[Unofficial] 4x4 Layer By Layer in 1:48.74 - tips please!

[Briosheje]

Hopefully you'll enjoy the quality...

Anyway.. Can someone help me giving some tips? I really can't get sub 1:35 averages right now, even though i'm pretty sure it is possible.

Forgiving the bad LL (I could have done it in one algorithm, but I completely forgot the commutator so I did it in 3 steps), is there something where i can get better at?

 

[vcuber13]

i would suggest learning k4, its quite similar to what you are doing.

here is a thread with a video of me doing a 1:10 average of 5
http://www.speedsolving.com/forum/showthread.php?28440-1-09.77-K4-Average-of-5

the main way to get faster (especially with big cubes) us to just practice.

also theres a better way of inserting the edges:
something like R U' r U R' U' r'

 

[qqwref]

First, basic stuff: practice more until you can lookahead since there are pauses almost every step, work on LL (http://snk.digibase.ca/k4/7.htm), work on turnspeed on the commutators/algs.

Bigger stuff: I think you can improve this method by changing it up a little. Try this change:
1L centers -> 1L edges -> 1L corners -> 2L centers -> 2L edges...
into: 1L centers -> 1L edges -> 2L centers -> 1L corners + 2L edges.
I like the way you insert the 3L center pairs (with the F move), but make sure to do an adjusting u move after the 2nd layer if it helps. In this video it looked like a U2 would have been good.
As for the last layer, try to move around and solve paired edges first, because they're the easiest to deal with (3x3 algs). Then you can finish up the last few pieces with easier recognition.

 

[Briosheje]

First, basic stuff: practice more until you can lookahead since there are pauses almost every step, work on LL (http://snk.digibase.ca/k4/7.htm), work on turnspeed on the commutators/algs.

Bigger stuff: I think you can improve this method by changing it up a little. Try this change:
1L centers -> 1L edges -> 1L corners -> 2L centers -> 2L edges...
into: 1L centers -> 1L edges -> 2L centers -> 1L corners + 2L edges.
I like the way you insert the 3L center pairs (with the F move), but make sure to do an adjusting u move after the 2nd layer if it helps. In this video it looked like a U2 would have been good.
As for the last layer, try to move around and solve paired edges first, because they're the easiest to deal with (3x3 algs). Then you can finish up the last few pieces with easier recognition.

I agree about almost all you said but what i wanted to show in this video is a layer by layer, and since it's Layer By Layer I do one layer at time... I mean that it wouldn't be layer by layer anymore if i will do 1L centers -> 1L edges -> 2L centers -> 1L corners, since before completing the first layer i'm already doing the second one, which is like doing F2L instead of 1st layer edges, corners and then 2nd layer edges in the 3x3...

I'm however pretty confused about the third layer... Instead of using the method i'm using (which is like k4 centers), is there any other "faster" way to make it better? I mean, I'm actually using some centers commutators (some basic centers commutators to be honest) but I actually don't know if they are faster than the way I'm using... Do you know any other way to do that?

Thanks,

Briosheje

 

[qqwref]

As for the first layer: OK, fair enough that you don't want to save the corners for later. You should at least insert them with block building though (for instance: make one edge, then when you make the 2nd edge you also add the corner in between, then one more corner with the 3rd edge, and the last corner with the 4th edge).

For the 3rd layer: There aren't really too many nice ways to do it. One thing that is helpful is commutators like rUr' u rU'r' u' - you can solve two centers of different colors at once that way.

 

[coinman]

I'm also solving big cubes in a sort of layer by layer style, a completely homemade method except for the parity's. But with one big difference from what you are doing, I solve al centers first and then a cross on white. Continuing with doing CE pairs like on a 3x3 with one edge (it doesn't matter if it's the upper or lower edge under the corner that connects) and solving the missing edge with keyhole-method. The last edge that cant be solved with keyhole I have some special tricks for. Last layer I solve by first orienting the edges and then permute them into pairs (if i know the case I do both simultaneously). Then I orientate the corners and do a PLL.
This method works ok for me, my PB is 1:34.80 but I turn mush slower then you and maybe a third of the speed of the fastest guys so i know someone fast cud do well under a minute with my method.

Why do I solve layer by layer? This is what i cam up with on my one and I'm simply to lazy to learn something else

I will try to make a video sometime to show people what exactly i'm doing and have some suggestions on to improve.

 

[cubeflip]

I would suggest the Yau Method, a version of the reduction method: http://www.youtube.com/watch?v=mpMpc8gmDdE

 

[Briosheje]

i would suggest learning k4, its quite similar to what you are doing.

here is a thread with a video of me doing a 1:10 average of 5
http://www.speedsolving.com/forum/showthread.php?28440-1-09.77-K4-Average-of-5

the main way to get faster (especially with big cubes) us to just practice.

also theres a better way of inserting the edges:
something like R U' r U R' U' r'

Ok, maybe i didn't explain well what I mean by "tips". I already know almost the entirely K4, but my goal actually is to get better time with LAYER BY LAYER method, not k4 or not any other method. My goal is to get sub 1:30 averages with this method, by being as fair as possible according to the layer by layer rules.
I know k4 is faster than layer by layer and I'm averaging around 1:10 too with k4, but it's not the subject of this video

As for the first layer: OK, fair enough that you don't want to save the corners for later. You should at least insert them with block building though (for instance: make one edge, then when you make the 2nd edge you also add the corner in between, then one more corner with the 3rd edge, and the last corner with the 4th edge).

For the 3rd layer: There aren't really too many nice ways to do it. One thing that is helpful is commutators like rUr' u rU'r' u' - you can solve two centers of different colors at once that way.

Ok then I'll try this blockbuilding way instead of the one I'm using and also thanks for the third layer centers commutator I was just working on to find something like that!

I'm also solving big cubes in a sort of layer by layer style, a completely homemade method except for the parity's. But with one big difference from what you are doing, I solve al centers first and then a cross on white. Continuing with doing CE pairs like on a 3x3 with one edge (it doesn't matter if it's the upper or lower edge under the corner that connects) and solving the missing edge with keyhole-method. The last edge that cant be solved with keyhole I have some special tricks for. Last layer I solve by first orienting the edges and then permute them into pairs (if i know the case I do both simultaneously). Then I orientate the corners and do a PLL.
This method works ok for me, my PB is 1:34.80 but I turn mush slower then you and maybe a third of the speed of the fastest guys so i know someone fast cud do well under a minute with my method.

Why do I solve layer by layer? This is what i cam up with on my one and I'm simply to lazy to learn something else

I will try to make a video sometime to show people what exactly i'm doing and have some suggestions on to improve.

That's far more better than the layer by layer method I'm using basicly because, in my hopinion, you do centers first... Your method is quite similar to reduction, though, and i'm pretty sure you can get sub 1:10 averages with it

I would suggest the Yau Method, a version of the reduction method: http://www.youtube.com/watch?v=mpMpc8gmDdE

As what I said to vcuber, i'm not looking for a faster method (and yeah, i know yau method too) (and my main method is a variation of the Yau method), i'm just looking for some tips to get faster with a "pure" layer by layer method, not because I want to use it in competition or somewhere, but just because I want to see if it actually is possible to do sub 1:30 averages with this method.

 

[Kirjava]

Of course it's possible to be sub1:30 with pure LBL. Sub1 is probably possible even.

It's like intentionally doing LBL instead of F2L though. Possibly cool but rather pointless.

 

[MrIndianTeen]

I would suggest getting better at the last layer and maybe some of your centers.

 

[Briosheje]

I would suggest getting better at the last layer and maybe some of your centers.

Uhm... Okay for the last layer, but what about the centers?

Of course it's possible to be sub1:30 with pure LBL. Sub1 is probably possible even.

It's like intentionally doing LBL instead of F2L though. Possibly cool but rather pointless.

In fact it IS pointless, and it's the main goal of the method. What i mean by that is that doing F2L actually is absolutely faster than LBL, and my goal actually is to see what times i can get doing LBL instead of F2L...
Anyway.. Do you have any tip for the third layer? I'm getting pissed off by it...

 

[Kirjava]

1:17.08, 1:07.73, (1:05.93), (1:24.42), 1:17.84
current avg5: 1:14.22 (σ = 4.60)

Did a few solves; this method sucks so hard.

Anyway, why are you doing edges then corners for first layer? Why not do them both at once?

 

[Briosheje]

1:17.08, 1:07.73, (1:05.93), (1:24.42), 1:17.84
current avg5: 1:14.22 (σ = 4.60)

Did a few solves; this method sucks so hard.

Anyway, why are you doing edges then corners for first layer? Why not do them both at once?

Because i fail at blockbuilding... I'm working on it anyway... Even though I'm pretty slow...

Anyway 1:14.22 is INSANE. How much time does the last layer take to you?

 

[Kirjava]

LL is like 10 seconds.

 

[Tim Major]

Gah, coinman, that's exactly how I solve, the F3L, but I use yau for the centres, to do cross at the same time. Then I do CO, CP, EO, then very inefficient EP. My best is 1:09, the EP was just a U-perm.
What do you mean by parity? The way I solve EO solves OLL parity.

 

[Briosheje]

Gah, coinman, that's exactly how I solve, the F3L, but I use yau for the centres, to do cross at the same time. Then I do CO, CP, EO, then very inefficient EP. My best is 1:09, the EP was just a U-perm.
What do you mean by parity? The way I solve EO solves OLL parity.

Coinman?.. Anyway why dont you do EO and EP together? (and same for corners, even though CLL are quite hard)

LL is like 10 seconds.

Then I think I'm supposed to get far more better at it too, since it seems it's taking me 20 or more seconds...

 

[rachmaninovian]

only advice i can give is centers...
think of this approach...1 entire center at a time...maybe have a systematic approach or something..
this should decrease reaction time. i can't think of any way to adapt my avg4-look last 4 centers system to this right now though..

 

[Briosheje]

only advice i can give is centers...
think of this approach...1 entire center at a time...maybe have a systematic approach or something..
this should decrease reaction time. i can't think of any way to adapt my avg4-look last 4 centers system to this right now though..

This is Layer By Layer. 1 entire center at time means reduction / k4. This is not k4 and not even reduction All what I did in the video is intentional, i'm not looking for a better way to solve it, I'm just looking for some better "tips" for getting faster at every step of the method

 

[rachmaninovian]

This is Layer By Layer. 1 entire center at time means reduction / k4. This is not k4 and not even reduction All what I did in the video is intentional, i'm not looking for a better way to solve it, I'm just looking for some better "tips" for getting faster at every step of the method

sorry i didn't clarify.. up to F2l it was okay. for the 3rd layer centers, perhaps finish each entire center 1 at a time, before f3l edges? eg, orange - blue - red - green, and free last center. I noticed a lot of lagging, so I would think that more spamming without thinking would be better..

 

[Briosheje]

sorry i didn't clarify.. up to F2l it was okay. for the 3rd layer centers, perhaps finish each entire center 1 at a time, before f3l edges? eg, orange - blue - red - green, and free last center. I noticed a lot of lagging, so I would think that more spamming without thinking would be better..

Ok I think I got it... But I just don't get how I am supposed to do one center at time... you mean doing something like F / F' setup then r U2 r' or something like that to pair up centers and put them to their correct spot? (to complete one center at time)