• Welcome to the Speedsolving.com, home of the web's largest puzzle community!
    You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

    Registration is fast, simple and absolutely free so please, join our community of 35,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!

The method debate thread

Joined
Nov 23, 2019
Messages
21
I have. But he's an exception; he doesn't use double flicks so you shouldn't either according to that logic. Also, he still knows more algs than just OLL/PLL.
Yeah he knows more, but he doesn't use them. He only uses oll and pll in actual speedsolves.
 
Joined
Nov 23, 2019
Messages
21
World class ZZ solvers use ZBLL, and it's currently the best form of ZZ. There's no reason not to include it
yeah well it is insanely difficult to learn, and once you learn all of it, it still wouldn't be faster than oll and pll because recognition takes much longer. in theory you could get it faster, but here is the thing. It would take tons and tons of practice. and there are about 500 zbll cases, and you average about getting one new zbll every 500 times, soooo, since there are about 70 something, so you get to practice one 35 times because you always use each in every solve. For zbll, you only use each once in every solve, and you have to do 10 times the practice to get just as good. so yeah, cfop is better.
 

ProStar

Member
Joined
Oct 27, 2019
Messages
3,521
Location
An uncolonized sector of the planet Mars
WCA
2020MAHO01
yeah well it is insanely difficult to learn, and once you learn all of it, it still wouldn't be faster than oll and pll because recognition takes much longer. in theory you could get it faster, but here is the thing. It would take tons and tons of practice. and there are about 500 zbll cases, and you average about getting one new zbll every 500 times, soooo, since there are about 70 something, so you get to practice one 35 times because you always use each in every solve. For zbll, you only use each once in every solve, and you have to do 10 times the practice to get just as good. so yeah, cfop is better.
You're recognizing wrong if it takes more time than OCLL->PLL
 

efattah

Member
Joined
Feb 14, 2016
Messages
554
In terms of getting fast singles from skips, CFOP variants are hard to beat as PLL skips save a ton of moves. Probably the second best type of skip is a corners first solve where you get a 2-3 move solve on the corners. In this fashion, during the inspection, you can look ahead well into the edge solving phase. This is big, if you consider that a PLL skip in CFOP does not allow additional lookahead beyond that step. In my case all my fastest LMCF solves were with 2-4 move CLL solves and looking ahead well into the edge solving phase. Still, I think CFOP is the best in terms of fast singles with the PLL skip (or VLS and a type of LL skip).
 

maticuber

Member
Joined
May 18, 2020
Messages
73
WCA
2009MACA01
Don't use Max Park as an example for good methods, not sure about now, but he used to use his 2H algorithms for OH, and we all know that OH specific algorithms are faster than 2H algorithms for OH.
 
Joined
Oct 7, 2018
Messages
222
WCA
2017OWEN01
yeah well it is insanely difficult to learn, and once you learn all of it, it still wouldn't be faster than oll and pll because recognition takes much longer. in theory you could get it faster, but here is the thing. It would take tons and tons of practice. and there are about 500 zbll cases, and you average about getting one new zbll every 500 times, soooo, since there are about 70 something, so you get to practice one 35 times because you always use each in every solve. For zbll, you only use each once in every solve, and you have to do 10 times the practice to get just as good. so yeah, cfop is better.
This is just wrong it’s not that hard to learn zbll which is why lots of people have done it and many more people are learning right now. The algs can easily be executed quickly and there are multiple trainers to help you recognise it quicker. I guarantee Tao yu’s ll is quicker than yours.
 

maticuber

Member
Joined
May 18, 2020
Messages
73
WCA
2009MACA01
This is just wrong it’s not that hard to learn zbll which is why lots of people have done it and many more people are learning right now. The algs can easily be executed quickly and there are multiple trainers to help you recognise it quicker. I guarantee Tao yu’s ll is quicker than yours.
I've heard from a lot of people that sticking to one LL method is a bad idea. A better method is to learn as many different LL methods as possible, but only drill the easy cases and use them if you happen to get them. For example COLL, is worth to learn them but you don't necessarily use all of them, most people don't learn sune/antisune just because sune/antisune+PLL is faster than COLL+EPLL. OLLCP is another great example, you can learn only a handful of algs to avoid bad PLLs and try to force PLL skip. At the end of the day you'll know a bunch of algs, probably more than full ZBLL, but you'll know a lot of methods and ways of solving the cube, and you'll get "lucky" more often. We are at a point were forcing skips and "luck" is the ultimate method.

I have a PDF with 170 1LLL cases that are really easy to recognize and really fast to execute, the NR in my country is 5s, it was done by a dude that avgs 9ish seconds and he happened to get one of those 1LLL cases and solved the LL in 9 moves, others used the std OLL and got a Y-perm.
 

brododragon

Member
Joined
Dec 9, 2019
Messages
1,976
Location
Null Island
tbh "the best method" is just a combination of a bunch of methods.

The purest form of cfop (doing cross, then F2L, then OLL and finally PLL) is a really, really bad method, but if you add more stuff to it, like COLL, 2GLL, winter variation, OLLCP, ZBLL, 1LLL, keyhole, x-cross, multi-slotting, edge control, etc it gets a lot better.

No other method has that many variations, so maybe roux can be extended, or maybe a new, crazier methods gets invented, who knows.
Max Park averaged very fast times on pure CFOP.
 
Joined
Mar 16, 2020
Messages
598
Location
well, a country.
Roux and Hawaiian Kociemba are better than CFOP here are my arguments for now.

for HK:

HK has EOArrow which makes F2L completely rotationless and more fun that Cross

HK is far more efficient than CFOP, averaging around -40 to 50.
HK is more intuitive than CFOP, the user can choose what to do and what not to do.

HK's last layer (or last step) is more algorithmic than CFOP, and algs are sometimes faster than thinking

HK's solve can be rotationless.

HK's LL is heavy on MU gen which makes it debatably faster

for Roux:

First block and second block is flexible

MU gen is debatably faster

Roux is more intuitive and the solver can do all kinds of tricks in SB and LSE.

Roux has less algorithm than CFOP making it easier to learn.

Roux's first block and second block are rotationless and so is CMLL and LSE.
 
Last edited:

PapaSmurf

Member
Joined
Jan 4, 2018
Messages
597
WCA
2016TUDO02
YouTube
Visit Channel
yeah well it is insanely difficult to learn, and once you learn all of it, it still wouldn't be faster than oll and pll because recognition takes much longer. in theory you could get it faster, but here is the thing. It would take tons and tons of practice. and there are about 500 zbll cases, and you average about getting one new zbll every 500 times, soooo, since there are about 70 something, so you get to practice one 35 times because you always use each in every solve. For zbll, you only use each once in every solve, and you have to do 10 times the practice to get just as good. so yeah, cfop is better.
It's not insanely difficult to learn. The recog+execution is not slower than OLL/PLL, it's around 2 seconds compared to about 2.5-3 in most cases. Also, it takes tons of practice to get fast with any method, so I don't see the problem with that. Also, have you heard of alg trainers? They completely bypass the problem of 'this only comes up once every 500 solves'. Anyway, I seem to know over 50% of them comfortably, so that's not an issue I have.
 

Owen Morrison

Member
Joined
Aug 16, 2019
Messages
1,190
Location
Tennessee
WCA
2017MORR06
YouTube
Visit Channel
Roux and Hawaiian Kociemba are better than CFOP here are my arguments for now.

for HK:

HK has EOArrow which makes F2L completely rotationless and more fun that Cross

HK is far more efficient than CFOP, averaging around -40 to 50.
HK is more intuitive than CFOP, the user can choose what to do and what not to do.

HK's last layer (or last step) is more algorithmic than CFOP, and algs are sometimes faster than thinking

HK's solve can be rotationless.

HK's LL is heavy on MU gen which makes it debatably faster
Sorry, but these are some of the worst arguments I have ever seen. First, you say that HK is better because it is intuitive, and then you say it i better because it is algorithmic...

And then you say that HK CAN be rotationless. Same with CFOP. the only thing that matters if it is rotationless 100% of the time.

Also, MU is not faster than RU, it isn't a bad moveset, but there is no way that MU is faster.
for Roux:

First block and second block is flexible

MU gen is debatably faster

Roux is more intuitive and the solver can do all kinds of tricks in SB and LSE.

Roux has less algorithm than CFOP making it easier to learn.

Roux's first block and second block are rotationless and so is CMLL and LSE.
Honestly the only things bad about these arguments is that MU is not faster than RU, and that you are arguing for Roux's intuitiveness while in your argument for HK you said it was better because it is algorithmic.

Also the fact that Roux has less algorithms just means that there aren't as many ways to improve.
 
Joined
Mar 16, 2020
Messages
598
Location
well, a country.
Sorry, but these are some of the worst arguments I have ever seen. First, you say that HK is better because it is intuitive, and then you say it i better because it is algorithmic...

And then you say that HK CAN be rotationless. Same with CFOP. the only thing that matters if it is rotationless 100% of the time.

Also, MU is not faster than RU, it isn't a bad moveset, but there is no way that MU is faster.

Honestly the only things bad about these arguments is that MU is not faster than RU, and that you are arguing for Roux's intuitiveness while in your argument for HK you said it was better because it is algorithmic.

Also the fact that Roux has less algorithms just means that there aren't as many ways to improve.
the pro of HK's intuitiveness is that you can make it more efficient by what you see, for example, if you see the edges oriented, you can take advantage of it. and I didn't say that the whole HK method is algorithmic and only the last step is important to be algorithmic because there are only so few cases to solve but Roux is an exception because intuitiveness benefit Roux last step, LSE.

I say MU is debatable faster because my MU turning TPS is personally faster than my RU turning TPS.

sorry I meant if you want a method that is rotationless 100% of the time, use HK because EO makes only the F2L oriented so F2L is rotationless. During EO you don't need to rotate since it's just a single F or B moves.

There are many ways to improve with intuitiveness, you can get creative with intuitive solving which leads to efficient solutions, algorithmic step just restricts you to what case it is and recognition is worse than an intuitive step if you have lots of cases, OLL has 57 cases. and there's no way to make the FB and SB step algorithmic.
 
Top