Fastest Method

[CubeLord]

What is the fastest method that isn't freestyle?

 

[StachuK1992]

BH edges and corners?

 

[CubeLord]

bh is freestyle

 

[porkynator]

Turbo for edges is very fast, in my opinion.

 

[JonnyWhoopes]

bh is freestyle

What? If BH is freestyle, than so is pretty much every other method that uses three cycles of any kind.

 

[Kirjava]

What? If BH is freestyle, than so is pretty much every other method that uses three cycles of any kind.

"Freestyle" is very meta.

 

[Erdos]

Considering "standardized" freestyle methods (that is, any method that more than, say, 3 people do and which have tutorials and premade algorithms), BH for corners. For edges, it's disputed between M2, TuRBo, and BH. However, by the time you get to BH/whatever, you'd probably have a unique set of algorithms you built on the way, slowly digressing into your own "non-standardized" freestyle method.

In terms of classification, I would classify all BLD methods as freestyle, since each basically do a particular set of commutators, 2-gen, etc. To separate the common ones from the 'self-improved' ones, I would introduce the term "standardized" and "non-standardized". So M2/R2, BH, OP, 3OP, etc. are all standardized freestyle methods, and self-improved ones are non-standardized. However, there is of course still some discrepancy between the terms, but not clearly as much as simply using the term freestyle.

 

[Kirjava]

People don't call 3OP freestyle.

 

[Erdos]

People don't call 3OP freestyle.

Meh. That leads to a criteria for what makes a method freestyle and what doesn't make a method freestyle. And for simplicity, I'd just say all methods are freestyle, just that some people popularized their freestyle method.

 

[Jorghi]

Ya but most color neutral people don't know how to use it O_O

 

[MaeLSTRoM]

Ya but most color neutral people don't know how to use it O_O

Talking about BLD solving here. Colour neutral can save you time but not much. Recog can be made faster for one colour only than all 6.

 

[Erdos]

Talking about BLD solving here. Colour neutral can save you time but not much. Recog can be made faster for one colour only than all 6.

Or just 2 colors (opposite ends). Some people who do multi-color tend to recog by pairs (2, 4, or 6).

 

[cmhardw]

We need a term for "BLD methods which work to solve position and orientation of pieces at the same time." I remember a debate about this happening a few months ago, but was a term ever decided on for this? I think having a clear term for this would help to alleviate a not insignificant amount of the terminology confusion and miscommunication in BLD solving right now.

 

[Kirjava]

Meh. That leads to a criteria for what makes a method freestyle and what doesn't make a method freestyle. And for simplicity, I'd just say all methods are freestyle, just that some people popularized their freestyle method.

I choose correctness over simplicity. We already have criteria for it by the way - it's quite annoying that you'll ignore it for the sake of using the terms how you would like to instead of the general convention.

Also, if all methods are classified as freestyle the term becomes meaningless and redundant. You're being a bit stupid.

We need a term for "BLD methods which work to solve position and orientation of pieces at the same time." I remember a debate about this happening a few months ago, but was a term ever decided on for this? I think having a clear term for this would help to alleviate a not insignificant amount of the terminology confusion and miscommunication in BLD solving right now.

Yah, I remember. We decided that freestyle was the 3-cycles of orientation and permutation comm-heavy style of method.

 

[cmhardw]

Yah, I remember. We decided that freestyle was the 3-cycles of orientation and permutation comm-heavy style of method.

I like this, but what would we call the "Classic Pochmann" method? It too solves position and orientation of pieces at the same time, but is not strictly freestyle by the definition you listed.

Clearly freestyle and Pochmann have the "solve positions and orientations of pieces at the same time" part in common though. What would we call this property specifically? Of course there are many options, but I think this should be a standard definition that all people use such as to avoid confusion.

If I remember correctly the phrase "direct solve methods" or "direct solving methods" was mentioned before. Would people agree on this as a viable option?

So freestyle as Kir defines it and Classic Pochmann are both direct solve methods. That would be the main property that the two methods have in common.

Does the term sound good? Thoughts? Comments?

 

[Kirjava]

I like this, but what would we call the "Classic Pochmann" method? It too solves position and orientation of pieces at the same time, but is not strictly freestyle by the definition you listed.

Clearly freestyle and Pochmann have the "solve positions and orientations of pieces at the same time" part in common though. What would we call this property specifically? Of course there are many options, but I think this should be a standard definition that all people use such as to avoid confusion.

If I remember correctly the phrase "direct solve methods" or "direct solving methods" was mentioned before. Would people agree on this as a viable option?

So freestyle as Kir defines it and Classic Pochmann are both direct solve methods. That would be the main property that the two methods have in common.

Does the term sound good? Thoughts? Comments?

Ok, I like the term for "direct solving methods" as being "non-orient first" or "non-redux" or whatever type methods.

I wouldn't call M2 or Classic Pochmann freestyle though... because they simply aren't. There's more to freestyle than just direct solving. (Not much, mind )

 

[cmhardw]

Ok, I like the term for "direct solving methods" as being "non-orient first" or "non-redux" or whatever type methods.

I wouldn't call M2 or Classic Pochmann freestyle though... because they simply aren't. There's more to freestyle than just direct solving. (Not much, mind )

I'd be fine with either "non-orient first" or "non-redux". Of those two I personally like "non-redux" the best.

I agree that Pochmann should not be classified as "freestyle", but I don't know why I agree. I don't think it's because of the 2-cycles, but perhaps that is the reason Pochmann doesn't count as freestyle? Is it a necessary requirement that a freestyle method use 3-cycles (a.k.a. "Freecycle" methods)?

I also take issue with the definition that freestyle is the "absence" of method. Even if you make up algs on the fly, you will eventually fall into a pattern of using certain algs in certain situations, which becomes a method (In case A I do either X, Y or Z - which is a method if you do this for all possible cases, in my opinion).

So perhaps freestyle is: "Using a mix of commutators and non-commutator algorithms to perform 3-cycles in a non-redux method."

This would exclude all 2 cycle methods, and for the time being I feel like I am ok with that. I can't easily think of a less structured method than Pochmann that would also solve via 2-cycles. Can anyone else? Does anyone disagree with the exclusion of 2-cycle methods from the freestyle definition?

 

[Kirjava]

I'd be fine with either "non-orient first" or "non-redux". Of those two I personally like "non-redux" the best.

I agree that Pochmann should not be classified as "freestyle", but I don't know why I agree. I don't think it's because of the 2-cycles, but perhaps that is the reason Pochmann doesn't count as freestyle? Is it a necessary requirement that a freestyle method use 3-cycles (a.k.a. "Freecycle" methods)?

I'm not sure. 3-cycles seems a little like an arbitary rule, but it seems like all the direct solving methods that don't feel like freestyle methods are the ones that use 2-cycles.

I also take issue with the definition that freestyle is the "absence" of method. Even if you make up algs on the fly, you will eventually fall into a pattern of using certain algs in certain situations, which becomes a method (In case A I do either X, Y or Z - which is a method if you do this for all possible cases, in my opinion).

Naaaah. I never saw it like this at all. Freestyle isn't a specific method or the absense of one, it's just a label that applies to a group of methods. It's a meta-method, like FreeFOP.

So perhaps freestyle is: "Using a mix of commutators and non-commutator algorithms to perform 3-cycles in a non-redux method."

If you like. It certainly seems to match how we're using the word.

This would exclude all 2 cycle methods, and for the time being I feel like I am ok with that. I can't easily think of a less structured method than Pochmann that would also solve via 2-cycles. Can anyone else? Does anyone disagree with the exclusion of 2-cycle methods from the freestyle definition?

The only ones I can think of are M2 and Classic Pochmann and they certainly don't fit into the "freestyle" group. I was happy with our definitions until I wrote the next two paragraphs.

Personally, I don't mind a bit of ambiguity involved in the "freestyle" definition. If the definition is a bit less clear we can sort of fudge methods that feel incorrectly labeled due to strict definitions into "correct" groups.

Lets do an experiment and make up random BLD method #324; Solving with 3cycle algs by doing a 3cycle then the inverse to solve one piece at a time. Fake 2cycles. Technically it's freestyle, but we both know that it's not truly freestyle.

 

[Zane_C]

Interesting conversation.

As for the original poster; it's quite accepted that the fastest 'ideal' BLD execution method is "speed-optimal 3-cycles".

So perhaps freestyle is: "Using a mix of commutators and non-commutator algorithms to perform 3-cycles in a non-redux method."

This sounds good. It condenses something like: "Combination of x, y and z" into just one term.

So "freestyle" could essentially apply to: "Solving 2 pieces at a time with a mixture of commutator and non-commutator 3-cycles?"

However, TuRBo by this definition is classified as 'freestyle', which I don't believe it should be.

 

[Marcell]

Personally, I don't mind a bit of ambiguity involved in the "freestyle" definition. If the definition is a bit less clear we can sort of fudge methods that feel incorrectly labeled due to strict definitions into "correct" groups.

That sounds nice, but I simply doesn't seem to work in practice. I think the term "freestyle" has caused too much confusion and debate to be considered useful. We've been using it for quite a while but we still don't know what we really mean when we say it (see this topic for example).
I think that instead of trying to define "freestyle" for the hundredth time we should just stop using it - and then come up with a couple of consciously phrased and well-thought-out criteria that puts all the popular methods in reasonable categories.