OCLL vs COLL

[jms_gears1]

ive been wondering for some time now why COLL is frowned upon with fridrich,

considering it orients and permutes corners and has a better chance for a PLL skip then, OLL+PLL, and it can, if you would like to take it that far, lend itself nicely to learning ZBLL.

I dont use fridrich so i could be wrong however, wouldnt it make sense to do F2L minus one pair, make the last F2L pair, then EO (which can be done completly intuitivly) and then do COLL?

also im pretty sure this is the sub-forum it would go in, although its asking a question i think its more of a discussion...

 

[Lucas Garron]

I dont use fridrich so i could be wrong however, wouldnt it make sense to do F2L minus one pair, make the last F2L pair, then EO (which can be done completly intuitivly) and then do COLL?

It wouldn't. You'd be taking 3 steps and making them into 4, for a 1/12 chance at a skip.

COLL is nice for getting better LLs and more likely PLL skips, but only when you can make EO work out nicely.

Anyhow, someone who actually uses COLL would be able to tell you better; I would rarely have a use for it.

 

[Escher]

I know about half, and it sucks.
(my) EPLLs are not fast enough on average to make up for the worse algorithms and recognition time.

The PLL skip advantage is kinda nice, but I still don't think it outweighs the general slowness of it.

However, knowing what COLL case each of your OCLLs solve and what gives you a diagonal corner swap (or E perm) would be a good idea, just so you have an edge on PLL recognition. I do this with pretty much every OLL and it really helps identifying between V/As etc.

Intuitive EO/VHF2L also sucks unless you know or are learning ZBLL.

To any Fridrich users reading; if you are interested in learning methods in addition to Full Fridrich, I would recommend MGLS (at least learn cases I and Im). Winter Variation is another option, but it's only worth it if you use Petrus or ZZ, and even then it isn't that great.

 

[blah]

just so you have an edge on PLL recognition

Me loveses cube punses.

 

[Escher]

just so you have an edge on PLL recognition

Me loveses cube punses.

I'm so happy someone noticed that xD

 

[LarsN]

I use COLL when ever I get all edges oriented OLL's. I learned COLL before OLL, and I just continued using them after learning OLL. The question is would I have better times if I stopped using COLL? I don't know. My PB lucky time is with an COLL + PLL-skip. My PB non-lucky time is COLL + U-perm. But then again I've had +50 official solves and no PLL-skips (COLL or OLL).
I've thought about quitting the COLLs but I couldn't decide which ones to use for the OLL cases, so I sticked with the whole set.

Oh, and I don't use the sune COLLs, except for Niklas.

 

[deadalnix]

I use partial edge control, so I get all edges oriented more than I should have. in these cases I use COLL for nice cases or ROLL for others.

 

[ErikJ]

I try to use COLL as often as I can. there are a few cases that I cannot find any good algorithms for so I just use a different COLL alg that solves the OLL and moves the corners in such a way that it wont give me a PLL with a diagonal swap like Y, N, E, and V cases. COLL + EPLL is pretty fast.

 

[Kian]

I'm in the process of learning COLL/deciding what situations I plan on using it in. The H case is great, considering all the algs are quick, but some (like Sune and Antisune) hardly seem worth it to me. I will get back to you when I do a bit more research, but I think you should try and see what works best for you.

 

[cubeninjaIV]

I know full COLL although i don't use it as much as i used to. Mostly i only use it when the set up for EO is an R' F R F' or requires no additional AUF oand even then i don't use 4 of those cases.

EDIT: COLL on big cubes FTW!!

 

[a small kitten]

I know full COLL but I use it with ZZ. I would say that using it as a main method for anything that doesn't preorient edges makes you think more than you should.

 

[waffle=ijm]

I know full COLL for Petrus and K4. Very very effective in petrus since you already have EO dealt with.


and full CMLL for Roux (I don't COLL for Roux)

 

[Lucas Garron]

just so you have an edge on PLL recognition

Me loveses cube punses.

I'm so happy someone noticed that xD

They can be kinda corny.

 

[rubiknewbie]

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.

 

[anythingtwisty]

just so you have an edge on PLL recognition

Me loveses cube punses.

I'm so happy someone noticed that xD

They can be kinda corny.

win

 

[cmhardw]

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

 

[waffle=ijm]

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.

how fast is you N-perm or E-perm or V-perm or and other permutation when corners aren't solved?

when you do COLL, you always end up with a U, H, or Z perm, which are fast...like sub-1 fast for some people.

EDIT - I can sub-2.7 all my COLL. I'm not that fast and I know there are people who are faster and can get sub-2

 

[4Chan]

i think coll is a prety cool guy, eh solves corner orientation and permutation and doesnt afriad of anything

 

[Weston]

I use COLL whenever I don't have to work for the edge orientation and it just so happens that they are all oriented. I chose to do this because the recognition is the same as CLL and I practice CLL a lot. For me, I feel like it helps a lot, especially since I know my U perms from 3 angles.

I don't use the sune ones except the Niklaus one if i see it fast enough.

I highly recommend learning the H and Pi case. The algs are pretty good and recognition is super easy.

 

[rubiknewbie]

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.

how fast is you N-perm or E-perm or V-perm or and other permutation when corners aren't solved?

when you do COLL, you always end up with a U, H, or Z perm, which are fast...like sub-1 fast for some people.

EDIT - I can sub-2.7 all my COLL. I'm not that fast and I know there are people who are faster and can get sub-2

I like U perm but my H and Z perm are not much faster if at all than other PLLs (like V perm) so there is only 50% benefit for me.

Another thing I consider is that there are quite a lot of nice PLL cases (like U, T, R, J etc) that occur frequently and "bad" PLL cases of N and E actually have a lower probability of occurring. Often I get U without working for it anyway .

Yet another thing is that many of what I thought were "bad" PLL cases are actually not so bad after learning new algorithms for them (E, N, F, V etc) or learning how to execute them better. Everyday I am hating less and less PLL cases to make me want to do more things to avoid them.

On the other hand I think of Sune and Anti-sune as like 1+ sec algorithm and the other edge-oriented cases are all pretty smooth, so it is more difficult to justify complicating them to get some benefit. In fact, ZBF2L actually seems like a more viable option though it is too early for me to learn it.