The WaterRoux 3x3 method thread

[efattah]

The same scramble using my personal preference for CLL finishes in 41:

Scramble: L2 U2 B2 D2 B' U2 R2 U2 L2 U2 F2 L B R D R2 F U B2 R2 F2
D' R2 y M2 B//FB
U' R' U' R2 // lower right corners
U R U' R U' R' U R' F R2 F' // CLL pi case
U' M' U' r2 U M' U' R2 U M U' r // E2L
M' U M U2 M U // LMCF L5E BDR set
E2 M' E2 L' // permute midges
41 STM

 

[Neuro]

Yeah I figured it could've been under 40, I guess I need to work on E2L. But thanks for showing me that, I'm sure it'll be useful!

 

[efattah]

With all the talk of TEG1, I realized TLEG1 could be a game changer for WaterRoux. With TLEG1 it becomes possible on many solves to see the entire first block AND the corner solve in 1-look. It still takes an expert to do so, but this would finally resolve the last hurdle of Roux which is to put the CMLL recognition into the inspection phase; while still maintaining a lower move count than regular Roux, by a quite significant amount.

Better still, you don't need full TEG1 for WaterRoux since the left block is always solved, so you only need four TLEG1 sets, L+, L-, R+, and R-, as well as normal LEG1 and CMLL optimized for an unconstrained left block. So this would be 252 corner algorithms in order to be 'in range' of a 1-look solve of the left block and all the corners. An example:

Scramble:
B2 R' D2 B2 R2 F2 D F2 U B' R' F' R2 B F U2 B' R' U'
y M' S U S' // FB (4/4)
R2 U' R' // yellow corners (or JUST R' if you solve the white corners)
// Now solve corners with TLEG1, L+ set
[I would comment that if you are good enough to solve with non-matching corner sets, it only takes 1 move to put the white corners in, meaning just 5 moves to 1-look the corners and first block]

In the above example we need to see 7 moves into the future in order to 1-look the yellow corners which is hard but within range of a master (or 5 moves to 1-look the yellow/white not matching corners). After that, there are just 5 L/R edges left, of which we need to solve 3 in order to reach LSE. However half the time one is already solved at random, meaning we are one E2L pair away from LSE. Allowing no constraints on where this pair is, sometimes this pair solves the UL slot and a random redge, putting us into a Waterman L6E situation. However, I have started optimizing the Waterman L6E sets, and Waterman Set 2 is the one that occurs most often by far, and I have so far optimized it to an amazing 9.4 moves while improving the speed and ergonomics of the algorithms. So even in the situation where you end up with UL solved, it means in most cases you can solve the last two redges and orient the midges in less than 10 moves, or 11 moves if you include 1 set up move.

If you 1-look the first block and corners, this is around 15-17 moves at essentially max TPS since you know exactly what to do; so this could be 1.7 seconds. Then you solve one E2L pair in 1 second, then you finish with either Roux EOLR, LMCF L6E, or Waterman L6E, the worst case being around 2 seconds for that operation. So a 1-look solve would be around 4.7 seconds for an expert.

As WaterRoux-TLEG1 cuts at least 1-2 more moves off the average move count, then it will be even lower than regular WaterRoux, which was already lower than Roux. But the bigger advantage is being able to 1-look the corners.

 

[Thermex]

@efattah in response to what you posted on the new concept thread.. you're totally right, that is the main difference to our methods. To test out which method of ours is better, I took the TEG hammer set I made yesterday and applied it to my method (preserving DL, DB and DR edges while solving corners) and your method (preserving FL, DL and BL while solving corners) to see if the cases would preserve our pre-solved edges. Only one or two TEG algs preserved the 3 ledges in your guys' method, and almost all of the TEG algs preserve the d-layer edges used in my method. My method is also almost completely rotationless if you solve FB on the d-layer (which is realy really easy) and use a TLEG alg+three u-layer edges. The only rotation would be a z rotation after these for L6E. I also feel like the third option you gave for L9E (three redges+LMCF L6E) is the best way to go until a better way of doing L7E is developed. Could you leave me a link to the place somerandomkidmike has these algorithms or talks about them? I'm very curious. And yes lets move this discussion over to the waterroux thread, you can write your reply to this post there.

 

[crafto22]

@Thermex I have just finished developing a method that solve L7E in 18 moves on average, although lucky sub-15 move singles are very common. I can share my document with you as soon as I have added the finishing touches.

 

[Thermex]

@Thermex I have just finished developing a method that solve L7E in 18 moves on average, although lucky sub-15 move singles are very common. I can share my document with you as soon as I have added the finishing touches.

Oh cool! Show me it when you're done. Today I just finished one of the 42 alg TEG sets, I'll be showing that in the new method/concept thread later today once I'm done revising it. But isn't 18 moves for L7E higher than the average for the method Effatah came up with for solving L7E (placing redges/ledges, permuting midges)?

 

[efattah]

Oh cool! Show me it when you're done. Today I just finished one of the 42 alg TEG sets, I'll be showing that in the new method/concept thread later today once I'm done revising it. But isn't 18 moves for L7E higher than the average for the method Effatah came up with for solving L7E (placing redges/ledges, permuting midges)?

We have done extensive tests & comparisons and Crafto's L7E method averages 18 moves with around 100 algorithms, whereas my method average 18.5 with more than 350 algorithms. So it seems his method is way better. There is still some question about how fast the recognition is, but by the sound of it, pretty good.

His new L7E method should be a game-changer not only for WaterRoux but for LMCF, as LMCF 'intermediate' will experience a decrease of around 3-5 moves per solve with a drastic reduction in the number of algorithms ultimately needed.

 

[crafto22]

Oh right, I forgot to mention I'm working on a recognition system atm that works really well! It relies on the fact that only 3 or 4 edges out of 6 are required to recognize the L6EP algs, so you just ignore the last redge entirely. In cases when L6EP must be solved intuitively due to a setup being very inefficient, one can either ignore the last redge or imagine it is the edge in front or behind it. Sorry if that makes little to no sense, details will be in my document.

 

[Thermex]

@crafto22 I kinda get what you mean and it sounds pretty cool, but of course seeing the document would help. Btw the link to the TEG- PDF is attached at the bottom of this post.

Like I said in the new method/concept thread, all the TEG algs are really good, but the gun set kinda sucks. I can sub 2 all of the cases and sub 1.5 most of them. A lot of them (~15%) were 2-gen and the average movecount in an algorithm was about 8.6 moves. I was too busy to make pictures for the cases, so I just wrote them out in the same order they are on cyotheking's website: http://www.cyotheking.com/tcll-1/
And when there's an "a" alg and then a "b" alg, those are two equal algorithms that can be used for one case.

@efattah I thought your original L7E method was like 15 moves on average or something, but whatever. Crafto's method sounds pretty cool anyway and can hopefully be a game changer for upcoming methods. So now that I'm done with the FR- set I think I'm going to move on to the FL- set! I'll post those algorithms at the end of the month once I'm done. Tell me if you find any algs that can be done more efficiently/ergonomically, especially the 4th gun case, that one SUCKS. Also do you think there should be a TEG thread? Seems like we talk about it a lot.

 

[crafto22]

Here is the link to the complete WaterRoux L7E spreadsheet. I will be uploading an in-depth tutorial to my Youtube channel soon.

 

[efattah]

@crafto22 I kinda get what you mean and it sounds pretty cool, but of course seeing the document would help. Btw the link to the TEG- PDF is attached at the bottom of this post.

Like I said in the new method/concept thread, all the TEG algs are really good, but the gun set kinda sucks. I can sub 2 all of the cases and sub 1.5 most of them. A lot of them (~15%) were 2-gen and the average movecount in an algorithm was about 8.6 moves. I was too busy to make pictures for the cases, so I just wrote them out in the same order they are on cyotheking's website: http://www.cyotheking.com/tcll-1/
And when there's an "a" alg and then a "b" alg, those are two equal algorithms that can be used for one case.

@efattah I thought your original L7E method was like 15 moves on average or something, but whatever. Crafto's method sounds pretty cool anyway and can hopefully be a game changer for upcoming methods. So now that I'm done with the FR- set I think I'm going to move on to the FL- set! I'll post those algorithms at the end of the month once I'm done. Tell me if you find any algs that can be done more efficiently/ergonomically, especially the 4th gun case, that one SUCKS. Also do you think there should be a TEG thread? Seems like we talk about it a lot.

Isn't the DFR- set thet same one Cale generated? What was his case for the 4th gun case?

BTW my L7E was 15 moves for the cases where the cube was already 'set up' with the two redges on the R layer. It is pretty bad when neither redge slot has any redge in it.

 

[Thermex]

@efattah crap, I assumed that the algs Cale S made were only stollery cases (he said they could be solved by doing F R U' R', which is the stollery set) but he might have made the full FR- set, I'll have to ask him (I'll actually do it this time). Either way, it still wasn't a big waste of time since it was only a week and I got to know cube explorer a lot better, plus since I spent a lot of time on some of the algs they might be a little better then the ones he generated, but I would still like to see what he had for some of the gun and pinwheel poser cases as some of those kind of sucked.

Okay, three quick questions, first off, should one of us start a TEG thread, and if so, who? Second, do you think over the summer we should generate TEG 2 algs? It seems like they should exist as anyone crazy enough to learn all 320 TEG 1 cases probably won't mind adding an extra 80 algs to save half a second or so every ten solves. Finally, something I've been wondering about for a while: do you think for the BL and BR TEG sets the twisted corner should be held in the back or the bar should be held in the back? Which is better? So yeah, hopefully you can answer those questions, I'll ask Cale S for his algs and start making the FL- set tomorrow.

As for L7E, I like this improvement Crafto made as it consists of several short algs which are fun to use and is also pretty fast with not too many algs. I think it's a great method for intermediate solvers, but I still think some sub-15 move L7E method needs to be developed in order for the "solve two u-layer edges then L7E" strategy to be fast and sub-40 moves, since solving two u-layer edges is about 6 moves. My favorite WaterRoux variation is still the one I came up with a while ago (block, TEG, TEUL and L6E) and I'll try to generate the TEUL algs with someone over the summer break.

 

[crafto22]

@Thermex I spent literally hours upon hours thinking of a way to solve L7E, so what I am saying comes from a lot of experience with L7E. My method is the absolute best I could come up with using less than 100 algs. However, I know for sure that L7E can be solved in 15 moves or less every time if we remove a limit on alg count. If we eliminate the O2E step from my method, essentially turning OL5E into OL7E and add 43 new L6EP algs, we get a 14.5 move L7E with about 210 algorithms. I can start generating algs for that if anyone really believes saving 3.5 moves is worth 100 new algs and bad recog. Honestly, efficiency is overrated. Fast recog + ergonomic/small algsets > extreme efficiency. That is my opinion, of course. I think we have all become so caught up with creating the most efficient method possible that we have forgotten that the current "best" 3x3 method in terms of real-life results in terms of speed is CFOP, despite the average move count surpassing 60. WaterRoux is very particular. It has a very low move count, but also has excellent recog for almost all the steps. My L7E method has pretty good recog in my experience, but I find I get FAR better times using LMCF L5E, and this method is ALMOST as efficient as L7E. The major difference is that using LMCF L5E produces a solve with extremely good recognition, unlike anything I've ever seen. Pieces are extremely easy to track during ERL and L5E recog is very nice. Same goes for all the corner algs. So really, what I'm saying is that maybe L7E could be extremely useful for very advanced WaterRoux/LMCF solvers. However, I don't see the use in pushing the limits of move count so far when the real objective here is speed. Sorry for this ramble XD. Just my personal opinion, of course.

 

[Thermex]

@crafto22 I see your point, me and others really get way to caught up in thei move count and sometimes forget about ergonomics and recognition. I'm not sure if it's worth generating those algs, but I still think some super alg-heavy way of doing L7E with decently good recognition could be developed, but of course you know way more about this than I do.

 

[efattah]

I agree 100% with crafto. The more I cube, the more I realize the benefit of using smaller sets with good ergonomics and super recognition. There seems to be a certain critical number of algorithms, for me it seems to be around 250, and going above that, the cases just don't come up often enough. This means that solving cubes (fast or slow) is no longer sufficient to maintain familiarity and speed with the algorithm sets. What that means is that cubing becomes a chore; a beneficial cubing session ceases to be solving but instead becomes repeating long stressful slide shows, drilling algorithms that you haven't seen in thousands of solves. And no amount of slide-show drilling ever seems to create the grace and fluid familiarity of executing a case that you have seen in thousands of solves. Executing a case you have drilled thousands of times but only seen a handful always seems slower and more erratic.

I think one reason vanilla CFOP (78 LL + 42 F2L) and Roux (even with EOLR) are so successful is because the algorithm sets are just the right size to be small enough to allow real mastery and each case comes up often enough that solving alone is good enough practice. The cubers that use full OLLCP and ZBLL are forced to spend huge amounts of times doing drills, and the ones that are diligent in doing that gain some advantage, but the fun factor drops.

 

[Thermex]

@efattah I see your point, you're totally right. Unlike you guys I've been cubing for a very very short time (~6 months) and haven't had to learn any huge algsets or anything, so I don't know much about this stuff. The two sort of distict types of methods are the hardcore low movecount alg-heavy needs-a-lot-of-practice method for people who want to put in a LOT of time in practicing and learning algs and getting faster (ex: ZBLL), and a simpler low algorithm fun version of that method for the general public (ex: 2-3 look CFOP LL). I really feel like method designing can be split into those two categories, and people's opinions on "what's a good method" often depends on which of those two examples they're going for. So far pretty much all the variations of WaterRoux we're making are definitely geared toward the first audience I talked about, the people that are willing to learn a TON of algs to save a few moves. And I really think you're right, the problem with my WaterRoux variation is the fact that althogh TEG and L6E aren't that alg-heavy (250 & 89 algs), the third step (TEUL) would be over 300, which is learnable but as you said not fun and would probably have some crappy recog. This weekend I'm going to look into some slight less alg-heavy ways of doinng L9E in WaterRoux, maybe something kind of styled like Crafto's method but with 9 unsolved edges would be cool, but idk.

 

[Neuro]

Here's a solve proving the true potential of WaterRoux. The solve was fairly average and still got under 40 moves while over half of the solve was pseudo 2 gen and over 75% of the solve was simply algorithmic.

U R' D2 L B L' y//FB (6/6)
R2 U y' U R' U2 R' F R F' R y//Corners (10/16)
M U2 M2 U R2 U M2 U'//ERL (8/24)
r U' M U R2 M U M U2 M U' M2 U2 R'//L7E (14/38)

I'm working on a special ERL set and will probably get with somerandomkidmike about cases where no redges are in UL or M slice.

 

[crafto22]

I wasn't sure where to post this so I'm putting it here. I found a pretty good way to solve the L5C when FB and a 2x2 at BDR are solved.
First, you orient all the corners using one of 16 algorithms, most are 2-gen and super fast, average is 5 moves. Next, you permute the corners using one of 6 algorithms, average is 7 moves. This would have L5C solved in roughly 12 moves using only 22 algorithms (more like 14 because 6 of the cases are essentially intuitive). This is very good considering using regular TCLL is about 11 moves and requires 84 algorithms. Seems like saving 1 move probably isn't worth the 70 extra algs if ya ask me!

 

[crafto22]

So the average moves is actually about 13, but considering we're talking about only 14 algs versus 84, I'd say it's worth 2 moves. This could be used to solve in one of two ways:
1. FB (8)
2. DRB corner (1.5)
3. L5C (13)
4. ERL (7)
5. L7E (17)
Total: 46.5

1. FB (8)
2. Square (7)
3. L5C (13)
4. L7E (17)
Total: 45
Second option is only slightly more efficient, probably due to more blockbuilding.

 

[Rcuber123]

I know this is closed to Roux but it's a cool idea I thought of while reading crafts post.

FB
Square
Make a pair without inserting it
Orient 3 top corners while preserving the pair
Insert pair with block commutater to solve permutation of corners (9.5 I think)
Lse

Block commutaters give good edge influence possibilities so ppl could almost always solve at least an edge leading to l5e. EO influence can also be intresting