CFOP Cross Efficiency, Intuitive F2L, and LL Recognition Questions.

[Ryntak94]

Hi guys I've been cubing for about 3 years now. I've been serious about being able to solve different types of puzzles and of varying sizes, but not speed.
This only being because I didn't want to take the time to sit down and memorize the OLL's so I decided to do it over time. I finally finished a couple days ago.

So I've spent the last couple days solving as fast as I can and analyzing what I could improve on. (Obviously everything seeing as I don't hold any world records nor am I even close) Before I get into everything I currently average around 35 seconds and use a stickerless ZhanChi. Color neutral.

By far the slowest part of my solve is the cross. It's not efficient and it takes at east 4-5 seconds, if not longer.
I just watched this video that I found on a different thread

And so I've decided the best thing for me to do will be to sit down and scramble the cube, then look for the most efficient cross solve and solve it as fast as possible.

My F2L is very slow as well when I'm timing myself I usually finish my F2L around 18 seconds. The problem I see with my F2L is that with my ituition what I always do the same thing. Let's say that I'm solving the Red Green edge/corner pair with white on bottom. I'll always get the red or green of the edge paired with it's center and get the corner with white facing up and I'll turn the edge away from where it's supposed to go and put the corner on top of it to pair it then I'll put the pair in. Every. Single. Time. This is what I do intuitively. I'm wondering if looking at what the algorithms do and working those out intuitively for situations where my intuition isn't the most effective is worth it. Also I don't have any look ahead so I've decided I'll work on look ahead after I speed up my cross. Doing this by looking at it then trying to solve a pair with my eyes closed and then when I get good at that move to two pairs then to three then to all four and then start trying to look for as many pairs as I can while solving one or something to that effect.

When it comes to the last layer my recognition for most of the OLL's is slow and my recognition for the E-perm and all of the G-perms is pretty bad. recently sped up my A-perms though. As far as speeding these up I have no idea how I can actively do this without just plain practicing because if I set up the specific OLL case then I KNOW what the case will be before it happens, therefore there's no help. And I'm clueless for the PLL as well.
Recognition is by far the most difficult thing for me that takes the most time. My execution of many of the OLL's is pretty bad too but I KNOW that will get faster with time. The fastest part of my solves is the execution of the PLL's

Is there anything else that I could do to speed up my solves that I haven't thought of yet? Also my ZhanChi didn't come with torpedos, so could somene link me to a thread or give me some advice about tightening/loosening the screws?
Also lubricants. I've in the past used Heavy Duty Silicone Spray but unfortunately it dissapeared and I can't what I used in the past. And everything I find in the stores contains petroleum distallates so I haven't bought anything.

I know this is a lot to take in at once, but I've been watching the forums trying to find my answers and maybe I didn't look hard enough or in the right places but as I was looking I had more and more questions and concerns about my technique and I refrained from asking them because I didn't want to ask and get those "Your answer is already in this thread" or whatever responses...

There's one other thing. I'm doing a demonstration speech in my speech class. I'm planning on attempting to teach everyone how to solve rubiks cube. While I know that the chances of anyone actually memorizing any type of method and being able to succesfully solve a rubiks cube from then on is far less than slim to none, I would like to be able to speak to them in a way in which a few people will succesfully be able to solve the cube following my instructions. (I'm going to have around 18 3x3's in the class for the other students to use). My normal approach this (1 on 1) would be to figure out how a person learns best and then teach to them that way, however teaching a group is completely different than teaching an individual. Do you guys have any suggestions? I'm already planning on having many visuals including my pictures of situations on the cube along with definining algorithm and commutator and explaning notation.

Thanks in advance guys for all the help! I know it's a lot to take in. Happy cubing!

 

[ncube]

Pretty much, you shouldn't be trying to solve as fast as possible. And to address all other questions, I'd like to refer you to my friend badmephisto on youtube. Watch his f2l tutorials and cross tutorial, and when you understand all the concepts, you should LOOKAHEAD instead of trying to turn quickly. Speed comes over time, but the main thing for you is to turn slowly and get used to looking for the pieces involved in the next step while solving your current step.

The actual cube you're using is irrelevant until you are sub-15 at most, then it's more about having a comfortable cube you can rely on than having a cube that turns super fast. Would you rather lock up like crazy all the time or have a cube you can control and rely upon?
To prove my point about turning speed, let's look at turns per second. A typical CFOP solve has around 60 moves or less, so if you are solving in 30 s, you're only really doing 2 tps. Try doing the sexy move exactly one time every second. (4tps) That speed seems slow, but it's enough to get you times of 15 s if you can maintain that speed during your whole solve. So yeah, focus on fludity and lookahead most of all, and practice and give it time.
edit: your idea of lookahead isn't really practical. After your f2l is less than 30 moves, just focus on finding the easiest next pair durring cross and f2l, and create a seamless transition. Over time, you may be able to track several pieces at once, but it isn't really possible to be able to plan the whole f2l with one look.

 

[uniacto]

A nice way to transition from cross to the inserting F2L pairs is to build the cross on the bottom efficiently that can be moved into the correct position with a D or D'. While you're doing the D or D' or D2, you can use that time to find the first two pieces for the first pair.

 

[Ramo]

Look at Badmephisto's cross stuff. Those are some helpful videos!

 

[Smiles]

For me cross came naturally. I cubed for a long time before taking speedcubing seriously, and by then I naturally built some kind of cross efficiency because of my knowledge about how the cube works. At some point you'll be so experienced with piece relationships while building the cross that you won't even really have to think.

Learning anything new can be hard to do, and transitioning to regular CFOP F2L from a beginner method / intuitive method can be annoying. It takes thinking, and who wants to do that?
Even now, I'm often changing OLL algorithms and such. When I encounter a case that I'm suddenly not familiar with because I changed my algorithm, I never worry about speed. Just worry about doing the right thing and know that eventually you'll get it. The more you do it right, the better you'll get at it.
Do not worry about look ahead until you can do every F2L case smoothly and without looking. If you start looking ahead too early, it'll mess up the pair you're working on since it's hard to focus on 2 different things at once. When you're good at the F2L cases, then it's your muscle memory solving the pair and your eyes/brain looking for the next pair.

For LL case recognition, it's just practice and repetition. For me, the hardest perm to recognize is R perm because it looks like a G perm from one angle and kind of like a T perm too. E perm is pretty easy, basically all the edges are correct OR you look around for a pattern and you don't see one.
For G perm, you need to come up with your own strategy. The way I do it is I look at the 1x1x2 bar (2 stickers of 1 colour) and find the 1 remaining sticker of that colour. Based on its position relative to the 1x1x2 bar, I know the case. Mirror of that for recognition for the other pair of G perms.

Slow OLL recognition is natural, since there are so many angles you can see it from. Again, idk if you'll find this is a good or bad thing, but experience and repetition is what will improve you the most on this. Eventually you'll be able to look ahead to OLL (especially edge orientation) to have a good idea of what case is coming up, or even predict it. But it's hard to start thinking until your muscle memory fully remembers every alg and the case it corresponds to, so again, just practice practice practice.

As for teaching, I have taught many friends 1 on 1 how to solve the cube and I find it's really hard to teach certain people. The most important thing is that not everyone understands these concepts the same way, so I tend to keep things simple and teach the easiest possible way to solve each step (someone told me to speak as if my audience is a 2 year old, and i felt stupid but it worked).
One important thing I do is teach the sexy move as a finger trick, not an algorithm, and incorporate it into the algorithms as much as I can. No one likes doing algorithms, but doing the sexy move is a fun thing everyone will start doing. for example, for teaching beginner's method 2nd layer I tend to teach this algorithm:
URUR'U'yL'U'LU. But I dont write it down. I teach it as "the edge needs to go to the right, so flick U with right hand, right sexy move, turn to right, then left sexy move." I managed to teach someone how to do the entire cube in 2 hours without even teaching notation or writing anything down at all.

 

[elrog]

I got an idea that I think would be more efficient as far as F2L goes, thuogh I'm not sure it fits into this thread because its technically not CFOP. (I consider the F2L more efficient because other things are solved at the same time) To do an extended cross with every solve, it is better to fin a way to create a 2x2x2 block during recognition as adding on the 2 extra cross pieces is shorter than an average F2L slot.

I think you could also orient the top 4 edges at the same time as doing a pair as is done in the ZB method. I was thinking that this should be done with the second to last pair rather than the last. The goal would be to get the pair solved and get all the edges oriented correctly so that no flipping of edges are required to solve the last pair. This would use all the same algorithms. The edge that goes into the last slot may be counted as a good or bad edge depending on making the number of good and bad edges even for the ones your working with, or by looking at how the edge in its place is flipped. If the edge that goes in the last slot is in the right place and flipped correctly, do the step normally. If the edge is in the slot incorrectly, turn it up to the top layer (the cross doesn't need to be fixed until after you orient the edges). I was also thinking that the cases for this may be reduced by how you put the pair before it into its slot just as in avoiding the OLL dot cases.

As for the last pair, I was thinking that you could orient the corners while placing it and finish with PLL. This would take 28 algorithms.

You could also place all of the edges at the same time as the last pair as is done in the 2nd approach for the Heise method (your pair would always be the F2L slot) and finish with L4C requiring 84 algorithms. Note that the 24 L3C cases can be done intuitively and you will memorise the cases and their quickest solutions with practice as in F2L, and learning the algorithms isn't necessary. This would still take more algorithms than doing the PLL approach, but also a few less moves on average. I'm not sure what the average recognition time is for L4C.

A third approach would be to solve a 1x2x2 block in the top layer and finish with 53 of the L5P algorithms (there are 107 total). Creating such a block with the F2L pair could be done easiest by creating a 1x2x2 block that fits into the top layer and placing it with its complementary edge while matching and placing the F2L pair. Getting the 1x2x2 block for the top layer would be easier than getting the last F2L slot matched because you have more options. This would require 72 algorithms with edges already oriented. I could generate these algorithms if needed thugh it may take some time because I'm working on some other ones atm.

I know that ZB is said to average 40 moves and around 47 for speedsolving, but this is only because of the 1 look last layer that uses tons of algorithms. I think this reduces recognition time more than it lowers move count, and is not worth it. It is also just a pain to memorise that many algs. I wander how my 3rd approach would compare in move count. I have also though about doing something similar to what is done in the synder method, but I think my 3rd option would be quicker. I'm open to any comments about which of the 3 last F2L slot options I should go with. I'm also sorry if this should be posted in a different thread.

 

[JianhanC]

A shorter suggestion than those above: for cross, remember your colour scheme (position of colours relative to each other) and spend a lot of time learning to do the cross in as little moves as possible, and don't look at it. Over time you'll want to track F2L pieces during cross execution but it's not that important now. For F2L don't stop turning and keep looking for other pieces. Also learn to insert pairs without looking. For OLL don't fall back on 2 look, keep using the OLLs you learnt, it's the best way to reinforce them. And practise turning fast for PLL, it's the last stage anyway.