why is there no best method?

[genericcuber666]

firstly please keep this thread objective opinions will ruin it!

theirs so many of us cubers who all use different methods and so many people making new methods but why isn't there an OBJECTIVELY better method? i know people have different turn styles but with most competitive things everyone has the same "method" (running cycling tennis etc.) why not cubing?

 

[Ordway Persyn]

In my opinion there may be a couple things as to why
A: Cubing hasn't got fast enough to the point that there is a clear superior method. we still can improve and so we haven't gotten to the limits of the methods we already use.
B: The big 3 all have multiple advanced all sets and more keep getting invented. With all of these, It would be really hard to Know how fast the base method is.
C: This is a weirder one that popped into my head but puzzle to puzzle transition. CFOP based reduction methods for big cubes /megaminx are arguably the best (or at least mostly used) while Roux or ZZ based Reduction methods are less explored and in turn no one has really gotten really fast with them. ZZ is very good for OH, And I would think that having Roux as your main method on 3x3 sighted would give you an edge in blind.
D: It may come to personal preference, some people are really comfortable with M slices and others are not. some people find learning algorithms easy and others do not so it may come down to personal preference. The big 3 have their respective strengths and weaknesses.

Thats just what I thought in response to this and note that I am no expert at all when it comes to different methods.

 

[shadowslice e]

Short answer: There is: Zeroing.

Longer answer: at the moment, there is a trade off between lookahead efficiency and ergonomics. My personal point of view is that this deadlock will continue perhaps indefinitely until either A) a new concept for methods to exploit is created or B) people stop cubing.

 

[bubbagrub]

firstly please keep this thread objective opinions will ruin it!

theirs so many of us cubers who all use different methods and so many people making new methods but why isn't there an OBJECTIVELY better method? i know people have different turn styles but with most competitive things everyone has the same "method" (running cycling tennis etc.) why not cubing?

I think it's because the "perfect" method (e.g., a method that would always result in 20 moves or fewer) is not possible for humans. This means that every method is a compromise and different people want to compromise on different things. People who use CFOP, for example, sacrifice move efficiency for easy recognition and relatively small numbers of algorithms. But even within CFOP, you have things like Winter Variation, ZBLL, etc., and no-one could learn all of those, which means everyone chooses which sets / subsets to learn, based on personal preference.

But really, I suspect this is true of most competitive activities as well. I doubt that all runners use exactly the same method, for example (e.g., food, training regime, gait, etc.)

 

[newtonbase]

I think part of the reason is that there's not enough people testing them out. Most of us just pick a method then try to get faster.

 

[One Wheel]

Cubing is more complicated than the other competitive things you mention. cycling, for example: everyone who is competing in standard cycling (leaving aside things like the Special Olympics, although that's fundamentally similar) has two legs that work in the normal way, as well as arms, back, lungs, a heart, etc. and everyone has the goal of using those tools to exert as much force as possible into moving forward, as efficiently as possible. There are small differences that might not be noticeable to an untrained eye, but to an aficionado they are significant. These differences are attempts to reach the same goal while optimizing techniques for different body types and abilities. Similarly in cubing, different techniques cater to different strengths. CFOP is conceptually simple, but requires more moves than other methods, making it more fit for someone whose strength is high TPS. Roux is probably better for someone with lower TPS but better lookahead and intuitive understanding of the cube. Finally, ZZ is probably best for someone who can easily memorize a lot of algorithms (I don't know. I want to learn ZZ, but I guess I don't have a solid grasp on why). You can pick apart my rudimentary understanding of differences between methods all day, but the point is that different people have different abilities and strengths, and different methods will work better for different people. It seems like for the most part the really fast people use CFOP, which makes sense because with their turning speed a) they can just turn faster, so the disadvantage of more moves is negated and b) they probably need the simplified lookahead/blockbuilding of CFOP, because the human brain can only process visual information so fast.

A better sports analogy might be American Football: the NFL is made up of the players and coaches who are the best of the best at American Football. Why would they not settle on either running the ball or throwing the ball? One way must be better than the other right? and they get tens of millions of dollars to figure out the best way to do it. There, the best way is a mix. It is possible that if cubing were to turn into a professional sport, with the very best cubers getting paid millions of dollars to compete, a few of the very best would take the time to figure out their optimal method, and the methods used by those few would likely converge on a near-optimal method, likely with hundreds, if not thousands, of algorithms, and move counts averaging in the low-30s, but even there they might just learn to recognize which method or combination of methods is best for a given scramble, and use something like what we see now.

 

[GenTheSnail]

Finally, ZZ is probably best for someone who can easily memorize a lot of algorithms (I don't know. I want to learn ZZ, but I guess I don't have a solid grasp on why)

#CFOPHasMoreAlgsThanZZ
You don't see ZZ solvers learning ZBLL, OLLCP, Ell, ELLCP and VLS. You do see CFOP solvers learning those alg sets though.

ZZ focuses on high turning speed, ergonomics and efficiency: F2L uses only <R, U, L> (which is a very fast moveset once you get used to shifting between R/L moves and also requires no rotations), F2L is block built, and LL is supposed to be 1-Looked with only ~200 cases and even less algs, ZZLL (phasing+ ~half of ZBLL).
The theoretical maximum number of algs you'll need to learn is only <500 if your going to go with full ZBLL and no mirrors or inverses.

It seems like for the most part the really fast people use CFOP, which makes sense because with their turning speed a) they can just turn faster, so the disadvantage of more moves is negated and b) they probably need the simplified lookahead/blockbuilding of CFOP, because the human brain can only process visual information so fast.

I think it's because that was the easiest extension from LBL, and no one ever bothered to switch.

 

[One Wheel]

#CFOPHasMoreAlgsThanZZ
You don't see ZZ solvers learning ZBLL, OLLCP, Ell, ELLCP and VLS. You do see CFOP solvers learning those alg sets though.

ZZ focuses on high turning speed, ergonomics and efficiency: F2L uses only <R, U, L> (which is a very fast moveset once you get used to shifting between R/L moves and also requires no rotations), F2L is block built, and LL is supposed to be 1-Looked with only ~200 cases and even less algs, ZZLL (phasing+ ~half of ZBLL).
The theoretical maximum number of algs you'll need to learn is only <500 if your going to go with full ZBLL and no mirrors or inverses.



I think it's because that was the easiest extension from LBL, and no one ever bothered to switch.

Easiest extension from LBL is the reason I'm using CFOP, but I believe that Feliks and Lucas use CFOP, no? They are arguably the best of the best right now, and I find it hard to believe they were simply too lazy to switch.

And you can always learn extra alg sets but at a basic level CFOP is complete with 21 PLLs and 57 OLLs. Anything else is extra. That's a lot less than 500.

ZZ does make sense to me, really I'm just trying to guess why it's not more popular than it is, especially among the fastest cubers.

 

[JTWong71]

Easiest extension from LBL is the reason I'm using CFOP, but I believe that Feliks and Lucas use CFOP, no? They are arguably the best of the best right now, and I find it hard to believe they were simply too lazy to switch.

And you can always learn extra alg sets but at a basic level CFOP is complete with 21 PLLs and 57 OLLs. Anything else is extra. That's a lot less than 500.

ZZ does make sense to me, really I'm just trying to guess why it's not more popular than it is, especially among the fastest cubers.

Well, the 500 Algorithms in ZZ is if you want a 1LLL, otherwise you just need 7 O(C)LL's and 21 PLL's for the equivalent 2LLL when comparing to CFOP's 2LLL, which is 28 Algorithms (ZZ) to 78 Algorithms (CFOP).

In the Reddit Forum Post Felik's says that <R, U, L> Gen isn't very good for 2H, so that may add a little reason why he doesn't use ZZ over CFOP, and because he has been using CFOP for so a long time now.

 

[GenTheSnail]

I believe that Feliks and Lucas use CFOP, no? They are arguably the best of the best right now, and I find it hard to believe they were simply too lazy to switch.

“I use CFOP for 3x3 solving because it's the first method I learned, and I've practiced very much with it, and I think at this point there is no point in switching methods unless there is a major breakthrough ultra amazing method.” - Antoine Cantin

And you can always learn extra alg sets but at a basic level CFOP is complete with 21 PLLs and 57 OLLs. Anything else is extra. That's a lot less than 500.

At a basic level, ZZ is complete with 7 OCLLs and 21 PLLs. ZBLL isn't necessary. That's a lot less than 78.
The point I was trying to make it that the number of algs a ZZ user needs to learn in order to 1LLL without any Last-slot influence is <500.
In order to 1LLL a CFOP solve every time, you'll have I dont even know how many cases. It would be something around 4,000 algs.

ZZ does make sense to me, really I'm just trying to guess why it's not more popular than it is, especially among the fastest cubers.

I think lack of resources for ZZ and the illusion of difficulty the EOLine. A lot of people at the few comps I've been to either don't know what the heck ZZ is or they have the illusion EOLine is super difficult.

And, some people just don't feel like switching.

EDIT:

In the Reddit Forum Post Felik's says that <R, U, L> Gen isn't very good for 2H

Oh, but <U, L, R, F, y> is? That doesn't make sense. (Not at you @JanW , at feliks)

 

[One Wheel]

Well, the 500 Algorithms in ZZ is if you want a 1LLL, otherwise you just need 7 O(C)LL's and 21 PLL's for the equivalent 2LLL when comparing to CFOP's 2LLL, which is 28 Algorithms (ZZ) to 78 Algorithms (CFOP).

In the Reddit Forum Post Felik's says that <R, U, L> Gen isn't very good for 2H, so that may add a little reason why he doesn't use ZZ over CFOP, and because he has been using CFOP for so a long time now.

Does Feliks use ZZ for OH?

 

[JTWong71]

Does Feliks use ZZ for OH?

Nope, he uses CFOP with Edge Control during F2L, and all those other crazy subsets he knows.

 

[CubeGasm69]

I think it's because the "perfect" method (e.g., a method that would always result in 20 moves or fewer) is not possible for humans. This means that every method is a compromise and different people want to compromise on different things. People who use CFOP, for example, sacrifice move efficiency for easy recognition and relatively small numbers of algorithms. But even within CFOP, you have things like Winter Variation, ZBLL, etc., and no-one could learn all of those, which means everyone chooses which sets / subsets to learn, based on personal preference.

But really, I suspect this is true of most competitive activities as well. I doubt that all runners use exactly the same method, for example (e.g., food, training regime, gait, etc.)

This is the best answer so far, imo

 

[tseitsei]

firstly please keep this thread objective opinions will ruin it!

theirs so many of us cubers who all use different methods and so many people making new methods but why isn't there an OBJECTIVELY better method? i know people have different turn styles but with most competitive things everyone has the same "method" (running cycling tennis etc.) why not cubing?

Please tell me what is the optimal method for tennis. There are many different players that play with different styles. If you do not understand that you probably don't follow tennis very closely at all.

Or please tell me the optimal "method" for playing football. ice hockey, volleyball or many other sports... There are literally guys that are earning millions of dollars a year trying to find that optimal way of playing football for example.

Of course some sports tactically simpler sports (like running or cycling) have much clearer "best strategy", but IMO we have not yet found that objectively best strategy for cubing (just as we haven't found it for tennis or football or ice hockey or many other sports). One day we may find a "perfect method" that provides (near) optimal solutions that a human can find in 15 seconds but I really doubt that, just as I doubt that we will find the optimal way of playing football.
We'll just have to go with what we think is currently the best way of doing things. And that is why we have different methods for solving cubes and different strategies for playing football. And IMO that's a really good thing and just makes the sport in question more enjoyable and versatile...

 

[Smiles]

I also strongly agree with bubbagrub.
I will take a more theoretical point of view on the topic.

There are good reasons why "best" needs to be defined.

Example world:
When using method A, I average 10.10 in general, and my Ao50s vary from 9.75 to 10.45 (+- 0.35)
When using method B, I average 10.00 in general, and my Ao50s vary from 9.80 to 10.10 (+- 0.20)
For argument's sake, just pretend that all the top cubers also get the same results, and ignore that the numbers are far slower than actual top-cuber averages.
Here, method A is more likely to set world records by getting lucky, even though method B is faster on average.

Method A is more prone to lucky solves (in other words, has higher variation) and method B is more consistent. Which is the best? That depends on your definition of best.

Issues with variation:

There may be a different method that is the best for each event. Each event is actually 2 different events (Ao5 and single) in which you can simultaneously perform differently (you could get a better Ao5 than somebody else, who got a better single than you). Of course, Ao5 should take precedence because that's what you use to win competitions and decide placings, but single is also important if you want a world record. If (obviously won't happen, but if) Ao100 was what decided the competition winner, and Ao100 world records were used, then method B would be better than it is in our current Ao5 system. In fact, the greater the number of solves per average, the more likely method B wins, since it is more consistent and slightly faster. But just looking at Ao5, method A would hold the world record, given enough time for a super lucky average to happen (all it takes is one lucky average). If method B got just as lucky, the times wouldn't be able to get as low as method A could. However, method B would help you win more competitions, simply because it is usually faster. The best method depends on what your goal is, as well as the system we use in competition.

Additionally, the metric doesn't need to be how well a single person does in competition, you could only look at competition winners. If half of the people in finals use method A, while the other half use method B, then it only takes one of the people using method A to get really lucky, and they win the whole thing. The people using method B would place higher on average, but the winner will often be that lucky person using method A. So if method A is seen to win more competitions and hold all the world records, but method B gets better placements on average, which is the best method?

Moving on:

Okay now, I'll admit I did make an assumption that methods have different amounts of variation, and that it was enough to make a difference in competition. While different methods definitely have different amounts of variation, maybe the difference is negligible. In that case, method B is the best because it is faster on average! Problem solved? No, there are still other ways you can define what is the best.

Pretty important (relating to bubbagrub's post):

The best method may be different for each person because of differing strengths (some methods benefit the most from high TPS, some benefit most from good lookahead, some benefit most from efficient visualization of piece relationships). So is "best" defined as the best for human potential (say, somebody who was genetically modified to be good at cubing, or alternatively, just the best cuber we have now or will have in the future)? Or is the best defined as the method that is the best for the most people?

As an extreme example, say that somebody has superhuman visualization of piece relationships, and their TPS is good, but not superhuman. This person could use a super-efficient method averaging 33 moves, and with 7 TPS, they average sub-5. Meanwhile, somebody with an average of 11 TPS and superhuman lookahead abilities could do CFOP in just under 60 moves and average mid-5. So obviously the 33-move method will hold world records because of that one person, but CFOP would definitely be more popular in this scenario (nobody else can even do the other method quickly).

The example is very extreme and maybe it is hard to imagine because right now CFOP holds world records and is also the most popular. But imagine that right now someone does a sub-4 official single and sub-6 official average with Petrus, multiple times. Does that make it the best method? Or does CFOP's popularity mean it should be the best, and we ignore the outlier person?

Finally, there is also a difference across actual events. Looking at current strengths/weaknesses of each method across different events, ZZ has strengths in OH, Roux gets worse on big cubes, and CFOP is good all around. You could focus your definition of "best" on each individual event, or decide that the best one should be good at all events, since mastering multiple methods is difficult.

 

[stoic]

Please tell me what is the optimal method for tennis. There are many different players that play with different styles. If you do not understand that you probably don't follow tennis very closely at all.

Or please tell me the optimal "method" for playing football. ice hockey, volleyball or many other sports... There are literally guys that are earning millions of dollars a year trying to find that optimal way of playing football for example.

Of course some sports tactically simpler sports (like running or cycling) have much clearer "best strategy", but IMO we have not yet found that objectively best strategy for cubing (just as we haven't found it for tennis or football or ice hockey or many other sports). One day we may find a "perfect method" that provides (near) optimal solutions that a human can find in 15 seconds but I really doubt that, just as I doubt that we will find the optimal way of playing football.
We'll just have to go with what we think is currently the best way of doing things. And that is why we have different methods for solving cubes and different strategies for playing football. And IMO that's a really good thing and just makes the sport in question more enjoyable and versatile...

I'd be interested to know if there are any/many sports that DO have a prescribed "correct" method. Golf springs to mind, everyone seems to agree on how to grip the club I think?
I follow motorsport, and there is massive variation in that - even in how to turn a corner. Some drivers brake in a straight line, turn in to the apex and apply the power early as they exit. Others brake later on the way in, turn and brake at the same time and get on the power later but carry more speed all the way through. One of those methods is objectively longer, in terms of distance travelled, but the net difference over a couple of miles has to be measured in tenths of a second. It's accepted that different drivers have different styles. In fact, it's a beautiful thing to watch once you learn to appreciate it.

 

[Rubix Cubix]

I think part of the reason is that there's not enough people testing them out. Most of us just pick a method then try to get faster.

I think you're right, the fact that it takes a long time to get good with a method, let alone world class with it means that once someone has spent a significant amount of time using one method they don't see it being worth it switching. I think the best method for one person is the one you're best with

 

[genericcuber666]

I think you're right, the fact that it takes a long time to get good with a method, let alone world class with it means that once someone has spent a significant amount of time using one method they don't see it being worth it switching. I think the best method for one person is the one you're best with

so if you only know lbl is that the best for you

tau yu is sub 10 with 4 methods

 

[shadowslice e]

so if you only know lbl is that the best for you

tau yu is sub 10 with 4 methods

No he isn't. He's sub-10 with maybe 3 but not the "big 4".

 

[JTWong71]

No he isn't. He's sub-10 with maybe 3 but not the "big 4".

Here He is Sub-10 with the Big-4.