Another last slot method: F2L to bar OLL case

[mark49152]

Here’s another idea for a last slot method:-

1. Solve cross and first 3 slots of F2L,
2. Solve last slot and force a “bar” OLL case, in one step,
3. Do one of the 15 “bar” OLLs then PLL to finish.

By “bar” case I mean two opposite LL edges correctly oriented, and the other two flipped.

The reason I think this has potential is that solving the last slot to a bar case requires knowing only 3 extra algs of max 5 turns, one of which is sledgehammer, plus mirrors. It’s so easy that it quickly becomes intuitive and cases are easily seen during F2L lookahead without major additional recognition cost. The pair must be connected first and the average turn cost is 3.3 more than a simple insertion of the pair.

Why bother? Well as a relative noob my first motive was to minimize learning OLLs, so I started looking for ways to control which OLL cases come up. The prospect of learning 15 algs for OLL is more palatable than 57 for me, and I’d rather practice and get fast at 15 than try to remember 57 and practice each of them a quarter as much. This suggestion isn’t aimed at sub-10 cubers who are happy to learn large alg sets to shave off fractions of a second – I already realize that if you know all OLLs you might as well save those 3.3 turns.

Why the bar case? Because of symmetry. While messing with various last slot methods, I noticed that solving last slot to a bar case is easier because (i) there are two possible bar orientations and you can arrive at either, and (ii) the same alg will work for a flipped case, e.g. the alg for 3 good LL edges can also be used for 3 bad edges. At first I thought this property would be useful just to reduce alg count, but it turned out to be even nicer because the algs needed are so short.

The symmetry of the bar also means that there are fewer “special case” algs to learn for the difficult last slot cases. I’m currently generating algs for F2L cases 23-24, 32 and 38-41 from the wiki (all those F2L cases that require more than one setup move when done intuitively). These seem to be short enough that learning them could cancel out those 3.3 extra turns for any solve where a bad F2L case arises, even if you deliberately wrong-solve the pair early (assuming you don’t care about the extra cube rotation). For example, cases 38, 40 and 41 including all LL edge orientations can be covered with only 3 algs total, of 8-9 turns each. That’s the same number of algs and turns as the F2L special cases that ignore EO, so you get the bar LL case for free.

I would really like to get feedback from the community, most of whom have a lot more experience than me. Is this worthwhile? Has it been done before? I’d be happy to write up and publish the cases, algs and math if anyone is interested.

Thanks for reading. Oh and I’m a noob on these forums so go easy on me please!

 

[Forte]

If this is being used as some sort of stepping stone towards full OLL, I'd rather someone just learn all bar cases, and then just reduce the last layer to a bar with F R U R' U' F' instead of compromising their F2L. I wouldn't like them spending a great deal of time recognizing F2L edge orientation cases when they could be learning to recognize OLLs.

Mostly I just wouldn't want someone to go ahead and learn all of this stuff, then realize that they want to learn full OLL and see that they've wasted time doing this

Oh also welcome

 

[Kirjava]

It's good that you realise that this isn't better than straight up OLL/PLL.

As a learning tool, eh.

 

[Escher]

Probably a reasonable stepping stone, learning that influencing LL early is good. Imo there's more utility in using it as a stepping stone for OLLCP :3

 

[mark49152]

Thanks for the feedback. Yes it’s a stepping stone, but more than just a learning aid. Of course there are other ways to add turns to force OLL cases if all you want to do is practice them. Instead I see this as an alternative to 3LLL as an intermediate method, because it effectively gives you 2LLL with half the algs of full Fridrich, at the expense of some really quick and easy edge control and about 3-4 more turns on average.

Actually it’s the easiness of the edge control that is the main point. I agree with Forte that I wouldn’t want to spend a load of time learning lots of last slot cases and algs instead of spending it on OLL. That’s what put me off VHLS and made me explore this in the first place. It’s so ridiculously easy compared to VHLS that I learned it in about an hour, and was surprised not to find it already written about somewhere.

Anyway here are the cases/algs. This is all of it. Each alg applies to two cases, which differ by having the 3 LL edges flipped, making them easy to remember - I learned them by the "clock" position of the 1-edge case, but they become intuitive quickly. Mirrors needed too of course.

(U) R U’ R’

R' F R F'

(U) F' L' U' L F

(U) R B U’ B’ R’

 

[qqwref]

Haha wow, that's way easier than I expected. I think it definitely has potential as a reasonable thing if someone isn't interested in actually learning all the OLLs. I would like to suggest that the left URBU'B'R' case be done with URU'R' as well, because corner-only OLLs are pretty fast and you should probably know them anyway (that is, I'm suggesting you allow either bar or solved edges).

PS: One of my least favorite OLL cases (#55) is a bar

 

[mark49152]

The same set of algs can be used to reduce every case to either all LL edges oriented or all mis-oriented. The same symmetries apply. That was my original plan - it still requires learning 7+8=15 OLL algs, and some of the corners-only cases I know already. I changed plan because the dot cases are longer and people seem to dislike them.

Although it would make it slightly harder to learn (as it breaks the symmetry), the above could be modified so that all the 2/3 edges cases go to a corners-only OLL and all the 0/1 edge cases go to a bar OLL. That would require knowing 7+15=22 OLL algs but might be preferable for those who already know or prefer the corners-only cases.