Yet another OLS subset: OLS-FE

[Escher]

Last night I had a couple of solves where the f2l ended with the really crappy f2l case with a placed and solved corner and placed but flipped edge. So I decided to generate a bunch of algs.

This f2l case usually takes 11 moves to solve and really sucks. Apparently all of the 1FE cases take around 11-12 moves too, but also solve CO

The 3FE cases aren't great, but imo still faster than LS->OLL->PLL.

Regardless, there are 27 cases for each of the two sets, and all of them are fairly intuitive and easy to learn. I have flagged up a few bad algs (I generated this set very very quickly), if anybody is interested and can find better please let me know here.

So yeah, the site.

Enjoy,
Rowan

 

[Kirjava]

hm! most algs seem really good

 

[Bob]

It's an interesting idea. What I like most about it is that if you notice this case but are not on your last pair, you can just save this pair for last and be sure to get the OLS.

Is it just my machine, or are the images not working?

 

[Escher]

It's an interesting idea. What I like most about it is that if you notice this case but are not on your last pair, you can just save this pair for last and be sure to get the OLS.

Hah, yeah I made a joke about this last night on IRC... Something about calling it RKF2L because it can be used like 'EJF2L'

Not sure if that's the most efficient approach if you have 3 slots left besides that case, but I reckon it is if you only have 1.

Edit: it's actually best when you have one slot left besides this case - with an extra move or two you can almost always force the 1FE case, which saves at least 2 moves over the better 3FE cases. Thus saving you even more moves than 1FE should already save you

Is it just my machine, or are the images not working?

Working fine for me, though I did just make the site... Until it works, the cases are displayed in the order H, T, U, L, Pi, Sune, AntiSune.

 

[Bob]

Hah, yeah I made a joke about this last night on IRC... Something about calling it RKF2L because it can be used like 'EJF2L'

Not sure if that's the most efficient approach if you have 3 slots left besides that case, but I reckon it is if you only have 1.

Edit: it's actually best when you have one slot left besides this case - with an extra move or two you can almost always force the 1FE case, which saves at least 2 moves over the better 3FE cases. Thus saving you even more moves than 1FE should already save you



Working fine for me, though I did just make the site... Until it works, the cases are displayed in the order H, T, U, L, Pi, Sune, AntiSune.

It was my browser. It was asking me if I wanted to HIDE content that is not secure whereas I thought it was asking if I wanted to SHOW it, so I was clicking yes instead of no.

 

[rowehessler]

these are awesome! I was waiting for someone to generate these . its nice cuz if you learn all of these, you will have 8 3-4 move inserts now instead of 4. Like, instead of this: d R' U2 R U2 R' U R for your last pair, you can do R U R' and then one of these algs. You can also do it for the R U2 R' d R' U' R case.

 

[Escher]

these are awesome! I was waiting for someone to generate these . its nice cuz if you learn all of these, you will have 8 3-4 move inserts now instead of 4. Like, instead of this: d R' U2 R U2 R' U R for your last pair, you can do R U R' and then one of these algs. You can also do it for the R U2 R' d R' U' R case.

That would definitely be worth it sometimes, maybe not all the time (at least until I have better algs).

Thinking about it now, it's probably quite simple to turn those easy insert cases into OLS cases, simply by looking ahead for those 3 moves and seeing if it will be a case that cancels...

 

[sa11297]

Can some one explain what all these methods are? I am confused.

 

[Escher]

Can some one explain what all these methods are? I am confused.

This is just a subset of a higher group of algs that would solve the last slot of the f2l, along with the orientation of all the LL pieces, known as OLL-LS (OLS for short).

By learning a few subsets such as this (where you have a bad case that you want to make decent) or RV (where you have a decent case and want to cancel a bunch of moves and get good singles), you can improve your average times overall.

These are only recommended if you already know OLL and are very comfortable with it; they're quite advanced sets.

 

[sa11297]

This is just a subset of a higher group of algs that would solve the last slot of the f2l, along with the orientation of all the LL pieces, known as OLL-LS (OLS for short).

By learning a few subsets such as this (where you have a bad case that you want to make decent) or RV (where you have a decent case and want to cancel a bunch of moves and get good singles), you can improve your average times overall.

These are only recommended if you already know OLL and are very comfortable with it; they're quite advanced sets.

thank you, and yes this does sound VERY advanced. what is the total number of algorithms? for the OLS and its subsets?

 

[Escher]

what is the total number of algorithms? for the OLS and its subsets?

Hard to say really, easily in the thousands - it would be a complete waste of time to learn it all.

RV covers a bit, MGLS a few more, this a little... There are only 3-5 more f2l cases I would even consider generating and learning besides.

 

[sa11297]

Hard to say really, easily in the thousands - it would be a complete waste of time to learn it all.

RV covers a bit, MGLS a few more, this a little... There are only 3-5 more f2l cases I would even consider generating and learning besides.

well yes, I just thought that someone had the exact number, anyway, it does not really matter. Either way full OLS would be impractical to say the least.

 

[Divineskulls]

I like all the last slot/oll options that are popping up. Very good to know for the more advanced cubers that want an extra edge for every few solves. Thanks for posting Rowan.

well yes, I just thought that someone had the exact number, anyway, it does not really matter. Either way full OLS would be impractical to say the least.

I believe a lot of this is still being experimented with. Someone (I think it was aronpm) posted a list of LS->LL possibilities, and and that is, I believe, a more general term than OLS. Someone should create one master diagram of all the methods/subgroups/steps/pseudo-steps. I think I might do that, actually... Sorry, I'm getting off on a tangent.

 

[Bob]

thank you, and yes this does sound VERY advanced. what is the total number of algorithms? for the OLS and its subsets?

thank you, and yes this does sound VERY advanced. what is the total number of algorithms? for the OLS and its subsets?

I did a quick calculation on this. If you take each F2L case, you just have to look at the number of distince edge orientations and the number of distince corner orientations.

For most F2L cases (36 of 41), there are 8 distinct edge orientations each with 27 distince corner orientations.

For three cases, there are only 2 distinct edge orientations (because the edge and corner are both permuted, and the edge is flipped in the F2L), and those have 27 corner orientations each.

There are also two F2L cases in which the edge and corner are permuted but the corner is misoriented. This allows (assuming I have counted symmetric cases correctly) only 57 cases for each of these last two F2L cases.

Therefore, we have 36*8*27 + 3*2*27 + 2*57 = 7776 + 162 + 104 = 8042 distince OLS cases (not including regular OLL).

In other words, OLS on its own is extremely impractical. I think the only practical subsets would be RV (and its mirror), the RV equivalent of Summer Variation (and its mirror), the OLS-FE cases, the CLS cases from MGLS (the ones with corner permuted but twisted would be okay with flipped edges because of the limited number of cases--the ones with the corner in the LL would probably not be worth it except for when all edges oriented), and perhaps the case with a flipped edge and twisted corner both permuted. The last case has the advantage of having only 54 cases because the F2L pieces are both in the F2L. In general, if you have any LS piece in the LL, it makes for a bad OLS case except for WV/SV cases because it's natural to turn those other F2L cases into these.

 

[Rubiks560]

So, I've learned about 3-4 sets of 1FE, and I'm trying to figure out when exactly this should be used. Do you think it's something that should be used every solve? Or just when it's easy to insert the flipped edge pair?

 

[Robert-Y]

I would personally only use it when the flipped edge pair is already in place.

 

[Bob]

I have actually liked this case for a last slot for a long time. There are only two ZBLS algs required for it because you always have 1 or 3 flipped edges. You just need to know where to put the 1 good/bad edge. Man, there's so many things I want to learn right now. :/

 

[Robert-Y]

Bump.

I've added almost 45-50 new and mostly better algorithms. Enjoy.

 

[Ranzha]

For A-1FE, the alg I use is (R2 U R' U') r' U2 (R U R U') (R2' U2' M').

A lot of these algs are fantastic! =O

 

[advincubing]

F2L edge flip

So, I asked this on the one-answer-thread (here), and got great answers already. But that's not a great place for follow-up. So, here's a thread dedicated to it.

I was looking for a better F2L edge flip algorithm than this: (R U' R') d (R' U2 R) U2' (R' U R). Based on the responses, it seems that this is the most efficient/common, and is the first listed on the wiki (#38):

(R' F R F') R' U2 R2 U R2' U R

Spoiler: OTHER OPTIONS ARE

Optimal Algorithms (9 HTM):
F2 U2 R' F2 R U2 F U' F
F' U F' U2 R' F2 R U2 F2
R U' R U2 F R2 F' U2 R2
R2 U2 F R2 F' U2 R' U R'

RrU Algorithms (10 HTM):

R U' r' R' U2 R2 U R2 U r
r' U' R2 U' R2 U2 r R U R'

RFU Algorithms (9 HTM):

F2 U2 R' F2 R U2 F U' F
F' U F' U2 R' F2 R U2 F2
R U' R U2 F R2 F' U2 R2
R2 U2 F R2 F' U2 R' U R'

I'm having a really tough time figuring out a rhythm/finger placement, etc. for (R' F R F') R' U2 R2 U R2' U R. I looked on Youtube for videos on this algorithm, but haven't found any. Anyone have good suggestions on technique/finger-tricks? Or a video?

Thanks.