L2Lk Help Needed

[TheNextFeliks]

So I decided screw CFOP, L2Lk is pro! So I am goin to switch. I find the approach very good and most of the algs pretty nice. Anyway, I would like your input.

What is the best approach to learn the algorithms? I obviously can't learn 100+ algs at once. Which order should I learn? What should I do for the other steps while I don't know the algs? I guess, what is the best beginners approach?

 

[Hypocrism]

Look gimmicky, with too many inefficient steps. Breaking up the stuff that's there is bad, which is why F2L is used.

 

[TheNextFeliks]

Look gimmicky, with too many inefficient steps. Breaking up the stuff that's there is bad, which is why F2L is used.

I am not asking your opinion on the method. I am asking for help.

 

[Renslay]

So I decided screw CFOP, L2Lk is pro! So I am goin to switch. I find the approach very good and most of the algs pretty nice. Anyway, I would like your input.

What is the best approach to learn the algorithms? I obviously can't learn 100+ algs at once. Which order should I learn? What should I do for the other steps while I don't know the algs? I guess, what is the best beginners approach?

Where do you have a full algorithm list?

By the way, I think the order of learning algorithms of a database:
1) Easiest ones / 2-look substeps
2) Hardest ones
3) Nicest ones: nice/sort ones, similar ones and/or those who belongs to a group. (By group, I mean similar cases, like W shapes in OLL.)
4) The rest, one by one. Try to group them by similarities, either by recognition/shape or their execution algorithm.

EDIT: if you know their probabilities, learn the more probable earlier.

 

[StachuK1992]

Email me at [email protected]

If you're serious, I'll relearn with you.

When I knew all of the algs, I was averaging about 23 with it, when I typically average 18 with CFOP.

With more practice, good times (sub15) are definitely doable.

EDIT: As the only person to ever use L2Lk, I figured I've give some opinion here, where criticisms arise:

L2Lk has a higher lower bound than CFOP or Roux. I'd say it has the same 3x3 speed potential as Petrus.
Move counts are pretty similar to CFOP, if not a few moves lower. I did the full analysis a while back, but I'll post some more concrete thoughts on the method in a blog post soon.

 

[TheNextFeliks]

Where do you have a full algorithm list?

By the way, I think the order of learning algorithms of a database:
1) Easiest ones / 2-look substeps
2) Hardest ones
3) Nicest ones: nice/sort ones, similar ones and/or those who belongs to a group. (By group, I mean similar cases, like W shapes in OLL.)
4) The rest, one by one. Try to group them by similarities, either by recognition/shape or their execution algorithm.

EDIT: if you know their probabilities, learn the more probable earlier.

The algs can be found on Stachu's website.

Email me at [email protected]

If you're serious, I'll relearn with you.

When I knew all of the algs, I was averaging about 23 with it, when I typically average 18 with CFOP.

With more practice, good times (sub15) are definitely doable.

EDIT: As the only person to ever use L2Lk, I figured I've give some opinion here, where criticisms arise:

L2Lk has a higher lower bound than CFOP or Roux. I'd say it has the same 3x3 speed potential as Petrus.
Move counts are pretty similar to CFOP, if not a few moves lower. I did the full analysis a while back, but I'll post some more concrete thoughts on the method in a blog post soon.

Hey look! It's a statue. I was hoping you would post here. I am serious about learning it. Do you want to learn it again? If not, don't feel obligated. But if so, it would be nice to learn it with someone. I also feel sub-15 is doable. Depending on how fast the algs and first layer are, I could see it getting close to 10. So yeah like Petrus times. What do you think kept you sup-20? Any major problems?

 

[Tim Major]

I'd say it has the same 3x3 speed potential as Petrus.
Move counts are pretty similar to CFOP, if not a few moves lower. I did the full analysis a while back, but I'll post some more concrete thoughts on the method in a blog post soon.

This doesn't make sense for me. For an "algorithmic method", if it has less than or equal to the same average moves as CFOP yet without intuition, it should be very similar, if not faster in speed.

I thought the problem with L2Lk was the movecount was high, and CFOP, whilst having "intuitive" CF, it turns into algorithms with enough practise, so L2Lk was basically pointless.

Intuition is slow. With enough practise of F2L, it isn't intuitive it's just recognise case, solve case.

If speed optimised L2Lk is actually more efficient than CFOP it should be a superior method.

What do you think kept you sup-20? Any major problems?

Same thing that stopped the user "4chan" (don't remember his name) from being fast with ZB. Case recognition/alg recall, and TPS/pauses due to lower practise than sub 10 solvers.

 

[StachuK1992]

I'm just a bad cuber. I recognize slow. I turn slow.

That, and the first layer is HARD.

 

[TheNextFeliks]

I'm just a bad cuber. I recognize slow. I turn slow.

That, and the first layer is HARD.

Ok. I am not that great either (avg like 16-17). And layer looks like a pain but with practice it could probably get faster.

 

[Kirjava]

layer is ezmode :3