How to break the 20 second barrier?

[dvtpgm]

Hi guys,

I've cubed for around 8 months (2 months beginner, 6 months Fridrich). Currently, my best single solve is 19.46 sec with best average of 24.18 sec, which isn't any impressive, given the fact that I know all the OLLs and PLLs.

So far, I've only managed to get sub-20 on 2 occasions. Also, most of the time, my individual solves range between 24 to 29 seconds.

I did a break down of my solves, and hereby determined that my cross + F2L is holding me back from sub-20. So I have a few questions regarding them.

Cross wise, is practice the only way to improve? Many times, no matter how hard I look, I cannot ever see the 8 move cross. Also, even if I do see it, I do not know what's the best way to execute the moves. I just hate doing Ds and cube rotations. Currently, my cross averages around 6 seconds. I remember there's this site with a Javascript cross solver, but it seems to have disappeared.

F2L wise, I know what my bad points are. My look ahead is pretty bad. Usually, I will only be able to trace 1 corner at a time. Also, I rely on algorithms. Hence, I always do redundant setup moves to set up the cases exactly like the way I learn them. Also, I only can comfortably solve pairs to the FR or BR slots. Must I really switch to "intuitive" mode in order to improve? Must I really learn to slot pairs from any angle? How do I go about learning them?

Sorry for asking too many questions. Just pretty eager to go sub-20. Thanks.

EDIT: To see my current solve, simply go to my youtube channel listed on the left.

 

[juone95]

Hi guys,

I've cubed for around 8 months (2 months beginner, 6 months Fridrich). Currently, my best single solve is 19.46 sec with best average of 24.18 sec, which isn't any impressive, given the fact that I know all the OLLs and PLLs.

So far, I've only managed to get sub-20 on 2 occasions. Also, most of the time, my individual solves range between 24 to 29 seconds.

I did a break down of my solves, and hereby determined that my cross + F2L is holding me back from sub-20. So I have a few questions regarding them.

Cross wise, is practice the only way to improve? Many times, no matter how hard I look, I cannot ever see the 8 move cross. Also, even if I do see it, I do not know what's the best way to execute the moves. I just hate doing Ds and cube rotations. Currently, my cross averages around 6 seconds. I remember there's this site with a Javascript cross solver, but it seems to have disappeared.

F2L wise, I know what my bad points are. My look ahead is pretty bad. Usually, I will only be able to trace 1 corner at a time. Also, I rely on algorithms. Hence, I always do redundant setup moves to set up the cases exactly like the way I learn them. Also, I only can comfortably solve pairs to the FR or BR slots. Must I really switch to "intuitive" mode in order to improve? Must I really learn to slot pairs from any angle? How do I go about learning them?

Sorry for asking too many questions. Just pretty eager to go sub-20. Thanks.

EDIT: To see my current solve, simply go to my youtube channel listed on the left.

try this video for the cross http://www.youtube.com/watch?v=hY_FbpdNRCQ


for F2L try intuitive.

 

[DAE_JA_VOO]

Intuitive F2L. Definitely.

 

[qqwref]

A few ideas:
1) For cross, until you get better at doing fingertricks with D, try executing the cross with the bottom layer on F or even U. If you can spend two seconds instead of five or six executing the cross, it's okay to waste one second flipping the cube over. Later you can work on bringing the cross down to D while you're executing, which is what I think Jason Thong used to do (he was well under 15 at the time, mind you).
2) For F2L, definitely try to learn how to solve each case from a few different positions. Sometimes you can even use free slots to insert (for example stuff like B' U B R U R' can be helpful). If you want to learn a few more ways to put slots together, just spend some time doing slow intuitive F2L solves, or do a lot of practice scrambling and solving with only L, U, and R. Don't use what you know, just try to solve things in as few moves as you can.
3) About tracing pieces - trying to trace a piece through the entire algorithm isn't always helpful. Personally I find that I try to watch where a piece goes, and if I see it going near another piece, I sort of look at what case it is as it's happening. If you put too much effort and thought into lookahead, it will just slow you down.

 

[Neroflux]

instead of using D, why not Uw?

 

[Dene]

Practise. Just practise.

 

[amostay2004]

I've tried learning to do Uw, but it totally confuses me as it shifts everything on the top..and i have to stop for 2 secs to think how to perform the moves i planned....guess i'm just not used to it.

 

[brunson]

I've tried learning to do Uw, but it totally confuses me as it shifts everything on the top..and i have to stop for 2 secs to think how to perform the moves i planned.

Do it more. ;-)

 

[Neroflux]

I've tried learning to do Uw, but it totally confuses me as it shifts everything on the top..and i have to stop for 2 secs to think how to perform the moves i planned.

Do it more. ;-)

erm i had this problem too. i solved it by planning out the Uw turns clearly during inspection.

dvtpgm actually just focus on recognition..... you took a few seconds to see a u perm that was AUFed(there is no such term i know).

 

[dvtpgm]

I've tried learning to do Uw, but it totally confuses me as it shifts everything on the top..and i have to stop for 2 secs to think how to perform the moves i planned.

Do it more. ;-)

erm i had this problem too. i solved it by planning out the Uw turns clearly during inspection.

dvtpgm actually just focus on recognition..... you took a few seconds to see a u perm that was AUFed(there is no such term i know).

Recognition wise, I can recognize any perm from any angle, except U, H and Z. I always have to AUF to see the case. Yes this definitely needs improvement.

 

[Escher]

to sum up, give up those f2l cases. use what you have learn from them and go intuitive. it means that you will be able to do any case quickly and from many different angles. it will help teach you how to 'pop' difficult cases out while solving another, and will improve your lookahead.
to practice f2l, do smooth solves a LOT. not slow, smooth. there should not be any gap between turns. while executing one turn you should anticipate the next one, two or more.
spam your PLLs. go through different alg pages and decide if there are any you prefer. learn 2 more Us, so that AUF never has to be more than a U/U'
what is your OLL recognition like?

 

[PCwizCube]

Practicing Cross:
-For memorizing the cross solution, first practice without a time limit. If it takes you 10 minutes to memorize it, do it. After you can memorize it, do it blindfolded. Then GRADUALLY keep on lowering the time limit to around 12 seconds or less.
-And if you don't like D turns or cube rotations, Uw is the move for you!

F2L Improvement:
-Definitely learn intuitive. It will help you get tremendously better at F2L because you will understand a lot more and will be able to do a lot faster things.

I don't know what your last layer is like, but you should give us a breakdown of your times for each step. For example, I average around 19 seconds for Fridrich, and my break down is:
-Cross = 3 seconds
-F2L = 10 seconds
-2 Look OLL = 2.5 seconds
-PLL = 3.5 seconds

 

[Escher]

your f2l pairs could be faster rubiks cube fan...
but yes, everything he said is correct, especially about the cross. you need to be able to memorise it and place edges without looking.

 

[PCwizCube]

your f2l pairs could be faster rubiks cube fan...
but yes, everything he said is correct, especially about the cross. you need to be able to memorise it and place edges without looking.

Yeah I know. I've been trying to get a faster F2L for a long time but it's really hard! That's why I haven't started learning the rest of the OLLs.

Quote from Tyson Mao at the Newark Open 2008:

You should learn your OLLs

 

[Escher]

and he is correct.
i think that training f2l and learning algs are completely unrelated in terms of priorities... in fact, if you just did solves before practising each OLL, you'd be doing f2l over and over again. which is never a bad thing.
do some slow solves to practice look ahead
i find a good 'drill' is to do 5 smooth/slow, 5 speed, and repeat 'til you get at least two avgs of 20

 

[Hepheron]

do Uw it will take some time to get used to but in the long run it will help out a ton

PRACTICE PRACTICE PRACTICE

 

[dvtpgm]

Practicing Cross:
-For memorizing the cross solution, first practice without a time limit. If it takes you 10 minutes to memorize it, do it. After you can memorize it, do it blindfolded. Then GRADUALLY keep on lowering the time limit to around 12 seconds or less.
-And if you don't like D turns or cube rotations, Uw is the move for you!

F2L Improvement:
-Definitely learn intuitive. It will help you get tremendously better at F2L because you will understand a lot more and will be able to do a lot faster things.

I don't know what your last layer is like, but you should give us a breakdown of your times for each step. For example, I average around 19 seconds for Fridrich, and my break down is:
-Cross = 3 seconds
-F2L = 10 seconds
-2 Look OLL = 2.5 seconds
-PLL = 3.5 seconds

Breakdown wise, forgotten, seriously. But as a sample, you can take a look at my youtube videos (left icon).

Thanks for all the replies. Yes I will work harder and practice hard. Although F2L wise, I foresee a tough time transiting to intuitive, but won't give up.

 

[EmersonHerrmann]

Alrighty -

For the cross you should learn to solve is blindfolded/eyes closed, so that when you go to solve you can do the cross with little or no thought about it.
Next, you should learn intuitive F2L and do many slow solves to improve your look ahead.