Square-1 Mehta APDR Method Now Available!

[OreKehStrah]

Hi all,

I've always been interested in block-building based methods for square-1, and after hearing people talk about adaptations of the Mehta method for square-1, I decided to look into it since it sounded interesting. Unfortunately, it was only an idea people had discussed, but never developed.

Last week, I decided I would start learning how to gen square-1 algs, and I genned the full set of algs needed for the APDR step of the method. The link to my document containing all the algs will be down below.

For those unfamiliar with the method, this is a 4 step method that goes as follows: 1. CS or CSP, 2. Mehta-Style FB on the D layer (Lin FB with DF and DB also solved), 3. APDR, 1 alg to solve the last 3 pieces of the D layer, 4. (Parity) PLL.

I'm not sure how well this method can compete against VB and Lin on its own, but I can see a future where people who use Lin currently, will actually learn a few different block-based methods, such as this to take advantage of the different blockbuilding opportunities. This would be similar to how 2x2 solvers use lots of different approaches based on the scramble. I have another blockbuilding method in the works with 2 variants that I think could be pretty cool to learn as well.

Square-1 APDR Algs v1 Sheet

docs.google.com

Like I said, this was my first attempt at genning algs for square-1 so the algs might not be 100% optimized (I'm almost 100% sure someone more experienced will be able to find better algs for some of the cases), but they should be fairly decent, and I plan to continue to work on refining the alg list over time. If anyone wants to contribute better algs, please drop them in a reply to this thread!

 

[Cubing Forever]

wow I never though anyone would take this seriously lol
(I know that this was discussed a lot but we thought the algs would be bad. These do seem good though. 6 slice average is already pretty great since you have DF and DB solved(which is a hindrance imo))
about being as good as VDB and Lin, I do have my doubts(like parity PLL would be pretty bad(both for recog and exec))

(that being said, nothing can beat 3-look Vandenbergh(CSP-OBL-PBL))

 

[OreKehStrah]

wow I never though anyone would take this seriously lol
(I know that this was discussed a lot but we thought the algs would be bad. These do seem good though. 6 slice average is already pretty great since you have DF and DB solved(which is a hindrance imo))
about being as good as VDB and Lin, I do have my doubts(like parity PLL would be pretty bad(both for recog and exec))

(that being said, nothing can beat 3-look Vandenbergh(CSP-OBL-PBL))

What I envision the future of non-VB solving is using a few different block patterns and just knowing a bunch of blockbuilding algs. For example, imagine a Lin user builds the left block, and in the process DF and DB are solved. If they knew these algs, they just do the alg and immediately get PLL vs building the right block and then setting up into PLL/PLL+1.

But yeah, when the algs didn't exist I was like, "Fine, I'll do it myself" lol

 

[OtterCuber]

There is some serious hard work here (although I don't understand it yet). Great job, OreKehStrah!

 

[Cubing Forever]

What I envision the future of non-VB solving is using a few different block patterns and just knowing a bunch of blockbuilding algs. For example, imagine a Lin user builds the left block, and in the process DF and DB are solved. If they knew these algs, they just do the alg and immediately get PLL vs building the right block and then setting up into PLL/PLL+1.

Oh wow this is just like top-first pyraminx where you learn a bunch of different methods like oka, 1 flip, nutella and WO and do what's best depending on the scramble.

 

[SuperDuperSir]

Oh wow this is just like top-first pyraminx where you learn a bunch of different methods like oka, 1 flip, nutella and WO and do what's best depending on the scramble.

squan mehta-os
sounds so cursed

 

[IsThatA4x4]

How would other mehta variants hold up as a squan method? e.g 6CO, DCAL, etc.
It could work like mehta-OS does for 3x3, as you said OS could be quite a viable option in the future for Lin users.

 

[Reirto-RRNF]

How would other mehta variants hold up as a squan method? e.g 6CO, DCAL, etc.
It could work like mehta-OS does for 3x3, as you said OS could be quite a viable option in the future for Lin users.

What i can think is

6CP variant:
CS/CSP>Mehta-Style FB on the D layer>6CO(Put the rest of the corners to it layer)>6CP(Put the rest of the corners to it place)>L5E(Solve the last 5 edges)
Or
CS/CSP>Mehta-Style FB on the D layer>6C(Solve all corners)>L5E(Solve the last 5 edges)

CDRLL variant
CS/CSP>Mehta-Style FB on the D layer>DCAL(Put D layer corners to it place)>CDRLL(Solve all last layer corners)>L5E(Solve the last 5 edges)

 

[OreKehStrah]

How would other mehta variants hold up as a squan method? e.g 6CO, DCAL, etc.
It could work like mehta-OS does for 3x3, as you said OS could be quite a viable option in the future for Lin users.

DCAL could be useful potentially, but I don’t see a reason to no use the fastest alg to solve D corners regardless of U layer CP if you already know PLL+1.
In similar fashion, 6CP is not going to useful over doing normal FB then SB in terms of efficiency or speed in most cases.

 

[Reirto-RRNF]

Well should it named APDR method or Mehta method?

 

[Swagrid]

when mehta loses relevancy it just jumps over to a different event.

pyra mehta when?

 

[DuckubingCuber347]

when mehta loses relevancy it just jumps over to a different event.

pyra mehta when?

CubeRoot

www.cuberoot.me

ML4E is as close as you'll get.

 

[TipsterTrickster]

Cool method, I will definitely check it out as I am a lin user currently.

Also I noticed that there were a quite few cases that have much better algs for them.
6-5 alg 1, here is a 4 slicer: 1,3/2,-1/-2,-2/0,3/ (originally a 6 slicer)
2-6 alg 3, here is a 3 slicer: 4,3/-4,-1/4,1/ (originally a 6 slicer)
2-6 alg 5. here is a 3 slicer: 0,-1/0,3/0,-3/ (originally a 4 slicer)
5-2 alg 2, here is a 3 slicer: 6,-4/4,1/-4,-1/ (originally an 8 slicer)
5-2 alg 4, here is a 3 slicer: 4,0/-3,0/3,0/ (originally a 7 slicer)

there a probably some more that I didn't spot.

 

[eeeeeeeee]

sorry for the HUGGGGEEEE BUMP but what is the average slice count for this entire method?

 

[abunickabhi]

sorry for the HUGGGGEEEE BUMP but what is the average slice count for this entire method?

I guess you can write a small python code and find that out!