Cubing method debate.

[AJT17]

There is actually a solution that makes Roux 2-Look viable. Pinkie pie. You do a weird form of ZBLS & ZBLL during Second block to CMLL transition which makes LSE easily predictable and the solve is 2 lookable. It also slightly reduces the move count and requires less M moves.

I'm aware that Pinkie Pie acts as an improvement to Roux, but it's still not viable for consistent 2-look solving that ZB has.

I think this video does a good job of offering alternatives to Roux to improve it. And as he highlights in the video, OLLCP is not a good alg set and acts as a brick wall, making it not viable.
And another thing that I forgot to mention, which he talks about in the video, is grip shifts. Roux requires a grip shift to do MU moves, which are inherently slower than RUF/RUD/RUL turning that comprises most of ZB. Having a required grip shift, especially to MU moves, is just such a disadvantage that a lower move count just can't make up the gap. The most likely variant of Roux that can potentially come close to being 2-look viable is ZBRoux, although that still isn't as consistent as ZB and has more variance with the grip shifts that ZB just doesn't suffer from.

 

[Roux]

I agree with you, I just don’t want to

 

[Roux]

I have an overt favouritism to Roux and I can’t explain why

 

[AJT17]

I agree with you, I just don’t want to

I have an overt favouritism to Roux and I can’t explain why

Yeah, that's completely fair. Roux is definitely a good method, and it might have the ability to compete with CFOP or APB at the highest level once people start to get as good as Yiheng is at CFOP or Xuanyi at ZB. I personally have a bias towards APB and think it could be one of the best methods. But objectively ZB has shown to be the best, and any new methods are slowly converging into ZB variants, or at least in the way of ZBLL.

 

[Martyt7777]

I give up I’m just an idiot. I also think that my account has to be of a certain age before I can.

Does this show up in the top right of the 1st post?

 

[Roux]

Nah don’t know why not

 

[Swagrid]

I'm aware that Pinkie Pie acts as an improvement to Roux, but it's still not viable for consistent 2-look solving that ZB has.

I think this video does a good job of offering alternatives to Roux to improve it. And as he highlights in the video, OLLCP is not a good alg set and acts as a brick wall, making it not viable.
And another thing that I forgot to mention, which he talks about in the video, is grip shifts. Roux requires a grip shift to do MU moves, which are inherently slower than RUF/RUD/RUL turning that comprises most of ZB. Having a required grip shift, especially to MU moves, is just such a disadvantage that a lower move count just can't make up the gap. The most likely variant of Roux that can potentially come close to being 2-look viable is ZBRoux, although that still isn't as consistent as ZB and has more variance with the grip shifts that ZB just doesn't suffer from.

All due respect, I just disagree with a lot of the takes you've put up. I think Pinkie Pie is barking up the wrong tree, the rouxers have considered it worse than baseline roux for a long time. The issue I have is how you're defining looks. If you can lookahead into something flawlessly, consistently, does it count as a look? Frankly I'm not sure how to define it, I just want to put up the argument for pauseless Roux. With Straughan recognition, you only need to find 4 stickers to recognise all of CMLL, including preauf. With top Roux solvers currently capable of (almost) always planning FB+SS, we can assume that if a Rouxer were to train their skills like the young stars that can plan F2L do, that they would be able to plan F2B. It's fewer pieces and fewer moves than F2L, after all. With F2B planned, they have the entire exec time of F2B to find the 4 stickers needed for CMLL, which can be executed at full speed since it is an algorithmic step. During CMLL, you just need to catch the EO and find two opposite edges to get your EOLR case, which is technically an intuitive step, in the same way that last pair is, if you know what I mean. From there, you just have to watch the cube during EOLR to foresee your 4c case. You only need to know where 3 edges are gonna be, and there's only like 4 possibilities.

 

[Roux]

Yeah I suppose

 

[Roux]

Pinkie pie isn’t that good I guess

 

[Roux]

Unrelated question; do you have a phone based ZZ YouTube channel

 

[airandwavebbs]

me like roux. roux good. roux fun.

 

[Roux]

me like roux. roux good. roux fun.

Agreed!!!

 

[CrossPetrus]

I think that in theory, APB is the best method (though ZB is probably easier because you don’t have to get good with blockbuilding). But I think that the current methods are not the best we’ll have. I think that in around 2-3 years there will be a new hype method that will in theory be the best (but of course it probably won’t gain too much traction in practice).

 

[Swordgamer969]

I would say that it depends simply on what your used to, and methods can be for others. ZB is good if you're willing to learn over 800 algs, Roux is fine especially if you're good at M moves, and also Roux is more optimized, but CFOP is basically ZB but easier so its WAY more accessible

 

[ClockIt]

Technically roux is the very move efficient and if M moves were the same speed or faster then R and L moves then Roux would no doubt be the best method, so maybe if there is a dude out there who can do the World's fastest M moves then Roux would be unstoppable for that guy

 

[Martyt7777]

Technically roux is the very move efficient and if M moves were the same speed or faster then R and L moves then Roux would no doubt be the best method, so maybe if there is a dude out there who can do the World's fastest M moves then Roux would be unstoppable for that guy

There was a prototype cube like half a year ago that made slice moves require the same force as outer face turn cubes on regular cubes (to clarify, in regular cubes, slice turns are 2x harder than outer face turns, while the prototype cube is the opposite

 

[Sion]

Tbh, almost every "new" method is some variant of "Blockbuilding step -> EO -> Arbitrary Algorithmic Step that leads to Oriented Edges and unsolved corners -> ZBLL." I just call this the "People keep Reinventing Petrus" method family.

My interest will be more piqued when there's a method besides roux isn't just another arbitrary Freefop+EO Lead-up to ZBLL.

 

[Caleb Valenzuela]

Tbh, almost every "new" method is some variant of "Blockbuilding step -> EO -> Arbitrary Algorithmic Step that leads to Oriented Edges and unsolved corners -> ZBLL." I just call this the "People keep Reinventing Petrus" method family.

My interest will be more piqued when there's a method besides roux isn't just another arbitrary Freefop+EO Lead-up to ZBLL.

I mean, that is pretty much the easiest way to make a good method for today’s standards

Variability in methods is becoming much less in recent years though.

 

[Mr.Ducky]

I mean, that is pretty much the easiest way to make a good method for today’s standards

Variability in methods is becoming much less in recent years though.

After looking into alottt of methods, yea pretty much all of them are basically petrus and roux varients(solving blocks then doing eo in some way or have to use zbll at the end). anyway I really liked rhe ideas of mehta method, tripod, heineas (or whatever its called) and how roux works in last six edges. These were truly unique, but "can we really make such mind boggling methods today that aren't copies of other methods?" Is the true question today

 

[Sub 20 Cuber]

ZB’s the best method in my opinion