Pointless algorithms? (ZBF2L++ vs F2L++)

[jacob15728]

So I've been thinking.

If you want to solve your last layer and one f2l pair you can do either of the following:


F2l algorithm + OLL algorithm + PLL algorithm

OR

ZBf2l algorithm + COLL algorithm + PLL algorithm

Wouldn't the second way be slower? Or am I missing something?

 

[Ellis]

yes you are....

The second PLL should be replaced with EPLL (quicker) with a greater (1/12) chance of PLL skip. Also with the possibility to replace the COLL+EPLL with 1LLL

 

[jacob15728]

Doesn't 1lll require several thousand algorithms? That doesn't seem feasible

 

[ThatGuy]

I can't explain but the way it is set up it is reduced

 

[jacob15728]

OK, thanks for your help. I just watched a video teaching 8 simple ZBF2L cases which could be useful if I ever run into them, do you think it's worth learning them? It's only 8 algorithms after all so it can't hurt

 

[ThatGuy]

here's some stuff you can look at if you want to pursue:
http://www.speedcubing.com/chris/zbf2l.html
I'm not sure what you're looking at but you may want to learn cases that would normally require a long last pair insertion. You could go ahead and learn them if you want but because there are a lot of cases you may only rarely run into those 8 cases and you'll end up forgetting them.

Someone tell me if I'm pointing jacob in the completely wrong direction.

 

[TimeFreeze]

Was it weston's? Because I'm pretty sure his has 8 algs on ZBF2L.

 

[Ellis]

OK, thanks for your help. I just watched a video teaching 8 simple ZBF2L cases which could be useful if I ever run into them, do you think it's worth learning them? It's only 8 algorithms after all so it can't hurt

imo, you'll be wasting more time checking to see if you have those cases when you don't than the time you'll save on the rare occasion you do get them. EO should be less than a second anyway if you're not doing full OLL.

Someone tell me if I'm pointing jacob in the completely wrong direction.

yes you are, maybe this is better?

 

[Cride5]

F2l algorithm + OLL algorithm + PLL algorithm

OR

ZBf2l algorithm + COLL algorithm + PLL algorithm

Wouldn't the second way be slower? Or am I missing something?

COLL/EPLL takes less moves. It is believed to have easier recognition than full OLL/PLL, but don't use it yet so don't take my word for it. However, comparing them directly is unfair, unless the work required to achieve LL edge orientation is taken into consideration. ZBF2L probably isn't worth it if you're going for COLL/EPLL, but if you're thinking about ZBLL then that's a different story!

Doesn't 1lll require several thousand algorithms? That doesn't seem feasible

1LLL with LL edges pre-oriented (which is what ZBF2L does) has 494 cases, solvable using 177 algs, assuming the cuber is able to mirror/inverse them.

 

[Johannes91]

1LLL with LL edges pre-oriented (which is what ZBF2L does) has 494 cases, solvable using 177 algs, assuming the cuber is able to mirror/inverse them.

If the cuber is capable of using setups like R <some-alg> R', the number goes down a bit more.

 

[Lucas Garron]

So I've been thinking.

If you want to solve your last layer and one f2l pair you can do either of the following:


F2l algorithm + OLL algorithm + PLL algorithm

OR

ZBf2l algorithm + COLL algorithm + PLL algorithm

Wouldn't the second way be slower? Or am I missing something?

Hey! Where's ELS algorithm + CLS algorithm + PLL algorithm?

 

[ManasijV]

I always thought it was just ZBF2l + ZBLL . ZBLL requires 21X8 algorithms I think.

 

[ThatGuy]

ZBF2L: 306
ZBLL: 494

from this thread: http://www.speedsolving.com/forum/showthread.php?p=167575

 

[Cride5]

@ThatGuy: Remember, you don't always need a separate alg for each case, some are solvable by applying mirrors, inverses, and (as Johannes says) conjugates to other algs. 800 for ZB is the number of cases, not necessarily algs.

 

[Haste_cube]

ZBF2L+ZBLL vs F2L+OLL+PLL vs F2L+COLL+EPLL?

 

[Nukoca]

COLL/EPLL takes less moves.

C+P from here: http://www.ai.univ-paris8.fr/~bh/cube/

Algorithms for Jessica Fridrich's method :
OLL (41 configurations, average : 9.22),
PLL (14 configurations, average : 11.21).
The positions are in the same order as in Jessica Fridrich's site.

Other 2 stages solutions :

permutations of corners and edges (14 configurations, average : 8.10),
then orientations of corners and edges (30 configurations, average : 12.80).
permutation and orientation of corners (25 configurations, average : 9.18),
then permutation and orientation of edges (19 configurations, average : 11.27).
permutation and orientation of edges (13 configurations, average : 7.87),
then permutation and orientation of corners (42 configurations, average : 11.73).
Algorithms for Lars' method (assumes edges already have correct orientation) :
Step 5 (permutation of corners) (3 configurations, average : 6.17).
Step 6 (orientation of corners) (6 configurations, average : 9.41).
Step 7 (permutation of edges) (4 configurations, average : 8.75).
COLL (25 configurations, average : 9.78).
Step 6 and 7 simulaneously (35 configurations, average : 11.72). The positions are in the same order as in Lars' site.
(so in fact I give a list of 50 configurations, since it doesn't take into account inversing the positions)
Step 5, 6 and 7 simulaneously (177 configurations, average : 12,08).

I believe this does not include mirrors.

 

[blah]

So I've been thinking.

If you want to solve your last layer and one f2l pair you can do either of the following:


F2l algorithm + OLL algorithm + PLL algorithm

OR

ZBf2l algorithm + COLL algorithm + PLL algorithm

Wouldn't the second way be slower? Or am I missing something?

Hey! Where's ELS algorithm + CLS algorithm + PLL algorithm?

Wha...?! Haven't you heard? Everyone's on an anti-Lucas strike right now!

 

[Cloud_9ine]

But it should be obvious that we don't include EJF2L (since it in itself is awesome) there so I wouldn't see a problem with incl using CLS