A beginner's view of EO

[eo]

I'm a beginner, just learned to finish the cube with a traditional CFOP. I'm trying to learn a new method and just have started to look at the zz and Roux tutorials. Don't know how i collected enough courage to make a post here, maybe just can't help to show my ignorance I guess...

Anyway, here it is.

No matter what methods in use, they all have to deal with edge orienting. As far as I learned:


In CFOP:
- cross, 4 pillar inserts, LL 4eo, LL corners, LL permutes...
Here, 4eo done in cross, another 4 in pillar building, and the last 4 was done explicitly as a standard step.


In Roux:
- left block, right block, top corners, LM 6eo, 2 LL edges, M permutes...
Here, 3eo done in first block, another 3 in second, and the rest 6 were done explicitly later as a standard step too.


In ZZ:
- 12eo, cross, 4 pillar inserts, LL corners, LL permutes...
Here, all eo were done in the very beginning, and no worries for them all the way to the end.


One thing I noticed during my learning is that, strangely enough, those standard EO dealing (4eo in CFOP, 6eo in Roux, 12eo in ZZ) are all results around an average moves about 6. (I know, it's not an accurate statement, definately not. Just did encounter quite some scenarios of that...)

So, will this results in a little bit of advantage of ZZ? Since it uses same amount of moves to eliminate all eo turns throughout the rest solving?

I'm not a speed solver, not fast at all. So below is what I image how you guys feeling about this:

As a speed solver, those 4 and 3 eos during the early stages of CFOP or Roux could probably absorbed in solving *without* extra overhead. I mean, the next eo, pair matching/block building are planned out during the finger movements of previous inserts. With sharp eyes and quick pattern recognition, it really doesn't matter if the next edge block is oriented or not...

And even it causes extra turns, it may result in some extra 0.1 secs I guess. In the end it's very likely an absurd fingering stuff comparing to the heavy process of ZZ pre-solving inspection.

And the reality might be on the contrary. Without the fixed barrier of the EOline, F2L might just been easier. Many shortcuts could be used to build out those blocks. (Well, ZZ claims the same thing though, I think it's the shortcuts in some cases in which we don't have to do the cross before inserts...)

So perhaps ZZ is really shining for those top cubers only. Essentially, it means stealing the (Look+Plan) time of the LL/LM edge orienting out of the solving. Also, since there are no EO in the LL, it reduced the LL cases to a number of sub 500, which is somehow falling within the human memory boundaries. A few people could ultimately able to achieve the amazing "1LLL"...

So in the cost of some extra brain cells to burn, it might help to improve the world records.

But for beginners like me, it's definately not a "squeeze out anthor half second" method which could bring me the best time...

The reason why I still decide to keep investing my time and training on ZZ is it's fun(when you are not being pushed by a ticking timer), elegant design(consistancy during the F2L, no need to treat differenly on any F2L blocks or corners), and improval potential in long run(every steps of it could be settled from very easy steps, all the way up to an extremely hard but high efficient level... for example, eo-cross-F2L could be optimized to eoLine-F2L, and might even go to eoBlock-F2L, all depends on your ability of imagenations...)

Hope this post is useful to ... er, obviously, beginners like me.

For you guys I admired here, please correct me for the wrong, and kindly bare me for the rest low level common sense garbage stuff ...

: o)

 

[teller]

Welcome to the forums! You seem like a thoughtful cuber...

Unfortunately I have little to say about ZZ. Blockbuilding is fun, but to have F, B, M, and S removed from my F2L? Eeeek!!!

 

[5BLD]

Welcome to the forums. Your username is awesome- EO rocks.

BUT, the flipside to all your talk about EO is... Sometimes EO gives restrictions, so although more ergonomic, less efficient.
Look at the Petrus method. After EO you've got an RU restriction, so in a case where MUrU'R' would suffice, you'd have to use a really long (albeit fingertricky) 2-gen alg.

In Roux the second block has the advantage that there's no EO and the M slice is free.

This is not to say that ZZ sucks.. Of course it doesn't. It's a perfectly good method like any other. But this is one of the cons of EO, and why EOCross you mentioned is not good.

 

[Kirjava]

ZZ has disadvantages that you seem to be ignoring.

 

[asportking]

ZZ has disadvantages that you seem to be ignoring.

What are they? I know it can be difficult to plan ahead very far during inspection with ZZ, but that's all I can think of.

 

[Kirjava]

<L,U,R> is not a fast movegroup to use. Thinking it is seems to be a common misconception.

 

[asportking]

True, but, at least for me when I solve using ZZ, it's not really <L,U,R>. It's more like I use L and U, with the occasional R, for the first 1x2x3 block, then only R and U for the last block.

 

[5BLD]

<L,U,R> is not a fast movegroup to use. Thinking it is seems to be a common misconception.

You could do a z then do RUL, cuz LUR is annoying. But then the view is kinda horrible

 

[eo]

Wow, didn't expect some of the ace cubers here replied... forgive me for continuing this pupil to professors talking...

I guess in most cases, those "magic turns" during F2L (F/B/M and l/r) are beautiful shortcuts to solve the bad EO and matching them with the corners in one shot.

I'm really ignorant on this: if you are solving an all-edge-correctly-oriented cube, is there still a lot chances to use those magic turns for the F2L? (Since no matter what, EO has to be retained, so for any magic turns that messes up the EO, it has to restore it back sometime later...)

So, the real advantages for a free-style F2L, is it doesn't have to maintain the 4EO belongs to top layer. This could produce some neat tricks for F2L. But just like I learned, it leaves an extra step in the beginning of the LL. (which is moved out from solving time by ZZ's pre-inspecting.)

Without the 15 secs preview time, ZZ could be the slowest methods we discussed here, i guess. (suppose a person always be faster recognize wrong edges among 6 rather than among 12s...)

 

[Tim Major]

You could do a z then do RUL, cuz LUR is annoying. But then the view is kinda horrible

I assume you mean RUD, and that's worse. RUD algs usually only contain few D moves. Doing one side UR instead of LU does not compensate for doing DR instead of RU. I suppose you could do the right block then z then left block.

 

[Kirjava]

True, but, at least for me when I solve using ZZ, it's not really <L,U,R>. It's more like I use L and U, with the occasional R, for the first 1x2x3 block, then only R and U for the last block.

...

That's <L,U,R>.

 

[eo]

thanks, didn't put too much thoughts on registering. usually a forum doesn't allow such a short user id... ; o )

yes, totally agreed with the comments. In fact, I thought Roux is the best design for an instant-on solving. It's essentially a layer by layer, but smartly leave out the M as the last step.

According to the fact: "on the cube there are turns affecting edges without touching corners, but there are NO turns affect corners only without touching edges", leave an all-edges layer as the last step looks the most efficient way on preserving the previous built work...

 

[eo]

ya, you are right. As a beginner, when I do CFOP or ZZ, i have to constantly switching hands to hold the cube...

I see many comments on ZZ including the word: ergonomic. But really don't get what it means...

 

[eo]

sort of feeling like doing the last step of F2L in Petrus method twice.

 

[Godmil]

Excellent first post. I think you may be overlooking something in CFOP/Fridrich when you move past the beginner versions the whole orientation of the top layer is done in one Alg. Some of those algs are shorter/faster than some of the ones that have the edges oriented. So having a whole extra orientation step, like in ZZ can feel like a wasted step.

 

[asportking]

...

That's <L,U,R>.

Yes, it is, but I'm not constantly switching between L and R. I'm assuming you were saying <L,U,R> isn't very good because you have to sort of regrip when you switch between L and R, but correct me if I'm wrong.

 

[Kirjava]

<L,U,R> just isn't very good generally.

 

[eo]

yup, i did looked over that. thanks for pointed out. the more steps combined together, the more chances we could find an optimized shortcut...

basically, it's OCLL vs. OLL.

 

[Rpotts]

For starters, congratulations on making a well-thought-out first post. That truly is a rarity among new members here, I think I can speak for all of us when I say we appreciate you spending time and effort on a thought-provoking post.

Now to touch on a few of the points made here -

I agree with Kir et al: LUR is not a particular nice movegroup, regripping between left hand grip and right hand is annoying, especially when both of the first two blocks in ZZ require both L and R moves, even worse if you need both L2s and R2s. It has some advantages, but, in my opinion, EO is usually the easiest substep of any method , and doing all the work of EOLine and then being restricted during F2L to only have LL edges oriented is not worth it.

Also, as Godmill pointed out - most CFOP users eventually learn full OLL (57 algs) instead of using 2 look OLL, where you start with solving EO. Now, with practice, most OLLs are every bit as fast as OCLLs, and some even faster, so there ends up being almost no difference between fast solvers OLL times compared to their OCLL times.

Solving EO in advance does increase the chances of getting skips though : in a normal CFOP solve with no edge control, your chance of an OLL skip is 1/216. If you preorient the edges like in ZZ, your probability increases by a factor of 8, to 1/27 (the number of unique OCLL cases with AUF taken into consideration.) The reason is there are 8 possible LL EO cases (four adjacent flips, two opposite flips, one 4-flip and solved) and ZZ restricts the LL to only the solved case. This means that you're substantially more likely to get an OLL skip, giving you a good chance at a fast single time.

Also, preoriented edges open up your options to using more LL systems, such as COLL, where you solve the LL corner orientation and permutation, while preserving EO, leaving just 4 EPLL cases. If you use ZZ with COLL, you have a 1/12 chance of a 1LLL, these cases are when the COLL alg you use happens to be the corresponding ZBLL. Of course you could use full ZBLL and have a 1LLL all the time, but so far no one has finished it and then practiced enough to get faster than their times with OLL/PLL. Jason Baum tried way back in 2008 or so, but lost interest in cubing and never finished. Chris Tran did finish ZBLL, and learned much of ZBLS (formerly referred to as ZBF2L) but couldn't get his recognition and alg recall down quick enough to surpass his previous times. This, along with pressures from school/work/etc lead him to disband ZBLL and forget most of what he learnt quickly. Currently, Ville Seppanen is the closest to using full ZBLL, knowing most of it other than I believe the Sune and Anti Sune set, and he's able to average ~12 seconds with it (not quite as quick as his CFOP times but promising nonetheless.)

anyway, keep cubin, and know that if you're cubing, and timing yourself / making an effort to get faster : you are a speedcuber. Doesn't matter how fast you are yet. Good luck, keep practicing, welcome to the forums.

 

[Lucas Garron]

yup, i did looked over that. thanks for pointed out. the more steps combined together, the more chances we could find an optimized shortcut...

basically, it's OCLL vs. OLL.

CLS.