2-gen methods

[Devagio]

What are some amazing 3x3x3 2-gen methods you can come up with?
By 2-gen methods, I mean if a 3x3x3 cube has been scrambled using only say R and U moves, and needs to be solved using only R and U moves. You need to mention an estimate of the number of algorithms required to solve using your method, and whether inspection is necessary. The method can be as absurd and creative as you like.

Here are some common examples:

1: Square > F2L > 2GLL - Intuitive first two steps, 84 algs for the last step. Doesn't necessarily need inspection.
2: Square > LSLL - Intuitive first step, ~7000 algs for last step. Doesn't necessarily need inspection. (Edited)
3: 3 style - ~200 + ~150 algs for corners and edges. Inspection would help greatly.

 

[Imam Alam]

The method can be as absurd and creative as you like.

you may want to have a look at the second half of this method

a beginner version of this method would be to simply do CO and then EP

You need to mention an estimate of the number of algorithms required to solve using your method, and whether inspection is necessary.

number of algorithms: 1~4, depending on how you count (for me the number of algorithms is 0, since this method is intuitive for me)

inspection: I usually solve the entire cube with this method, so that the 2-gen part is in the middle of a solve, that means I don't have extra inspection time for the 2-gen part of the solve

edit: this was supposed to be the first reply in this thread haha, but it was kept in waiting for moderator approval for some reason

 

[mukerflap]

tripod

 

[Devagio]

tripod

I assume that means
Square > Square > tripod
How many algs is that?

 

[Username: Username:]

a

I assume that means
Square > Square > tripod
How many algs is that?

according to the wiki, it's 58 algs. https://www.speedsolving.com/wiki/index.php/Tripod_Method

 

[Devagio]

a

according to the wiki, it's 58 algs. https://www.speedsolving.com/wiki/index.php/Tripod_Method

That’s when CP is not done. In RU gen scrambles, the number of cases will decrease.

 

[ThisNameIsAlreadyTaken]

Not for speedcubing, just for fun:

1. Solve the remaining F2L pairs (or other "pairs" which are okay for the next step)
2. Orient U corners so that the cube is solvable by just doing U2's, R's (R, R', R2) R2's and U's (U, U', U2), using Sune and/or other algorithms.
That is to say, the colors on the U side must all be or U or D colors.
3. Intuitively solve the cube to a state where you can just use U2 and R2 to solve the cube and/or
4. Finish solving the cube intuitively

Number of algorithms: from none to 42 (COLL)

 

[Devagio]

Not for speedcubing, just for fun:

1. Solve the FR and BR pairs (or other "pairs" which are okay for the next step)
2. Orient U corners so that the cube is solvable by just doing U2's, R's (R, R', R2), using Sune and/or other algorithms.
That is to say, the colors on the U side must all be or U or D colors.
3. Intuitively solve the cube to a state where you can just use U2 and R2 to solve the cube and/or
4. Finish solving the cube intuitively without breaking the orientations

Number of algorithms: from none to 42 (COLL)

The number of algs will be 7, since for RU gen scrambles, COLL reduces to OCLL.
Still not sure how the method works though the idea seems quite interesting; can you explain deeper?

 

[ThisNameIsAlreadyTaken]

The number of algs will be 7, since for RU gen scrambles, COLL reduces to OCLL.
Still not sure how the method works though the idea seems quite interesting; can you explain deeper?

The idea is to go from <R, R2, U, U2> to <R2, U, U2> to <R2, U2> (spotted an error in my post: in step 2 you orient it so that you only have to do R2, U, U2). Those are different states of the cube which can be solved by using only the moves in "<>".
A way you can do that is how I explained. For the intuitive parts, you can first solve the corners (because they are more tricky), then solve the edges. For those you can use the simple (U2 R2)x3 "algorithm" to swap two opposite edges.

 

[Deleted member 51471]

EOCP223 + RB + 2GLL
EOCP223 would be really hard to plan though

 

[Devagio]

EOCP223 + RB + 2GLL
EOCP223 would be really hard to plan though

EOCP223 is already done because it’s a 2-gen scramble.
So your method is RB > 2GLL.

 

[Deleted member 51471]

EOCP223 is already done because it’s a 2-gen scramble.
So your method is RB > 2GLL.

Oh ok sorry I misunderstood the first post.

How about Square + Make Last Pair + 1LLsLL?

 

[I'm A Cuber]

lol is @Devagio just looking for faster ways to end a YruRU solve?

 

[Devagio]

Oh ok sorry I misunderstood the first post.

How about Square + Make Last Pair + 1LLsLL?

Nice.
By pair I guess you mean RB minus R U’ R’. That’s ~300 algs as far as I can see.

lol is @Devagio just looking for faster ways to end a YruRU solve?

Lol, that’d be a great takeaway.
I was just trying to search for any optimisations on RU gen scramble solving to help with YruRU, and to my surprise I found no documentation of any method or discussion about it.
Thought I’d take ideas because I feel my thinking is kind of restricted when it comes to RU solving. I’m sure there’s much more creative ways to do it than simple F2L>LL; the more people think about it, the more creative solutions we’ll get.
Plus, it’s really fun to think about it.

 

[Deleted member 51471]

Nice.
By pair I guess you mean RB minus R U’ R’. That’s ~300 algs as far as I can see.

Yeah.

 

[Devagio]

Not for speedcubing, just for fun:

1. Solve the remaining F2L pairs (or other "pairs" which are okay for the next step)
2. Orient U corners so that the cube is solvable by just doing U2's, R's (R, R', R2) R2's and U's (U, U', U2), using Sune and/or other algorithms.
That is to say, the colors on the U side must all be or U or D colors.
3. Intuitively solve the cube to a state where you can just use U2 and R2 to solve the cube and/or
4. Finish solving the cube intuitively

Number of algorithms: from none to 42 (COLL)

Step wise Reduction to HTR, interesting. Though in your specific approach, this is what I understood:
Step 1 is solving Right block.
Step 2 is doing OCLL
Now all that remains is EPLL, 4 cases. You’re saying “solve EPLL intuitively”

 

[WoowyBaby]

Domino Reduction !

 

[Username: Username:]

Domino Reduction !

Isn't Domino Reduction an FMC technique? it's not its own fully-fledged method right?

 

[Devagio]

Domino Reduction !

Looking for a hard-code way to do it, not an AI way; if that makes sense.
Maybe break it down into cases to achieve DR and solve further; shouldn’t be too hard since we are talking about such a restricted set of states.

 

[ThisNameIsAlreadyTaken]

Step wise Reduction to HTR, interesting. Though in your specific approach, this is what I understood:
Step 1 is solving Right block.
Step 2 is doing OCLL
Now all that remains is EPLL, 4 cases. You’re saying “solve EPLL intuitively”

You could do it like that, but not per se...
Step 1 is solving the FR and BR edges in the correct layer and simultaneously solve two corners in FRD and BRD so that the sticker colours match or are opposite to all the adjacent face colours.
Step 2 is OCLL
Step 3 is intuitive edge solving

BTW what is HTR?