[Help Thread] CFOP Discussion

[uberCuber]

Yeah he's not doing 2-Look OLL in that video.

Technically he is, it definitely took 2 algs before the last layer was fully oriented. He just happened to solve the permutation at the same time.

But yes, a couple things are definitely true:
1. Feliks does know full OLL.
2. Sub-10 is possible without full OLL.

 

[SoKeT]

Yeah he's not doing 2-Look OLL in that video. One of the solves he does edge orientation (4th solve) but that's because he's setting up the case for ZBLL, which is far more advanced (hundreds of algorithms instead of the "mere" 57 for OLL).

I concur though that you don't need to learn full OLL. Definitely finish those PLLs though and work on LL since 8 seconds with 3LLL can be improved (Feliks can do 3LLL in probably like 4 seconds or less). Good to hear that you think Cross + F2L is the most important, I agree with that too. I think it's the most interesting and fun aspect of improving as well. I've worked primarily on looking ahead and that has gave me the biggest improvement.

Average ~13.5 sec.
Cross ~1.5 sec
Rest of F2L ~6.5 sec
1/2 Look OLL + PLL ~5.5 sec
I know around 25 OLLs and Full PLL

Yeah he's not doing 2-Look OLL in that video. One of the solves he does edge orientation (4th solve) but that's because he's setting up the case for ZBLL, which is far more advanced (hundreds of algorithms instead of the "mere" 57 for OLL).

I concur though that you don't need to learn full OLL. Definitely finish those PLLs though and work on LL since 8 seconds with 3LLL can be improved (Feliks can do 3LLL in probably like 4 seconds or less). Good to hear that you think Cross + F2L is the most important, I agree with that too. I think it's the most interesting and fun aspect of improving as well. I've worked primarily on looking ahead and that has gave me the biggest improvement.

Average ~13.5 sec.
Cross ~1.5 sec
Rest of F2L ~6.5 sec
1/2 Look OLL + PLL ~5.5 sec
I know around 25 OLLs and Full PLL

Those are the times I'm trying to achieve in the next two months. I forgot to mention I already know around 20 OLLs, so it's not like I know the crosses and nothing else.

One question if you don't mind, how long has it been since you had 20 second averages? I know the learning process varies from cuber to cuber but just to have a general idea.

 

[SoKeT]

Yeah he's not doing 2-Look OLL in that video. One of the solves he does edge orientation (4th solve) but that's because he's setting up the case for ZBLL, which is far more advanced (hundreds of algorithms instead of the "mere" 57 for OLL).

I concur though that you don't need to learn full OLL. Definitely finish those PLLs though and work on LL since 8 seconds with 3LLL can be improved (Feliks can do 3LLL in probably like 4 seconds or less). Good to hear that you think Cross + F2L is the most important, I agree with that too. I think it's the most interesting and fun aspect of improving as well. I've worked primarily on looking ahead and that has gave me the biggest improvement.

Average ~13.5 sec.
Cross ~1.5 sec
Rest of F2L ~6.5 sec
1/2 Look OLL + PLL ~5.5 sec
I know around 25 OLLs and Full PLL

Those are the times I'm trying to achieve in the next two months. I forgot to mention I already know around 20 OLLs, so it's not like I know the crosses and nothing else.

One question if you don't mind, how long has it been since you had 20 second averages? I know the learning process varies from cuber to cuber but just to have a general idea.

 

[Lchu613]

If your F2L needs faster, just turn super slow without pausing. Its a good drill.



It would help a bit with recognition but probably not much.

Uh yeah that actually really helps.
F2L without pauses is vital. Get familiar with pairing positions until the intuitiveness becomes drilled into algorithms.
LOOKAHEAD. Make sure you see that next pair while you're doing the first one. Might take a while. That's where going slowly comes in; instead of doing each pair as fast as you can, do it slowly and give yourself time to figure out the next pair.

Cross too is important, you can watch some videos just to see some cross tricks. Remember that any cross case can be solved in 8 moves or less. Try to hit that mark.

Learn full PLL and 2-look OLL (although I think you already have 2-look OLL and 2-look PLL). One thing I learned is that you really have to get the recognition of the cases down. Just stare at them from all sides, figure out how each PLL is special. Eventually your goal should be to be able to recognize any PLL case without cube rotations (looking at all the sides of the cube).

Also you don't need to time every solve. I really almost never time myself and never used a computer timer until I was sub-30. Just practice, you should be able to tell when you get more efficient.

 

[Randomno]

Uh yeah that actually really helps.
F2L without pauses is vital. Get familiar with pairing positions until the intuitiveness becomes drilled into algorithms.
LOOKAHEAD. Make sure you see that next pair while you're doing the first one. Might take a while. That's where going slowly comes in; instead of doing each pair as fast as you can, do it slowly and give yourself time to figure out the next pair.

Cross too is important, you can watch some videos just to see some cross tricks. Remember that any cross case can be solved in 8 moves or less. Try to hit that mark.

Learn full PLL and 2-look OLL (although I think you already have 2-look OLL and 2-look PLL). One thing I learned is that you really have to get the recognition of the cases down. Just stare at them from all sides, figure out how each PLL is special. Eventually your goal should be to be able to recognize any PLL case without cube rotations (looking at all the sides of the cube).

Also you don't need to time every solve. I really almost never time myself and never used a computer timer until I was sub-30. Just practice, you should be able to tell when you get more efficient.

I have improved a lot from practising since my last post. I think my main point of improvement is the cross. I always get worried during inspection as I try to plan out some moves to get the first two pieces in. Often I notice two or three moves to put in half the cross and start the timer. I can place in edge pieces on the opposite colors (e.g. putting the white-blue piece between the white and green centres) but not adjacent. Sometimes I put two opposite pieces the wrong way round which I only notice during F2L (= whole solve messed up). I don't really check the time I'm at by the end of the cross.

My F2L is still pretty horrible. I don't have a bad color scheme or light but I still sometimes confuse orange and red. I want to learn a few algorithms for tricky cases (I know R U R' U' R U R' U' R U R' but I've only used it once in a solve).

My OLL is okay though I don't like some of the algorithms. Probably better sticking to them though, they're not that bad.

Like I said I use 2-look PLL. I don't know why you need algorithms like the E Perm if there are always two corners solved. I know H, J(b), T, U(a), U(b), and Y, but I can't recognise T and Y because I use them to permute the corners, and J because I know it from the T perm.

I'm planning on giving another method a go soon but I can't tell if I'll like them before I spend a lot of time learning them.

 

[TDM]

Sometimes I put two opposite pieces the wrong way round which I only notice during F2L (= whole solve messed up).

Rotate so the two edges are in the front and back and do M2 U2 M2.

I don't know why you need algorithms like the E Perm if there are always two corners solved.

Do Lw U' R' D R U R' D' R U R' D R U' R D' F x. Two corners are solved, but to solve that case you need U' then E perm. You do need every PLL, as unless you do an OLLCP you could have any of them.

I can't recognise T and Y because I use them to permute the corners, and J because I know it from the T perm.

Y perm:

Quite easy to recognise from that angle.
Jb perm:

Three on the left, two on the right. Don't confuse it with this!
T perm:

Easy to recognise. I'm bad at describing it, but you have two pairs opposite each other. Also, if you are looking at it from this angle, you have to solved corners and the edge between them goes opposite them, and one of the corners has a solved edge next to it.

 

[Randomno]

Rotate so the two edges are in the front and back and do M2 U2 M2.

Thanks.

Do Lw U' R' D R U R' D' R U R' D R U' R D' F x. Two corners are solved, but to solve that case you need U' then E perm. You do need every PLL, as unless you do an OLLCP you could have any of them.

I don't know what you mean. All I'm saying is that after OLL either two adjacent or opposite corners will be in the right place.

 

[uberCuber]

I don't know what you mean. All I'm saying is that after OLL either two adjacent or opposite corners will be in the right place.

If you are just doing 2-look PLL, then yes, you only need either the Y perm or the E perm, not both. I find the E perm faster, myself.

 

[Randomno]

If you are just doing 2-look PLL, then yes, you only need either the Y perm or the E perm, not both. I find the E perm faster, myself.

The E perm swaps two sets of corners though, though one set will already be solved after OLL.

 

[uberCuber]

The E perm swaps two sets of corners though, though one set will already be solved after OLL.

If there are diagonally swapped corners, doing an E perm will solve corner permutation.

 

[Randomno]

If there are diagonally swapped corners, doing an E perm will solve corner permutation.

Huh. Weird.

 

[uberCuber]

Huh. Weird.

Do your Y perm or something to a solved cube to get it to have a diagonal swap. Now, from any angle, do your T perm to swap two adjacent corners. This will cause those two corners to now be solved relative to each other. Since, on the diagonal swap case, doing an adjacent swap from any angle will give you two solved corners, it follows that if you do another adjacent swap on the opposite side, those two corners will be solved, too, giving you a total of all four corners solved. This is how the E perm works, by doing both of those adjacent swaps at the same time.

In fact, just look at the diagonal swap case. Pick any two adjacent corners, and look at the stickers on each piece that are opposite each other (for example, the L sticker of the ULF corner and the R sticker of URF). Those two stickers are the same color, so you should be able to visualize those two pieces swapping places and ending up with those matching stickers on the same face.

 

[kcl]

Hello, I used to solve the 3x3 in 16 seconds on average 3 years ago, but I stopped cubing completely until about a month ago. I picked up a lot of the knowledge I already had but something I need to work on is looking ahead since I completely lost that skill. Right now I'm very inconsistent, sometimes I solve the cube in 18-22 seconds, sometimes in 26-30, most of the time in between those times. I've done 5 sub 16 times I believe and my PB since I came back is 14.XX (it was 12.8 before, PLL skip, though the 14.XX was an OLL skip). What I'd like to achieve again is getting sub 20 times consistently, and then I can work on becoming sub 15. Now that I've introduced myself I'm gonna talk about my current times a little bit more in depth.

I always solved the cross on top, but since I came back I'm solving it on bottom and so far it's really hard to transition to the first F2L pair. I solve it in under 4 seconds on average.

When my look ahead was pretty decent, I remember solving cross + F2L in under 10 seconds, but right now it takes me 16 seconds for that.

OLL + PLL takes me another 7.97 seconds on average (150 solves on LLtrainer). I do a 2 look OLL, but that's actually not the problem since I recognize and execute the cases decently fast. Only problem is when I have a dot case (no edges oriented) since I have to do 3 algorithms, but I'll learn those cases soon. Then there's PLL. I know 18/21 right now, but the 4 G's I learnt them yesterday, so my recognizion takes as much time as my execution on those cases. I'll learn the remaining 3 cases in the following 2 days. As a side note, before I stopped cubing I knew around 12 PLL's, so most of the time I had to do a 3-4 look last layer and yet I managed to average ~16 seconds. This leads to something I already knew: Cross + F2L is the most important part, and at the moment that's what I'm practicing most, with the metronome method.

I'm certain I'll be solving everything faster, specially the last layer, once some new cubes arrive home (HuanYing, Aurora, Zhanchi, Cyclone Boys). My only speed cube today is an old Guhong that locks up a lot and whose springs sound rusty. It's completely dry after 2 years left alone, so yeah, I'm impatient for my new cubes, which should arrive in the next week.

So after all of this info about me, it would be nice to be given some advices by the most experienced cubers, and some info about yourselves aswell:

PB:
Ao5:
Time to solve cross:
Time to solve F2L + cross:
Time to solve LL:
Amount of OLLs known:
Amount of PLLs known:
Years you've been speedcubing:

I would really appreciate if you could answer these questions, and if you'd like, give some advice you think helped you achieve your current times.

Thanks.

Hmmm
PB: 6.92
ao5: 9.74
Time for cross: ~2 unless an Xcross
time for f2l+cross: 7-9
LL: 3-5 sec
OLLS: All 57 plus some alternate algs
PLLS: all 21 plus some alternate algs
Years speedcubing: coming up on my first year this month

 

[SoKeT]

Hmmm
PB: 6.92
ao5: 9.74
Time for cross: ~2 unless an Xcross
time for f2l+cross: 7-9
LL: 3-5 sec
OLLS: All 57 plus some alternate algs
PLLS: all 21 plus some alternate algs
Years speedcubing: coming up on my first year this month

Woah, I'm jelly. One year after I learnt to solve the cube I was solving it in 40 seconds or so. Congrats on your amazing progression!

 

[kcl]

Woah, I'm jelly. One year after I learnt to solve the cube I was solving it in 40 seconds or so. Congrats on your amazing progression!

Thank you haha

 

[TDM]

So after all of this info about me, it would be nice to be given some advices by the most experienced cubers, and some info about yourselves aswell:

CFOP average: 17 (I switched method a few days ago)
PB: 9.23
Ao5: 12.84
Time to solve cross: 3
Time to solve F2L + cross: 10
Time to solve LL: 7
Amount of OLLs known: 31
Amount of PLLs known: 21, plus U perms from two different angles, A perms from 3, L perm from 3, F perm from 2 and Y perm from 2
Years you've been speedcubing: Exactly 1 year and 1 week at the time of switching methods (28/12/13).

 

[Ipawds]

How do I improve my times?

Right now, I am using CFOP, and I am averaging around 29 seconds, but I really want to get sub-20 average.

My average times for each section are:

Cross: 5.574 seconds

F2L(Pairing pieces up + apply algorithm + 80% look ahead): 13.454 seconds

OLL (2 look): 4.94 seconds

PLL (I know all but 1 G-Perm) : 5.762 seconds

Please tell me which specific sections I could improve on. Thanks in advance

 

[IanTheCuber]

Definately practice that cross.

If you go at full speed as long as possible, slow down. Try using a metronome and slowly increase your turning speed. Control > speed.

 

[Dene]

You should focus on improving cross and PLL. Mostly at your stage it's just a lot of practise. My best advice is to set yourself a specific target. My motivation to get to sub20 was my decision to reach it in 6 months (which I did). I was also motivated by an article I saw about Dan Knights getting sub26 in 4 months (I think those were the numbers). I was determined to be as good as him so I made sure I did the same

 

[SoKeT]

I recorded myself doing an average of 12 in order to be aware of all or most of my flaws by getting people better than me watch me solve. I'm aware of some because they are pretty obvious but I don't want whoever decides to watch my video to be biased, so if you spot a weakness in my solves please don't hesitate to tell me.