# Zeroing LSLL (BT+): How to skip LL every time

#### Swagrid

##### Member
Very fun that the vast majority of cases for the ending step take two commutators, with an outstanding case or two that takes three

#### IsThatA4x4

##### Member
This looks awesome!
Can you post some example solves?

#### Swagrid

##### Member
This looks awesome!
Can you post some example solves?
Here's a quick one I whipped up

Got some poor cases through the solve, oh well. The OCELL here was lucky but as of rn they are quite underdeveloped, we only have old boomer RU gen algs that are quite long.

#### TheCubingCuber347

##### Member
Very fun that the vast majority of cases for the ending step take two commutators, with an outstanding case or two that takes three
Could you please point to the case that takes 3 commutators to solve? I fail to see that.

#### IsThatA4x4

##### Member
Could you please point to the case that takes 3 commutators to solve? I fail to see that.
With 5 corners left:
A pure 5 cycle needs 2
A 5-twist (impossible with this method) needs 4
Everything else (3c2t / 2c2c1t) needs 3 (1 to solve any twisted corners and 2 to solve the resulting 5 cycle for example)

#### Swagrid

##### Member
Could you please point to the case that takes 3 commutators to solve? I fail to see that.
Turns out I was just an idiot and it can be solved in two.

#### IsThatA4x4

##### Member
This could be used for megaminx right? Just would need mega OCELL algs and then the last step is intuitive and still requires a max of 2 commutators!

#### DarthDK

##### Member
A
This could be used for megaminx right? Just would need mega OCELL algs and then the last step is intuitive and still requires a max of 2 commutators!
Actually now that I think about it.....it could work

#### abunickabhi

##### Member
BT+ is a cool concept. Are you going to expand on the documentation? If yes, looking forward.

#### DarthDK

##### Member
A

Actually now that I think about it.....it could work
Just spent a couple minutes toying around with this on the megaminx....and it looks like it could be pretty viable. Unfortunately, looks like 3 commutators are possible because stuff like y perms take 3 times to return the cube back to a solved state

#### IsThatA4x4

##### Member
Just spent a couple minutes toying around with this on the megaminx....and it looks like it could be pretty viable. Unfortunately, looks like 3 commutators are possible because stuff like y perms take 3 times to return the cube back to a solved state
After OCELL, you are left with 6 corners. Because of the zeroing pair and the fact that all but 2 corners will be oriented (1 being the corner for the LS which cannot be in the correct position), no twisted corners can occur. That means you are left with:
3c3c (2 3 cycles = 6 corners), 2 commutators needed
5c (1 5 cycle), 2 commutators needed
2c2c (a double 2 swap = 4 corners), 2 commutators needed
3c (1 3 cycle), 1 commutator needed.

Correct me if I'm wrong (because above is how L6C would work on 3x3), but it should be the same for megaminx

#### zzoomer

##### Member
There are cases where you can have 3 comms on megaminx.

#### GenTheThief

##### Member
I still prefer zz-0

I am excited for this variant though - L3C reduction is a promising theory and this looks to be one of the better approaches.

#### PiKeeper

##### Member
Is this as fast as normal oll+pll?

#### GenTheThief

##### Member
Is this as fast as normal oll+pll?
No. LS + OCLL + PLL is still a faster variant. However, it's less creative and novel than this approach.
While BT+ isn't awful, and is actually one of the better variants IMO, there are definitely better and more utilitarian approaches if your end goal is speed.