Should I learn Roux or ZZ?

Status
Not open for further replies.
i thought CubeRoots has already switched to Roux by now.

ZZ is definately the closest of the methods to CFOP, but expect a mildly steep learning curve. it's not really that bad, but it's a learning curve nonetheless, because EOLine is deceptively difficult, more the EO than anything else. F2L is pretty straightforward, but doing it efficiently might take a bit of time to get your head around, but that's also mildly the case with Roux, and even just blockbuilding in general because the mindset's a bit different than CFOP-style F2L.

in the end it's really your choice to make, since you're the one that's gonna be solving with your choice, but try both for a bit, see which one you like more. If you're not good with M turns but choose Roux, your M slice moves will improve by default really, there's no real way around not having good M slice turns with roux.

I'm not the fastest at ZZ, but i don't really personally regret switching, in fact i'm enjoying it quite a bit despite i barely ever really practice or try to lower my times anymore QQ. My real big choice for ZZ over say Roux or Petrus was the fact that ZZ seemed like the closest to CFOP which i was already doing anyway.

tl;dr, Roux is cool, ZZ is cool, whichever way you go, is also cool

ps. did i mention cool?
 
I'm not sure if was mentioned already, but OH and big cubes (4x4+) will be much more difficult with Roux if you use reduction method on them. Not impossible, but you need to be more than a little crazy to make it work. ZZ isn't so bad on big cubes because with practice you can see bad edges without thinking about it, and orient them pretty fast intuitively while looking for the DB and DF edges. Throw those in place, and then the rest of the solve is really nice from there. With Roux you can throw the first block together pretty fast with practice when you get to the 3x3 stage, but from there the M and r movements get more awkward the larger the cube is. On a 7x7 you have to be completely nuts to seriously use Roux with it, although again I don't want to say it's impossible, just making things harder than they need to be.

For OH ZZ seems like it's hailed as king because of the <R, U, L> set, and that's been my experience for sure. With 2H 3x3 there are no real objective advantages to ZZ or Roux. It's comes down to preference.

In a way CFOP (Fridrich) is kind of the catch all of cubing. You don't need to learn block building, the inspection is simple but extendable with things like XCross. Color neutral CFOP is easier relative to other methods I think. It's also fast for transitioning to the 3x3 stage of big cubes. Look ahead is arguably the easiest of all speed methods because you can easily get away with "see case, solve, see case, solve, etc". Also the finger tricks are very straight forward. It's obviously just fine for OH. Not to mention the huge amount of documentation and guides/tutorials/examples out there for it, I think it's pretty clear why CFOP is the most popular.

Just please choose a method that you find the most interesting and fun. If you're deciding between ZZ, Roux, or even CFOP, you really can't make a bad choice, you just need to make it's the right one for you.
 
Dominate said:
I'm not sure if was mentioned already, but OH and big cubes (4x4+) will be much more difficult with Roux if you use reduction method on them. Not impossible, but you need to be more than a little crazy to make it work. ZZ isn't so bad on big cubes because with practice you can see bad edges without thinking about it, and orient them pretty fast intuitively while looking for the DB and DF edges. Throw those in place, and then the rest of the solve is really nice from there. With Roux you can throw the first block together pretty fast with practice when you get to the 3x3 stage, but from there the M and r movements get more awkward the larger the cube is. On a 7x7 you have to be completely nuts to seriously use Roux with it, although again I don't want to say it's impossible, just making things harder than they need to be.

For OH ZZ seems like it's hailed as king because of the <R, U, L> set, and that's been my experience for sure. With 2H 3x3 there are no real objective advantages to ZZ or Roux. It's comes down to preference.

In a way CFOP (Fridrich) is kind of the catch all of cubing. You don't need to learn block building, the inspection is simple but extendable with things like XCross. Color neutral CFOP is easier relative to other methods I think. It's also fast for transitioning to the 3x3 stage of big cubes. Look ahead is arguably the easiest of all speed methods because you can easily get away with "see case, solve, see case, solve, etc". Also the finger tricks are very straight forward. It's obviously just fine for OH. Not to mention the huge amount of documentation and guides/tutorials/examples out there for it, I think it's pretty clear why CFOP is the most popular.

Just please choose a method that you find the most interesting and fun. If you're deciding between ZZ, Roux, or even CFOP, you really can't make a bad choice, you just need to make it's the right one for you.
Click to expand...

Thank you for the advise, I already know CFOP, I am going to try zz and roux to see which one i like better.
 
I recommend learning roux because one of the fastest people in the world uses roux (5BLD) he subs 8. If you are going for OH learn zz but, roux really really pays off.
 
This has already been stated, but ZZ does have more last layer options thatn Roux. I'd suggest you expiriment with Roux, ZZ, Fridrich and their different last layer approaches, and experiment with some other Fridrich approaches, then see what you want to do. Some things not metnioned here are:

CPLS and 2GLL- This would work well with methods that already have edges oriented after you get to the last F2L slot. You can solve the edge that goes in the last slot, then place the corner that goes in the slot while permuting top layer corners and preserving edge orientation. This is called CPLS. It has 26 algorithms and its recognition method is not very straight-forward. 2GLL may then finish the cube by solving the permutation of edges and orientation of corners. This can be done with only R and U moves (great for OH, a small kitten knows and uses a portion of 2GLL for OH ZZ solves), thus it is done very quickly. It is probably less common than OLL and PLL because it has 84 algorithms (110 with CPLS), and recognition takes time to get used to.

COLL and EPLL- This also is used for methods with the last layer edges already oriented. You first orient and permute corners using 42 algorithms. Then you use EPLL (PLLs that deal with only edges) to finish the solve. I don't honestly see why people learn it when the 7 corner OLL cases and PLL are quicker recognition (in my opinion) and require less algs.

CLL and ELL - This is a last layer method that could be used well with Roux or Fridrich. You first solve the corners (orientation and then Permutation) using 42 algorithms, and finish with ELL (solves edge orientation and permutation) which has 29 algs. In Roux you can use slightly shorter CLL algs because you don't need to preserve the M slice as it is done afterwards. Without preserving the M slice, this is called CMLL. I think this has just as much potential as doing CLL and PLL due to move count, and the ability to use with shorter algs in Roux. I think it is less popular because OLL and PLL have easier recognition.

ZBLL - This is not really practical because it has 493 algorithms. It solves the corners of the top layer while permuting the edges. Basically, it is a 1 look last layer solve after edges are oriented.

ZZLL - This is similar to ZBLL, but requires only 167 algorithms because it requires that you have atleast 2 opposite edges permuted in the top layer. Goo cubers can solve 2 top edges with insertion of the last F2L pair.

L5P - This solves the whole top layer after you have a 1x2x2 block solved in it. It is called L5P because there are 5 pieces in which it solves (3 corners and 2 edges). This requires 107 algorithms if you don't have edges already oriented. It is not used by speedsolvers mainly because the high alg count and having to intuitively create a 1x2x2 block in the top layer first. It is the final step of the Tripod method.

OLLCP - This is just like regular OLL, but you place the corners at the same time. This requires over 300 algorithms. Most cubers who use this just learn some ertain cases with easily preformable algs and that are easily recognizable. Basically, they just learn a couple different OLL algs for the OLL cases to be able to permute corners at the same time as OLL every so often (whenever they get a case they know).

That pretty much sums up any last layer options you might consider. There are some others out there, but I don't think I even need to mention them because they are not for speedsolving or are just inefficient in general. There are some other useful steps that can be done to affect the last layer before you get to it. Here are some of them:

MGLS - You first place the edge to the last slot while simutaneously oriented edges with 21 algorithms. This step of MGLS is called ELS. For methods that pre orient edges, you just have to place the correct edge in the slot. After ELS, you do CLS which solves the corner of the last slot while orienting the last layer corners. This uses 104 algorithms. Afterwards, only PLL remains.

ZBLS aka ZB F2L - This is used mainly for Fridrich solves. It orients the last layer edges at the same time as solving the last F2L pair. This uses 125 agorithms without mirrors or inverses. It is possible to do this by orienting 2 top layer edges on the 2nd to last slot and finishing the edge orientation with the last slot using 86 of the ZBLS algorithms. You could also use this a slot early making this be done on the 3rd to last slot and 2nd to last slot enabling you to use a different last slot substep. Very (I repeat - VERY) experienced solvers can intuitively solve edge orientation throughout doing F2L and have edge orientation complete by the time they reach the last slot.

WV and SV aka Winter Variation and Summer Variation - Winter Variation is the solving of the top layer corner orientation while placing the last corner edge pair. Summer Variation is the same thing, but it solves the F2L cases where the edge is opposite of the corner (the case solves with R U R' or L' U' L). These both use 27 algorithms individually and can be recognized by looking at the F2L pair and the other 3 corners in the top layer. These both preserve edge orientation.

Out of all the different ways to solve the last slot and the last F2L pair, MGLS is definently one of the best. Many good cubers know some of CLS if not all. You do not have to learn all of these and try them to see which ones you like, just look though them and see what interests you. Anything that you think would suit you well, I'd suggest trying out. Sorry for the long message and I just how It is understangable. I've been told that It is hard to read my ideas without being confused...
 
I'd pick a method to focus on, and just work on being very efficient and fast at that before delving into the world of LL options. Otherwise, you're just learning algorithms and advanced edge control techniques without having the basics down.
 
I am going to learn both of them, get to where I have mastered their basic, then start on the other. Which will I start first? Roux, mostly because more use it, so I will be able to find more help. Plus, I love me some m slices. After I dabble in both, and can get a decent sub 40 with each, I will decide which route to take. (The one that is more fun)
 
If you wanna start practicing OH then I highly recommend ZZ method because it uses only L,R and U moves after the EO line and after finishing the F2L the last layer edges are oriented

Well! I will recommend Roux because it consists of block building and gives a cool look in the last step because it uses a lot of M slices which gives a cool look. Alexander Lau is one of the fastest roux solver. You can watch his video from his youtube channel. http://www.youtube.com/user/5BLD . He made some roux tutorials. ONE MORE THING! It uses less number of algorithms and less number of moves. There is no cube rotations in roux. I think there are 42 CMLL (orients and permutes the corners). Only 42 algorithms is needed to learn in total. The first, second and forth step is intuitive. In my opinion roux method is highly recommended with practice you may get sub 10, :)

I would recommend ZZ for one handed solves because it uses only L, R and U moves after the EO line . There is no cube rotations in ZZ. ZZ method is all about efficiency. There are about something like 400 algorithms for the last layer. It uses less number of moves. All in all I highly recommend this method for one handed solving. :)
 
Last edited by a moderator:
Roux is not harder it feels harder because you started practicing CFOP earlier and roux you just started. Keep practicing with roux it's so cool
 
ZZ is vastly better for FMC and OH, but I think both methods offer unique learning experiences and that you should try both.
 
Status
Not open for further replies.

Similar threads

HiroshiCubes
Should I learn Roux?
2
Replies
33
Views
3,105
Dutch Speed
D
C
Should I learn CFOP, Roux or ZZ?
2 3 4
Replies
75
Views
22,483
Burrito
Burrito
Woofi
So um what order should i learn CLL?
Replies
6
Views
721
Renslay
Renslay
E
What should I learn next
Replies
3
Views
481
Jared Burkhart
Jared Burkhart
LukeKing67
How should I learn blindcubing?
Replies
5
Views
729
Caleb Valenzuela
Caleb Valenzuela
Back
Top