Theoretical New Alg Set Similar to WV That Leads to ELL

[Alex3]

Before we start, if you take a look at my post count you will see that I'm actually new, so if this is in the wrong place I'm really sorry.
I've looked through all of the possible places I could put this and here seemed the most logical.

Anywho, I've been using Winter Variation and COLL in my solves whenever applicable. I find that during solves I'm sometimes hesitant to use WV because there's a chance my corners won't be permuted, so I sometimes forgo the WV case and solve into a COLL so that I can get a nice EPLL. I was just wondering if there was an alg set that solved for both orientation and permutation of the corners (possibly while ignoring edges?) which would then direct the solve into a final ELL alg. If this set were to ignore the orientation of the edges, that would mean that this alg set could be used regardless of whether there is a cross solved on the top, which would be a nice plus for CFOP solvers who don't pre-orient edges. The cons however are certainly present. The alg count by my estimates would be plenty above 100 and the recognition would probably be tricky, but for someone who likes using ELLs, this could be a fun alternative to WV.

I've searched the forums and Google for "Winter Variation + CP" and other similar search terms and I couldn't find anything so I thought I would mention it.
If it turns out this method does exist under a different name, then again, I'm really sorry.
And if this is a new method, I'd have no idea what to call it (Spring Variation?) so suggest naming ideas if you'd like.
Also, if any of you guys could share links to this alg set or even want to generate your own algs, that would be super kind of you.

Thanks so much for taking the time to read all this!
-Alex

 

[maps600]

This actually sounds like a good idea for Roux solvers, in fact. (CMLL skip) For CFOP it may be useful at times if you like recognizing ELLs and such, however it would be more logical to apply this to Roux. Interesting idea. We'll see what comes out of this.

 

[yoinneroid]

hmm, 27*6 doesn't seem that much

 

[Robert-Y]

I'm not really sure such a set would be worth learning. I don't think ELL is that nice as far as I know, but we *know* PLL is quite decent overall. Just for the U R U' R' insert alone, there are 27*6 = 162 algorithms I believe. So I think there are 162*2 = 324 algorithms for this entire set.

Off-topic: What method do you use and what do you average?

 

[TheNextFeliks]

Similar to bbf2l check it out.

 

[Coolster01]

This would be interesting to use for 2x2, can't believe nobody has thought of this yet.

 

[Tim Major]

This would be interesting to use for 2x2, can't believe nobody has thought of this yet.

I know! Chris should really make some sort of page on his site to make algs for this method. (maybe keep it private for now?)

 

[kcl]

I know! Chris should really make some sort of page on his site to make algs for this method. (maybe keep it private for now?)

lol there's a reason for the secret page. I had the idea a while ago and had no clue there were actually some algs for it.

 

[Coolster01]

I KNOW THIS POST IS OFF TOPIC DON'T FWEAK OUT PLZ

lol there's a reason for the secret page. I had the idea a while ago and had no clue there were actually some algs for it.

I feel like it wouldn't be this though, after further thought.. EG-1 saves 7 moves to get to CLL (R U' R2 F R2 U' R'), TCLL saves 7 moves (R U' R' U R U' R' / mirror), and EG-2 saves 3 moves, but 6 QTM (R2 F2 R2). R U' R' is pretty darn fast compared to R2 F2 R2 imo. He said it's definitely not TEG. An idea I thought about was 3/4 layer with the last piece on the U face. That would save 5 moves / 8 QTM (R2 U R2 U' R2) and some good algs could probably come from that. Personally, I'm still 120 algs behind this Chris child, so if he ISN'T doing my suggestion, then I'd probably not get into that

 

[pewpewrawr]

People have thought of this, the reason why nobody has used this idea is because it's not worth it; it makes recognition considerably harder, there are far more cases than COLL/CMLL, and it would only possibly save you a few moves at most. It would be far easier to just insert the pair and recognize from there. Wouldn't really work for roux either because you would have twice as many cases, there are two ways the pair can be oriented in U, and I feel like it would be too restrictive of block building. Can't see it being better than EG or CLL for 2x2, you would have to build a layer minus one corner and recognize from there, and the corner would have to be oriented in a specific way unless you want to learn 3x the amount of algs.

I'm really tired so I probably sound like a d**k but whatever.

 

[guysensei1]

Why not ignore CP and EO, and just solve the last slot + CO, then finish with MBLL or antiZBLL or whatever that's called?

 

[Hypocrism]

Or WV with edge phasing, leading to ZZLL?

 

[TDM]

People have thought of this, the reason why nobody has used this idea is because it's not worth it; it makes recognition considerably harder, there are far more cases than COLL/CMLL, and it would only possibly save you a few moves at most. It would be far easier to just insert the pair and recognize from there. Wouldn't really work for roux either because you would have twice as many cases, there are two ways the pair can be oriented in U, and I feel like it would be too restrictive of block building. Can't see it being better than EG or CLL for 2x2, you would have to build a layer minus one corner and recognize from there, and the corner would have to be oriented in a specific way unless you want to learn 3x the amount of algs.

I agree entirely with this post. Yes, it's been thought of many times before; I've seen it more than once in this thread. You may have to look through it a bit, but there has been quite a bit of discussion about this substep's advantages and disadvantages there.

 

[Bindedsa]

Or WV with edge phasing, leading to ZZLL?

That doesn't make sense? WV with edge phasing would just give you a limited subset of PLL.

 

[elrog]

I don't think learning 162 algorithms compared to the 42 is worth it in this case because you are only a few moves away from the pair being inserted.

Rather than make a new set for cases when you have an easy F2L insert, choose a hard F2L case. This would turn the worst case into one of the best.

I personally think it would be better to learn a CLL or CMLL that inserts the FR edge or twists the FRD corner.

 

[Hypocrism]

That doesn't make sense? WV with edge phasing would just give you a limited subset of PLL.

Ah I was using WV to mean edge control, rather than corner control. I get the seasonal terms all mixed up.

 

[Alex3]

I use both Fridrich and ZZ and I average about 15s
Also thanks for all the input everyone. I'll certainly look into all of the alg subsets you have recommended.

 

[goodatthis]

I personally have thought of this idea before as well, and other variants (CP whilst orienting edges, to do 2GLL, basically CPEOLS) and I don't think it's very practical. WV isn't used that often, so I think that learning that many algs for such hard to recognize CP cases as well as not being able to use it every solve is impractical.

Plus, I think the reason why OLL/PLL is so popular is because it's a 2LLL system that is evenly balanced in terms of algs and speed, meaning each step is roughly similar in speed (well PLL is slower) and each have a reasonable amount of algs each. But something like EO/ZBLL is very disproportioned, you have a ton of algs on one end, 3 on the other, yet it's still a 2LLL system that is only slightly faster than OLL/PLL. And I think the same thing goes for LS methods as well, you could learn full VLS and basically cancel or save a few moves by doing OLL while inserting your last pair. But it's still ton more algs, and not much save in speed. The same thing goes for this, you can learn just 4 EPLLs and use this WVCP, but there's more efficient ways to do a faster 2LLL with less algs.

But in theory, the idea is great! I always like playing around with theoretical LL systems, it's quite fun!

 

[kcl]

lol if somehow olscp is possible that's a 1/12 ll skip

 

[TDM]

lol if somehow olscp is possible that's a 1/12 ll skip

The highest LL skip chance I can think of for something that's realistic, after ZZ-WV/SV with a 1/72 skip chance, is ZZ-d with phasing (1/108 skip chance). It has a slightly lower LL skip chance, but it does have advantages over ZZ-WV/SV; unlike ZZ-WV/SV, you only need to learn two algs (for CP), plus 30 for LL (including H/Z perm). WV/SV requires 54 algs, plus 21 for LL. There's also a lot of 2-gen, which is great for OH. The only thing you during your last slot is phase, which is easier than WV/SV as well. CP isn't too hard either. I used to use this for OH, but when I realised I would never bother to learn 30 algs I gave up the phasing part and did suneOCLL -> EPLL.

(E: unless I've missed another quite obvious alternative that has a low LL skip chance?)