Partial Corner Control during F2L

[AHornbaker]

If you choose your F2L algs carefully, you can influence pairing and multislotting as well as things like the orientation of edges (as shown by R U' R' vs. Sledgehammer insertions) and the orientation of corners (as in Winter's Variations) Although there have been algorithms made for each pairing case and last slot case, a set of algorithms encompassing F2L and forced OLL would be too large to be useful. The line has to be drawn somewhere before we create alg lists for every possible scramble. I have created this thread because I am interested in learning how to intuitively force OLL while doing F2L in the most efficient and effective way. Instead of employing the use of algorithms, I propose melding together the steps of pairing, pair insertion, edge orientation and corner orientation. By sharing lists of "operators" for each step, we can analyze how each one affects these steps. Some chunks of our beloved algorithms are very common, take R U R' U' for example, but how many of us know what it does? By exploring each of these operators, we can then determine which combination of these would work best in each solve, rather than mindlessly memorizing scores of algorithms. In theory, this will slowly get my move count down, improve my F2L and OLL times drastically, and give me a higher insight into how algorithms work.

Remember- the Rubik's cube was solved intuitively before algorithms ever existed. The embedded links should help give this concept a kick-start.

Thanks in advance for any feedback.
-Austin

 

[Escher]

http://cubefreak.net/speed/advancedf2l/mw.php

http://rowe.cubing.net/rls/

Any site still listing ZBF2L would also be somewhat useful in building a good picture of how forcing OLL skips work.

I don't think it would be too difficult, if one practised LS+OLL for long enough, to be able to intuitively solve both in one look. From there mixing in new cases to look for for pair + LS would be fairly easy.

 

[Sa967St]

I made this video a few months ago:

I go over some common OLL skip techniques and rant about them. The audio is low unfortunately, but I've added captions (under CC).


edit:

Spoiler: transcription of relevant parts

If you decide to learn any algorithms from these subsets, you should know that they can’t always be used in every solve. You can’t decide at the beginning of a solve that you’re going to skip OLL this time, because you might not have a good opportunity to do it. Even when you have the ability to set up certain kinds of cases, if you go too far out of your way, it’s not efficient at all, which completely defeats the purpose of forcing an OLL skip. When you force an OLL skip, it’s supposed to be efficient; it’s supposed to be faster than solving the Last Slot in one look then
the OLL in one look.
Also, there are so many different LS+OLL cases, and so anyone can just take a group of them that have something in common and call it something.
It’s very unrealistic to learn lots of different subsets that force OLL skips because there are thousands of these cases. If you like getting OLL skips and want to learn techniques to help you get them more often, you should learn the sets that have the nicest recognition and nicest algorithms.

 

[Kirjava]

Instead of employing the use of algorithms, I propose melding together the steps of pairing, pair insertion, edge orientation and corner orientation.

So what are your techniques for doing this?

I see a post that lacks substance.

 

[aznanimedude]

so instead of using algorithms...you use algorithms? is that right?

 

[Escher]

So what are your techniques for doing this?

I see a post that lacks substance.

I don't disagree, but I still think that a more intuitive approach to forcing OLL types/skip frequencies is the way forward - so it's at least good that people are thinking about ways to implement it, even if it is entirely speculative.

Which reminds me, I still ought to finish that guide to pair selection/manipulating case frequencies...

 

[AHornbaker]

Any site still listing ZBF2L would also be somewhat useful in building a good picture of how forcing OLL skips work.

This is Lars' ZBF2L page, with 210 different distinct algorithms for LS including flipped algs. Takes care of pairing, insertion, and edge orientation.
http://www.cubezone.be/zbf2l.html

If you decide to learn any algorithms from these subsets, you should know that they can’t always be used in every solve. You can’t decide at the beginning of a solve that you’re going to skip OLL this time, because you might not have a good opportunity to do it. Even when you have the ability to set up certain kinds of cases, if you go too far out of your way, it’s not efficient at all, which completely defeats the purpose of forcing an OLL skip. When you force an OLL skip, it’s supposed to be efficient; it’s supposed to be faster than solving the Last Slot in one look then
the OLL in one look.
Also, there are so many different LS+OLL cases, and so anyone can just take a group of them that have something in common and call it something.
It’s very unrealistic to learn lots of different subsets that force OLL skips because there are thousands of these cases. If you like getting OLL skips and want to learn techniques to help you get them more often, you should learn the sets that have the nicest recognition and nicest algorithms.

I do see your point, and my goal is to use alternate F2l algs to make cases like these occur more often (ideally every time) I already use different algs to break up pairs and group more than one pair at a time rather than focusing on just one pair at a time. Why cant it be taken one step further by controlling the orientation of the LL pieces? Edge pieces can be controlled easily enough by using Sledgehammers leading up to Winter's variations for LS, and I cant imagine it would be much harder to make a 2x2 for the Wondeful alg set.

Bob Burton said on his ZBF2L page "Almost every F2L algorithm inserts the pair after first transforming the case into [an R U' R' or R U R' move]" My idea isnt to wait until the LS to choose from thousands of different algs for a specfic OLL skip, but to intuitively build up to a LS (R U' R', R U R', R' F R F', etc.) that is an OLL skip.

 

[Kirjava]

My idea isnt to wait until the LS to choose from thousands of different algs for a specfic OLL skip, but to intuitively build up to a LS (R U' R', R U R', R' F R F', etc.) that is an OLL skip.

but how?!

having ideas is great and everything, but they're meaningless if you have no idea how to actually accomplish them

 

[AHornbaker]

EDIT:
A more appropriate title for this thread might have been "Partial Corner Control during F2L" After a while of searching around I discovered this concept and it seems like what I am looking for, unfortunately i couldn't find many pages on it.

Right now, I am trying out different variations on the Petrus method to (possibly) make it more efficient. I was hooked by the Winter's/Summer variations, but I'm not too excited about learning all the algs, so instead I want to learn how they influence the corner orientation with intuition (theoretically it's possible with just R and U moves, so it cant be that hard)

A current variation I am looking at is as follows:

Variation 1-
Step 1: Two opposite 2x2x1 blocks
Step 2: EO + Complete 2x2x3 block
Step 3a: 2x2x1 block on R face
Step 3b: Partial corner control + LS (OLL)
Step 4: PLL + AUF

Obviously step 3b needs either algs or substeps, and i'd prefer substeps.

Skimmed over this recently, sort of the idea im looking for.
http://www.speedsolving.com/forum/showthread.php?15094-Partial-Corner-Control-for-COLL-OCLL-Users

I think in the end it's probably simpler (and easier) to just learn the Winter's/Summer variations, but i think it would be more efficient to learn how to set up a really easy insertion case while pairing the LS pair.

 

[cubacca1972]

EDIT:
A more appropriate title for this thread might have been "Partial Corner Control during F2L" After a while of searching around I discovered this concept and it seems like what I am looking for, unfortunately i couldn't find many pages on it.

Right now, I am trying out different variations on the Petrus method to (possibly) make it more efficient. I was hooked by the Winter's/Summer variations, but I'm not too excited about learning all the algs, so instead I want to learn how they influence the corner orientation with intuition (theoretically it's possible with just R and U moves, so it cant be that hard)

A current variation I am looking at is as follows:

Variation 1-
Step 1: Two opposite 2x2x1 blocks
Step 2: EO + Complete 2x2x3 block
Step 3a: 2x2x1 block on R face
Step 3b: Partial corner control + LS (OLL)
Step 4: PLL + AUF

Obviously step 3b needs either algs or substeps, and i'd prefer substeps.

Skimmed over this recently, sort of the idea im looking for.
http://www.speedsolving.com/forum/showthread.php?15094-Partial-Corner-Control-for-COLL-OCLL-Users

I think in the end it's probably simpler (and easier) to just learn the Winter's/Summer variations, but i think it would be more efficient to learn how to set up a really easy insertion case while pairing the LS pair.

You should check if there is a significant move count advantage over Petrus for steps 1 and 2 as outlined. If not, consider Petrus + WV. Be sure to force the last CE pair to RF so that you don't have to bother with mirrors of the WV set.

 

[Cruzer50]

I personally do not think this will help. I think it would be a waste of time, I am not totally sure I may be wrong but wouldn't you waste time with special algorithms?