• Welcome to the Speedsolving.com, home of the web's largest puzzle community!
    You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

    Registration is fast, simple and absolutely free so please, join our community of 40,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!

Is 1LLL possible?

Aurichalcite

Member
Joined
Apr 30, 2015
Messages
20
Location
United States
WCA
2015ROSA08
I know it's getting harder and harder to improve from our current 3x3 world record. Feliks is probably one lucky solve away from a record that will stand for years. But what if you can shave off more time by 1LLL? It is possible, or too far-fetched?
 

TorbinRoux

Member
Joined
Nov 18, 2014
Messages
99
Location
Wisconsin
WCA
2014KITT01
Its definately possible, using vls or some edge control you can get cross everytime, which shortens the algorithm count for 1lll by a ton. peole are already learning 1LLL

yeah, that doesn't count as 1LLL because you don't actually know all of the cases, you just force it to a case that you do know.
 

Bindedsa

Member
Joined
Sep 8, 2013
Messages
2,035
Location
New York
WCA
2014NURU01
YouTube
Visit Channel
Can be done but I think it will not be any faster than oll/pll because recognition and alg recall will obviously be much harder and slower...

You're just wrong here. 1LLL recog is not that harder and not that much slower than the combined OLL + PLL recog and the algs are far faster. ZBLL is clearly faster than OCLL PLL, why would the same not be true for 1LLL?
 

tseitsei

Member
Joined
Jan 12, 2012
Messages
1,374
Location
Tampere, Finland
WCA
2012LEHT01
why would the same not be true for 1LLL?

Because alg/case recall gets harder and harder as the size of the set increases. And recognition is also obviously harder because you have more things to look at (in pll you need to know permutation of edges and corners. In zbll you need to know that as well as orientation of corners. And in 1lll you need to worry about eo as well...)

But yeah. That was just what I think. There really is no way of knowing if that's true or not until people actually learn 1lll
 

Bindedsa

Member
Joined
Sep 8, 2013
Messages
2,035
Location
New York
WCA
2014NURU01
YouTube
Visit Channel
Because alg/case recall gets harder and harder as the size of the set increases. And recognition is also obviously harder because you have more things to look at (in pll you need to know permutation of edges and corners. In zbll you need to know that as well as orientation of corners. And in 1lll you need to worry about eo as well...)

But yeah. That was just what I think. There really is no way of knowing if that's true or not until people actually learn 1lll

Recall is just a matter of practice, you had trouble recalling OLLs after you first learnt them. ZBLL did take a long time to get used to and as I am still changing algs around, recall is still an issue sometimes, but there is no metric where OCLL or COLL is faster. Even without EO skips, ZBLL makes up a disproportionate amount of my above average solves. All of my fullstep sub 6s have been 1LLL.

By the same logic, Recognizing OLL is harder than OCLL. You don't think damn, I didn't get solved EO this is going to be hard to recognize. It's possible that these cases are harder to recognize, but you wouldn't say don't learn OLL, he recog is terrible.

Your opinion is your opinion and I can see the argument that it's not "worth" learning, because that is dependent on how much you dislike learning algs. As someone who uses 1LLL on close to a fourth of my solves, I'm telling you it is faster.
 

Christopher Mowla

Premium Member
Joined
Sep 17, 2009
Messages
1,184
Location
Earth
YouTube
Visit Channel
I'm surprised that no BLD solvers posted that they solve the entire cube with "1-look".

So besides that "trivial" answer...

If you define 1LLL as executing a single algorithm on a cube in which its first two layers are solved but its last layer is scrambled to:
A) mess up the first 2 layers
B) restore the first 2 layers and solve the last layer (without first making the last layer more solved for which then you would need to apply one or more algorithms to repeat the process until the last layer is completely solved), then it is possible in two ways:

Assuming you do not solve the first two layers in any special fashion:
1) You memorize 1211 algorithms and what specific case they apply to.
2) You use my (unpublished) mathematical human method. (Technically, my method is 1-look entire nxnxn cube/minx^n).

However, as others have said about option #1, it's probably not going to be that much faster (if at all) than doing speed optimized OLL and PLL.

I'm still trying to find a "smart" way to make option 2 a reality, but the disadvantage of using it is that it's pretty slow (not speedsolving material).

In short, we've pretty got the best speedsolving last layer breakdown right now.
 

Bindedsa

Member
Joined
Sep 8, 2013
Messages
2,035
Location
New York
WCA
2014NURU01
YouTube
Visit Channel
I'm surprised that no BLD solvers posted that they solve the entire cube with "1-look".

So besides that "trivial" answer...

If you define 1LLL as executing a single algorithm on a cube in which its first two layers are solved but its last layer is scrambled to:
A) mess up the first 2 layers
B) restore the first 2 layers and solve the last layer (without first making the last layer more solved for which then you would need to apply one or more algorithms to repeat the process until the last layer is completely solved), then it is possible in two ways:

Assuming you do not solve the first two layers in any special fashion:
1) You memorize 1211 algorithms and what specific case they apply to.
2) You use my (unpublished) mathematical human method. (Technically, my method is 1-look entire nxnxn cube/minx^n).

However, as others have said about option #1, it's probably not going to be that much faster (if at all) than doing speed optimized OLL and PLL.

I'm still trying to find a "smart" way to make option 2 a reality, but the disadvantage of using it is that it's pretty slow (not speedsolving material).

In short, we've pretty got the best speedsolving last layer breakdown right now.

What can I do short of learning learning full 1LLL, which I plan to, to prove to you that it is faster than OLL PLL?
 

Christopher Mowla

Premium Member
Joined
Sep 17, 2009
Messages
1,184
Location
Earth
YouTube
Visit Channel
What can I do short of learning learning full 1LLL, which I plan to, to prove to you that it is faster than OLL PLL?

Why did you ask?
it's probably not going to be that much faster (if at all) than doing speed optimized OLL and PLL.
I didn't rule out the possibility because I know that it's reasonable to assume that 1LLL will be faster for a subset of cases, whereas OLL+PLL will be faster for (for probably most) others. In real life, it's never all or nothing. There are always exceptions and patterns do not always exist/or are clear.
 

Bindedsa

Member
Joined
Sep 8, 2013
Messages
2,035
Location
New York
WCA
2014NURU01
YouTube
Visit Channel
Why did you ask?
I didn't rule out the possibility because I know that it's reasonable to assume that 1LLL will be faster for a subset of cases, whereas OLL+PLL will be faster for (for probably most) others. In real life, it's never all or nothing. There are always exceptions and patterns do not always exist/or are clear.

OLL + PLL being better han 1LLL is extremely rare, in over a thousand cases that I use in solves, Ive preferred OLL PLL maybe 2 dozens times and it's more likely that I just need to search more to find something worth using.

I just don't get why people are in anyway confident 1LLL won't work. I've learn maybe 1.5k algs in the two years, I think that is enough to prove that 1LLL can be memorized in a reasonable amount of time when you consider how long most of the high level cubers have been at it. The set that I have finished, ZBLL and have yet to finish refining already clearly proven to be faster than the OLL PLL alternative, with VHLS as a LS method it is already extremely close to full OLL PLL

In short, we've pretty got the best speedsolving last layer breakdown right now.

Of course I can't prove it without a doubt, but what would it take to at least start to show you the potential of 1LLL?
 
Top