How to be sub 20

[Competition Cuber]

So i'm currently getting 25 second solves on average and my pb is 16.35. I need some tips to get sub-20 or at least better than 25.
Thanks!

Do you know 4-look last layer? If not, learn it, it will help greatly

 

[Cogman09]

Do you know 4-look last layer? If not, learn it, it will help greatly

yes, i do

I'm just wondering about any f2l tricks that would help
and i want to learn 2-look last layer (are there any good algs?)

 

[Daxton Brumfield]

cube skills typically has very good algs for oll and pll, but algdb.net is also great

 

[Astral cubing]

Just learn full pll which is 21 algs and if u know 4 look then you at least know like 3 of 21.
Learn a alg every other day and drill your algs. Hope this helps

 

[jronge94]

So i'm currently getting 25 second solves on average and my pb is 16.35. I need some tips to get sub-20 or at least better than 25.
Thanks!

I'm going to be brutally honest for a second here. We get the I'm x speed how do I become y speed waaaaaaaaay too often and it's a really annoying question if you don't provide any information. You could also probably find any answers we'd give you anywhere else. There's a lot of good resources out there.
If you still don't know what to do next, fine feel free to ask a question and I'll happely answer to the best of my knowledge.

(I'm going to make examples from my own solving)
Example of a bad question: "I'm currently almost sub-14, how do I become sub-12????" No one knows anything, what sets do I know I haven't provided any information on what I can improve one

Example of a somewhat decent question: "Hey I'm currently averaging a little over 14 seconds with pure CFOP (full OLL and PLL) and am dual color neutral (which I don't intend to change). My splits are give or take 1.5/8/1.75/2.75,what do you guys think I should work on"

A good question is combining the above question and providing a video of an ao5 or some example solves so people can get some actual footage that provides waaaay better information on how you can improve.

Quickly editing this in: I'm not trying to discourage you from asking questions, but encourage you to go exploring a bit first. Findig stuff on your own also makes you retain things better. Don't take this information in a bad way, but see it as opportunity to learn.

 

[hellfireofdoom]

Hey guys, I've been cubing since almost a year now, and I'm looking to get sub-20. I currently average around 25 seconds and I've learnt full PLL and on the verge of learning full OLL (still have like a few cases to learn). I don't know what I'm supposed to do to decrease my times, so I'm searching for guidance from greater cubers. I have this checklist of things I want to do on my road to sub-20 and I was wondering if you could give some feedback on it!

ROAD TO SUB 20

1. L̶e̶a̶r̶n̶ ̶f̶u̶l̶l̶ ̶P̶L̶L̶ ̶a̶n̶d̶ ̶h̶a̶v̶e̶ ̶g̶o̶o̶d̶ ̶m̶u̶s̶c̶l̶e̶ ̶m̶e̶m̶o̶r̶y̶ ̶a̶n̶d̶ ̶f̶i̶n̶g̶e̶r̶t̶r̶i̶c̶k̶s̶ ̶f̶o̶r̶ ̶t̶h̶e̶m̶
2. Learn full OLL and be familiar with all the cases(almost done)
3. Learn more F2L algorithms for different slots and improve look ahead(to improve)
4. Practice planning first F2L pair during inspection
5. Learn full COLL(40 algorithms)
6. Learn Winter Variation(27 algorithms)

Of course, I will try to improve my lookahead throughout the road, it's just that I find difficult finding the pieces. I watched this video of cyotheking talking about F2L lookahead and it recommanded doing slow solves for a week, so I think I'm going to do that soon. Thanks for anyone trying to help me, and have a good day!

 

[Daxton Brumfield]

the number one thing you should try and do is just work on efficiency. It is tough to get sub 20 when you are averaging 100 moves or so per solve so watch lazer monkey's video on f21 cases done efficiently. You also have to learn how to drill ll algs. I would try and get most plls sub 2, and all of them sub 2.5. Don't worry so much about full oll, and focus on getting sub 1 execution on all the cross cases, and implement all the easy one like the t cases. Besides the one with headlights in the front this should be easy. combining that with the time it takes you to get to a cross case, and pll, your ll should be sub 7. That leaves you with around 12-13 seconds for cross and f21. plan cross in inspection, and solve f21 efficiently and you are set. hopefully that helps.

 

[hellfireofdoom]

the number one thing you should try and do is just work on efficiency. It is tough to get sub 20 when you are averaging 100 moves or so per solve so watch lazer monkey's video on f21 cases done efficiently. You also have to learn how to drill ll algs. I would try and get most plls sub 2, and all of them sub 2.5. Don't worry so much about full oll, and focus on getting sub 1 execution on all the cross cases, and implement all the easy one like the t cases. Besides the one with headlights in the front this should be easy. combining that with the time it takes you to get to a cross case, and pll, your ll should be sub 7. That leaves you with around 12-13 seconds for cross and f21. plan cross in inspection, and solve f21 efficiently and you are set. hopefully that helps.

Thanks, I'll watch the video you recommanded. My last layer really isn't that good, so I should probably practice that as well.

 

[Daxton Brumfield]

yeah if you want sub 20 try and get at least sub 7 LL

I think you can probably get sub 6 or even 5 if you practice recognition and drilling algs

 

[hellfireofdoom]

I think you can probably get sub 6 or even 5 if you practice recognition and drilling algs

Yeah probably, certain PLL cases are just so obnoxious with their fingertricks

 

[WombatWarrior17]

• Learn full COLL(40 algorithms)
• Learn Winter Variation(27 algorithms)

Don't worry about WV that much. There are only a few cases worth learning. And don't learn Sune and Anti-Sune cases for COLL.

 

[Skittleskp]

If you want to find out what you need to improve at specifically than try this:
this is what percent of your solve time should be taken up by each step

cross: 10%
F2l: 60%
Oll:15%
Pll: 15%

this is just aproximate guidelines. I average 15 my f2l is pretty fast and my last layer takes up a lot

 

[hellfireofdoom]

Don't worry about WV that much. There are only a few cases worth learning. And don't learn Sune and Anti-Sune cases for COLL.

Got it, I don't think WV is that important for now, can probably help me later on but it prob won't be necessary for sub-20. For COLL, I learned 2 cases for the H OLL case ( the ones with two pieces of the same color next to each others) since they're quite easy.

If you want to find out what you need to improve at specifically than try this:
this is what percent of your solve time should be taken up by each step


F2l: 60%

Yeah I think F2L is probably the most important step of your solve since it's the longest part of your CFOP solve. I find mindblowing that top solvers can do F2L in under 5 seconds, it's incredible.

 

[samath]

I have been trying so hard in the past 6 months to get sub 20 average.
But i am still average about 21 to 23 secs.
Is there any use full tips on how to be sub 20?

i have pretty good f2l
ok cross
and my last layer sucks
i know 17 plls
and i dont know how much olls i know
my algs are a bit long




And my cube sucks
a super moded rubiks brand
(it is pretty smooth)

 

[Wish Lin]

I have been trying so hard in the past 6 months to get sub 20 average.
But i am still average about 21 to 23 secs.
Is there any use full tips on how to be sub 20?

Can you be more specific, like your f2l time ,how many OLL/PLL algs you know or something? A video will be the best.

 

[Llewelys]

how to be sub 20?

You can answer your own question:

my last layer sucks

What do you mean by that ?
- If you lock up a lot, drill your algs
- If it takes time to find the right alg / you hesitate during its execution, drill your algs
- If it takes time recognising cases without AUFing for nothing, learn 2 side recognition (quite easy for OLL, a bit harder but doable for PLL)

i know 17 plls

Learn full PLL, it's not that hard

my algs are a bit long

Change them. There are plenty of great ressources out there, like algdb.net

my cube sucks

There are great cheap speedcubes. You might want to check out the "Which cube should I get ? Up to date recommendations" thread.

 

[Atomix]

Hello everyone as the title states, I cannot get sub 20s often. my PB is like 14.xx (very lucky solve) but my AO5 is around 24 so here are the facts of my solves.

CFOP
Cross: 3-5 (seconds)
F2L: 10-13
OLL:3-4
PLL:~5 seconds.

I know a few oll algs.
I know full pll
I know Intuitive f2l (should I get to algorithmic f2l?)

obviously I need practice but anything else that you guys can help me with?

If you guys and gals want a video I can try to get one out but im not sure if I actually can thanks to everyone and have a wonderful day!!

I have the exact problem. I avg about 22. The thing that I am trying is advanced F2L(algorithmic approach for bad cases) and starting to Look ahead along with better cross solutions (which isnt going that well) as cross should take about 5% of your total solve time.

 

[PetrusQuber]

I have the exact problem. I avg about 22. The thing that I am trying is advanced F2L(algorithmic approach for bad cases) and starting to Look ahead along with better cross solutions (which isnt going that well) as cross should take about 5% of your total solve time.

Nice bump dude. Eh, I guess you’re continuing the current discussion.

 

[ProStar]

Ok. I see all the posts, but I'm going to sum up all of them:

Practice.

That's it. Literally what everyone is saying is either

"You need to be able to do PLL faster"

or

"Lookahead"

Both of those just come with doing hundreds to thousands of solves. There really isn't something in particular you need to do, just solve a bunch.

 

[chron0s]

Not disagreeing with anyone here - this is all great advice. My times are similar to the original poster - one thing I'm currently using to improve PLL recognition and execution is JPerms PLL trainer (jperm.net). I don't have a lot of time to practice and its a great way to isolate PLL cases that I'm looking to improve and drill them in a way that seems to translate well to real solves.