Why the conventional 2-look OLL?

[Turn n' burn]

I, like a lot of forum users use 2 look oll, and I have to say, I think it could be improved on. I'm sure I'm not the first to think of this, but why orient from the cross? I know OCLL algs are typically somewhat easy to memorize/execute, but I don't think it outweighs the potential benefit of what I'm thinking of. From what I've found when I hit last layer, it is very rare that I encounter a cross-case. In fact, most of the time, it's dot, bar, or L. I'm fine with the F sexy/antisexy F'+OCLL most of the time, but I think we can all agree that the dot case EOLL (4 gen) is pretty tricky to do in a timely manner. (I know, cut to the chase)

The plan is to learn all the dot case OLL's (of which there are 8) and increase the chances of a 1-look oll, or faster 2-look. More on 2-looks later. Now I know what you're thinking. Why not just learn full OLL? Well, there are only 8 dot cases (as opposed to 57) and at that point (especially compared to the potential benefits in store) you would only be saving a second or two. The whole idea is to orient the whole layer off of a much more common case, and by doing so, increase the chance of a 1-look OLL, which is especially beneficial when all of the dot cases (apart from the x case) are 3 gen and seemingly simple to execute. This would be especially helpful for me personally on some OCLLs I use that are utterly terrible yet easy to memorize.

Having addressed some potential issues, this brings me to when this strategy isn't as shiny, and that's when you don't encounter a dot case. The good news is that a dot can be reached from any other case via F sexy/antisexy F'. Unfortunately, if you reach a cross you are restricted to a 4-gen or F sexy F' U2 F antisexy F'. However, this is only necessary if you are really pursuing a dot and don't know OCLL.

Let me know what you think of this strategy (if you could even call it that), and wether or not it could help shave some times. I will be trying it for myself as soon as I finish pll.

 

[guysensei1]

Dot cases are:
1) longer in movecount
2) harder to execute (generally)

And besides, they have the same probability as a cross case, so why go to dot cases when cross cases are better?

 

[AlphaSheep]

Just a note on terminology - you mean OCLL - orientating the corners of the last layer. COLL is something else (completely solving corners of the last layer, preserving edge orientation)

Secondly, I don't know because I'm not a CFOP user, but I thought the dot cases OLLs were supposed to be the worst cases, both in terms of recognition and execution. Also, all OLLs can be done 3 gen, and the OCLLs can all be done 2 gen.

 

[shadowslice e]

1) Aren't all OLLs 3-gen?
2) I think you mean OCLL rather than COLL as the OCLL ignores permutation
3) the dot cases are usually the worst to the point where most people actively avoid them by doing a different insertion.
4) the cross cases are definitely some of the best OLLs and can all be done 2-gen
5) so why not do partial edge control and get cross cases more often (which would be at least as easy to do as forcing dot cases and have the same probability of coming up).

 

[stoic]

I think edge control is the obvious answer to your query; alternatively if you're planning to learn full OLL it used to be standard advice to learn the dot cases first.

 

[jaredye]

I don't get why people who avoid/delay learning OLL. Obviously you want to improve, right? Learning 1 alg a day is refreshing and fun. It takes 5 minutes to drill so that it gets in your muscle memory. Probably another 5 or less to get down the recognition. Then if you want to learn the mirror case you can spend another 5 minutes and now you've learned 2 algs in a day. That way you can finish OLL within a month.

 

[Turn n' burn]

I think edge control is the obvious answer to your query; alternatively if you're planning to learn full OLL it used to be standard advice to learn the dot cases first.

I think for now, edge control is a good option.

Just like to note that this is nothing more than an idea, and I am glad to see such courteous and objective consideration of it. I still plan on learning all the dot case algs in conjunction with OCLL. I think that both sets combined would be pretty effective, and from what I've found, most dot cases are just variants of sune and FsexyF'.

As for ditching all of this, and just learning full OLL, I just don't think time in = time saved, especially considering where I am as a solver (low 30s still learning pll). At this point, it may seem ridiculous that I would even attempt something like this, but I think my last layer suffers a lot more than it should. This doesn't have to be learned as a full fledged alternative to OCLL, but I think the two could work in conjunction to one another within a second or two of a one-look strategy.

 

[Chree]

The argument for learning Full OLL isn't necessarily that it saves more time than some other practice technique. The argument is that it saves time... period. So even if people might still use some 2-Look down to sub-14 global averages, there will still be some ground to gain by finishing the set and integrating it into your solves.

I support the "Learn Dot Cases First" approach. On average, doing a Dot Case Alg will save you ~7 moves over 2-Look OLL. So depending on your TPS, that's a good amount of gain. For me, as a beginner with CFOP that had no concept of what Edge Control was, doing 1 alg was still always faster that frurusruruf-ocll.