[Help Thread] Improving 3x3

[thylacine]

I've been casually cubing, with several different puzzles for a few years now. I've decided to try and speed up my 3x3 times. As of now, I average 30-35 seconds with a sort of mashed together method which I'll try to explain best I can. I use CFOP, but here's how:

Cross: I'm practicing solving yellow side up, I used to solve white up then flip over cube.
F2L: I use Fridrich for F2L, I'm currently working on improving my lookahead for less pauses.
OLL: I have a sort of 2 look OLL, where i build the cross, then know all the cross OLLs
PLL: I use a T-Perm to orient the corners, and am learning more efficient algorithms for edge cycling. My main weakness is having to use two T-Perms if there aren't existing headlights.

My question is: Should I keep improving my lookahead and edge cycling, or take the dive and learn full PLL/OLL? And if so, should I learn full OLL or PLL first?

note: my PB is 24.6
EDIT: my new PB is 19.88

 

[shadowslice e]

Right now, I would suggest you have a look at this and decide if you really think CFOP is best suited for you.

After that, some tips:
1) Be as CN as possible (fully for CFOP, x2,y for Roux and ZZ or most other methods).
2) Lookahead is the single most important part of speedcubing. It will dictate how fast you can turn and how efficient you are.
3) Build up your intuition. Better to understand something than to learn a series of algs (such as in F2B & F2L).
4) With that said, do not be afraid to learn algs if it is better than intuition.
5) Remember the corners of modern speedcubing: intuition, algorithms, lookahead and fingertricks.
6) With regards to fingertricks, Try to have several ways to do each move so you can regrip and move your hands less.

Good luck in your progress!

 

[thylacine]

Right now, I would suggest you have a look at this and decide if you really think CFOP is best suited for you.

After that, some tips:
1) Be as CN as possible (fully for CFOP, x2,y for Roux and ZZ or most other methods).
2) Lookahead is the single most important part of speedcubing. It will dictate how fast you can turn and how efficient you are.
3) Build up your intuition. Better to understand something than to learn a series of algs (such as in F2B & F2L).
4) With that said, do not be afraid to learn algs if it is better than intuition.
5) Remember the corners of modern speedcubing: intuition, algorithms, lookahead and fingertricks.
6) With regards to fingertricks, Try to have several ways to do each move so you can regrip and move your hands less.

Good luck in your progress!

1)Color Neutrality is something I could definitely practice, although I personally don't think it's as important as some other parts of CFOP I should be improving.
2 & 3) It was my understanding that CFOP was based largely on a knowledge of several algorithms, and that methods such as Roux, ZZ or Petrus are more intuitive. I understand very well how pieces move about on the cube from experience, but not so much perhaps that I would commit to an entirely new method after so long.
5) Will do!
6) Hand positioning is a problem I often face on new puzzles, and on 3x3 there are some algorithms I have changed because I felt I was having to regrip the cube too often. These changes probably haven't siginificantly impacted my time, but they make solving more comfortable for me personally.

 

[gateway cuber]

I've been casually cubing, with several different puzzles for a few years now. I've decided to try and speed up my 3x3 times. As of now, I average 30-35 seconds with a sort of mashed together method which I'll try to explain best I can. I use CFOP, but here's how:

Cross: I'm practicing solving yellow side up, I used to solve white up then flip over cube.
F2L: I use Fridrich for F2L, I'm currently working on improving my lookahead for less pauses.
OLL: I have a sort of 2 look OLL, where i build the cross, then know all the cross OLLs
PLL: I use a T-Perm to orient the corners, and am learning more efficient algorithms for edge cycling. My main weakness is having to use two T-Perms if there aren't existing headlights.

My question is: Should I keep improving my lookahead and edge cycling, or take the dive and learn full PLL/OLL? And if so, should I learn full OLL or PLL first?

note: my PB is 24.6View attachment 6813

I recommend that you...
1. Solve your cross on bottom
2. work on look ahead
3. Learn full PLL
4. get you last layer below 6-7 secs on avg
5. 2-look OLL is fine as long as it take no more than 3.5 secs...

If you have any questions fire away

 

[JustinTimeCuber]

Cross:
Color neutrality is good if you are really devoted but you can be moderately fast without being color neutral (I'm color neutral and moderately fast so yeah)
Try to do it in 10 or fewer moves every time. When you're faster you want to cut that down to 8 (8 is optimal) but 10 is ok for now
Definitely do cross on bottom, and if you kinda want to be color neutral then be "x2" neutral where you can solve yellow and white cross, because they have the same F2L colors and it's way easier- I am slightly slower on yellow cross but a nice XCross can make me do yellow cross.

F2L:
Lower movecount = better*
More lookahead = better
More TPS = generally better**

*A flowing, fast 8 move alg is better than an awkward 6 move alg.
**Don't spam TPS until you have good lookahead and make sure there's a balance between the two.

LL:
Your OLL is fine as 2-gen for now, but learning full OLL gradually can be a good thing. By my first comp I was ~5 seconds faster than you right now, and I knew a few easy OLLs. You can see in this video that I use the 1-look OLL at around 0:26. (My first ever solve)

 

[thylacine]

UPDATE: I've broken down my 3x3 times as follows:
C: ~5s
F: ~15s
O: ~6s
P: ~6s

I memorized L, J and Y perms as well as one of the X perms, but I'm avoiding learning G perms for now.
I think my biggest area for improvement is my F2L. Thoughts?

EDIT: As a side note, my main cubes are:

1. White Dayan Zhanchi, with black opposite yellow. I have a tight tension on it, but well lubed.
2. White Gans 365s, with white opposite yellow. Looser tension, but smaller and harder to recognize pairs for me sometimes.
3. Fangshi Shuangren, clear with black caps with colored stickers, 53mm if I remember correctly, tiny cube. VERY fast, but have trouble sometimes with its size.

 

[Jake Donnelly]

for f2l u can do slow solves to work on look ahead cyotheking has a good video on his channel about look ahead which is a great video to watch( pro tips for look ahead)

 

[DGCubes]

UPDATE: I've broken down my 3x3 times as follows:
C: ~5s
F: ~15s
O: ~6s
P: ~6s

I memorized L, J and Y perms as well as one of the X perms, but I'm avoiding learning G perms for now.
I think my biggest area for improvement is my F2L. Thoughts?

EDIT: As a side note, my main cubes are:

1. White Dayan Zhanchi, with black opposite yellow. I have a tight tension on it, but well lubed.
2. White Gans 365s, with white opposite yellow. Looser tension, but smaller and harder to recognize pairs for me sometimes.
3. Fangshi Shuangren, clear with black caps with colored stickers, 53mm if I remember correctly, tiny cube. VERY fast, but have trouble sometimes with its size.

You should always be working on F2L, but your OLL and PLL seem to be a bit of a problem. For alg-based steps, I really recommend learning better finger tricks and drilling algs to get them fast. It's not too hard to get PLL sub-2 or sub-2.5 with a little practice, and full PLL will definitely help you with that. Don't be scared of G-perms, by the way. I never understood all of the hate they get; the algs are actually pretty nice and it's easy to recognize. Just learn them throughout learning the rest of your PLLs instead of saving them all until the end. They come up a lot and it feels good when you know them.

Also, for cross, I think you should work on developing more efficient solutions. You should be able to see your whole cross in inspection, blindfold yourself, and do it. With a little more practice, I think you could get cross sub-3.

 

[thylacine]

You should always be working on F2L, but your OLL and PLL seem to be a bit of a problem. For alg-based steps, I really recommend learning better finger tricks and drilling algs to get them fast. It's not too hard to get PLL sub-2 or sub-2.5 with a little practice, and full PLL will definitely help you with that. Don't be scared of G-perms, by the way. I never understood all of the hate they get; the algs are actually pretty nice and it's easy to recognize. Just learn them throughout learning the rest of your PLLs instead of saving them all until the end. They come up a lot and it feels good when you know them.

Also, for cross, I think you should work on developing more efficient solutions. You should be able to see your whole cross in inspection, blindfold yourself, and do it. With a little more practice, I think you could get cross sub-3.

My PLL single look algorithms are 2.5-3s, I did a set of 15 or so PLLs to try and practice them. The problem is when I combine two 2.5-3s PLL algorithms, along with recognizing a secondary case, I end up with ~6 times. I'm working on memorizing full one-look PLL, but I do have a hard time recognizing G perms currently, which is why I'm putting them off for a bit and using two look.

Blind Cross is definitely something that sounds like it would help a lot, and I'll start practicing with it ASAP. I'm about to start an ao25 because I get the feeling that I'm averaging much slower when I don't hand scramble, which is most of my solves. This will hopefully get me a better idea of my actual speed.

EDIT: Finished my ao25, with an average 33.48, but the times did vary quite a bit.

 

[Smiles]

you can focus on lookahead and learn new LL algs at the same time

learn Y perm or some other diagonal swap so you don't have to do 2 T perms when there are no headlights

 

[thylacine]

you can focus on lookahead and learn new LL algs at the same time

learn Y perm or some other diagonal swap so you don't have to do 2 T perms when there are no headlights

I've learned Y perm and Nb perm for no headlights

 

[thylacine]

UPDATE: I'm going to take advantage of labor day to experiment with Roux method, and see how I like it. Hopefully I can learn enough about it to be able to solve using a cheat sheet