OCLL vs COLL

[Ranzha]

If you do this, know U-perms from each angle without AUF.
Loljk.
Making things harder than they actually are ftw!

 

[Escher]

Learning COLL was what allowed me to go from 17 second averages to sub-17. I now use maybe 80%-80% of COLL only because I do not use hardly any sune cases or some other slow cases. I like COLL, but some cases I think have slow enough execution as to not be worth doing.

Chris

Haven't you heard of CLLEF? It is just like COLL, but it works on the cases with no edges correctly oriented
I think some algs are on the wiki.

 

[LarsN]

Learning COLL was what allowed me to go from 17 second averages to sub-17. I now use maybe 80%-80% of COLL only because I do not use hardly any sune cases or some other slow cases. I like COLL, but some cases I think have slow enough execution as to not be worth doing.

Chris

Haven't you heard of CLLEF? It is just like COLL, but it works on the cases with no edges correctly oriented
I think some algs are on the wiki.

I've added all the algs to the wiki. But I didn't think anyone actually used it.

 

[Rubixcubematt]

I think it's too sune for me to learn coll. I don't know when I will learn it. Maybe someday, but probably not. Coll is great for big cubes though.

 

[4Chan]

I think it's too sune for me to learn coll. I don't know when I will learn it. Maybe someday, but probably not. Coll is great for big cubes though.

>>Too sune

I see what you did there.
I lol'd (x

 

[Rubixcubematt]

I wanted to join in on the punny act .

 

[jms_gears1]

i find that i like most COLL cases, for the most part the algorithms seem, i guess fluents the right word...

and again recog is insanely easy, its just hard to learn at first, and then you have to take it slowly so you dont get cases mixed up.

other than that COLL is nothing

 

[eastamazonantidote]

Using COLL for Fridrich is rather a waste, in my opinion. The concept is solid, but you will take too much time learning cases that rarely appear. For pure Fridrich, maximizing your OLL and PLL potential is far more important than complicating your LL.

However, with EO methods, COLL becomes the logical next step to the (current) Holy Grail of speedcubing - ZBLL. Though I have found the VH system to be slower than plain Fridrich, COLL is a great stepping stone to ZBLL. Also, COLL gets your EPLLs down to sub-2 when all your other PLLs are hovering around 3 or 4 seconds.

I have mixed feelings about COLL, despite the fact that I am using VH as my main method as I move on to ZBLL.

For the record, my best time with COLL was X-Cross + F2L + EO skip + a Pi COLL + EPLL skip = 17.37.

My best time was on this scramble: B F L2 F' L' B' F2 L2 B' F2 L' R B D2 F2 D B' L R2 B R U2 D2 L R2 where I had Double X-Cross + F2L + f' L' U' L U f + Y permutation = 15.31. Later I tried this scramble with VH and got a 20 ish solve.

Though I love VH (well, the COLL part), I have to say that learning pure Fridrich is a very acceptable way to get faster at cubing. If you want to learn COLL, go for it!

Here are my algs, if you guys want (and some cheesy bit about recognition at the bottom): http://nomiswanson.fileave.com/COLL.Algorithms.pdf

 

[powershotman]

pure fridrich is powerful indeed..

 

[Lt-UnReaL]

COLL is good if you have fast recognition and fast EPLLs (and it's faster on bigcubes - especially 4x4). It's also good as long as you aren't doing too many moves to orient edges, these cases are the main ones to know, most other cases are too long: http://jmbaum.110mb.com/f2l.htm (scroll down to partial edge control)

 

[Cride5]

The thing I don't like about COLL is they only work on OLL cases that are fast so you make fast OLLs slower. If they work on crappy OLL cases I might consider learning it.

Haven't you heard of CLLEF? It is just like COLL, but it works on the cases with no edges correctly oriented
I think some algs are on the wiki.

^^ what he said. I'm no expert, but for Fridrich users it seems much like there would be more value in learning CLLEF than COLL. OCLL's are pretty fast (personally faster than EPLL on average).

As a ZZ user, I'm happy that most of my OCLLs are sub-1-able (on a good day). I only use COLL for 4/6 of the Pi cases. My H-perm is pretty fast, but I think I might learn COLL for that too since its only 3 more algs, with good recog/execution..

 

[jms_gears1]

^^ what he said. I'm no expert, but for Fridrich users it seems much like there would be more value in learning CLLEF than COLL. OCLL's are pretty fast (personally faster than EPLL on average).

As a ZZ user, I'm happy that most of my OCLLs are sub-1-able (on a good day). I only use COLL for 4/6 of the Pi cases. My H-perm is pretty fast, but I think I might learn COLL for that too since its only 3 more algs, with good recog/execution..

What is the diffrence between OCLL and COLL?

COLL is good if you have fast recognition and fast EPLLs (and it's faster on bigcubes - especially 4x4). It's also good as long as you aren't doing too many moves to orient edges, these cases are the main ones to know, most other cases are too long: http://jmbaum.110mb.com/f2l.htm (scroll down to partial edge control)

Complete edge control is actually reallllly easy, when doing a fridrich solve i can 100% of the time, use one more second than i would have to orient edges and insert the pair, this leads to w/COLL only 44 algs to learn instead of over 70 with OLL and PLL, or you could learn 18(?) algs for full PLL.

 

[miniGOINGS]

COLL only works well when:

1. It's on a big cube.
2. Edges are preoriented (ZZ, Petrus...)
3. You can do fast EPLL's.
4. Your awesome.

 

[sz35]

I know around 65% of ZBF2L, when I practice ZB I use COLL+EPLL always and it's really really good, when I use Fridrich and I get a case with all edges oriented I use COLLs only in cases that the alg is good (for example: almost all of sune-COLLs suck so I use the regular sune alg when practicing Fridrich)

 

[falcon71]

COLL for Friedrich?

Hi!

I've now learned all OLL algorithms and I'm thinking about how to expand Friedrich.
I'm learning XCross at the moment, and for algorithms I think I could learn COLL.

While it is said that VH+COLL is slower than normal Friedrich, what about the simply lucky cases where all edges are already oriented? Would it then make sense to use COLL+EPLL, or do you think that OLL+PLL is still faster?
I'm not sure if recognition takes longer and if the algorithms are slower compared to the very fast OLL algorithms.

Please share your experiences!

 

[Rpotts]

COLL recognition takes longer than OLL recognition for most, and since most people can sub2 all their PLLs nowadays the time wasted executing the longer COLL (versus super easy OCLL) and recognizing it likely won't be worth the nicer EPLL versus the regular PLL you'd get without it.

 

[abctoshiro]

There are two sides to this:

1. After F2L, you got all edges oriented or,
2. You got misoriented edges.

I suggest you could learn the easy to execute COLL for case 1. For 2, well, regular OLL/PLL.

I forgot to suggest, you could learn CLL/ELL if you can't keep some edges oriented after F2L. Chances are, low.

BTW, that was a suggestion.

 

[LarsN]

I use COLL on my fridrich solves. It takes a lot of practise for recognition to be nearly as fast as OCLL though, but I like EPLL and the improved chance of a skip. There are so many places to find COLL algorithms and it's not hard to find some that are for most of the cases just as fast as OCLL. I don't use COLL for the sune-cases though, except for niklas.
I don't think learning COLL will make you faster, but it's fun to know more tricks.

 

[Akuma]

X-Cross sounds useful but in fact the majority of times you don't use it.
X-Cross only happens if you have an easy pair already but in the majority of times you simply just don't.
One would argue that X-cross is only suitable if you are 100% colour neutral but still one can easily claim that X-Cross is only used practically once every ten solves.

VHF2L is kind of pointless if you already know full OLL.

 

[irontwig]

Planning an entire x-cross is quite hard most of the time, building a 2x2x2 however is not and is imho more useful than a cross. I also find it easier to track cross pieces while building a block than to track a pair while building a cross. COLL and some OLLCPs give you fewer bad PLLs and more EPLLs and skips. It might not save too much time but yor solves at least get less bland.