The New Method / Substep / Concept Idea Thread

OKay I have revised a method, and have now used it, and had to look up the algs, and it does work.
Anyway...
Step 1-Solve a 2x2x3 block as in petrus
Step 2-Fix bad edges
Step 3-Solve your first layer using only R and U moves
Step 4-COLL
Step 5-Do a U perm to permute the edges to correct spot so that your B,L and D faces are all solved
Step 6- Do an edge cycle from blind solving to solve the cube (note: If done correctly, you only need to use an alg once).
Example Solve:
Scramble:U F2 U L2 B' L' R B2 L U' F B L F' U2 F' L' R' D' F2 B' D R F2 U
2x2x3 block- F' D' B D B D L U' L R U' R' U' L U L2 U2 L d
Bad edges- (I use a 2-look version) R2 U R U R' U' M' U R U' r
Solve first layer- R U R' U' R U' R' U R U' R'
COLL- (once again, 2-look) U' R U R' U R U2 R' U R' U L' U2 R U' R' U2 R L U'
Solve B,D and L-R2 U2 R2 U2 R2 U2
Solve the cube- R2 U R U R U' R' U' R' U' R'
Moves to solve-89 (I believe)
How is this? Im sure people could lower the movecount, but eh, I'm not there yet
 
Last edited:
2180161 said:
OKay I have revised a method, and have now used it, and had to look up the algs, and it does work.

How is this? Im sure people could lower the movecount, but eh, I'm not there yet
Click to expand...

It's good! Another advantage is that you can do step 3 on the OTHER first layer too (at the top left of the cube, instead of the bottom right.). This flexibility is a very good thing! nice idea
 
Thanks, It came to me last night, and wanted to see if it was possible first. Thanks for the feedback. The example solve works right? I did it again and couldn't but eh. BTW you only need 13 algs for a 2 look version of the last 2 steps(8 OLL, 2 CPLL). For bad edges, you dont need much. So is it worth it to do so, or are there too many moves?
 
Last edited:
It's a pretty good idea. you can always finish the first two steps in under 20 moves (easy).

And the third step is VERY fast.

I personally know full COLL, so step 4 can be done sub-2 if you know all the algs.

But steps 5 and 6, I wouldn't recommend breaking up the way you do. i would just combine them into one step (solving edges with commutators). because trying to solve three edges that are THAT close together is pretty awkward. (I see that you use 2-gen moves, which helps a bit).

Also, you don't actually have to orient ALL of the edges during step 2 (if you think about it). you only need to orient the F2L edges and the edges on B/D/L. then, at step 6, you will just have a few additional cases to deal with (none of which are difficult to solve with commutators or algs).
 
Step 1- Solve F2L -1 pair. Rotate so the unsolved slot is in RF
Step 2- Fix bad edges
Step 3- Solve first layer
step 4- COLL
step 5- Ua, Ub, H, or Z perm to solve L and B faces (AFTER using a BLD alg, but only if necessary)
Step 6-Using BLD solve algs, solve the cube. (You should only need to use 2 different algs, 3 for speed)
 
You could also try:

Step 3: Finish first layer
Step 4: COLL+1 (#algs ~= #COLL algs * 2)
Step 5: L5E (I haven't calculated the number of algs required for this step nor how decent this step would be)
 
2180161 said:
Step 1- Solve F2L -1 pair. Rotate so the unsolved slot is in RF
Step 2- Fix bad edges
Step 3- Solve first layer
step 4- COLL
step 5- Ua, Ub, H, or Z perm to solve L and B faces (AFTER using a BLD alg, but only if necessary)
Step 6-Using BLD solve algs, solve the cube. (You should only need to use 2 different algs, 3 for speed)
Click to expand...

bumping my idea, simply because of the lack of replies, and no other bumps on this thread for six days. Trying to keep it alive. Although it probably doesnt need help but... eh
 
2180161 said:
bumping my idea, simply because of the lack of replies, and no other bumps on this thread for six days. Trying to keep it alive. Although it probably doesnt need help but... eh
Click to expand...

I just dont see how this is more effective than something like mgls, petrus, heise, or OLL/PLL. Petrus fixes bad edges earlier when you have more freedom, and leaves you with less unsolved pieces to do intuitively. mgls just uses algs for the last slot, which is faster than intuition (even if it's possibly less cool) and leaves you with just a PLL case. Heise takes advantage of the empty slot to solve all edges and one or two corners, whereas you solve one corner and edge permutation. OLL/PLL is just fast, even if it requires a lot of algs. Basically, I don't see how your method is somehow more effective than any of the others I just listed. They all require fewer looks and are probably just always going to be faster. That being said, if you want to use this method, go ahead. It's not horribly inefficient, I just don't see it offering anything the other methods don't. Sorry.
 
2180161 said:
Isnt there a way to possibly lower movecount however in order to make it more efficient?
Click to expand...

Yes there is!

combine steps 1 and 2 (per petrus/heise)
combine steps 3 and 4 (so ~125 algs to solve 5 corners)
combine steps 5 and 6 (~100 algs to solve 5 edges, can be 100% 2 gen)

This might actually be a good method, if you're willing to learn ~225 algs.

Basically you solve EOF2L-1, then use 2 algs to solve the cube.

Better than CFOP? yes.

Better than Petrus? no.

Better than Roux? obviously.
 
You must log in or register to reply here.

Similar threads

D
The physics of Rubik's Cube to Substep Method Platform
2
Replies
33
Views
1,795
Christopher Mowla
Christopher Mowla
D
Incomplete Substep Method LL Megaminx
Replies
2
Views
304
Deleted member 27249
D
Dane man
Calculating Method/Substep Efficiency
2
Replies
27
Views
8,659
brododragon
brododragon
M
What method is this?
2
Replies
24
Views
6,126
~Adam~
~Adam~
AK51
2.51 world record used 33 moves, it shall be 28 moves using AK-cubie method
Replies
0
Views
236
AK51
AK51
Back
Top