecount, and there are no pauses as you execute the algorithm. With a 1-look 2 alg system, you're still getting the latter benefit, and considering the speed of EPLL, I'm not convinced that its higher movecount (in comparison to ZBLL) is that much of a hindrance.

Could 4Chan maybe shed some light on this? As in, provide LL times, and maybe an opinion as to which of these aspects (movecount or fluency) of ZBLL give him the bigggest advantage?

 

[nck]

My point was simply that overall, I think that in most situations OLL + OLL is not as good as COLL + EPLL. I understood fully what he was saying. I happen to disagree, but I understand what I'm disagreeing with. I don't think that COLL + EPLL could cover every case better than two OLLs, all I'm saying is it's likely to be better able to cover a larger portion of ZBLL efficiently in terms of memorization and recall.

Edit: Rubiknewbie, I actually don't think that COLL + EPLL could ever be optimal. Both COLLs and EPLLs are mostly pretty long.
And to your point on 1-look vs 2- you might be right, but I'm not so sure. All you have to do after quick recognition of the COLL case is look at two edges, right? If I understand correctly, ZBLL is theoretically superior to a normal LL system composing of either OCLL/PLL or COLL/PLL because it does two things: it lowers movecount, and there are no pauses as you execute the algorithm. With a 1-look 2 alg system, you're still getting the latter benefit, and considering the speed of EPLL, I'm not convinced that its higher movecount (in comparison to ZBLL) is that much of a hindrance.

Could 4Chan maybe shed some light on this? As in, provide LL times, and maybe an opinion as to which of these aspects (movecount or fluency) of ZBLL give him the bigggest advantage?

What this is I don't even.

 

[Anonymous]

11 moves for the Us, 12 for the H, 15 for Z, HTM. Even if you don't consider those long, what I meant was that they're long compared to what the optimal solution would be. I highly doubt that ~10 move COLL and an (at least) 11 move EPLL will even come close to being an optimal solution for a ZBLL. However, if you read my post, you would realize that I'm not going for optimality.

 

[4Chan]

Lol, my LL varies from sub-second(sune and a few 2GLLs) to 6 seconds(Bad cases).
It all depends on the case.

Also, you imply that algorithms with low movecounts can't be fluent.
You just choose algorithms that are both low movecount and fluent.

Cube Explorer can spit out a lot of algorithms within 3-4 moves of optimal. You can even choose which faces to omit, and even slice turns.
(If you know how to use it.)

Pic related: They're 3-gen RUF T perms within 14 and 15 moves.

 

[Kirjava]

I understood fully what he was saying. I happen to disagree, but I understand what I'm disagreeing with.

If you understood, you would not disagree.

I'm done wasting my time on you.

 

[Anonymous]

4Chan: By fluency, I meant the ability to not have to pause between algorithms, which can be acheived by using ZBLL or some other 1LLL. My main point was that fluency meant in this way is acheived with both 1-look 2-alg combinations and true, direct, 1-alg solves. The latter, however, usually has a lower movecount. My question to you was mostly whether its the movecount or lack of pauses in the LL which makes you think that ZBLL is good.

Kirjava: If you'll bother reading this post, can I just try to clarify myself one more time? I agree that there will be cases where OLL/OLL is better, and that there will be cases where there simply isn't a good OLL combination. I got sidetracked from my main point, which wasn't even that COLL/EPLL is necessarily better than OLL/OLL. All I was trying to say was that recall would be easier with COLL/EPLL if you were to attempt to cover a large portion of ZBLL with 2-alg combinations.