• 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!

A Roux/ABC Hybrid with a very low movecount and few algs

crafto22

Member
Joined
Sep 4, 2015
Messages
303
Location
Canada
WCA
2014ADAM03
Hi everyone, here is a Roux and Adam-Briggs Columns hybrid method I thought of and I think it has potential

1. Solve FB, but the E-slice edges only need to be oriented E-slice edges, they don't need to be permuted.
Movecount: 6-8
2. Solve SB with E-slice edges oriented and in the E-slice but not necessarily permuted and up to one corner misoriented.
Movecount: 7-9
3. Use an EJLS alg to orient corners.
Movecount: 8-9?
4. Solve SB whilst permuting the corners and fixing the E-slice
Movecount: 13-15
5. L6E
Movecount: 12

Total moves: ~48 STM

Tell me what you guys think.
 

Okboyyyy

Member
Joined
Apr 15, 2016
Messages
6
This method is just as efficient as roux, with harder look ahead. How many algs?

Also permuting corner while SB should require intense recognition right?
 

wir3sandfir3s

Member
Joined
Mar 24, 2016
Messages
538
Location
Under the hanging piano.
YouTube
Visit Channel
Seems somewhat similar to a method developing in the New Methods/Ideas thread. Cool idea, definitely potential here. Add more variations?
Also, I think it would be more efficient to just solve the E slice in the beginning, effectively making it Roux... So, yeah.
 

crafto22

Member
Joined
Sep 4, 2015
Messages
303
Location
Canada
WCA
2014ADAM03
Seems somewhat similar to a method developing in the New Methods/Ideas thread. Cool idea, definitely potential here. Add more variations?
Also, I think it would be more efficient to just solve the E slice in the beginning, effectively making it Roux... So, yeah.
Well I guess it is similar to what I proposed in the new method concept idea thread since that is what this is.
This method is just as efficient as roux, with harder look ahead. How many algs?

Also permuting corner while SB should require intense recognition right?
16 algs, and it's actually really easy I'll do an example solve tomorrow.
Seconded. This is not intuitive at all. but I do like the method overall
How is it not intuitive? I can do it intuitively.
 

theradhaxor

Member
Joined
Apr 5, 2016
Messages
32
I have a proposal
Why not
1. Roux FB
2. 2x2x1 in DB, but the corner piece can be any remaining white corner piece, just oriented
3. While inserting the last corner-edge to finish the second block with misplaced corners, orient the top corners as well
4. Permute conversion both layers using 5 square one algs
5. LSE
 

Teoidus

Member
Joined
Feb 11, 2016
Messages
583
Location
Char
DB -> "down-back". Since you're talking about finishing SB and ending with LSE i think you mean DR here.

I'm not sure it's worth the extra freedom you'll get from being able to misplace corners (as in, i don't know if it'd save moves) compared to normal Roux.
 

theradhaxor

Member
Joined
Apr 5, 2016
Messages
32
DB -> "down-back". Since you're talking about finishing SB and ending with LSE i think you mean DR here.

I'm not sure it's worth the extra freedom you'll get from being able to misplace corners (as in, i don't know if it'd save moves) compared to normal Roux.
Yes I meant DR thanks for correcting me. But I'm not sure you read the whole thing. You use F2Ll like SLS or WV to orient all corners while inserting, which takes 5 moves average. Then you have a 1/5 chance for a skip here, but you permute both corners using 5 algs. Even if this is just as effective(it should be moreffective ) the algs needed are in the 30s, compared to roux's 42
 

Teoidus

Member
Joined
Feb 11, 2016
Messages
583
Location
Char
I read the whole thing, I still don't think it's worth the extra fancy components you've added. First of all, you don't always want to build a 1x2x2 in DR, because that further restricts freedom of blockbuilding that gives Roux it's efficiency.

If you want to be able to build a 1x2x2 in front instead of in back and then use WV, I believe you'd then have to mirror algs - unless you can do this on the fly, that's already twice as many as you projected, and I doubt they're particularly ergonomic. At times it may also be rare, but you can get SB by building a line and keyholing in the last edge--this option is also lost with your method.

I find it very dubious that less efficient SB + WV + Sq1 alg saves any significant number of moves on average compared to normal SB + CMLL.

On top of that since you are using SQ1 algs to permute corners, which don't affect EO (i think, since ECE uses these and its first step is EO), you can't influence EO while solving corners like you could by learning multiple CMLL subsets or OLLCPs--because of the complexity you added, you lose room for more specialized alg sets that would make things even more efficient.

ALso, this definitely should be in the method/substep thread, not this one specifically devoted to the roux/abc hybrid
 

theradhaxor

Member
Joined
Apr 5, 2016
Messages
32
First, I see your point on DR limit is blockBuilding, but keep in mind that WV can be mirrored to BR slot as well, allowing you to build the block In DFR. Maybe instead of only doing this, speeds overs who already use roux could see which one is faster during the solve? Also, doing am M or M' or M2 usually helps before doing the Square one algs to improve edge orientation, the problem is that sq1 algs rotate bottom edges as well, so a new alg set might be required. Maybe this can be an addition to the roux method, is SB is a very bad case and would require more moves.
 

crafto22

Member
Joined
Sep 4, 2015
Messages
303
Location
Canada
WCA
2014ADAM03
Okay, here is a modification to the original proposal with an example solve:

1. Solve a 2x1x1 in DBL as a part of your Roux block
2. Solve the remaining FB corner with an oriented E-slice edge
3. Solve the rest of the E-slice and CO with EJLS (16 cases)
4. Permute the E-slice whilst solving the SB
5. CP + EO (11 cases)
6. L6EP

Example solve:
Scramble: F2 U B2 U F2 L2 U L2 U2 R2 U' B L' B2 R U L' F' L' R' U2

z2 // inspection
F U B' x // Square [3]
U' F U' F2 // Corner + oriented E-slice edge [4]
R' U F R F' R U R U' R // E-slice + CO (E-slice edges + cancel into an L OCLL) [10]
U2 R r U R' r' u' R2 u // SB + E-slice (completely pseudo 2-gen) [9]
M' L F' R' F' M' D2 B' U L U' B D2 x' // CP + EO [13]
U' M U2 M U2 M2 U' M' U2 M // L6EP [10]

Total moves: 49 STM
 

crafto22

Member
Joined
Sep 4, 2015
Messages
303
Location
Canada
WCA
2014ADAM03
IMO roux is more efficient and ergonomic (roux avgs 48 STM)
I got unlucky with the CP + EO, normally the movecount would be better, plus the SB is better recognition-wise and the L6E step is better IMO. As for CP + EO, it is about the same as CMLL, but the FB is undoubtedly better in this method.
 

Shiv3r

Member
Joined
Mar 16, 2016
Messages
802
Location
San Diego or thereabouts
WCA
2016LEWI04
YouTube
Visit Channel
Yes I meant DR thanks for correcting me. But I'm not sure you read the whole thing. You use F2Ll like SLS or WV to orient all corners while inserting, which takes 5 moves average. Then you have a 1/5 chance for a skip here, but you permute both corners using 5 algs. Even if this is just as effective(it should be moreffective ) the algs needed are in the 30s, compared to roux's 42
I use Roux and I've only needed 10 algorithms for the whole thing
 
Top